BEYOND

FREUD
A Study of Modern Psychoanalytic
Theorists

Edited, by
JOSEPH REPPEN, Ph.D.
Copyright © 1985 Joseph Reppen

e-Book 2013 International Psychotherapy Institute

All Rights Reserved

This e-book contains material protected under International and Federal Copyright Laws and
Treaties. This e-book is intended for personal use only. Any unauthorized reprint or use of this
material is prohibited. No part of this book may be used in any commercial manner without
express permission of the author. Scholarly use of quotations must have proper attribution to
the published work. This work may not be deconstructed, reverse engineered or reproduced in
any other format.

Created in the United States of America

To Kyra, Eve, and Alexander

www.freepsychotherapybooks.org 4
CONTENTS
PREFACE

1 JOHN BOWLBY: AN ETHOLOGICAL BASIS FOR PSYCHOANALYSIS

Victoria Hamilton

2 GEORGE S. KLEIN: PSYCHOANALYTIC EMPIRICIST

Frederic J. Levine Ph.D. and Joseph W. Slap M.D.

3 ROY SCHAFER: SEARCHING FOR THE NATIVE TONGUE

Donald P. Spence Ph.D.

4 BENJAMIN B. RUBINSTEIN: CONTRIBUTIONS TO THE STRUCTURE OF

PSYCHOANALYTIC THEORY
Morris N. Eagle Ph.D.

5 EMANUEL PETERFREUND: THE INFORMATION REVOLUTION

Stanley R. Palombo M.D.

6 MERTON M. GILL: A STUDY IN THEORY DEVELOPMENT IN PSYCHOANALYSIS

Irwin Z. Hoffman Ph.D.

7 ROBERT LANGS: THE COMMUNICATIVE APPROACH

Zvi Lothane M.D.

8 HEINZ KOHUT: BEYOND THE PLEASURE PRINCIPLE, CONTRIBUTIONS TO

www.freepsychotherapybooks.org 5
PSYCHOANALYSIS
Hyman L. Muslin M.D.

9 MARGARET S. MAHLER: SYMBIOSIS AND SEPARATION-INDIVIDUATION

Anni Bergman and Steven Ellman Ph.D.

10 OTTO KERNBERG: PSYCHOANALYSIS AND OBJECT RELATIONS THEORY; THE

BEGINNINGS OF AN INTEGRATIVE APPROACH
Monica Carksy Ph.D. and Steven Ellman Ph.D.

11 WILFRED R. BION: AN ODYSSEY NINTO THE DEEP AND FORMLESS INFINITE

James S. Grotstein M.D.

12 PAUL RICOEUR: REPORTING, READING, AND INTERPRETING

Robert S. Steele Ph.D.

13 JACQUES LACAN: PSYCHOANALYST, SURREALIST, AND MYSTIC

Jeanine Parisier Plottel Ph.D.

14 ADOLF GRÜNBAUM: PSYCHOANALYTIC EPISTEMOLOGY

Barbara Von Eckhardt Ph.D.

NOTES ON CONTRIBUTORS

ABOUT THE IPI

www.freepsychotherapybooks.org 6
PREFACE
The wellspring of ideas that originated with Sigmund Freud are today being expanded by the
intellectual vitality and energy of a host of creative psychoanalytic thinkers. This volume presents the

work of 14 modem analytic theorists. The clear influence of Freud’s ideas is deeply reflected in various

ways throughout this book and, although many of the theorists presented are at varying degrees of
agreement with each other and Freud, they are all basically informed by the original genius of Sigmund
Freud. Indeed, the title of this volume, Beyond Freud, intends in no way to disparage the originality of

psychoanalysis. Instead, it intends to demonstrate how Freud’s thinking and how the Freudian text
have been used to expand ideas beyond Freud. That the work of two philosophers who have been

attracted to Freud is included is a living testament to that profound genius and vision.

Beyond Freud grew out of my interest in the evolution of psychoanalytic theory, the history of

ideas, and in my study of comparative psychoanalysis. This interest was expanded by my editorship of

the Review of Psychoanalytic Books. As editor of the Review I have witnessed an even greater expansion
of Freudian thought as well as an enormity of work in applied psychoanalysis informed by a Freudian

perspective.

The inclusion of these 14 theorists is not intended to diminish the contributions of others. Erik
Erikson, Melanie Klein, Anna Freud, the interpersonal school, many mainstream Freudians, the major

ego psychologists, Winnicott and other object relations theorists are not included in this volume

because their writing is either not recent or there is already a considerable body of literature on their

work. Not every scholar/theorist included in this book is a practicing psychoanalyst, but all are

modem, vital, and informed; and, most importantly, their work continues to enlighten, enrich, and
influence younger analysts and students. In fact, these 14 theorists are of such considerable intellectual

influence, an influence beyond clinical analysis, that my choice was made quite easily. Each theorist

www.freepsychotherapybooks.org 7
provides a unique vision of contemporary psychoanalysis that should endure for some time.

In most instances, the author of the chapter on the psychoanalytic theorist has known the

theorist about whom they write or has been deeply involved in their work as a part of their own

professional life. In one case, the author is a co-worker, in another a former analysand, and so on, so
that there is an intimate and deep connection.

While this book was in its planning stages, an Epilogue was to be included to show where the
future of psychoanalytic theory might be headed. However, in thinking about this over a long time I

decided to exclude an Epilogue and to leave speculations about future directions up to the reader. I felt
that it would be presumptuous to speculate as to where psychoanalysis is heading. This book is thus an

introduction to modern psychoanalytic theorists who have gone beyond Freud and an opportunity for

the reader to draw conclusions of their own. I hope that the reader will read this book in an open-
minded way, not as advocacy but as information.

Certainly, there is a clash of ideas and theories in this volume, and the welter of schisms,

schools, and factions in psychoanalysis are well presented. The natural science/hermeneutic debate
can be clearly viewed in this work. The Epilogue might have expressed a hope for a more pluralistic,

integrative psychoanalysis. Psychoanalysis is a new science compared to the older natural and physical

sciences, and, as an infant science, it is still evolving. Psychoanalytic theory fifty years from now may be

quite different than it is today, informed perhaps by computer models and other discoveries as yet
unknown in science and philosophy.

Finally, I would like to gratefully thank each of the authors who has written a chapter in this
book. They have done so with enthusiasm and intelligence, and represent the highest levels of
psychoanalytic scholarship. They are scholars writing about scholars. That there are so many
independent lines of thought within psychoanalysis is proof, perhaps, of all the possible ways in which

man may be viewed and to the incredible richness of the Freudian endeavor which inspires in so many
ways. Perhaps the seeming confusion we see may one day develop into a more integrative, informed,

www.freepsychotherapybooks.org 8
sophisticated, pluralistic psychoanalysis that might return us more deeply to the original power of

Freud within a truly modern context. In the meantime, I hope Beyond Freud will be informative of the

present provocative clash of ideas and of the dialogue and dialectic of Freudian inspired thought.

Joseph Reppen, Ph.D.

www.freepsychotherapybooks.org 9
1
JOHN BOWLBY: AN ETHOLOGICAL BASIS FOR
PSYCHOANALYSIS

VICTORIA HAMILTON

The work of an original thinker often calls to mind a key idea: Darwin’s “survival of the fittest,”
Einstein’s “relativity,” or Freud and “sexuality.” We associate John Bowlby with his lifelong study of the

crucial role played by attachment and its corollorary, loss, in human development. He has assembled

his major work in three volumes entitled Attachment (1969), Separation (1973), and Loss (1980).
Bowlby’s ‘Attachment Theory,’ together with the view of separation and mourning that it incorporates,
is as novel to the study of human relationships as Darwin’s theory was to the study of evolution. Yet

Bowlby’s (1979a) work is based upon and reflects the most obvious features of everyday life.

Family doctors, priests, and perceptive laymen have long been aware that
there are few blows to the human spirit so great as the loss of someone
near and dear. Traditional wisdom knows that we can be crushed by grief
and die of a broken heart, and also that a jilted lover is apt to do things that
are foolish or dangerous to himself and others. It knows too that neither
love nor grief is felt for just any other human being, but only for one, or a
few, particular and individual human beings. The core of what I term an
“affectional bond” is the attraction that one individual has for another
individual. (p. 67)

www.freepsychotherapybooks.org 10
Few would disagree with this statement. And yet, as with many new and simple ideas, we

encounter considerable resistance to its implications. Bowlby is a psychoanalyst and psychiatrist who

was trained in the Freudian tradition of psychoanalysis. Since 1946, when he assumed responsibility
for the Children’s Department at the Tavistock Clinic,

London (swiftly renaming it the Department for Children and Parents), Bowlby has focused his
research and therapeutic skills on the study and treatment of young children and their families. This

experience has provided him with the basis for both his theory of normal infant and child development

and a new view of pathology and its treatment. Although his work is enriched by fields such as
ethology, cognitive psychology and systems theory, Bowlby’s preoccupation with the joys and sorrows,

the hope and despair, incurred in the making, sustaining and breaking of affectional bonds, places his

contribution squarely within the arena of psychoanalysis. More than any other branch of medicine and

psychology, psychoanalysis claims to investigate the emotional life of man. Nevertheless, despite over
thirty years of research and teaching, Bowlby’s conception of attachment has not yet been integrated

into the discipline and still remains foreign to the thinking of most psychoanalysts.

In this chapter, I shall attempt to supply reasons for the resistance of psychoanalysts to

Bowlby’s thesis. Indeed, by reference to some of his most basic assumptions about human psychology,

Bowlby himself offers various solutions. Throughout his work, he stresses the over-riding importance
of the parameter “familiar/strange” in the development of human beings from the cradle to the grave.

From infancy on, we tend to orientate towards the familiar and away from the strange, a trait that has

survival value for human beings and other species. We change our beliefs with reluctance and would
rather stick with the familiar model. Ironically, psychoanalysts do not recognize that this “cognitive

bias” (Bowlby, 1980) is functional and tend to regard the preference for the familiar as regressive.

The painful nature of the material that Bowlby presses upon us also elicits resistance. The
reading of Separation and Loss is a test of endurance since both volumes spell out the grief to which an

analyst must bear witness if he is to meet the pathologies of despair and detachment. To support his
view of attachment and the repercussions of a disruption of affectional bonds, Bowlby draws on

www.freepsychotherapybooks.org 11
personal accounts of bereavement, on observations of children who have lost their parents either
temporarily or permanently, and on works of literature. It is Bowlby’s (1980) belief and experience

that “He oft finds med’cine who his grief imparts” (p. 172) and that, in psychotherapy, “the deep vase of

chilling tears that grief hath shaken into frost” (p. 320) must break. The therapist, like the poet, must
have a capacity to endure and express the suffering that antecedes its cure.

The crux of Bowlby’s thesis is that the pains and joys of attachment cannot be reduced to

something more primary such as the sexual or death instincts. Just as a child’s love for his mother does

not result from the gratification of his oral desires, so the heart-rending expressions of grief quoted by
Bowlby do not denote destructive or guilty wishes that have been repressed. They may simply describe

the painful process of healthy mourning.

It is impossible to think that I shall never sit with you again and hear you
laugh. That everyday for the rest of my life you will be away. No one to talk
to about my pleasure. No one to call me for walks, to go “to the terrace.” I
write in an empty book. I cry in an empty room. And there can never be
any comfort again. (Carrington, in Bowlby, 1980, p. 229).

Although many analysts fail to comprehend the relevance of Bowlby to the consulting room, his

ideas are rooted in the Freudian context. Although he departs radically from parts of the Freudian
tradition, he develops many ideas that Freud held to be important (particularly in his later life).
Throughout his work, Bowlby acknowledges this debt and quotes passages from Freud’s later work to

support the theory of attachment. In 1938, Freud describes the relationship of the child to his mother
as “unique, without parallel, laid down unalterably for a whole lifetime, as the first and strongest love-
object and as the prototype of all later love relations-for both sexes” (p. 188). In the 1940s and early

1950s, when Bowlby first published his observations on disturbances in children and young people
who had been separated from their parents, Freud’s theories provided a stepping-stone away from the
then popular stress on constitutional and inherited factors and gave him a framework with which to

emphasize the importance of mother-child relations. Moreover, the effects of World War II upon both
bereaved adults and young children in care spelt out, to all, the stark realities of separation and loss.

www.freepsychotherapybooks.org 12
Dorothy Burlingham and Anna Freud (1942) had reported on the suffering of the children in their care
at the Hampstead Nurseries, London, and James Robertson, a psychiatric social worker familiar with

their work, had begun a series of studies of children separated from their parents who were living in

residential nurseries and hospitals. The plight of these children was unmistakable and terrible. The
World Health Organization was interested in the many thousands of post-war refugees and

approached Bowlby to write a report on the mental health of homeless children. This report, entitled

Maternal Care and Mental Health, was published in 1951. It was later popularized and reissued under
the title, Child Care and the Growth of Love.

Child Care and the Growth of Love is a refreshing and readable book, full of observations,

anecdotes and practical advice. Since all the heavy, statistical material is omitted in the popular

version, the hypotheses advanced seem almost naive when viewed from the context of the
sophisticated and well-documented model of attachment we have before us today. In this early work,

Bowlby’s basic insight into the origins of pathology stands out loud and clear: maternal care in infancy

and early childhood is essential for mental health. The importance of this discovery, Bowlby (1953)
felt,

may be compared to that of the role of vitamins in physical health (p. 69)
...The outstanding disability of persons suffering from mental illness, it is
now realized, is their inability to make and sustain confident, friendly, and
cooperative relations with others. The power to do this is as basic to man’s
nature as are the abilities to digest or to see, and, just as we regard
indigestion or failing vision as signs of ill-health, so have we now come to
regard the inability to make reasonably cooperative human relations (p.
109).

In the intervening 40 years, psychoanalysts of varying orientations—Freudian, Anna Freudian,

Jungian, and Kleinian—have responded to many of Bowlby’s ideas in a piecemeal fashion. All would
acknowledge the importance of his work and, with few exceptions, would claim that the nature of the

mother-child relationship together with the vicissitudes of separation and loss, have significant

implications for therapeutic intervention. Nevertheless, the proportion of practicing psychoanalysts

www.freepsychotherapybooks.org 13
who have been able to grasp the larger picture of human relationships and development outlined by
the theory of attachment remains small.

In addition to the painful nature and unfamiliarity of Bowlby’s point of view, the alienation felt

by many psychoanalysts may proceed from an ambivalent and even negative attitude towards research
in the behavioral sciences. Bowlby’s theory depends more upon direct observation of attachment and

separation behavior than upon inferences drawn from the analysis of adults. Freud himself waged a

comparable battle with the behavioral sciences of his day in his search for knowledge of man’s mental

life. But now that psychoanalysis has been established for almost 100 years, this posture amounts to
little more than prejudice and exacerbates the isolation of psychoanalysis from related branches of

human psychology and biology. Psychoanalysts often argue that research, based upon the observation

of “external reality,” is irrelevant to analytic work, the domain of which is the exploration of “inner” or
“psychic reality.” Some psychoanalysts even argue that the study of normal infant and cognitive

development would impede their “intuition” into the unconscious phantasy life of the patient.

In my view, neglect of research findings has led to a fixation in the psychoanalytic theory of
development. The Victorian picture of children, implicit in Freud’s theory, has changed very little in the

century since psychoanalysis began. A dominant feature of this picture is of a withdrawn, asocial,
narcissistic and egotistical creature. Young children must be socialized into affectionate relationships

with others and induced to learn about the outside world through the frustration of their wishes and

the civilization of their instincts. As Freud (1905) said, “All through the period of latency children learn
to feel for other people who help them in their helplessness and satisfy their needs, a love which is on
the model of, and a continuation of, their relation as sucklings to their nursing mother” (pp. 222-223).

Through her care and affection, the mother “teaches” her child to love. One of the leading child
psychoanalysts of today, Margaret Mahler (Mahler, Pine, & Bergman, 1975), describes the newborn as
little more than a vegetable. Only “by way of mothering ... the young infant is gradually brought out of

an inborn tendency toward vegetative, splanchnic regression and into increased sensory awareness of,
and contact with, the environment” (p. 42).

www.freepsychotherapybooks.org 14
This statement, based upon direct observation, is totally inconsistent with the body of infant

research that has been assembled by the disciplines of ethology, developmental psychology,

anthropology and pediatrics. The contrasting picture of the infant, to which Bowlby has made a large
contribution, is of an alert and curious creature who becomes intensely attached and most sensitively

attuned to his or her mother. The full impact of human attachment seems almost as unpalatable to
psychoanalysis today as was Freud’s discovery of childhood sexuality. Bowlby’s insight into the conflict

between the methods of traditional psychoanalysis and conventional scientific research is that, like

workers in many other disciplines, the psychoanalyst must be capable of assuming two roles that
require two very different mental outlooks. Whereas the scientific attitude discourages personal

involvement and advises emotional detachment as a requisite for rigor and objectivity, the art of
psychotherapy requires a capacity for immersion and imagination.

In order to delineate some of the major theoretical implications of Bowlby’s research for the
discipline of psychoanalysis, I will focus on four aspects of his theory of attachment. These are (1)

instinct theory, control theory, and evolution; (2) the nature and function of attachment behavior from

infancy to old age; (3) normal and pathological processes of mourning in response to separation and
loss; and (4) psychoanalysis as art and science.

INSTINCT THEORY, CONTROL THEORY AND EVOLUTION

All studies of human behavior, except those based upon the most extreme theories of learning

and conditioning, posit certain basic behavioral patterns, which have traditionally been termed

instincts. Although there is disagreement about the nature of these basic patterns, all agree that the

term “instinctive” denotes those behaviors that are common to the members of a species and that are
more or less resistant to environmental influences. Bowlby’s model of attachment is built upon a

theory of instinct that is widely accepted by biologists and physiologists but differs radically from that

of traditional psychoanalysis. There is disagreement not only over the kind of instincts deemed
common to man—for example, instincts for sex or self-preservation— but also over the meaning of the

www.freepsychotherapybooks.org 15
term “instinct” itself.

The psychoanalytic concept of instinct derives from Strachey’s translation of Freud’s trieb. Some

psychoanalysts now consider that the translation of trieb as “drive” is a more precise rendering of
Freud’s thinking. Omston (1982) has pointed out that Strachey “clustered and clumped” Freud’s
wording into single Latin and Greek terms, thereby losing the subtleties of Freud’s distinctions. Freud

himself used the term instinkt quite selectively. Instinkt was more of a technical term and referred to a

precisely determined activity. Trieb, on the other hand, was used to refer to a “surging and rather
undifferentiated need” (Omston, 1982, p. 416). Thus, problems of translation have compounded the

confusions arising out of the psychoanalytic view of the instincts and of the behaviors and emotions to

which they supposedly give rise.

Like Freud, Bowlby defines the concept of instinct precisely. The contemporary concept,

proposed by biologists and ethologists, offers an alternative account of human motivation that has not
yet been incorporated into psychoanalytic theory. Even critics of the traditional view seem unaware

that a coherent alternative exists. In accordance with the scientific framework of his day, Freud used

the term to denote an inner motivating force or drive that operates as a causal agent. An instinct is
activated from within by an accumulation of stimuli and is terminated when the energy aroused flows

away. For example, the oral instinct is aroused by hunger and, when a mother nurses her baby, she
reduces the amount of pent-up libido (energy) to a tolerable level.

Bowlby substitutes the phrase “instinctive behavior” for the more common noun “instinct.” The

adjective “instinctive” is intended to be descriptive and leaves open the question of motivation. Human
behavior varies in a systematic way, and yet, as Bowlby (1969) notes, there are so many regularities of
behavior and certain of these regularities are so striking and play so important a part in the survival of

individual and species that they have earned the named ‘instinctive’ ” (p. 38). Bowlby (1969) describes
four main characteristics of behavior that traditionally have been termed instinctive:

a. It follows a recognizably similar and predictable pattern in almost all

www.freepsychotherapybooks.org 16
members of a species (or all members of one sex);

b. It is not a simple response to a single stimulus but a sequence of

behavior that usually runs a predictable course;

c. Certain of its usual consequences are of obvious value in contributing to

the preservation of an individual or the continuity of a species;

d. Many examples of it develop even when all the ordinary opportunities

for learning it are exiguous or absent (p. 38).

This account shows that the ethological view of instinctual responses is based upon a very
different dynamic to the Freudian view. First, the term “instinctive” always refers to an observable
pattern of behavior, which is activated by specific conditions and terminated by other specific

consummatory stimuli. For instance, attachment behavior in a child is readily elicited under certain
environmental conditions such as cold, bright light, sudden darkness, loud noise, the appearance of

strange or unexpected objects and under certain internal conditions such as fatigue, hunger, ill health,

and pain. Nearly all the behaviors elicited by these conditions are terminated by contact with and
responsiveness from the mother. Second, instinctive patterns are usually linked together and do not
occur in isolation. This means that a particular behavioral pattern is not linked causally to one

motivating system, but results from the coordination—or the lack—of a number of instinctual
responses. Integration is often achieved through the avoidance of various hazards, such as cold
weather, sharp objects, loud and sudden noises, and so forth. Here, the care and protection afforded by

mother plays a unique integrating function.

Third, many attachment behaviors are reciprocal and only function effectively within a social
system. For instance, an infant’s proximity-seeking behaviors are matched by the mother’s retrieving

behaviors. The latter resemble the child’s attachment behaviors in their biological function-namely,

protection from danger and survival. Indeed, in Bowlby’s estimation, the feedback system involved in

watching and visual orientation is more important than the oral instinctual behaviors emphasized by

psychoanalysis. Many attachment behaviors only make sense within a social context and have been

www.freepsychotherapybooks.org 17
suitably termed social releasers and social suppressors. Babbling, for instance, is most readily released
and increased by human faces and voices, particularly by the sight and sound of the mother. In general,
friendly responses such as smiling and babbling are easily elicited and reinforced by human stimuli.

The situation is usually reversed with respect to crying. Here, social stimuli are the main terminators
or suppressors. For instance, picking up and holding the infant is the most rapid terminator of crying

from nakedness. Rocking and rapid walking is the most effective suppressor of crying from loneliness,
although not of crying from pain, cold or hunger.

A more thorough exposition of the new concept of instinctive behavior requires a review of

changes that have occurred since Freud’s day in two other disciplines: one, the new field of cybernetics
(also referred to as systems theory, information theory or control theory), and the theory of evolution.

Most psychoanalysts have not followed these developments and thereby compound their

misconception of Bowlby’s work.

Since most analysts are unfamiliar with control theory, they are unable to grasp that Bowlby

offers an alternative theory of motivation. According to cybernetic theory, behavior is organized

homeostatically into systems that are activated by certain signals and terminated by others. This
model’s characterization of causation calls into question methods used by psychoanalysts in

determining the source of a patient’s pathology. The analyst attempts to reconstruct past events that
overdetermine current behavior in the life of his patient. Cybernetic explanation, on the other hand, is
always negative. In cybernetic explanation, we do not look for the cause of an event. Instead, we first

consider alternative possibilities and then ask what knocked these other alternatives out of the
running. The negative nature of cybernetic explanation is conceptualized by the term restraints. When
we look at a particular behavior pattern, we ask, What were the restraints that excluded alternatives

from the system? An excellent example of this distinction between restraints that are negative and

clues that are positive has been given by the anthropologist Gregory Bateson (1967):

For example, the selection of a piece for a given position in a jigsaw puzzle
is “restrained” by many factors. Its shape must conform to that of its

www.freepsychotherapybooks.org 18
several neighbors and possibly that of the boundary of the puzzle; its
colour must conform to the color pattern of its region; the orientation of its
edges must obey the topological regularities set by the cutting machine in
which the puzzle was made; and so on. From the point of view of the man
who is trying to solve the puzzle, these are all clues, i.e., sources of
information which will guide him in his selection. From the point of view
of the cybernetic observer, they are restraints (p. 400).

Zoologists and ethologists working in the field have used this restraint model of explanation for
a long time. The ethologist Niko Tinbergen (1972) has described the life of animals observed in their

natural habitat as “a multi-dimensional tightrope act” (p. 200). The fittest are those life forms that are

not eliminated by environmental pressures. Animals survive, reproduce and evolve within the

restraints of many variables. Success depends upon their capacity to cope with a bewildering variety of
obstacles. However, the healthy and happy man balks at such a suggestion. He does not feel that

negatives have governed his success. But the cybernetic model does not imply a tragic outlook. It does

not seek to explain why people behave as they do but why, at any one time, an individual behaves one
way rather than another.

In accordance with the cybernetic model, Bowlby (1969) suggests that we call the successful
outcome of an activated behavioral system goal corrected rather than goal directed. Human beings

constantly revise, extend, and check their working models of the environment and adjust their

behavior accordingly. As with the system of negative feedback in cybernetics, goal corrected systems
are designed to control behavior so as to adjust any discrepancies between initial instruction and

performance. This approach further implies “that no single adaptation is viewed as ideal; it is always

the compromise result of many different, and often conflicting, demands. When we analyze human
behaviour, we usually study one behavioral characteristic and one environmental pressure at a time”

(Hamilton, 1982, p. 11). We lose sight of the broader context. We may not see the competition between

conflicting activities or that different environmental pressures are dictating incompatible responses.

An event is not the outcome of a number of causes but the end product of a process of elimination of

many factors, none of which may be causally related to the final outcome.

www.freepsychotherapybooks.org 19
Psychoanalysts are particularly interested in emotional ambivalence and conflict behavior, such

as that between approach and withdrawal. Bowlby points out that the activation of such conflicts often

will result in so-called compromise behavior. The individual plays out fragments of two different
systems. Within this class of compromise behavior I would include tics or stereotyped and

inappropriate gestures. An action may be dissociated from its context or cut across by a contrary
action. A person may signal his attraction to another only to negate his own initiative by rejecting the

other’s response. This compromise behavior represents an exchange between two people. Originally

the two incompatible sequences of behavior were enacted by two separate people-for instance a
mother and her child. Behavioral systems may also be “redirected” to another goal in the way that has

been traditionally described as displacement. Actions or feelings are, in Bowlby’s terms, redirected from
one person on to another person or object. We should not equate compromise behavior with neurosis,
however. Even a curious, securely attached child may exhibit both clinging and exploratory behavior in

a novel environment. Tinbergen (1972) discusses the compromises that birds must negotiate between
safety and nourishment. Camouflage protects the birds while they are motionless. However, they must
eat. As Tinbergen (1972) said: “While they could feed more efficiently if they never had to freeze, and

would be better protected against predators if they never had to move, they can do neither, and
selection, rewarding overall success rather than any isolated characteristics, has produced
compromises” (p. 154-155).

Both cybernetics and psychoanalysis concern themselves with the information carried by

events and objects rather than with the event or objects themselves. They do not investigate forces,

drives, impacts, or energy exchanges except as they confer meaning to concrete events. There is no

information or communication without context. A word acquires meaning in the larger context of the
utterance, which again has meaning only in a relationship. For instance, the schizophrenics’ “word

salad” becomes intelligible through study of the communicational patterns and relationships within his

family. Communication between psychoanalyst and client acquires meaning in the context of the

transference relationship.

www.freepsychotherapybooks.org 20
In addition to goal correction, systems theory discovers another restraint governing behavior.

“Nothing”—that which is not—can exert a powerful influence. Information theory refers to this as a

zero message. Zero messages, such as absence or unresponsiveness, may cause extremely strong
emotions. Bateson (1970) gives as an illustration of a zero cause “the letter which you do not write” (p.

452). This letter “can get an angry reply.” Increasingly, psychoanalysts now look at the negative
trauma, which is not an event such as incest, the birth of a sibling, or an aggressive attack, but rather is

a lack of psychological connection. This focus emerges from the many studies of the narcissistic

personality disorder over the past decade. A prolonged absence of connectedness and responsiveness
often lies at the root of the despair, apathy, and detachment that characterize attachment pathologies.

An evolutionary perspective is necessary to make sense of the last two characteristics of

instinctive responses listed by Bowlby (see p. 7): first, that the consequences of a sequence of

instinctive responses may contribute to the preservation of an individual or the continuity of a

species, second, an instinctual response may develop in an individual “even when the ordinary

opportunities for learning it are exiguous or absent” (Bowlby, 1969, p. 38). Clinicians usually do not

consider the evolutionary context. Frequently, their background is in medicine and they have not been
trained to interpret the behavior of individuals within the context of species survival. Moreover,

clinical practice does not provide much opportunity to acquire this perspective.

Consideration of the evolutionary perspective should affect psychoanalytic theory and practice.

What sort of inferences do clinicians make when they are unable to explain behavior in terms of the

individual, including his or her particular history and present environment? The practitioner usually
concludes that such behavior is caused by “constitutional” factors or that it is a bizarre externalization

of the patient’s phantasy life. Melanie Klein’s concept of persecutory anxiety, a state that gives rise to

all sorts of destructive phantasies and is itself consequent upon the workings of the death instinct,

exemplifies this sort of explanation. Bowlby’s interpretations of children’s fears and phobias spring
from the evolutionary view of attachment and entail a very different theory of explanation to that of

the death drive.

www.freepsychotherapybooks.org 21
The new concept of instinctive behavior, familiar to ethologists for many years, makes the

traditional antithesis between innate and acquired characteristics unnecessary. Every class of behavior

is a product of the interaction of genetic endowment and a specific environment. Although the human
species has a tremendous capacity for versatility and innovation, many behavioral systems only

operate in their environment of evolutionary adaptedness. Moreover, this adaptedness is a property

not only of the individual but of the population.

THE NATURE AND FUNCTION OF ATTACHMENT BEHAVIOR FROM INFANCY TO OLD

AGE

In 1958, Bowlby published “The Nature of the Child’s Tie to His Mother.” This paper marked the

second major juncture in Bowlby’s intellectual development and was pivotal to many of the ideas that

he pursued later. In this work, the somewhat anecdotal comments and observations of “Child Care and
the Growth of Love” (1953) coalesce into a coherent theory. He no longer underpins his argument with

references to Freud but rather to ethology and the new evolutionary point of view. Bowlby had not yet

incorporated the systemic approach, but his terms now belonged to that framework.

This paper confronted the various psychoanalytic schools with a direct challenge. Despite

subsequent developments in Bowlby’s attachment theory, this critique remains a valuable summary of
many of the major differences between the attachment and psychoanalytic viewpoints. Much of the

paper is devoted to an informative and incisive account of four traditional theories of the child’s tie to

the mother:

1. The theory of secondary drive. According to the view, the baby becomes interested in and

attached to his mother as a result of her meeting the baby’s physiological needs. In due
course, the infant learns that she is also the source of gratification.

2. The theory of primary object sucking. The infant has an inbuilt need to relate to a human
breast, to suck it, and to possess it orally. In due course, the infant learns that attached to

www.freepsychotherapybooks.org 22
the breast is a mother with whom he or she must develop a relationship.

3. The theory of primary object clinging. There exists an inbuilt need to touch and cling to a
human being, and this need is on a par with the need for food and warmth.

4. The theory of primary retum-to-womb craving. Infants resent their extrusion from the womb

and seek to return there.

In this early account of attachment, Bowlby includes the theory of primary object clinging. This
view had been proposed by Imre Herman in Budapest and adopted by Alice Balint and Michael Balint.

Together with W. R. D. Fairbairn and Donald Winnicott, they were to become prominent members of

the British Middle Group. This school of psychoanalysis, to which Bowlby belongs, shares with him an

emphasis on bonding and object relating over gratification or the avoidance of pain. Bowlby (1958)
lists five instinctual responses—sucking, clinging, following, crying and smiling. These five instinctual

responses “serve the function of binding the child to the mother and contribute to the reciprocal

dynamic of binding the mother to the child.... Unless there are powerful in-built responses which
ensure that the infant evokes maternal care and remains in close promixity to his mother throughout

the years of childhood, he will die” (p. 369).

Bowlby remarks upon the vast discrepancy between formulations springing from empirical

observation and those made in abstract discussions. He points out that leading child analysts with first
hand experience of infancy, such as Anna Freud, Dorothy Burlingham, Melanie Klein, Therese Benedek,

and Rene Spitz, are apt to describe such interactions in terms suggesting a primary social bond. In their

theorizing, however, they persist in describing social interaction as secondary.

Bowlby’s paper also challenges the traditional psychoanalytic view of orality. First, he
downplays both sucking and the primary orientation towards the mother’s breast. He argues that

psychoanalytic theory is fixated on this response and that clinging and following play a more central
role in later disturbance. Both Bowlby and Margaret Mahler emphasize the importance in the

www.freepsychotherapybooks.org 23
ontogenesis of pathology of disturbances arising during the second half of the second year. In Mahler’s
view, the rapprochement phase of the separation-individuation process is particularly stormy because
the child’s growing independence conflicts with the continuing need for mother’s care and control.

Bowlby focuses more upon the mother’s rejection of the child’s clinging and following. He also points
out than an infant’s oral behavior has two functions: attachment as well as feeding. Western culture

has overlooked the fact that the infant spends more time in nonnutritional sucking than in feeding.
Whereas traditional psychoanalysis views oral symptoms as regressive to an earlier, more infantile

stage of development, Bowlby interprets such disturbances as displacements. Within the context of

attachment, oral symptoms designate the substitution of a part for a whole. They chronicle the splitting
off of feeding from the rest of a relationship. Compulsive thumb sucking might express a frustrated

attachment or even a displacement of the nonnutritional aspect of feeding itself, rather than regression

to some autoerotic stage.

In similar fashion, Bowlby distinguishes sexuality from attachment in loving (traditionally

called libidinal) relationships. Although these two systems are closely related and share some of the

same patterns of behavior, they are distinct. Their activation varies independently of one another. Each

directs itself towards a different class of objects and is sensitized at a different age.

As already noted, Bowlby holds attachment behavior to be instinctual and on a par with the
pursuit of sex and food. He expresses his fundamental difference with traditional psychoanalysis most
clearly in his interpretation of the complex repertoire of behaviors with which the infant maintains

proximity to his or her caretaker. For Bowlby, the primary function of this behavioral system is to
insure the child’s survival and protection from predators. Most psychoanalysts do not think in such
terms. Although they do enumerate various primitive mechanisms of defense, none of these concern

the survival of the individual in his or her environment. The term “defense” is used to refer to

psychological processes, such as projection, projective identification, idealization, denial, splitting,

repression, and regression. Bowlby follows traditional usage by reserving the word “defense” for

psychological defenses and using the word “protection” when talking about the function of attachment

www.freepsychotherapybooks.org 24
behavior. Since this distinction does not exist in traditional theory, the child’s tenacious efforts to keep
close to his mother are not usually seen as related to a social system in which they elicit reciprocal
responses of retrieval and picking up. Rather, the child’s demands for closeness are interpreted

onesidedly as a denial of separateness or as an attempt to omnipotently control the “object” for the
fulfillment of narcissistic wishes. The infant is seen as using crying and clinging as weapons of control.

Some analysts even believe that the infant’s clinging and grasping and enjoyment of being held indicate

a wish for return to the womb.

In general, the evolutionary viewpoint leads us to interpret a great deal of human behavior,

whether of children or mature adults, as cooperative rather than self-seeking. Since the unit of

biological adaptation is the social group and not the individual, survival depends upon cooperation.

Psychoanalysis has concentrated on those behavioral systems that are limited by particular events,
such as orgasm, eating, or elimination, and has ignored systems such as attachment whose goal is a

constant state. Attachment theorists believe that only an indirect relationship exists between such

interactions as feeding, weaning and toilet training, and a healthy attachment. Attachment is neither a
developmental stage nor a system limited by an event. Its continuing set-goal is a certain sort of

relationship to another specific individual. Attachment is regarded as the product of a control system

that maintains homeostasis by means of behavioral rather than physiological processes. The
maintenance of proximity between child and mother is a kind of environmental homeostasis. As

Bowlby points out, there are many alternative ways of maintaining this homeostasis. However, the
organization that controls these behaviors is conceived as permanent and central to a child’s
personality. This organization is never idle. As Bowlby (1969) says: “In order for a control system to

perform its function effectively it must be equipped with sensors to keep it informed of relevant
events, and these events it must continuously monitor and appraise.” In the case of an attachment
control system, the events being monitored fall into two classes: one, potential danger or stress

(external or internal), and two, the whereabouts and accessibility of the attachment figure.

The distinction between behavioral systems that are limited and those that are ongoing affects

www.freepsychotherapybooks.org 25
the conception of development. As one would expect, current views of human biology and control
theory differ greatly from those of psychoanalysis. The traditional model implies that there is one
developmental line. Personality disorders derive their form from stages that were normal at some

earlier phase of life. In normal development, the individual is thought to progress through the oral,
anal, phallic and genital stages. If fixations occur, the person “regresses” back down the ladder. Thus,

the various disorders of later life repeat phases of healthy childhood. The diagnostician considers the

resolutions and fixations appropriate to each stage in order to decide whether the adult before him or
her suffers from a pregenital, anal-sadistic, narcissistic, borderline, oedipal or neurotic disturbance.

Bowlby’s model, drawn from control theory and ethology, proposes that at birth, there exists a

large array of potential pathways. Development progressively diminishes these alternatives. We

should look not for the cause of a fixation but at the restraints that lead an individual to choose one
alternative over another. Returning to Tinbergen’s analogy, healthy development resembles the

adjustments that a tightrope walker must make continuously in order to maintain his or her balance.

Either excessive sensitivity or insensitivity to environmental changes will cause the tightrope walker’s
downfall. In human development, sensitivity from birth allows for maximum adaptability to the social

environment. This biological perspective, which stresses the cooperative nature of human behavior, is

opposite to and contradicts Freud’s view that avoidance and withdrawal precede approach behavior.
According to Attachment Theory, avoidance and withdrawal are most readily activated when the infant

is able to discriminate the familiar from the strange.

Questions about the ontogenesis of mental disorder raise the problem of how to measure
attachment. Initially, theorists sought to measure normal or abnormal behavior by reference to the

strength of the attachment between the individual and his or her childhood attachment figures.
However, these reseachers soon noted that intense attachment did not necessarily indicate a good or
harmonious mother-child relationship. Paradoxically, attachment behavior can be most intense when a
mother discourages or threatens her child’s need for proximity. The traditional viewpoint might

diagnose such a child as perverse or masochistic. But the child’s stubbornness makes systemic sense if

www.freepsychotherapybooks.org 26
his or her instinctual apparatus is geared toward proximity as the means of survival. The threat of
withdrawal would redouble the child’s efforts. Fear stimulates attachment behavior. A victim will often
develop a strong attachment to the person who causes his or her suffering, especially if, as in the case

of a young child, there is nowhere else to turn. Loss of an attachment figure is the child’s foremost fear.

Research on attachment shows that the two most important variables in the creation and

maintenance of a secure attachment are the sensitivity of a mother’s responsiveness to her baby’s

signals and the amount and nature of interaction between the two. Degrees of security or insecurity
provide the yardstick by which we measure a healthy attachment. Consequently, Bowlby (1973) has

substituted the term “anxiously attached” for the traditional description of an insecure child as

overdependent. Clinging behavior, illustrative of anxious attachment, has often been described as

jealous, possessive, greedy, immature, overdependent, or intensely attached. Bowlby’s concept of

anxious attachment respects the natural desire for a close relationship without pejorative

connotations.

In addition to a child’s protest and upset over his mother’s departure, researchers now regard

various other correlations as indicative of the security of an attachment. Foremost among these are the
child’s behavior upon reunion with the mother, and comparison of his behavior at home with his or her

behavior in a strange (often experimental) setting. Anxiously attached children often fail to greet their
mothers upon return. Furthermore, they are less exploratory than their secure counterparts, not only

in a strange situation but also at home in their mothers’ presence. Ainsworth and Bell (1970a) have

correlated children’s ambivalence in a strange situation with general ambivalence in the home
environment. Ambivalent children tend to resist contact when picked up and to ask to be picked up

when they are set down. They do this whether at home or in a strange environment. Logically, one
might expect proximity-seeking behavior to be incompatible with exploration. However, Ainsworth,
together with other attachment researchers, have noted that most children do not explore
constructively when avoiding contact. Avoidant children tend to move around hyperactively or to

alternate uncomfortably between avoiding and seeking contact. In addition, children who resist

www.freepsychotherapybooks.org 27
contact, are often more angry, aggressive and disobedient than children for whom contact is
pleasurable.

Bowlby correlates the development of “puzzling phobias” (see Freud, 1962, p. 168) with

anxious attachment. When a child is unable to communicate directly his fears about separation, he may
try to redirect or displace onto animals or other puzzling objects the anxieties he feels in relation to his

parents. He may be furious and terrified that the parent will desert him, but he dares not express such

feelings lest by so doing he provokes that which he most fears. Instead, he complains about something
else, or he may have temper tantrums that express both rage and fear. Bowlby reinterprets Freud’s

case of Little Hans in this light.

In volume one of Attachment, Bowlby suggests that five main classes of behavior should be
considered in any attempt to assess the attachments of a child. These are:

1. Behavior that initiates interactions, such as greeting, approaching, touching, embracing,

calling, reaching, and smiling.

2. Behavior in response to the mother’s interactional initiatives that maintains interaction (all

the initiating behaviors plus watching).

3. Behavior to avoid separations, such as following, clinging, and crying.

4. Exploratory behavior, as it is oriented toward the mother.

5. Withdrawal or fear behavior, especially as it is oriented toward the mother.

None of these considerations fit the Freudian picture of the infant or young child, which
describes the infant as being enclosed in a state of primary narcissism, “shut off from the stimuli of the

external world like a bird in an egg” (Freud, 1911, p. 220). The child’s object relations are seen as

www.freepsychotherapybooks.org 28
minimal. The contrasting view of attachment theorists points to the quality of mother-infant
interaction, which is built up out of communication ‘games’ as well as proximity-maintaining
behaviors. The success or failure of this mutual endeavor is crucial to the arousal of a baby’s interest in

the first weeks of life. Indeed, Ainsworth and Bell (1970b) have correlated the attachment behavior of

1-year-old children placed in a strange situation with the extent to which they had been permitted to
be an active partner in the feeding situation as 3-month-old infants. Such findings suggest that the
mother’s ability to conceive of the relationship as a partnership affects the development of both

attachment and exploration.

One fascinating detail of this research, which again contradicts the primary narcissism

hypothesis, pertains to fluctuations in the responsiveness of each partner to the initiatives of the other.
The infants responded on every occasion when the mother initiated interaction. However, whereas

some mothers were encouraged by their baby’s social advances, others evaded them; where some
mothers were made more solicitous by their child’s crying, others became more impatient. By the time
the children’s first birthday was reached, the magnitude of the differences between one pair and

another could hardly be exaggerated.

Two other researchers, David and Appell (1969), describe, at one extreme, a pair who
interacted almost continuously throughout the baby’s waking hours, and, at the opposite extreme, a

pair who were hardly ever together, mother occupying herself with housework and largely ignoring
her daughter. In a third pair, mother and son spent much time silently watching each other while each
was engaged in some private activity. Such findings suggest that mothers play a much larger part in

determining interaction than do infants. For instance, although initially there is little correlation
between a baby’s crying and a mother’s responsiveness, by the end of the first year, a baby cared for by
a sensitive, responsive mother cried much less than one cared for by an insensitive or unresponsive

mother.

One of the strengths of attachment theory, initiated by Ainsworth (1982) and Bowlby (1982) is
that it has stimulated a very able group of developmental psychologists to make such empirical studies

www.freepsychotherapybooks.org 29
of socioemotional development. These studies would be extremely useful to psychoanalysts,
particularly those working with children and young people.

As is only too obvious to the layman, a child’s pattern of attachment usually correlates with the

way his mother treats him. By preschool age, this matrix will have become a function of the child
himself or herself. This internalization or, in Bowlby’s terms, “cognitive map” of attachment may also

correlate with the child’s participation in the regulation of his or her care and mothering. Bowlby

likens the regulation of mothering to the regulation of food. Both mothers and professional people

often ask whether or not a mother should meet her child’s demands for her presence and attention. If
she gives in on mothering, will this encourage the child to demand that she give in on everything else?

Will the child ever become independent? Bowlby (1969) responds with an answer which he tells us is

“now well known”:

From the earliest months forward it is best to follow a child’s lead. When
he wants more food, it will probably benefit him; when he refuses, he will
probably come to no harm. Provided his metabolism is not deranged, a
child is so made that, if left to decide, he can regulate his own food-intake
in regard to both quantity and quality. With few exceptions, therefore, a
mother can safely leave the initiative to him…Thus, in regard to mothering
—as to food—a child seems to be so made that, if from the first permitted
to decide, he can satisfactorily regulate his own “intake.” Only after he
reaches school years may there be occasion for gentle discouragement. (p.
356)

By 4 to 5 years of age, the child’s capacity to consider another person’s point of view provides

additional clues to the status of the child’s goal-corrected partnership. Another variable by which we
can measure the security of an attachment is a child’s resiliance. A child whose background state is one

of anxious attachment will have few resources to draw on when faced with untoward and stressful
circumstances. In conclusion then, the organization of attachment, which is initially labile, becomes

progressively more stable. This development may be cause for optimism or concern.

www.freepsychotherapybooks.org 30
Let us now consider what the attachment model implies for the growth of self-reliance.

Psychoanalysts have looked at development as a linear progression from a state of dependence to one

of independence. This has distorted our understanding not only of dependence in childhood but also of
independence in adulthood. For Bowlby, self-reliance goes hand in hand with reliance upon others.

Confidence in the attachment figure and in the self are built up together. Indeed, the capacity to rely on
others when occasion demands and to know upon whom it is appropriate to rely is essential for true

self-reliance. Many people have confused self-reliance with the kind of independence that Bowlby

characterizes as compulsive caregiving and compulsive self-sufficiency. The compulsive caregiver and
the fiercely self-sufficient person will experience their own needs for love and care through,

respectively, administering to others or apparently needing nothing. Bowlby believes that a person’s
success in finding appropriate people to help him or her through hard times depends upon childhood
experiences. This ability holds a special importance for dealing with a serious loss. A major

determinant of reaction to loss is the way the bereaved’s attachment behavior was evaluated and
responded to by the bereaved’s parents-whether they could share his or her fears, unhappiness, and
grief or whether he or she had to bear sorrows alone. The solitary child has a hard time finding a

comforting shoulder in later life. Such people shun the thought and disavow the need for solace. What
children learn to expect in the nature of comfort from their parents determines in large part whether,
as adults, bereavement will make them sad or whether it will overwhelm them with despair and

depression.

NORMAL AND PATHOLOGICAL PROCESSES OF MOURNING IN RESPONSE TO

SEPARATION AND LOSS

“The great source of terror in infancy is solitude” (James, 1890). A similar sentiment was

expressed indirectly in a poem quoted by Bowlby that was written by an 11-year-old girl whose

parents were abroad for some years:

The beauty of love has not found me

Its hands have not gripped me so tight

www.freepsychotherapybooks.org 31
For the darkness of hate is upon me
I see day, not as day, but as night.

I yearn for the dear love to find me

With my heart and my soul and my might
For darkness has closed in upon me
I see day, not as day, but as night.

The children are playing and laughing

But I cannot find love in delight
There is an iron fence around me
I see day, not as day, but as night.

Bowlby could not study attachment without encountering the suffering that ensues from the

breaking or disruption of affectional ties. In the years between the publication of “The Nature of the

Child’s Tie to His Mother” in 1958 and Attachment in 1969, Bowlby published five papers on
separation anxiety, grief and mourning in infancy and early childhood, processes of mourning, and

pathological mourning. The publication of Attachment was followed in a similar fashion by the second

and third volumes in the series, Separation (1973) and Loss (1980). The latter two volumes, based on
the attachment model, again provide a very different picture of human responses to separation and

loss than that of traditional psychoanalysis. Their central and simple thesis is that, just as attachment is
the primary source of well-being in human beings, so loss is the major source of suffering.

Bowlby looks at human loss and distress on two levels: first, the inevitable grief, anger, and

despair that result when ties are broken, and second, the ways we organize ourselves to deal with

these painful and often conflictual feelings. Just as in his study of affectional ties Bowlby first searched
for regularities in the attachment behaviors common to human beings, so Bowlby detects prototypical

responses to loss and separation. The uniformity of these responses makes sense in the context of the
theory of attachment and the evolutionary framework.

By the time Bowlby wrote Separation and Loss the most common successive responses to loss—

www.freepsychotherapybooks.org 32
protest, despair, detachment—had been well documented by other authors, foremost among whom
were James and Joyce Robertson. Although many psychoanalysts had recognized that separation from
loved ones is a principal source of anxiety, there was still considerable reluctance to assimilate this

simple formula into clinical practice. In addition, Freud’s influence had led to the belief that the
processes of both adult and childhood mourning and normal and pathological mourning differed

considerably. Bowlby pointed out, however, that, as in the case of attachment, there is considerable
similarity between the mourning of children and of adults and that many of the responses to loss that

had hitherto been regarded as neurotic were quite natural. Attachment, unlike dependence, remains as

an organizational system throughout life; so grief, even in its normal course, has a long duration. A
bereaved person may experience for a long time an insatiable yearning for, and an “irrational” but

natural striving to recover, the lost person. These feelings may return intermittently for the rest of the

individual’s life.

Although most attachment theorists would now characterize the three phases of protest,

despair and detachment as typical of normal mourning in both children and adults, in fact an additional

initial phase is usually described as well as—depending on whether the loss is final or temporary—a

fifth and final phase. Prior to the protest and angry attempts to recover the lost object, most people
experience a sense of numbness and disbelief. During this period, bereaved individuals must adjust all
their expectations and beliefs. Whereas psychoanalysis uses the term “denial” to describe the state of
disbelief, Bowlby renames it “selective exclusion.” The fifth stage, experienced only when loss is

temporary, is characterized by extremely ambivalent behavior upon reunion with the lost person. This
can be demonstrated by a lack of recognition and absence of all emotional affect at one extreme and, at

the other, by clinging, acute fear of being left, and bursts of anger lest the person desert again.

Bowlby links the three most common reactions to loss—protest, despair, and detachment—

with three processes, all of which contain considerable potential for future disturbance. These are
separation anxiety, grief and mourning, and defense. Separation anxiety is a reaction to the danger or

threat of loss; mourning is a reaction to actual loss; and defense is a mode of dealing with anxiety and

www.freepsychotherapybooks.org 33
pain. As with attachment, the outcome of these responses depends largely on the ways other people
respond to the feelings of the bereaved person.

Following Freud, most psychoanalysts have concentrated exclusively upon the last of the three

phases-detachment and defense. Although Freud and Melanie Klein accorded a central place to anxiety
in everyday life, neither recognized that separation anxiety was as primary as, for instance, castration

or persecutory anxiety. W. R. D. Fairbairn and Ian Suttie were the first psychoanalysts to assign a

primary status to separation anxiety. Not until Freud’s seventieth year, in Inhibitions, Symptoms and
Anxiety (1926), did he perceive that separation and loss were principal sources of psychopathology.

Hitherto, Freud had linked anxiety to fears of castration, to the harshness of the superego, to

aggression, and to the death instinct. Even analysts such as Anna Freud and Melanie Klein who

remarked on the universal distress shown by infants and young children when their mothers were
absent continued to ask, Why are they anxious? What are they afraid of? Many ingenious explanations

have been proposed to answer these questions: the birth trauma, signal anxiety, anxiety consequent on

repression of libido, persecutory and depressive anxiety, and guilt about aggressive impulses.

Bowlby has made various suggestions as to why psychoanalysts have found it so very difficult to
conceptualize in theory that which they so clearly observe. First, Bowlby makes the common

observation that the psychoanalytic theory of normal development is almost entirely based upon work
with adult patients. Obviously, in clinical practice, the psychoanalyst is constantly preoccupied with the

understanding of defenses that, although once useful for survival, are now obsolete. When these

findings are projected back onto the theory of infant and child development, we find an imbalance
towards a study of the mechanisms of defense, and an ignorance of the normal child’s expressions of

loss, grief, and anxiety.

Second, traditional psychoanalysis assumes that a child does not seek out other people for their
own sake but only as containers or modulators of tension, anxiety, aggression and so forth, or as

sources of gratification. This tenet discourages the idea that a child might react directly to the absence

of a loved one.

www.freepsychotherapybooks.org 34
Third, Bowlby believes that the lack of distinction between cause and function has not only

harmed psychoanalytic theory in general but also that this confusion particularly impedes its

understanding of the anger that so often follows a loss. This anger is caused not just by the separation.
Bowlby believes that its function is to recover the lost person. Not only do anger and reproach ensure

the person’s return, they also threaten him so that he or she dare not desert again. In a responsive
mother-child relationship, the child’s anger is often very effective. The aggressive wishes not only

express the simple desire to hurt the person who has inflicted pain and suffering, but they are also

intended to punish the person for desertion and to reinstate proximity.

Fourth, Bowlby makes another distinction between guilt and grief in response to loss. Freudian

and Kleinian theory lose track of the difference between these two responses. Grief and mourning are

expressions of depressive guilt. Guilt is a “natural” reaction to loss. For Bowlby, grief covers an

amalgam of emotions—anger, anxiety, and despair. Guilt, on the other hand, may often signify
displacement and may result from an angry reproach against the self instead of the lost person. When

the expression of natural feelings, such as yearning, anger, and reproach, are discouraged (which is

very often the case, particularly when the bereaved are young children), these feelings can be
redirected either to third parties or on to the self. When reinforced socially, these displacements can

generate various pathological behaviors, such as denial of permanent loss with sustained secret beliefs

in reunion or vicarious caregiving and sympathy for other bereaved persons. Repressed yearning can
lead to compulsive wandering, depression, and suicide. Depressed people often tend to idealize their

attachment figures. In traditional theory, idealizations are often thought to mask aggressive and

destructive phantasies. According to Bowlby, however, such depressed people, particularly children,

may be entertaining two completely incompatible models of relationship—their own and that of their
caretakers. When circumstances are favorable, however, anger, reproach, and yearning fade following

their expression to the appropriate person. The mourner finally accepts that his loss is permanent and

that his or her feelings are nonfunctional. These responses are then succeeded by a period of

disorganization and almost unbearable grief. However, if this grief is expressed to and understood by

others, it can lead to reconnection with the world and “a relieving sweet sadness may break through”

www.freepsychotherapybooks.org 35
(Bowlby, 1963, p. 7).

Fifth, Bowlby makes a crucial distinction, ignored in traditional theory, between “natural” and
“reasonable” fear. This distinction affects our understanding of separation anxiety and of the responses

to actual or threatened loss. Following Freud, psychoanalysts have concluded that when anxiety is not

related to real danger, it signifies a neurosis. Absence per se does not seem to threaten life or limb.

However, as we noted earlier, the zero message exerts just as much influence as its positive
counterpart. Even among mature adults, mourning often is mixed with acute and “irrational” terror.

Nearly all bereaved persons report symptoms such as insomnia and fear of being alone or of going to

strange places. All these feelings are natural to separated children. The loss of a secure base threatens

both children and adults as much as physical assault. This phenomenon, Bowlby (1973) notes, prompts
the psychoanalyst to engage in “a prolonged hunt for some primal danger situation” (p. 169). The

analyst concludes that the expressed fear is not the real fear. So many of the fear stimuli that affect us
seem inappropriate in the modem context. We don’t see too many saber-toothed tigers these days!

Nevertheless, it is perfectly natural for a young and vulnerable child to fear the existence of dangerous

creatures. All children exhibit some fear of the dark, of being alone, of loud or sudden noises, of bright
lights and of looming objects, particularly when these appear in combination. Bowlby points out that
these same phenomena frighten the same child much less if they occur when the child is with an older,

trusted person. All these fears are viewed by Bowlby and other ethologists as natural. They contribute

to survival in the environment of evolutionary adaptedness. As Bowlby notes, these fears still hold
their survival value. Although in the city we need not worry too much about wild animals, we still need

to remain alert to danger. City children are vulnerable to traffic accidents, for example, and many city

parents worry about the risk of criminal assault. Besides, the fear of wild and dangerous animals is still

reasonable in many parts of the world. Even in Los Angeles, a young child growing up in certain hillside
areas must treat his environment with some caution. Chances are they are sharing the hill with a family

of coyotes and the odd rattlesnake. Fears are often ordered hierarchically. For instance, children will

follow their mothers in the face of dangerous traffic rather than risk separating from her. When we

investigate a fear that has become unmanageable, we would consider its evolutionary context before

www.freepsychotherapybooks.org 36
making our interpretation. Otherwise, we risk taking up arms against a mechanism of survival.

Sixth, Bowlby’s concept of defense—renamed selective exclusion— also reflects the systemic
approach. In the normal course of events, we exclude a vast proportion of information from

consciousness. This protects our attention from distraction and overload. The selective exclusion of

information is as necessary and adaptive as the reduction of flexibility that follows from specialization.

Both contain the potential for maladaptation, however. Persistent exclusion is usually maladaptive; nor
does automatic attachment and attachment behavior necessarily contribute to survival. Change can be

economical, but it is difficult; and correction requires skilled attention. Bowlby (1980) also stresses the

diversionary role of defensive activity, “for the more completely a person’s attention, time and energy

are concentrated on one activity and on the information concerning it, the more completely can
information concerning another activity be excluded” (p. 66). Any activity—work or play—can be

undertaken as a diversion. The only psychological requirement is absorption. Much defensive

exclusion is related to suffering. A response is disconnected from its context in an interpersonal

situation and relocated upon the self. This gives rise to symptoms such as hypochondria, guilt and

morbid introspection. For

Bowlby, no system is more vulnerable to defensive exclusion than attachment. For instance,
pathology may develop if defensive exclusion continues beyond the initial stages of bereavement.

The evolutionary context makes Bowlby’s theory of attachment and mourning seem simple,

even blindingly obvious. Human beings come into the world genetically biased to develop certain

behaviors that, in an appropriate environment, result in their keeping close to whoever cares for them.
This desire for proximity to loved ones persists throughout life. Only when children feel secure in their

primary attachments can they go out with confidence to explore and make the most of their world.

PSYCHOANALYSIS AS ART AND SCIENCE

www.freepsychotherapybooks.org 37
If we consider the development of psychoanalysis over the past nine decades, we find that new

discoveries have rarely led to consolidation, let alone to critical discussion. Indeed, the extreme

subjectivism of many psychoanalysts generates one quarrel after another. Psychoanalysts avoid
rational methods of discrimination between rival hypotheses. It seems that any interpretation can be

supported from within the terms of any one theory. Relationships between analysts are not usually
built on the pursuit of a common, though tricky, endeavor, but on loyalty to a particular faith. Does

such and such an analyst believe in Freud, Jung, or Klein? Does this analyst practice “real” analysis?

And yet, the whole edifice of psychoanalysis—its theory of development and its theory of cure
—depends upon the assumption that adult pathology stems from problems, real or phantasied, in

infancy. An outsider might then assume that psychoanalysts would keep up with the findings of those

disciplines to which their field is most closely related-in particular, cognitive and developmental

psychology and human biology. Surely, analytic research would benefit from the study of infants and
children in natural settings. Instead, psychoanalysts tend to fall back on so-called veridical

reconstructions of infancy gathered from the clinical material of adults. The paucity of observational

studies diminishes the number of independent variables with which to correlate analytic
reconstructions. Inevitably, the psychoanalytic theory of normal personality development has

remained weak and open to criticism.

By contrast, a minority of psychoanalysts, such as John Bowlby, James Robertson and Christoph

Heinicke, are attuned to the methodological limitations of retrospective research. As a result, they have

undertaken various prospective studies that follow the behavior of children about to undergo
experiences of separation from their mothers. Since analysts of all orientations seem to agree that

separation in childhood plays an undeniable role in adult pathology, one might expect that the

observation of separation behavior in a variety of settings could offer a fertile ground for intergroup

study. In my opinion, those analysts who have undertaken such studies manifest greater agreement
over their findings, show less of a propensity for schisms, and have found it easier to maintain a more

open and scientific attitude towards the work of their colleagues. An analyst’s views of infancy

www.freepsychotherapybooks.org 38
crucially affects his interpretations of unconscious material. It makes a great deal of difference to the
patient whether the analyst sees him or her as a bundle of id impulses, a raging orally fixated infant, a
frustrated narcissistic self, a thinking and curious creature, or a victim of a broken attachment.

In a recent article, “Psychoanalysis as Art and Science,” Bowlby (1979b) draws attention to two
very different aspects of the discipline of psychoanalysis: the art of psychoanalytic therapy and the

science of psychoanalytic psychology. In so doing, he emphasizes on the one hand, the distinctive value

of each and, on the other, the gulf that divides them “in regard both to the contrasting criteria by which
each should be judged and the very different mental outlook that each demands” (p. 3). As Bowlby

observes, this distinction is not confined to psychoanalysis. “It applies to every field in which the

practice of a profession or a craft gives birth to a body of scientific knowledge-the blacksmith to

metallurgy, the civil engineer to soil mechanics, the farmer to plant physiology, and the physician to the
medical sciences. In each of these fields the roles differentiate. On the one hand are the practitioners,

on the other the scientists, with a limited number of individuals attempting to combine both roles. As

history shows, this process of differentiation often proves painful and misunderstandings are frequent”
(p. 3). Bowlby attributes much of the confusion in psychoanalysis to the lack of differentiation of these

two roles. He contrasts the roles of practitioner and research scientist under three headings and uses

the case of medicine as an example:

1. Focus of study. The practitioner aims to take into account as many aspects as possible of each

and every clinical problem with which he must deal. This requires him to draw on any

scientific principle that may appear relevant and also to draw on his own personal
experience of the condition in question. The research scientist must have a very different

outlook. He aims to discern the general patterns underlying individual variety and,
therefore, ignores the particular and strives for simplification. He also tends to
concentrate on a limited aspect of a limited problem.

2. Modes of acquiring information. In his role of giving help, the practitioner is permitted access

to information of certain kinds that are closed to the scientist. He is permitted to

www.freepsychotherapybooks.org 39
intervene and privileged to observe the consequences of such interventions. The
research scientist, however, has the advantage of enlisting new methods to cross-check

on observations made and on hypotheses born of older methods.

3. Mental attitudes—scepticism and faith. If he is to be effective, a practitioner must have faith.

He must be prepared to act as though certain principles and certain theories were valid.

He is likely to choose between various theories on the basis of his own experience. As

Bowlby points out, such faith is not a bad thing in clinical practice. A great majority of

patients are helped by the practitioner’s faith and hope. The very lack of these qualities
may make many excellent research workers ill suited to be therapists. The scientist, on

the other hand, must exercise a high degree of criticism and self-criticism. In his world,

neither the data nor the theories of a leader, however much personally admired, may be
exempt from challenge and criticism.

Bowlby (1979b) believes that it is only by recognizing these differences that the strengths of

each role can be used to fullest advantage “or that any person can occupy both of them with any hope
of success” (p. 5). The repercussions of Bowlby’s view are serious because it calls for a reversal of the

set adopted by a great number of psychoanalysts—namely, unquestioning faith in a theory and

scepticism in their practice. Bowlby’s cross-checking of the reports of adult patients with observations

of young children should reduce the analyst’s scepticism of his patient’s memory. For example, Bowlby

takes very seriously the reported threats of separation made by parents to their children. Not only
should the analyst cross-check his findings with those of neighboring disciplines, but he must be able
to review his work critically outside his consulting room, either by taking notes or by detailed

discussion of case notes or tape recordings with his colleagues. Bowlby proposes that analysts might
keep a detailed record of the responses of their patients before and after each weekend, each vacation,
and each unexpected interruption of the sessions, with an equally detailed record of how the analyst

dealt with them. This would enable the analyst to check the repertoire of responses a given patient
presents on these occasions, and also the changes in response the patient presents over time.

www.freepsychotherapybooks.org 40
In my view, the medical bias in psychoanalysis has led not only to neglect of the two roles

required of the research psychoanalyst, but also to an underestimation of the art of psychoanalysis.

Fearful of his emotional responses to the patient and of his imaginative powers, the analyst, aiming to
maintain a “scientific” attitude, may remain aloof, neutral and dissociated from the interactions with

his patient. The art of psychotherapy, according to Bowlby (1979b) requires “all the intuition,
imagination and empathy of which we are capable. But it also requires a firm grasp of what the

patient’s problems are and what we are trying to do” (p. 12). For instance, analysts who are not

prepared to meet the heavy burdens of dependence should be careful about their choice of patients. In
order to have such a firm grasp of the patient’s problems, questions of etiology and psychopathology

should be clarified and the practitioner should be informed of the whole range of family experiences
that evidence shows affect the development of the child. Although medical science is competent to deal
with this area of psychopathology, it eschews the use of imagination, and psychoanalysis has suffered

accordingly. In addition, analysts have followed Freud in his equation of imagination with phantasy
and the creative process with sublimation. Like Freud, many analysts continue to regard art as an
anarchic process motivated by sublimation. They continue to ignore the skills and rules that are

involved in every creative process.

Bowlby’s work is a testimony to the skills of imagination, immersion and objectivity. Not only

does his trilogy present a simple point of view based on the distillation of a vast array of research, it
also portrays the extraordinary depth of feeling of a unique individual. Few psychoanalytic books

evoke the utter grief, despair and loneliness that bereaved persons, particularly children, have

suffered. Most psychoanalytic texts prefer to discuss the stereotyped defenses against feeling—

aggression, projection, denial, and so forth. One might conclude that few psychoanalysts are
themselves capable of suffering the depths of anxiety and sadness that are only too painfully obvious to

all those who have worked with young children.

…Dick…told him about his own father’s death, which had happened when
Dick was a child at Dublin, not quite five years of age. “That was the first
sensation of grief,” Dick said, “I ever knew.... I remember I went into the

www.freepsychotherapybooks.org 41
room where his body lay, and my mother sat weeping beside it. I had my
battledore in my hand, and fell a-beating the coffin, and calling papa; on
which my mother caught me in her arms, and told me in a flood of tears
papa could not hear me, and would play with me no more...And this,” said
Dick kindly, “has made me pity all children ever since and caused me to
love thee, my poor fatherless, motherless lad.” (Thackeray, H. E., in
Bowlby, 1980, p. 265)  

REFERENCES

Ainsworth, M. D. S. (1982). Attachment: Retrospect and prospect. In C. M. Parkes & Joan Stevenson-
Hinde (Eds.), The place of attachment in human behavior (pp. 3-31). New York: Basic
Books.

Ainsworth, M. D. S., & Bell, S. M. (1970a). Attachment, exploration and separation: Illustrated by the
behavior of one-year-olds in a strange situation. Child Development, 41, 49-67.

Ainsworth, M. D. S., & Bell, S. M. (1970b). Some contemporary patterns of mother-infant interaction in
the feeding situation. In A. Ambrose (Ed.), Stimulation in early infancy (pp. 133-171).
London: Academic Press.

Bateson, G. (1967). Redundancy and coding. In Steps to an ecology of mind (pp. 399-410). New York:
Ballantine Books, 1972.

Bateson, G. (1970). Form, substance and difference. In Steps to an ecology of mind. New York: Ballantine
Books.

Bowlby, J. (1951). Maternal Care and Mental Health. Geneva: World Health Organization.

Bowlby, J. (1953). Child Care and the Growth of Love. Harmondsworth, England: Penguin Books.

Bowlby, J. (1958). The nature of the child’s tie to his mother. International Journal of Psycho-Analysis,
39, 350-73.

Bowlby, J. (1963). Pathological mourning and childhood mourning. Journal of the American
Psychoanalytic Association, 11, 500-41.

www.freepsychotherapybooks.org 42
Bowlby, J. (1969). Attachment and Loss: Vol. 1. Attachment. London: Hogarth Press.

Bowlby, J. (1973). Attachment and Loss: Vol. 2. Separation. New York: Basic Books.

Bowlby, J. (1979a). The making and breaking of affectional bonds. London: Tavistock.

Bowlby, J. (1979b). Psychoanalysis as art and science. International Review of Psychoanalysis, 6, 3-14.

Bowlby, J. (1980). Loss: Sadness and depression. London: Hogarth Press.

Bowlby, J. (1982). Attachment (rev. ed.). London: Hogarth Press.

Burlingham, D., & Freud, A. (1942). Young children in wartime. London: Allen & Unwin.

Burlingham, D., & Freud, A. (1944). Infants without families. London: Allen & Unwin.

David, M., & Appell, G. (1969). Mother-child relation. In J. G. Howells (Ed.), Modern perspectives in
international child psychiatry (pp. 98-120). Edinburgh, England: Oliver & Boyd.

Freud, S. (1905). Three essays on the theory of sexuality. Standard Edition, 7, 135-243.

Freud, S. (1911). Formulations on the two principles of mental functioning. Standard Edition, 12, 215-
226.

Freud, S. (1926). Inhibitions, symptoms and anxiety. Standard Edition, 20, 77-175.

Freud, S. (1938). An outline of psychoanalysis. Standard Edition, 23, 144-207.

Hamilton, V. (1982). Narcissus and Oedipus: The children of psychoanalysis (p. 11). London: Routledge &
Kegan Paul.

James, W. (1980). Principles of human psychology. New York: Holt, Rinehart & Winston.

Mahler, M., Pine, F., & Bergman, A. (1975). The psychological birth of the human infant (p. 42). New
York: Basic Books.

www.freepsychotherapybooks.org 43
Omston, D. (1982). Strachey’s influence: A preliminary report. International Journal of Psycho-Analysis,
63, 409-426.

Robertson, J. (1953). Some responses of young children to loss of maternal care. Nursing Times, 49,
382-386.

Tinbergen, N. (1969). Ethology. In The animal in its world—Explorations of an ethologist, 1932-1972

(Vol. 2) (pp. 130-160). London: Allen & Unwin.

Tinbergen, N. (1972). Early childhood autism—An ethological approach. In The animal in its world—
Explorations of an ethologist, 1932-1972 (Vol. 2) (pp. 175-199). London: Allen & Unwin,
1973.

www.freepsychotherapybooks.org 44
2
GEORGE S. KLEIN: PSYCHOANALYTIC EMPIRICIST

FREDERIC J. LEVINE, PH.D.
JOSEPH W. SLAP, M.D.

INTRODUCTION AND BIOGRAPHY

By the time of his sudden death in 1971 at age 53, George S. Klein had already made many

compelling contributions to psychology and psychoanalysis. He was then working on what was

probably his most important contribution, a reexamination and attempt at restatement of basic

psychoanalytic theory which was published posthumously as Psychoanalytic Theory: An Exploration of

Essentials (1976) under the editorship of Merton M. Gill and Leo Goldberger. Although this book was in

a still incomplete form, many consider that its daring sweep and powerful insights will have an

enduring impact on psychoanalysis.

That Klein was a man of great personal magnetism, energy, and leadership ability, as well as a
scientist and theoretician, is given ample testimony by the numerous activities his friends, students,

and colleagues have dedicated to his personal and professional memory. Preceding each meeting of the
American Psychoanalytic Association is a meeting of the George S. Klein Research Forum, dedicated to
the advancement of research in psychoanalysis; the main lecture at the annual meeting of

www.freepsychotherapybooks.org 45
Psychologists Interested in the Study of Psychoanalysis (a group of American Psychological Association
members) is the George S. Klein Memorial Lecture; and annually in his beloved Stockbridge,

Massachusetts, researchers and psychoanalysts meet in the George S. Klein-David Rapaport Study

Group. In addition, there have been a number of panels, symposia, and memorial publications
dedicated to the advancement of Klein’s work (Gill and Holzman, 1976; Mayman, 1982; Reppen, 1980).

George Klein was born in Brooklyn, New York, in 1919. He was educated at the City College of

New York and Columbia University, where he developed what was to be a lifelong interest in the study

of perception. After receiving his doctorate, Klein served during World War II in the United States
Army Air Force, where he performed statistical studies and co-authored several reports on selection,

diagnosis, and prediction of outcome in patients.

In 1946, Klein joined the staff of the Menninger Clinic in Topeka, Kansas, under the supervision

and instruction of David Rapaport. The few years he spent there were to have a profound impact on the
rest of his intellectual life. At that time, the Menninger Foundation provided the best available clinical

psychological training to nonphysicians and was alive with the clinical and theoretical ferment
stimulated by Rapaport and his colleagues. To this exciting atmosphere were attracted a number of

people, including Philip Holzman, Herbert Schlesinger, Lester Luborsky, and many others who have
continued to be among the foremost leaders in psychoanalytically oriented clinical psychology, as well

as a number of outstanding medical psychoanalysts. Klein quickly became an important figure and

guiding spirit in this group. He studied Rapaport’s diagnostic psychological testing procedures and
took some patients in psychotherapy, but his main contributions at the Menninger Foundation were in
the area of research. In this first major phase of his career, as a result of the joining of his experimental

background with his exposure to psychoanalysis, Klein and his colleagues produced a series of studies
of individual consistencies in perceptual and cognitive behavior that added the terms “cognitive
control” and “cognitive style” to the technical lexicon. The fact that Klein was the leader in this research

was obscured by the alphabetical listing of the authors of the major comprehensive publication on the
subject (Gardner, Holzman, Klein, Linton, & Spence, 1959; see also Holzman, 1982).

www.freepsychotherapybooks.org 46
In 1949, Klein began a personal psychoanalysis while still at the Menninger Foundation. In

1950, he went to Harvard as a visiting professor and continued analysis there. While at Harvard, Klein

became interested in an organismic view of psychology in which neurological mechanisms would be

included along with the psychological. He later came to explicitly reject this position and made

significant theoretical contributions in his attempt to disentangle the quasi-neurological speculations

of metapsychology from the clinical theory of psychoanalysis.

In 1952, with Robert R. Holt, a former colleague at the Menninger Foundation, Klein founded

the Research Center for Mental Health at New York University, where he remained for the rest of his
career.

This center became the heart of an outstanding graduate program in clinical psychology and
clinical research laboratory, producing research on the interface between psychoanalytic and

experimental issues. Klein also started his own clinical practice at this time and began to turn to more

purely psychoanalytic theoretical concerns in his writings. He wrote his paper on consciousness (Klein,
1959a), which developed yet another perspective on the ways in which ego processing of perceptual

and cognitive data can vary independently of drives and needs, and began to develop his theoretical

critique of the psychoanalytic drive theory. His clinical experience broadened as he became affiliated

with the Austen Riggs Center in Stockbridge, Massachusetts, and his efforts to distinguish the
psychological from the metapsychological within psychoanalytic theory increased.

During this time, Klein founded the Psychological Issues monograph series, which continues to

be the major publication vehicle aimed at fulfilling the goals that Klein (1959b) enunciated in his first

issue: “To develop its theoretical potentialities psychoanalysis must scrutinize data from all fields of

psychological and psychiatric inquiry” (pp. iii-iv). Klein also continued his experimental investigations
of such varied fields as dream content and the effects of drugs and cultural deprivation. He undertook a
formal psychoanalytic education and was graduated from the New York Psychoanalytic Institute.

There he was a leader in obtaining the right of nonmedical graduates to have full privileges to conduct
a psychoanalytic practice. During this final period of his life, Klein’s primary scholarly interests moved

www.freepsychotherapybooks.org 47
more fully in the direction of clinical psychoanalytic theory, and he produced the papers (and papers in
progress) that were published in the posthumous volume on Psychoanalytic Theory (1976).

In reviewing Klein’s odyssey from research to psychoanalytic theorist, certain themes

consistently appear. First is Klein’s energy, innovative talent, and leadership ability. A second theme is
the persisting influence of his academic psychological studies of perception. These became the vehicle

through which he first expressed his interest in the ego’s autonomous role in directing behavior, in

contrast to theories stressing needs or drives as the main controlling forces. This view, heavily

influenced by David Rapaport, ultimately evolved into Klein’s criticism of the drive theory of
psychoanalysis. Somewhat surprisingly, perhaps, along with this criticism Klein also produced a telling

critique of the tenets of ego psychology, many of which had been formulated by his mentor, Rapaport.

It may be said that as Klein absorbed the insights of ego psychology, he at first found them liberating,
but later began to find them constricting and limiting as his appreciation of clinical data and theory

grew. Even in his final works, however, Klein’s emphasis as a psychoanalytic theorist was always on

the broad group of phenomena that make up what had been called the ego—those autonomous
processes, structures, and motives that he believed were important contributors to personality

functioning independent of the driving power of sensual needs. In this respect, he was trying to
broaden the scope of psychoanalytic inquiry beyond the limited data base provided by the patient on

the couch.

A third major theme in Klein’s work is his consistent emphasis on theoretical rigor and
precision, which led him ultimately to see many flaws in the classical drive and structural theories of
psychoanalytic metapsychology. As part of his lifelong effort to integrate various branches of

psychology, Klein proposed substituting for parts of metapsychology several concepts from the work
of Piaget and cognitive psychologists, which he felt would simplify psychoanalytic theory and make it
closer and more responsive to the data of observation.

A further, minor trend was Klein’s enduring interest in creativity and the artistic process. His

wife, Bessie Boris Klein, is a painter and Klein, too, enjoyed painting at times. He periodically returned

www.freepsychotherapybooks.org 48
to a consideration of the interrelationship of needs and ego processes in the “reparative” work of the
creative act.

COGNITIVE CONTROLS: PSYCHOANALYTIC EGO PSYCHOLOGY IN THE RESEARCH

LABORATORY

When Klein went to the Menninger Foundation following World War II, a prominent trend in

research on perception was the “New Look” (Postman, Bruner, & McGinnies, 1948), a group of studies

demonstrating that motives or needs could influence and significantly alter the registration and
judgment of perceptual data. Prior to that time, experimental psychologists had investigated the formal

details of perceptual and thought processes, but had assumed that these functions operated stably,

regardless of the individual’s purposes and need states. In contrast, this new group of investigators
found that in some situations, drives would “sensitize” the individual to perceive stimuli related to the

drive state; in others, suppressive effects were observed as a result of “perceptual defense.” Data in

these studies were notoriously inconsistent, and individual differences in these effects were
conspicuous but had not been explained. In Klein’s (1958) view, these findings were having an

exaggerated impact:

There was at the back of our minds a feeling that while motivation-in-
perception studies were rectifying older sins of omission, they were also
assuming that if only a drive is intense enough it can bend any or all
cognitive structures to its aim. While no one committed himself blatantly
to such a statement, the drift of empirical work seemed to be moving
steadily toward it. Some way had to be found in theory of providing for
effective processing without renouncing the possible pervasiveness of
motivational influence upon thought [pp. 87-88].

Klein’s response, through his research, was to demonstrate that while it is true that needs and
motives influence perception, these influences vary from individual to individual, as a result of

differences in the preexisting structural characteristics of style or pattern of thinking. He and his group

identified a number of what they believed to be intra-individually consistent patterns of perceiving and

www.freepsychotherapybooks.org 49
thinking, which could be detected in the laboratory, and showed that these patterns shaped ways in
which motives or needs influenced or distorted perception. These cognitive structures—first called

cognitive attitudes and later cognitive controls—were thought of “as ways of contacting reality,

whereby one’s intentions are coordinated with the properties, relations, and limitations of events and
objects” (Klein 1958, p. 88). In an early influential study, Klein (1954) examined the effects of thirst on

perception of objects that had to do with thirstiness and drinking, comparing the performance of

thirsty and nonthirsty subjects on the same tests. He divided both groups according to their
performance on a cognitive test intended to detect contrasting ways in which people typically deal

with distracting, intrusive feelings, such as thirstiness. Klein found that the different cognitive attitudes

identified by that test did cause characteristically different kinds of distortions in thirsty subjects.1

Although the “New Look” studies constituted one of the first areas in which psychoanalytic
ideas (i.e., the importance of needs and motives) had an impact on research in perception, Klein

showed that impact to be one-sided-considering only the influence of drives (like the early

psychoanalytic id psychology) and not the role of the coordinating and controlling structures of the
ego. His research, which was conceptualized according to the ego psychology of Heinz Hartmann, Erik

Erikson, and particularly his teacher, David Rapaport, was intended to correct this imbalance.

With his collaborators, Klein investigated various aspects of the cognitive controls to flesh out

the understanding of their roles in ego functioning. Underlying the specific perceptual attitudes that

were initially identified, Klein (1958) believed, were ego structures of broad generality and
significance for personality functioning, which seemed “to reflect highly generalized forms of control as

likely to appear in a person’s perceptual behavior as in his manner of recall and recollection” (p. 89).

Studies of these control principles explored their possible relationships to other ego processes,
particularly the classical defense mechanisms, patterns of personality organization as identified in

projective testing, modes of handling stress, and learning and intellectual ability (Gardner, Holzman,

Klein, Linton, & Spence, 1959; Gardner, Jackson, & Messick, 1960; Holzman, 1962; Holzman & Klein,
1956; Klein and Schlesinger, 1951). Klein and his coworkers (Holzman & Klein, 1956) tended to

www.freepsychotherapybooks.org 50
assume that while each cognitive control might undergo an epigenetic development, these structures
probably had their roots in constitutional givens—what Hartmann (1939) called apparatuses of
primary ego autonomy. Thus, they speculated, cognitive control patterns might form an important part

of the constitutional matrix that determines the individual’s character structure, reliance on particular
defense mechanisms, and choice of particular symptoms and psychopathological patterns (see, for

example, Shapiro, 1965, pp. 13-14). Although Klein (1958) stated that “cognitive attitudes seemed to

resemble what psychoanalysts have called character defenses” (p. 88), he believed they were not
actually defenses, resulting from conflicts, but precursors or predisposing conditions, which

contributed to the choice of defenses. In any individual, the patterning or arrangement of cognitive

controls would constitute a superordinate structure, “cognitive style.”

The following cognitive control principles were studied by Klein and his group:

1. Leveling-sharpening. Consistent individual differences were found between people (known as

sharpeners) who tend to clearly distinguish newly perceived stimuli from their previous

experiences, and “levelers” who tend to show a high degree of assimilation between new

percepts and old ones, resulting in judgments of current stimuli as being similar to
previously perceived ones. Some data suggested that levelers might have generally

hysteroid personalities and favor the use of the defense mechanism of repression
(Gardner, Holzman, Klein, Linton, & Spence, 1959; Holzman, 1962). Klein and his co-

workers reasoned that a tendency to assimilate new events to existing schemata was

similar to Freud’s (1915) definition of secondary repression as “the attraction exercised

by what was primally repressed upon everything [in consciousness] with which it can

establish a connection” (p. 148).

2. Scanning. Individuals high on scanning were thought to “deploy attention to relatively many
aspects of stimulus fields...[they are] constantly scanning the field” (Gardner, Holzman,

Klein, Linton, & Spence, 1959, p. 47) and were also said to “narrow awareness and keep

experiences discreet; and...to separate affect from idea” (p. 46). These characteristics

www.freepsychotherapybooks.org 51
were considered cognitive analogues to the defense mechanism of isolation, which was
found to be correlated to scanning in some people, as rated in the Rorschach test

(Gardner, Holzman, Klein, Linton, & Spence, 1959).

3. Equivalence range. Equivalence range denoted a dimension of individual differences in

preference for using broad and inclusive versus narrow and precise categories in

classifying objects and events. No connection was suggested between this control

principle and defenses.

4. Tolerance for unrealistic experiences. People were found to differ in the flexibility and

efficiency with which they accepted and dealt with ambiguous situations, with situations
that “controvert conventional reality,” and perhaps also with affects (Gardner, Holzman,

Klein, Linton, and Spence, 1959).

5. Constricted-flexible control. Flexible controllers were considered to be individuals who

function efficiently on tasks even when experiencing strong drives, feelings, or other
distractions; constricted controllers’ adaptive functioning was impaired by these things.

Constricted controllers were thought to tightly suppress feelings and impulses, whereas

flexible controllers were freer and less compulsive. Here too, a theoretical connection
was made to the use of the defense of isolation of affect.

6. Field articulation. This is the field dependence-independence variable extensively studied by

Witkin (Witkin, Dyk, Faterson, Goodenough, and Karp, 1962). It is a thoroughly explored

dimension of individual differences in the tendency to focus on background versus

“figure” cues in many situations. Field articulation is connected to numerous aspects of

personality and cognitive functioning.

Klein conceptualized cognitive controls as quasi-motivational—he believed they direct

behavior, but, unlike drives, they do not lead to discharge or consummation. Like defenses, they shape

www.freepsychotherapybooks.org 52
the expression of drives and control drives, but he saw them as more general than defenses—as basic,
conflict-free, “positive” causes of behavior. In his research on cognitive controls, Klein attempted to
broaden knowledge of primarily and secondarily autonomous ego functions. He also began to

elaborate what was to become a continuing theme of his work-his view that psychoanalytic drive
theory, with its basis in physiological need states, was an inaccurate and inelegant way to formulate

human motivation. As he put it (Klein, 1958):

It seems more parsimonious to follow Woodworth’s (1918), Woodworth

and Schlosberg’s (1954), and Hebb’s (1949) lead and think of drive as a
construct which refers, on the one hand, to “relating” processes-the
meanings-around which selective behavior and memories are organized;
and in terms of which goal sets, anticipations, and expectations develop,
and, on the other hand, to those processes which accommodate this
relational activity to reality [p. 92],

CONTRIBUTIONS TO PSYCHOANALYTIC THEORY

A. STATES OF CONSCIOUSNESS

Klein expanded his research into the mechanics of thought and perception as “tools” of

adaptation after leaving the Menninger Foundation. Having first studied individual differences in

apparently enduring, relatively autonomous cognitive structures, his attention now turned to another
conflict-free area of the determination of perception and cognition: studies of the effect on thinking of

variations in states of consciousness—subliminal stimulation, dream research, and the influence of

drugs.

Very early in the development of psychoanalysis, Freud (1900) had pointed out that incidental
experiences perceived on the periphery of awareness are processed differently by the organism than

events of greater importance, which are dealt with in the focus of consciousness. Incidental
experiences tend not be remembered consciously but to contribute heavily to the day residue of

www.freepsychotherapybooks.org 53
dreams. Unlike the contents of focal consciousness, they are more subject to the primary process than
to secondary process modes of thinking. Studies of subliminal registration (by Poetzl, 1917 and Fisher,

1954 as well as by Klein, 1959a) produced findings that confirmed and elaborated Freud’s early

observations. Subliminal stimuli activated a range of conscious and unconscious meanings, which
could be discerned in dreams, imaginative products, and various indirect aspects of verbal and

nonverbal behavior but were not available to conscious recall. Thus, the state of consciousness of the

individual—as affected by attention, chemicals, and the sleep state—has significant effects on the
extent to which primary process, assimilatory cognition—as opposed to secondary process,

accommodative cognition—is active. This bridge between psychoanalysis and academic psychology
was significant to the latter as well. Laws of perception that were developed in the laboratory were
now seen as specific only to particular states of awareness. Since perception is a cognitive event, under

conditions in which reality content is not prominent registrations are recruited to very different, more
primitive conceptual schemata than those that are ordinarily dominant in focal attention (Klein, 1959a;
1966).

B. CRITIQUE OF METAPSYCHOLOGY

Having spent a lifetime doing both academic research and clinical psychoanalytic work, Klein
found many flaws in the ways in which existing theories from both vantage points had attempted to
account for the meanings of human behavior. He believed that academic psychology, in its efforts to

encompass the findings of psychoanalysis, had actually found ways to omit and ignore the most salient
features of Freud’s insights. By directing attention to generalized conceptions of the determining
influence of infantile experience on adult behavior, academicians had focused attention away from the

specific issue of infantile sexuality. Phenomenological and humanistic psychological theories, which
stress the present moment and conscious experience, have little place for the unconscious. This, of
course, is also true of behaviorism. Social psychiatry stresses environmental rather than intrapsychic

causation. But Klein saw classical psychoanalysis as also burdened by a mechanistic theory—
metapsychology—which is not only unnecessary but is actually harmful to the understanding of

www.freepsychotherapybooks.org 54
meanings and the practice of psychoanalysis. Other authors as well (see particularly Gill, 1976;

Guntrip, 1969; Holt, 1976; Schafer, 1968; Waelder, 1962), some of them earlier than Klein, have
pointed out that Freud produced both a clinical theory and a metatheory, at different levels of logical

analysis, and that the existence of these two theories has created many serious problems. Although

Klein was therefore not the first to espouse this point of view, his was a most compelling voice.

Klein’s first step toward rectifying this situation was to search for the “essential theoretical
understructure that constitutes ‘clinical psychoanalysis’ ” (1976, p. 1). Clinical theory, he specified,

attempts to organize and explain psychoanalytic data from the viewpoint of the patient’s experiences

and motivations, both conscious and unconscious. This approach contrasts to the body of theory that

attempts, as Klein (1976) put it, “to place psychoanalysis in the realm of natural science by providing
an impersonal, nonteleological view of the organism as a natural object subject ultimately to the laws

of physics, chemistry, and physiology. Teleological considerations—the patient’s standpoint—are

irrelevant to this level of explanation” (p. 2).2 For Klein (1976), purposive considerations were at the
very heart of clinical theory, which does not “distinguish sharply between description and explanation;

to describe a person’s intention or aim is to say that what a person is doing is also why he is doing it”

(p. 2). In view of these convictions, Klein attempted to carefully disengage the clinical theory from the
mechanistic metapsychological theory.

Like his predecessors, Klein attributed the existence of the two psychoanalytic theories to

Freud’s philosophy of science, which, in turn, was determined to a significant degree by the intellectual
climate of the late nineteenth century. As Klein understood this philosophy, it rejected concepts of

intention and meaning, considering them unacceptable terms of scientific explanation. Freud struggled

to construct a neuroanatomical-physiological model, the Project for Scientific Psychology (1895).

Although he was forced to abandon the Project, his search for neurophysiological levels of explanation

continued to manifest itself, for example, in the optical apparatus model in Chapter 7 of The
Interpretation of Dreams (1900) and in the instinctual drive theory of motivation, with its concepts of

energic excitations, cathexes, and reservoirs of energy.

www.freepsychotherapybooks.org 55
Klein considered the concepts of drive and energy to be the central flaw of metapsychology, a

flaw attributable to Freud’s reliance upon the Brucke-Meynert value system. These concepts, he

believed, are not only inherently implausible but also irrelevant to the clinical psychoanalytic
enterprise. Freud’s drive-reduction model, Klein (1976) stated, “is more appropriate to a rat than to a

human being, and is as congenial to violently antipsychoanalytical theories as to Freud’s metapsy-

chological ones” (p. 47). Metapsychological concepts are not verifiable by the clinical method and are

based on the reification of such hypothetical constructs as drives and the psychic structures of id, ego,

and superego. They do not deal with the basic intent of psychoanalysis—unlocking meanings—and
cannot substitute for terms that are descriptive of human experience and object relations. Nonetheless,

as a consequence of these dual theories, psychoanalysts have tended to assume that they actually work
on the basis of clinical concepts, which they take for granted as observable or inferable phenomena,
and to think of metapsychology as the underlying basic theory that explains those concepts. In

contrast, Klein considered that the concepts of the clinical theory, if correctly understood, were at an
appropriate level of abstraction to replace metapsychology. They are explanations in psychological
terms, are personal and purposive, and are not translatable into physical or neuroanatomical models.

For Klein (1976), “the phenomenological concepts, the logic of the analyst’s inferences, and the

extraphenomenological concepts of function, purpose, and meaning of experience and behavior make

up psychoanalytic theory” (p. 51). Within this realm he made a distinction between experiential and
functional concepts. Experiential concepts are the mental contents that the analyst attributes to the

patient, including both the patient’s conscious experiences and unconscious fantasies. These are

verifiable in the psychoanalytic situation. Functional concepts, such as projection, introjection, and

repression, are inferred processes connecting the accessible and inaccessible levels of experience, and
cannot be verified in the clinical setting. They “almost always have to do with purpose, function and

accomplishment” (Klein, 1976, p. 50). Using these conceptual tools, analysts are able to observe

regularities in behavior that are not recognized by other students of the mind.

Psychoanalysis, for Klein, deals with the histories of meanings throughout a person’s life. He

www.freepsychotherapybooks.org 56
understood personality as formed through syntheses that evolve out of conflicts-that is, points of crisis
in the individual’s life. Since there are commonalities in these crises, it is possible to specify critical
developmental periods. These phenomena are not reducible to physiological and neurological

processes and it would be fallacious to think that such processes are more real or valid as scientific
data, than observations of people’s intentions and meanings. As Klein (1976) said: “Statements of

purpose or meaning and principles of physiological regulation are two mutually exclusive ways of

being aware of our bodily activities” (p. 62).

1. Sexuality and Sensuality. Klein began his task of teasing apart the two psychoanalytic theories

by examining Freud’s two theories of sexuality. Freud was led by his clinical observations to expand

the meaning of sexuality from its ordinary use, referring to adult procreative ability, to a wide variety

of behaviors beginning in infancy and developing throughout life. The invariant factor in all sexuality
(i.e., in all those experiences which Freud referred to as libidinal) Klein (1976) felt, is “a capacity for a

primary, distinctively poignant, enveloping experience of pleasure” (p. 77). These experiences, which

Klein felt are best termed “sensual,” do not depend simply on the removal of “unpleasure,” but are
positive excitatory processes. Sensual pleasure has a number of special characteristics that cause it to

be a highly significant motivational force and that create a unique potential for conflict surrounding it.

It can serve various functions that are not originally primarily sexual; it can be experienced in organs
and activities that are usually nonsexual, and, conversely, sexual organs can lack erotic sensation at

times.

Recognition of the distinction between sensuality and sexual behavior was a fundamental
difference between Freud and other theorists. Sensuality has a characteristic development, which

interlocks with all other developmental areas. This made it possible for Freud to understand how
sexual development affected, and is affected by, a person’s symbolized cognitive record of
interpersonal encounters. In Klein’s view, sexual needs do not exert a driving force upon behavior;
instead, the experience of sensual pleasure acquires important developmental meanings, and these

lead to a craving for repetition of the experience. It is in this way, rather than because of biological

www.freepsychotherapybooks.org 57
drive, that sexuality attains its great motivational force. Since these experiences occur originally in
relationships with parents and other early objects, the sensual cravings take the form of specific object-
related desires. They also are highly subject to the inevitable contradictions and conflicts arising

between the plasticity of sensual arousal and the constraints of social guidance and expectations. It is
this plasticity and these unavoidable contradictions that create a potential for intrapsychic conflict

surrounding sexuality.

Sexuality feels as though it is a drive—in Klein’s (1976) words, “as if an alien pressure were
developing from within” (p. 96)—but it is neither necessary nor logical to assume that this feeling of

impulsion reflects the presence of an actual, concrete drive quantity or mechanism. In this respect, the

concept of libidinal drives is a theoretical reification of an experience—a “hybrid concept” (Slap and

Levine, 1978). The strength of a sexual motive is not based on some drive force, but on the functional
significance of the sexual appetite in cognitive terms-the meaning of the activated schema of

sensuality. In drive theory, said Klein (1976) the essence of sexuality “is not an experience, but a

contentless physiological event” (p. 110), and the clinical and drive concepts of sexuality imply
different biological formulations. In one it is a pleasurable experience that is sought repetitively in

order to reexperience the pleasure, and in the other it is a need to relieve pressure and tension. In the

clinical theory the focus is on sensuality as a higher mental function; in the metapsychological theory,
sensuality is a peripheral function. As Klein (1976) phrased it: “The critical difference of emphasis

between the two viewpoints…[is] that in the drive theory pleasure is derived not from the pursuit of
drive, but from the getting rid of it” (p. 119). Klein believed that the theoretical preeminence of the
drive model led to a failure to fully test the implications of the clinical theory and impeded efforts to

enlarge clinical understanding.

2. Ego Psychology. Klein traced the problems of the two psychoanalytic theories in his analysis
of the theory of the ego (1976), which he considered “a concept in search of an identity” (p. 121). Freud
had formulated the ego unsystematically as the representative of reality, essentially a set of functions

and processes standing for survival and self-preservation. As is still the case for many clinicians, Klein

www.freepsychotherapybooks.org 58
(1976) noted, Freud was “inclined to regard the ego mainly in the light of its participation in conflict
and in its partnership with instinctual drives pressing for discharge. [He paid] scant attention to the
adaptive functions of the ego as a system” (p. 130). This was satisfactory, in Klein’s view, within the

framework of the early psychoanalytic conception of drives as blind motivational forces controlled by
some structure. With the expansion of motivational theory initiated by the conception of the active role

of signal anxiety as a motivating, directive factor, however, Freud assigned an enlarged and elaborated

role to the ego, but left crucial issues unsettled. These included the questions of what energy is
employed by the ego in its activities; the energic basis of signal anxiety; and whether primary process

functions such as condensation and displacement are within the scope of the ego. In addition, Klein

believed that Freud had hinted at the possibility of autonomous structures and motivations in the ego,
and thus, Klein was convinced that this required further development. Freud fell short in that he did

not see the ego as “positive creator” of behavior, and did not confront “the crucial issue of the
independence of the ego processes from libidinal control” (Klein, 1976, p. 131).

Following Freud, ego psychological theorists, especially Hartmann, Erikson, and Rapaport,
proposed solutions for these problems. These solutions often involved a broadened focus of

psychoanalytic interest from psychopathology to general psychological questions, and to “all

psychological ‘disequilibria,’ whether specifically psychopathological or not” (Klein, 1976, pp. 145-
146). They placed greater emphasis on the adaptive point of view and on independent, conflict-free

motivations and structures within the ego. Klein characterized the changes in psychoanalysis as: (1)
from a narrow concern with conflict to concern with dilemma and crisis; (2) from concern with
defense to an interest in adaptational controls; (3) from concern with sexual and aggressive drives to

general motives such as mastery and the synthetic principle; and (4) from giving priority to the
unconscious to greater priority for conscious phenomena such as affects.

Klein saw these as salutary new emphases for psychoanalytic theory. However, he believed that
they were not done justice by the metapsychology of Hartmann and Rapaport, which is essentially an

expanded theory of ego controls, established to compensate for the deficiencies of the drive theory of

www.freepsychotherapybooks.org 59
motivation. Ego psychology tended to see the ego either as a reified entity or simply as an unsystematic
grouping of functions. (Klein considered this a throwback to faculty psychology.) The ego’s relationship
to drives remained ambiguous. Drive was spoken of as both independent of the ego and involved with

ego development. Drives develop and have structure, but it is not clear whether they are part of the
ego. Finally and centrally, like all metapsychological explanations and like academic psychology as

well, these theories suffer from a focus on process rather than motivation. According to Klein (1976),

“To the extent that we pursue the process explanation, we are distracted from the need to improve
upon psychoanalytic propositions regarding the aims, motives and goals of behavior” (p. 158).

Ego psychology is torn between two objectives of explanation: It offers a

half-hearted and half-annotated commitment to explanation in terms of
purpose on the one hand and to mechanism on the other. It faces the
choice either of trying to sophisticate a conception of a regulatory
mechanism, building into it ever more detailed assumptions of processes
to implement its heretofore implied reifications, or of frankly restricting
the scope and the terms of ego theory to the level of motivational
explanation that brought psychoanalysis into being in the first place—an
endeavor exclusively concerned with understanding behavior in relation
to psychological dilemma, conflict, task, and life history, an enterprise that
explores, in Waelder’s (1936) terms, the “multiple functions” of action.
This is the shoemaker’s last of the psychoanalyst. Sticking to it and
pointing his theorizing in this direction, the psychoanalyst would shed all
pretense of offering a nonteleological, mechanistic picture of ego
processes, such as is implied by such impersonal terms as homeostasis,
equilibrium, cathexis, energy and its modifications, and the like [pp. 159-
160].

REFORMULATION OF PSYCHOANALYTIC THEORY

A. THE EMERGENCE OF STRUCTURE FROM EXPERIENCED INCOMPATIBILITY: THE SELF SCHEMA

Klein’s critical analysis of metapsychology was virtually complete at the time of his death, but

www.freepsychotherapybooks.org 60
his attempts to reformulate psychoanalytic theory at a clinically relevant level were in a preliminary
stage. In this reformulation, he took as a central tenet of Freud’s genetic and structural theories the
idea that structural residues of past conflicts persist as organizing principles of behavior and thought.

He noted (1976): “The concept that intrapsychic order and the motives governing action arise from
experienced disorder is basic to psychoanalytic theory” (p. 165). Crisis and conflict, then, are not only

pathogenic, but also play a constructive role in personality development; and Klein felt that

psychoanalysis should be broadened to encompass “normal” as well as pathological development.

Klein proposed a number of modifications and elaborations of psychoanalytic theory. First,

although he recognized that intrapsychic conflicts over unacceptable wishes (those that conflict with

social constraints and expectations) have a uniquely important role, he believed that there are other

“incompatible experiences” (for example, the contradiction between old behavioral modes or attitudes
and new ones that occur at a point of conflict-free developmental change) that are of similar

developmental significance and should be dealt with by theory. All such incompatibilities present

“threats to the integration” of the self and are felt as painful experiences, such as “estrangement from
self.” Their resolution takes the form of a cognitive-emotional schema that frames later perceptions

and actions.

A key aspect of Klein’s thinking in this area is his introduction of the concept of the self, or self-
schema, which is certainly among the more elusive and controversial concepts of psychoanalytic

theory.3 Klein (1976) pointed out that, as had been recognized by psychoanalytic structural theory,
“conflict occurs only in relation to an integrating organization that is capable of self-observation, self-
criticism, and choice, and that can regulate emotional needs and their expression” (p. 171). However,

since classical theory at times sees the ego as the locus and resolver of conflict and at other times as a

party to conflict (e.g., between an “ego aim” and a drive), some organization beyond or supraordinate to
the ego must be conceptualized to provide for integration of aims and adjudication of contradictions.

Thus, in Klein’s words, “the notion of self…now seems indispensable” (p. 172). Klein cited Hartmann’s

(1950) concept of the self as a further “grade” of personality organization and H. Lichtenstein’s (1964)

www.freepsychotherapybooks.org 61
formulation that the self is the source of the experience of successful and failed integration.

Klein conceived of the self as active in regard to the problems it confronts—both in resolving

the demands made on it and in initiating purposes of its own. He cited as an early exposition of this
position Waelder’s view that the ego is not simply a passive, mechanistic switchboard, but has “its own
peculiar activity” (Klein, 1976, p. 47), that is, “an active trend toward the instinctual life, a disposition

to dominate or, more correctly, to incorporate it into its organization” (Waelder, 1936, pp. 47-48). For

Klein, the synthetic function has the purpose of helping the individual to maintain integrity among
conscious aims, motives, and values; and the self is the source of this feeling of integrity. The sense of

the self has two aspects in dynamic equilibrium. One is individuality—“an autonomous unit, distinct

from others as a locus of action and decision” (Klein 1976, p. 178); the other is ‘“we-ness”—“one’s self

construed as a necessary part of a unit transcending one’s autonomous actions” (p. 178). An example of
“we-ness” is oneself as part of a family, community, or profession. Klein follows closely upon Erikson’s

(1963) ideas here, particularly Erikson’s concept of the sense of identity, which implies an overall

continuity extending from the past into the future and from a particular place in the community’s past
into anticipated work accomplishment and role satisfaction. Thus, the parameters of the sense of self

involve conscious feelings of continuity, coherence, and integrity of thought in respect to both

autonomy and “we-identity.”

Having laid this groundwork, Klein proposed a redefinition of the concept of intrapsychic

conflict in the broader context of “synthesizing efforts” necessitated by crises that threaten the

coherence, continuity, and integrity of the self. These crises are of several kinds: wishes that are
incompatible with the self-concept (intrapsychic conflict proper); traumatic experiences in which the

person is passively overwhelmed; experiences inducing feelings of finiteness or loss of important roles
or objects; and developmental crises “when the adaptational modes of one stage no longer suit the
requirements of a new stage” (Klein, 1976, p. 190). Concomitantly, Klein (1976) reconceptualized
anxiety as a signal of threat to self-identity, a feeling of discontinuity in selfhood akin to helplessness or

meaninglessness:

www.freepsychotherapybooks.org 62
From such experiences of cleavage, whether of the nature of conflict,
trauma, or developmental crises, arise efforts at solution in the form of a
realignment of aims and goals. The more relevant a motivation is in
bringing about a solution to a crisis of selfhood, the more lasting,
generalized, and thereafter relied upon it is in the economy of personality.
Thus the resolution of experienced incompatibility is the major basis of
motivational structure. Motivations, rather than being regarded as arising
from a “parallelogram of impersonal forces” defined in terms of intensity,
are regarded as arising as resolutions of issues of self-integration and self-
perpetuation [p. 208].

These experiences of incompatibility, conflict, or cleavage in the self, and their resolutions, are

organized as cognitive structures with motivational effects. These structures, which Klein called
“schemata” (a term borrowed from cognitive psychology [Bartlett, 1932]), encompass the relevant

wishes, object representations, affects, and defenses-all of which together form the representation of
conflict. As Klein wrote:

“The component tendencies of a conflict are embodied as an unconsciously

sustained structure (unconscious fantasy) which may be repetitively
enacted throughout life ... [p. 185].

…The terms of incompatibility and the solutions adopted to reduce the

incompatibility are internal structures which state themselves as themes,
affective positions, and styles of action and thought. The structured
residues of incompatibilities are dynamisms, which organize the aims of
behavior. Structurally, they are meaning schemata ... [p. 193],

…Such internalized representations of conflict and their defensive aspect

are features of that created inner environment which serves as the
person’s notions of and dispositions toward the “real world,” providing the
means of encoding it and making it meaningful” [p. 199],

Klein saw the efforts to preserve self-identity as falling into two main categories. One is
repression, which he defined as a dissociation of the threatening conflict from the mainstream of the

www.freepsychotherapybooks.org 63
self-identity structure. The other is active reversal of passive experience, that is, repetition of events
experienced passively, yielding a sense of active mastery.

B. MOTIVATIONAL SYSTEM BASED ON PLEASURE-SEEKING

In order to replace the quasi-physiological concept of drives that originate in the soma and

“push” the psyche, and the tension-reduction model of the pleasure principle, Klein conceptualized

motivations as active strivings for experienced pleasures. He considered pleasure as an experience

within the province of the ego, just as anxiety is an experience and activity of the ego, and in his

formulations he relied heavily on parallels with Freud’s development of the concept of anxiety. In

Freud’s early model, anxiety was understood as a direct consequence of the disturbance in psychic
economy caused by failure of adequate discharge of libidinal energy; later, as signal anxiety, it was

conceived as an anticipatory reaction of the ego to danger. As a complementary concept to signal

anxiety, Klein proposed the existence of actively created, anticipatory “signal pleasures;” and just as
there are basic prototypical anxiety experiences (fear of separation, fear of castration, etc.), Klein

postulated six prototypical pleasure experiences. He saw these “vital pleasures” as innately given and

not reducible or analyzable to simpler components. Each pleasure was seen as having its greatest
impact on development at a particular phase of the life cycle; and Klein diagramed their probable

epigenetic development and interrelationships in an Erikson-like table. The six “vital pleasures” were

as follows:

1. Pleasure in reduction of unpleasant tension. Although Klein believed that Freud had

overestimated the significance of tension reduction as a motivational aim, he did nonetheless agree

that it was highly important. In addition to release of drive tension and reduction of anxiety, Klein also
included here numerous other experiences of relief from unpleasant feelings, such as experiences of

unfamiliarity or strangeness, lack of recognition of people or situations, and task incompleteness. In all

these situations, tension reduction yields positively pleasurable feelings and not just an absence of
unpleasure. Consequently the experience is actively and repeatedly sought.

www.freepsychotherapybooks.org 64
2. Sensual pleasure. This is the broad group of pleasurable psycho-sexual experiences, ranging

from genital sexuality to tickling and “contact comfort” that Freud recognized as interrelated and as

having a characteristic development. Here, too, Klein stressed active pleasure seeking, and not merely
the removal of unpleasure. Sensuality is characterized by plasticity—i.e., displaceability in terms of

zone, mode, and object choice.

3. Pleasure in functioning. Klein (1976) asserted that there is inherent pleasure in the exercise of

many conflict-free apparatuses and functions, which is sought for its own sake. “The gratification of

bodily needs does not account for activities that an infant engages in spontaneously:…there are times
when he seems to grasp, suck, babble, squeeze, and pull for no reason other than the pleasure of their

repetition” (p. 223). More broadly, this is a “pleasure [in]…efficient use of the central nervous system

for the performance of well-integrated ego functions” (p. 224).

4. Effectance pleasure. Klein (1976) said: “The component that distinguishes effectance from

pleasure in functioning is the pleasure in observing the successful correspondence of intention and
effect” (p. 225). This is pleasure in accomplishment and mastery, not merely in the exercise of

capacities.

5. Pleasure in pleasing. Klein was impressed by the infant’s early ability to know how to act

pleasing to the mother, and he believed that doing so was inherently pleasurable for the infant. As he

noted (1976): “Generating pleasure in another is... an occasion... for being affirmed in one’s being” (p.
228). This form of pleasure is the basis for the need for affiliation with other people, the need to

belong, which Klein considered a universal purpose, actively pursued: “The fact that pleasure arises

from such a source [pleasing others] tells us too that the affiliative requirement has roots just as deep

as those pleasure potentials that are more directly localized and originate in the “body ego’ ” (p. 229).

6. Pleasure in synthesis—aesthetic pleasure. In infant observation, animal research, and

observations of adult life, Klein (1976) saw many examples of the pleasurable effect of the “delighted
contemplation of restored or discovered order” (p. 229). He saw this as an important motive for play

www.freepsychotherapybooks.org 65
and creative artistic activity. Klein (1976) also viewed this pleasure as closely connected to the

principle of active reversal of passive experience and the “necessity in man…of having to create a self-

identity” (p. 230)—two of the main conceptual pillars of his formulations.

Although Klein (1976) saw these “vital pleasures” ordinarily as fundamental motivational

givens, he also recognized that there could, at times, be other unconscious motives for pursuing these

aims, noting that “inherently pleasurable activity can be extrinsicallly motivated as well” (p. 234).
Thus, pleasures can serve defensive functions; and behavior can at times be motivated toward

excessive pleasure seeking and at other times toward excessive avoidance of pleasure.

C. THE MAINTENANCE OF SELF-INTEGRITY

1. Repression. Klein delineated two broad categories of activity by which the coherence, identity,

continuity, and integrity of the self can be maintained in the fact of threat due to conflict,
developmental incompatibility, or trauma: repression, and active reversal of passive experience. As is the

case with other psychoanalytic concepts that have evolved over many years, repression is ordinarily

used with both clinical and metapsychological meanings and is consequently subject to some confusion
and lack of precision. Psychoanalysis did not begin with a drive theory. Rather, in his early
formulations Freud attributed psychopathhology to the preemptive power of unconscious memories

and ideas; that is, to mental contents dissociated from consciousness but nonetheless active. Because

the power of the repressed ideas appeared to derive from their sexual content, Freud eventually
altered his emphasis from the ideas themselves to the drives that he presumed to underlie them as the

sources of intrapsychic conflict. It is consistent with the clinical theory of psychoanalysis, however, to

understand intrapsychic conflict as occurring between opposing sets of ideas (with their associated
affects, object representations, and aims), or between the self-schema (the nonconflictual “main mass

of ideas,” feelings, attitudes, and aims) and a meaning schema that is contrary to the self, and hence

threatens its maintenance, integrity, and continuity. The understanding that conflict derives from a
clash between the ego and a drive is consistent only with the mechanistic concept of metapsychology.

www.freepsychotherapybooks.org 66
Klein saw repression, then, as one mode of coping with conflicting meaning schemata. He evolved a
unique understanding of this phenomenon based largely on his orientation, derived from academic
psychology, of seeing the mind as an apparatus for learning and adaptation.

Klein (1976) pointed out that repression does not necessarily operate by prohibiting the
individual from having any awareness of conflicts. Rather, it leaves a gap in comprehension of the

warded-off material, without impeding its behavioral expression. In fact, its function is to permit the

expression of the conflicted wish, while at the same time protecting the integrity of the self-schema by
denying it “the attributes of self-relatedness…[excluding it] from the self as agent, self as object, and

self as locus” (p. 242).

In repression, the threatening meaning schema is dissociated from the self and continues to
have a motivating influence on behavior and thought. It functions in the mode that Piaget designated as

assimilation; that is, it provides a code for understanding, reacting to, and internalizing new
experiences, while its own existence and effects are uncomprehended and unchangeable because no

feedback is possible about them. Repression can impede adaptation, growth, learning, accurate

perception, and cognition. Undoing of repression is not only recovering the memory of traumatic event
or conflictual idea, but also understanding its meaning, bringing about “comprehension in terms of a

previously uncomprehended relationship, the perception of a causal link to which the person had been
impervious” (Klein, 1976, p. 248).

Klein considered it more accurate to think of repression in terms of its mode of operation—the

splitting off of an organization of ideas that are threatening to the self and that then function in a
purely assimilative fashion—rather than in terms of the unconscious, whether conceived of as a system
or as a quality of experience.

2. Reversal of Voice. The second major strategy available for resolution of threats to self-

coherence, according to Klein, is the principle of reversal of voice, or active reversal of passive

experience. This concept has a long history in psychoanalysis. Its clearest statement occurs in Beyond

www.freepsychotherapybooks.org 67
the Pleasure Principle (Freud, 1920) in Freud’s description of a game played by his grandson. The same

concept is at the heart of the defense mechanism of identification with the aggressor. In Klein’s view, it

is also the essence of Freud’s description of signal anxiety—an instant, miniature act of reversal, an

active repetition of a trauma. Although the concept of active reversal has long been available and
allusions to it occur in various contexts, however, it is fair to say that it has not before been accorded a

central and important role and has not been well integrated into the main body of psychoanalytic
theory. Klein proposed that it is a principle of major importance, not a defense mechanism, coordinate

with and “equally vital” to repression.

In its basic form, active reversal is observed most clearly in children who respond to traumata

of various kinds by actively repeating the painful experience (usually in play or fantasy), or by doing to
another person what was done to them, so that they make it seem to occur under their control. In this

way, the painful experience is mastered and internalized, modifying and differentiating some aspect of

the selfschema by accommodation, to restore its harmony and integration. Accommodation is Piaget’s
term for a process complementary to assimilation, in which new data is recognized as different from

past experiences, and the schema is changed to encompass it. As Klein (1976) explained it: “The heart

of the principle [of reversal of voice] is that when a passively endured encounter or relationship is
affectively coded, a search for information is stimulated, towards two ends: a) to make the experience
understandable in relation to the self; b) to position the internalized relationship within the self-

schema as usable information related to the self-as-agent” (p. 285). In contrast to repression, which is a

regressive solution to incompatibility that restricts the personality, Klein (1976) saw active reversal as
a progressive mode, a “positive” mechanism, leading to “growth through reconstruction, innovation

and integration” (p. 196) and requiring advanced development. This concept, he wrote,

captures the essential distinction between activity and passivity which

Rapaport (1953) intuitively felt to be one of the fundamental foci of
psychoanalytic theory generally and of a dynamic conception of ego
organization in particular. I believe it is the essential dynamic aspect of
what is usually encompassed by the term “will.” The principle
encompasses such diverse phenomena as play, novelty, curiosity,

www.freepsychotherapybooks.org 68
repetitive working through of traumatic experiences, interruption
phenomena, and certain aspects of art-making. From a developmental
standpoint it encompasses…competence motivation; perhaps most
important of all, it provides a dynamic basis for identification. In
psychoanalytic therapy the positive or adaptive aspects of transference, as
Loewald (1960) has emphasized, are explainable in its terms [Klein, 1976,
p. 261].

In passing, Klein (1976) suggested that the principle of active reversal might also account for
aggression: “Activities of reversal of voice could be considered synonymous with the ‘aggressive

drive’—not in the sense of a specific aggressive motive to destroy but as manifestations of an

instinctual aggressive potential” (p. 264). This idea has much in common with the concept of

nonhostile aggression as employed by Marcovitz (1973). He conceived of aggression as consisting of a

spectrum of interpersonal behaviors ranging from simple activity at one extreme, through such

phenomena as self-assertion, dominance, and self-defense, to hatred at the other pole. This broad,

dynamic view of aggression is easily reconciled with the principle of active reversal of passive
experience.

D. DEVELOPMENTAL AND STRUCTURAL CONSEQUENCES OF REPRESSION AND ACTIVE REVERSAL

Klein conceived of the two mechanisms for maintaining self-integrity as a duality with
widespread consequences—sometimes interacting, sometimes contrasting—throughout psychological
life. He specifically called attention to the following:

1. The repetition compulsion. The motive to repeat phenomena can reflect either the continuing

activity of repressed meaning schemas or the active reversal and repetition of passively experienced
events. In one case, the repetition is assimilative and in the other, accommodative.

2. Internalization processes—fractionation and identification. Repression “fractionates” the self

by splitting off unacceptable mental contents. Often these contents consist of the schemata of

www.freepsychotherapybooks.org 69
interpersonal relationships, including an image of the object, an image of the self in interaction with
the object, and an affect image.4 According to Klein (1976), these split-off interpersonal schemata,

called introjects, preserve the threatening relationship within the personality, in dissociated form:

“From the subjective phenomenological (not necessarily conscious) point of view the introject is

experienced as an alien presence. ... It is felt as part of the body, or one’s thought, but not as part of the
self” (p. 295). The internalization process involving reversal of voice, on the other hand, is

identification. Through accommodation, the selfschema is modified, differentiated, and enlarged to

bring into it “the values, manners, and interpersonal modes of others” (p. 292). Experientially, when a
successful identification occurs, there is no felt separation between the newly internalized mental

contents and the self as a whole. In fact, ego identity is largely composed of identifications formed by

means of active reversal of voice. Similarly, the superego is the result of active incorporation of

prohibitions that were originally passively experienced. Introjection is a defense mechanism and
identification a nondefensive structuring process.

3. Pathology. At times, repression and active reversal operate simultaneously in regard to the

same objects, with pathological consequences. For example, a man may identify with certain aspects of
his father and repressively introject other, opposite aspects. The result may be ambivalence and

symptomatology.

4. Creativity and art. Klein (1976) considered the art-making impulse as an effort to remedy a
sense of “fracturing of selfhood and anxieties that herald…such a threat” (p. 206), primarily through

the operation of active reversal of voice. Here, in response to some painful failure or rejection, the
artist uses his talent to convert previously repressed, fractionated fantasies into creative products. In

this process, the artist is actively mastering (through repetition) both the early conflicts that had been
dissociated and the current traumata.

E. PSYCHOTHERAPY

www.freepsychotherapybooks.org 70
Klein considered classical psychoanalysis both inefficient and perhaps less effective than other,

modified forms of treatment. He believed its main value was as a method of training and research;

however, he felt its true potential as a naturalistic setting for data gathering was not being achieved, in
part because of a lack of systematic research approaches to analysis and in part because of the

stultifying effect of metapsychology on creative clinical thinking.

The theoretical formulations he proposed were partly intended as a remedy for this problem.

However, Klein said relatively little about the direct clinical application of his ideas—perhaps because

of their incompletely developed state.

Characteristically, Klein’s few direct comments about treatment concerned active, growth-

inducing aspects of psychotherapy, which he felt had received too little consideration in the past. He
suggested that the principle of reversal of voice contributes a new dimension to understanding

transference repetition. In addition to transference being a regressive expression of split-off conflicts,

Klein postulated that through active reversal in transference (or perhaps the “treatment alliance”),
direct, positive personality change is brought about. In effect, this appears to refer primarily to growth

through identification with adaptive functions of the therapist and the treatment process itself.

DISCUSSION: KLEIN’S CONTRIBUTIONS TO PSYCHOANALYTIC THEORY

George Klein turned his creative, inquiring intelligence to a remarkable scope of problems and

tasks. He was at the forefront of his field almost from the first. He was an innovator of research
methodology and had an ability to challenge theory and to see problems in new ways, both in the
laboratory and in his theoretical formulations. It is probably correct to view him, as Goldberger (1982)

has said, as primarily a psychologist and only secondarily a psychoanalyst. His work was always at the
interface of the two fields, beginning with explorations in the use of experimental methods to enhance
psychoanalysts’ knowledge of ego functioning, and culminating in the seeming paradox of his

proposals to use concepts from general psychology to create a clinically relevant theory for

www.freepsychotherapybooks.org 71
psychoanalysis, free of what he saw as the counterproductive burden of metapsychology. As
Goldberger (1982) points out, Klein was unusually talented at synthesizing concepts from many fields

—for example, making Piaget meaningful to psychoanalysts. He notes: “The gift that George Klein

evidenced was being able to cross conceptual and theoretical boundaries, a brand of creativity that
bespeaks a mature thinker.”

The manifest form of Klein’s work changed radically over time, not only in his shift of emphasis

from laboratory research to theoretical formulation, but also in his sudden change from leadership in

the expansion of Rapaport’s ego psychology to a diametrically opposite, clinical and phenomenological
point of view. Throughout his career, however, certain basic themes can be clearly discerned. From the

beginning, in perceptual research and then in psychoanalysis, he was dissatisfied with explanations

based, as he saw it, too heavily on drive causality and too little on structure. He was persistently and
articulately critical of the drive concept itself, considering it both inaccurate and logically unsound, and

he ultimately developed these ideas into his sweeping indictment of metapsychology. Finally, he

always advocated an enhanced role in psychoanalysis for “positive” growth potentials, conflict-free
motives, and autonomous structural characteristics. In all these areas he was constantly interested in

broadening the scope of psychoanalysis to encompass normal, conscious, and nonconflictual

phenomena as well as pathology, while at the same time he strove to maintain its fidelity to clinical

experience.

Klein’s impact as a leader, teacher, and pioneer in the study of unexplored territory was
unquestionably significant. Similarly, although Klein was not alone as a critic of metapsychology, his
careful dissection of the inconsistencies and logical defects of the two psychoanalytic theories had

considerable value, not only for theory building, but also in helping the psychoanalytic clinician cope
with the complex, layered conceptual heritage handed down by Freud and his followers. Previously,
analysts who were resistant to metapsychology were vulnerable to the charge of being “atheoretical.”

Klein, however, contended that the focus of attention should be on meaning, not mechanism; and that
clinical concepts, correctly framed, are sufficient to stand as the basic theory. They have the advantage

www.freepsychotherapybooks.org 72
of being closer to the clinical material, more responsive to pressures of the data.

These assertions had a stirring effect on those analysts who were defensive about their aversion

to metapsychology, and led one (Slap, 1980) to write:

Klein liberates such analysts. He confers upon us theories much as the

Wizard of Oz dispensed courage, heart and intelligence to creatures who
already had them. More than that, Klein congratulates us for our scientific
integrity, our willingness (we knew not what else to do) to stick with the
observational data rather than to fudge. Suddenly we are the purists and
the emperor has no clothes [p. 170].

In the words of another commentator (Gedo, 1977), Klein’s book on Psychoanalytic Theory

(1976) “lives up to its promise to explore the essentials of psychoanalytic theory with so much
authority that no future work in the field will qualify to be taken seriously which does not come to
grips with Klein’s arguments” (p. 320).

On the other hand, the usefulness and validity of many of Klein’s new formulations, created to

replace metapsychology, have been the subject of much disagreement. One reviewer (Loeb, 1977)

concludes that Klein “clearly separates data-related, clinical psychoanalytic theory from data-unrelated
metapsychological psychoanalytic theory”; he feels that Klein’s new model was derived from clinical

theory and “should be highly useful and relevant to both therapists and researchers” (p. 215). In
contrast, another critic (Chessick, 1980) expressed concern about the “radical nature” of Klein’s
proposed theoretical revisions; and Frank (1979) saw Klein’s entire enterprise as flawed: “It is difficult

to see where Klein’s basic principles would be useful in application to either the clinical or theoretical
psychoanalytic situation” (p. 193).

In his suggested revisions of clinical theory, Klein often struggled with major problems and
dilemmas of psychoanalysis, areas with which many were dissatisfied. Although many of his solutions

did not succeed in forming a model that is free of internal contradictions and logical flaws as well as

www.freepsychotherapybooks.org 73
consistent with clinical data, in our view his deliberations do have heuristic value. Included here are
such matters as whether psychoanalysis requires (or whether it can encompass) conceptions of
“active, positive” forces for growth and mastery; the related issue of the autonomy and conflict-free

status of various motives and structures; the nature and role of the “self”; repetitive “mastery”;
activity-passivity; and the role of conscious experience in psychoanalysis.

We see much merit in Klein’s delineation of a cognitive model of repression and in his
invocation of the Piagetian constructs of assimilation and accommodation to describe the different

modes of processing data used by dissociated (repressed) mental contents and those that are not split

off. Indeed, one of us has co-authored a paper that sought to bring this model to the attention of a
wider audience (Slap & Saykin, 1983). We see this model as embodying the advantages of Freud’s early

concept of repression as dissociation of a set of mental contents from “the main mass of ideas,” which

then remain active as an unconscious fantasy shaping behavior (by assimilation). This model of
repression is close to clinical experience, accounts in a superior fashion for the impact of current life

experiences on the mind, and absorbs and explains in an internally consistent way numerous

phenomena, including transference, repetition, and symptom formation (all of which reflect

assimilation). As Klein himself said: “Psychoanalytic understanding lies precisely in the recognition of
themes “which we have never lived down nor successfully outlived’ ” (p. 185). We find this model of
repression helpful in this clinical task.

However, Klein’s complementary principle of “reversal of voice”— although based on often-

noted clinical observations of undeniably real phenomena—stands on much shakier ground, as is the
case for many of the factors that he construes as “positive,” “growth-inducing,” and the like. We do not
consider it justifiable either on theoretical or clinical grounds to dichotomize behavior and experience

as Klein does so often into regressive-progressive, positive-negative, defensive-nonconflictual, and

sensual-autonomous. It is as much a misunderstanding of repression to view it as totally maladaptive,

regressive, and so forth as it is to view active, identificatory turning passive-to-active as entirely

nondefensive. After all, the ego expansion of latency is founded on repression of infantile sexuality; and

www.freepsychotherapybooks.org 74
in identification with the aggressor and many other instances, reversal of voice is used as a defense to
create for oneself the illusion of being aggressive and to deny passivity. Of course, in each of these
instances, the other side of the coin is also present—identification with the aggressor can be adaptively

useful.

Klein’s concept of the self-schema is more difficult to evaluate, because his use of it varied. At

times his “self” is a clinical concept—an active self-identity and a sense of self—which contrasts with

the abstract concept of the ego as a structure. It seems consistent with clinical experience to
conceptualize intrapsychic conflict as an internal struggle between repressed fantasy and the main,

integrated system of selfrepresentations, ideals, values, and wishes.

At other times, however Klein’s self-schema is a superordinate structure, with its own inherent
need for self-consistency that has the status of an autonomous motive. Further, the conception of

conflict is broadened and redefined in cognitive terms as a problem of resolution of “incompatible

tendencies”; and conscious experiences such as anxiety or feelings of estrangement are considered to

be direct reflections of intrapsychic realities—that is, of deficiencies in self-integrity. In these

formulations, Klein replaced the old metapsychology with a new and, if anything, more abstract one. It
is subject to the same criticism of inappropriateness in level of discourse and unresponsiveness to

clinical data that Klein leveled at the old metapsychology. In addition, it shares many of the difficulties
we have elsewhere found in Kohut’s self psychology (Levine, 1978,1979; Slap & Levine, 1978), such as

the reliance on hybrid concepts in which levels of data and theory are inappropriately mixed. Among

these is the concept that internal structural psychic conditions are directly reflected in conscious
experience.

Klein’s postulated series of vital pleasures, too, has both features that we find valuable and

problematic ones. The concept that pleasure is to be seen as within the scope of the ego and as a
positive experience of gratification and consummation rather than merely a tension release

contributes to the internal consistency of psychoanalytic theory. Klein did not claim priority for this

idea, which, he pointed out, is similar to suggestions made by Ludwig Eidelberg, Mark Kanzer, and

www.freepsychotherapybooks.org 75
Thomas Szasz. We have serious disagreements, however, with his list of vital pleasures because—as he
himself pointed out—just these conscious experiences and motives regularly occur as disguised,

derivative representations of unconscious conflict. In addition, his treatment of sensual pleasure leaves

one with the impression that such matters as gender identity, sexual appetite, and the procreation of
the species are essentially accidental. In his discussion of these motivational constructs, we believe

Klein fell prey to a number of fundamental fallacies that appear repeatedly in his work: the need to

dichotomize conflictual and nonconflictual forces; the enumeration of presumably autonomous

motives, without supporting data, resulting in a fragmented conception of human beings as extensively

“preprogrammed” (analogous to the instinctual patternings in lower animals), which is inconsistent

with the flexible nature of human adaptation; and the predilection to accept conscious mental contents
as basic, unanalyzable data. These problems, of course, might have been eliminated if Klein had been

able to subject his work to further revision.

Klein focused part of his critique of metapsychology on its presumably inappropriate avoidance

of “teleological” explanation. As defined by Webster’s New Twentieth Century Dictionary (25th edition,
1950), teleology is “the doctrine which asserts that all things which exist were produced for the end

which they fulfill.” Many of Klein’s own explanatory rubrics, in fact, fit this definition of teleology, with
its attendant implication of arbitrariness and untestability by any independent data source. Freud’s

clinical theory was not teleological in the same way. It is true that it did not have recourse to
extrapsychological realms of data, such as neurophysiology; however, it does involve a systematic
method of forming and testing hypotheses about a psychological realm (the unconscious) that is not

directly observable but that can be inferred from future behavioral observations. When Klein’s
conceptions of self and dissociated schema are used in a clinical sense—as referring to conflict related
psychic organizations with conscious and unconscious features—they are useful aids in organizing

data and inferences. If they are conceived of in reified, structurelike terms, they have the same
stultifying, counterproductive effect Klein saw in classical metapsychology.

In summary, we see Klein’s theoretical work as a valuable but not fully successful contribution

www.freepsychotherapybooks.org 76
to the effort that has been undertaken by many theorists to remedy the often-noted difficulties of
metapsychology-We find Klein’s analysis of these difficulties particularly cogent and valuable. His
proposed solutions contain many heuristically valuable elements, but also very significant weaknesses.

Klein’s earlier work on cognitive controls, through which he attempted to enrich psychoanalytic
ego psychology by research methods, similarly does not appear, at this point, to have fully achieved its

objectives. Although there was a great deal of interest in the study of cognitive controls and styles for

some time, it has not yet fulfilled its promise as a bridge between academic psychology and
psychoanalysis, or as providing a means of determining “constitutional givens” that contribute to the

formation of defenses and other personality structures. Further research did not always demonstrate

the postulated unitary character of the cognitive controls or the anticipated direct connections

between these structures and defenses. However, research has suggested that similar cognitive control
behaviors may reflect different personality determinants in different individuals; that controls may

differ in degree of relative autonomy; and that many other factors such as sex differences,

developmental variations and fine distinctions in measurement of these cognitive processes remain to
be understood (see, for example, Levine, 1966,1968; E. Lichtenstein, 1961; Spivack, Levine, and Sprigle,

1959).  

REFERENCES

Bartlett, F. C. (1932). Remembering: A study in experimental and social psychology. Cambridge:

Cambridge University Press.

Chessick, R. D. (1980). Critique: Psychoanalytic metapsychology. American Journal of Psychotherapy,

34, 127-130.

Eagle, M. N. (1982, August 24). Contribution to M. Mayman (Chair), George Klein: A seminal figure in
psychoanalysis. Scientific program presented at the annual meeting of Psychologists
Interested in the Study of Psychoanalysis, Washington, D.C.

Erikson, E. H. (1963). Childhood and society (2nd ed.). New York: Norton.

www.freepsychotherapybooks.org 77
Fisher, C. (1954). Dreams and perception: The role of preconscious and primary modes of perception
in dream formation. Journal of the American Psychoanalytic Association, 2, 389-445.

Frank, A. (1979). Two theories or one? Or none? Journal of the American Psychoanalytic Association, 27,
169-202.

Freud, S. (1895). Project for a scientific psychology. Standard Edition, 1, 295-397.

Freud, S. (1900). The interpretation of dreams. Standard Edition, 4-5.

Freud, S. (1915). Repression. Standard Edition, 14, 141-158.

Freud, S. (1920). Beyond the pleasure principle. Standard Edition, 18, 7-64.

Gardner, R. W., Holzman, P. S., Klein, G. S., Linton, H. B., & Spence, D. P. (1959). Cognitive control: A
study of individual consistencies in cognitive behavior. Psychological Issues, 1 (4,
Monograph 4).

Gardner, R. W., Jackson, D. N., & Messick, S. (1960). Personality organization in cognitive attitudes and
intellectual abilities. Psychological Issues, 2 (4, Monograph 8).

Gedo, J. E. (1977). [Review of Psychoanalytic theory: An exploration of essentials, by George S. Klein,

and Psychology versus metapsychology: Psychoanalytic essays in memory of George S.
Klein, edited by M. M. Gill and P. S. Holzman.] Psychoanalytic Quarterly, 46, 319-324.

Gill, M. M. (1976). Metapsychology is not psychology. In M. M. Gill & P. S. Holzman (Eds.), Psychology
versus metapsychology: Psychoanalytic essays in memory of George S. Klein.
Psychological Issues, 9 (4, Monograph 36, pp. 71-105).

Gill, M. M., & Holzman, P. S. (Eds.) (1976). Psychology versus metapsychology: Psychoanalytic essays in
memory of George S. Klein. Psychological Issues, 9 (4, Monograph 36).

Goldberger, L. (1982, August 24). George Klein’s contribution to psychology-A tribute. Paper presented
at the annual meeting of Psychologists Interested in the Study of Psychoanalysis,
Washington, D.C.

www.freepsychotherapybooks.org 78
Guntrip, H. (1969). Schizoid phenomena, object relations and the self. New York: International
Universities Press.

Hartmann, H. (1939). Ego psychology and the problem of adaptation. New York: International
Universities Press, 1958.

Hartmann, H. (1950). Comments on the psychoanalytic theory of the ego. In Essays on Ego Psychology
(pp. 113-141). New York: International Universities Press, 1964.

Hebb, D. O. (1949). The Organization of Behavior. New York: Wiley.

Holt, R. R. (1976). Drive or wish? A reconsideration of the psychoanalytic theory of motivation. In M. M.

Gill & P. S. Holzman (Eds.), Psychology versus metapsychology: Psychoanalytic Essays in
memory of George S. Klein. Psychological Issues, 9 (4, Monograph 36, pp. 158-197).

Holzman, P. S. (1962). Repression and cognitive style. Bulletin of the Menninger Clinic, 26, 273-282.

Holzman, P. S. (1982, August 24). The contributions from George Klein’s Menninger years. Paper
presented at the annual meeting of Psychologists Interested in the Study of
Psychoanalysis, Washington, D.C.

Holzman, P. S., & Klein, G. S. (1956). Motive and style in reality contact. Bulletin of the Menninger Clinic,
20, 181-191.

Kernberg, O. (1980). Internal World and External Reality. New York: Jason Aronson.

Klein, G. S. (1954). Need and regulation. In M. R. Jones (Ed.), Nebraska Symposium on Motivation.
Lincoln: University of Nebraska Press, pp. 224-274.

Klein, G. S. (1958). Cognitive control and motivation. In Gardner Lindzey (Ed.), Assessment of Human
Motives (pp. 87-118). New York: Grove Press.

Klein, G. S. (1959a). Consciousness in psychoanalytic theory: Some implications for current research in
perception. Journal of the American Psychoanalytic Association, 7, 5-34.

Klein, G. S. (1959b). Editorial statement. In E. H. Erikson, Identity and the life cycle: Selected papers.

www.freepsychotherapybooks.org 79
Psychological Issues, 1 (1, Monograph 1). pp. iii-v.

Klein, G. S. (1966). The several grades of memory. In R. M. Loewenstein, L. M. Newman, M. Schur and A.
J. Solnit (Eds.), Psychoanalysis-A general psychology: Essays in honor of Heinz Hartmann
(p. 377-389). New York: International Universities Press.

Klein, G. S. (1976). Psychoanalytic theory: An exploration of essentials. New York: International

Universities Press.

Klein, G. S., & Schlesinger, H. J. (1951). Perceptual attitudes toward instability: Prediction of apparent
movement experiences from Rorschach responses. Journal of Personality, 19, 289-302.

Levine, F. J. (1966). The relationship of constricted-flexible control to cognitive handling of impulses in

three vocational groups. Unpublished doctoral dissertation, University of Michigan.

Levine, F. J. (1968). Color-word test performance and drive regulation in three vocational groups.
Journal of Consulting and Clinical Psychology, 32, 642-647.

Levine, F. J. (1978). [Review of The restoration of the self, by Heinz Kohut.] Journal of the Philadelphia
Association for Psychoanalysis, 4, 238-246.

Levine, F. J. (1979). On the clinical application of Heinz Kohut’s psychology of the self: Comments on
some recently published case studies. Journal of the Philadelphia Association for
Psychoanalysis, 6, 1-19.

Lichtenstein, E. (1961). The relation of three cognitive controls to some selected perceptual and
personality variables. Unpublished doctoral dissertation. University of Michigan.

Lichtenstein, H. (1964). The role of narcissism in the emergence and maintenance of a primary
identity, International Journal of Psycho-Analysis, 45, 49-56.

Loeb, F. F. (1977). [Review of: Psychoanalytic theory: An exploration of essentials, by George S. Klein.]
American Journal of Psychiatry, 134, 215.

Loewald, H. (1960). On the therapeutic action of psycho-analysis. International Journal of Psycho-

Analysis, 41, 16-33.

www.freepsychotherapybooks.org 80
Marcovitz, E. (1973). Aggression in human adaptation. Psychoanalytic Quarterly, 42, 226-233.

Mayman, M. (Chair). (1982, August 24). George Klein: A seminal figure in psychoanalysis. Scientific
program. Presented at the annual meeting of Psychologists Interested in the Study of
Psychoanalysis, Washington, D.C.

Poetzl, O. (1917). The relationship between experimentally induced dream images and indirect vision.
Psychological Issues, 2, Monograph 7, 1960, pp. 41-120.

Postman, L., Bruner, J. S., and McGinnies, E. (1948). Personal values as selective factors in perception.
Journal of Abnormal and Social Psychology, 43, 142-154.

Rapaport, D. (1953). Some metapsychological considerations concerning activity and passivity. In M. M.

Gill (Ed.), The collected papers of David Rapaport (pp. 530-568). New York: Basic Books,
1967.

Reppen, J. (Ed.). (1980). Symposium on G. S. Klein’s Psychoanalytic theory: An exploration of essentials.

Psychoanalytic Review, 67, 161-216.

Schafer, R. (1968). Aspects of internalization. New York: International Universities Press.

Shapiro, D. (1965). Neurotic styles. New York: Basic Books.

Slap, J. W. (1980). Part 2. In J. Reppen (Ed.), Symposium on G. S. Klein's Psychoanalytic theory: An

exploration of essentials. Psychoanalytic Review, 67, 168-172.

Slap, J. W., & Levine, F. J. (1978). On hybrid concepts in psychoanalysis. Psychoanalytic Quarterly, 47,
499-523.

Slap, J. W., & Saykin, A. (1983). The schema: Basic concept in a nonmetapsychological model of the
mind. Psychoanalysis and Contemporary Thought, 6, 305-325.

Spivack, G., Levine, M., & Sprigle, H. (1959). Intelligence test performance and the delay function of the
ego. Journal of Consulting Psychology, 23, 428-431.

Waelder, R. (1936). The principle of multiple function: Observations on overdetermination.

www.freepsychotherapybooks.org 81
Psychoanalytic Quarterly, 5, 45-62.

Waelder, R. (1962). [Review of Psychoanalysis, Scientific Method, and Philosophy, by S. Hook.] Journal
of the American Psychoanalytic Association, 10, 617-637.

Witkin, H. A., Dyk, R. B., Faterson, H. F., Goodenough, D. R., & Karp, S. A. (1962). Psychological
Differentiation. New York: Wiley.

Woodworth, R. S. (1918). Dynamic Psychology. New York: Columbia University Press.

Woodworth, R. S., & Schlosberg, H. (1954). Experimental Psychology. New York: Holt.

Notes

1) These studies were, of course, not close experimental analogues of psychoanalytic propositions.
There is no reason to expect thirst to have similar motivational properties to those of the
“drives” with which psychoanalysis concerns itself because, unlike libido and aggression,
thirst is not ordinarily a focus of conflict and defense.

2) Klein used the word “teleology” at times as though it were synonymous with “purposiveness” or
“intention.” Although this usage is incorrect and, as will be seen, some of Klein’s
motivational constructs may be open to criticism as truly teleological, it will be retained
here in an effort to accurately represent Klein’s statement of his own ideas.

3) Klein’s work preceded the publication of Heinz Kohut’s self psychology and apparently was
developed completely independent of it. Eagle (1982) believes that Klein’s thinking on
this score did not share the problems and weaknesses of Kohut’s. Perhaps because of the
preliminary nature of Klein’s formulations, however, the two conceptions of self do, at
times, appear similar, and they may therefore be susceptible to many of the same
criticisms.

4) This is similar to Kernberg’s (1980) concept of self-object-affect “units” in the ego.

www.freepsychotherapybooks.org 82
3
ROY SCHAFER: SEARCHING FOR THE NATIVE
TONGUE

Donald P. Spence, Ph.D.

In taking a long look at the work of Roy Schafer and the major themes he has explored and
discussed, it is tempting to try to find a single thread that leads from his initial publications on

diagnostic tests to his more recent work on action language and narrative appeal. To search for such a

thread is, of course, to put into practice one of Schafer’s better-known claims. He has argued for the
central place of the narrative in the way we view someone’s life and works, and if I am able to make the
pieces of his own career fit together in a persuasive fashion, I can make his point even as I am

describing it. To find such a thread will, furthermore, help to uncover some of the similarities beneath
what seem like differences in his approach to psychoanalytic phenomena and also to show how each
phase of his career is, in a certain sense, a reaction to what had gone before. Such an attempt, it should

be noted, may also take advantage of hindsight, and as a result, what seems to be a smoothly flowing
progression of ideas may be, in fact, quite different from the way they were originally conceived.

Nevertheless, this newly discovered sequence may also reveal its own kind of truth, even though it may

not match the experience of the author.

But first I must back off and look at the central problem facing any follower of Freud. When

www.freepsychotherapybooks.org 83
Freud was alive and writing psychoanalytic theory, it was assumed that science was the only path to
the truth and that the mission of science was to discover the whole truth about the natural world. The

human observer was something apart from the thing observed, and any piece of reality was as much an

object of study as an apple or a raindrop. To see the world clearly (with an emphasis on the visual
metaphor) became the goal of science. Troublemakers such as Heisenberg, Heidegger, and

Wittgenstein were still over the horizon.

The visual metaphor and the clear separation between observer and observed are emphasized

in Freud’s conception of the process of free association and in his well-known metaphor of the patient
as passenger on the train, reporting the scene outside the window to a listening seatmate (the analyst).

Tangible reality was assumed to be either outside or inside the head (as in “reality testing” and

“psychic reality”); and in the metaphor of psychoanalysis as a kind of archaeology which uncovers
(reconstructs) the past, Freud called attention to the tangible nature of what had been-memory is laid

down in “mnemonic residues,” waiting to be uncovered and brought to light. The analyst, listening with

evenly hovering attention, was assumed to be the near-perfect observer who, because detached from
the subject, was in an ideal position to see and hear with maximum fidelity and minimum error. The

patient as observer of his or her inner life was the complement of the analyst as observer of the
patient. The symmetry of the two roles is brought out clearly in Freud’s (1912) statement that “the rule

of giving equal notice to everything is the necessary counterpart to the demand made on the patient
that he should communicate everything that occurs to him.” (p. 112).

The naive realism contained in this model always hovered in the background, despite the
gradual accumulation of findings to the contrary. Discovery of the transference was the most obvious

embarrassment to this point of view, because what is transference but the realization that reality is not
simply “out there,” waiting to be described, that what the patient “sees” is often a product of his or her
own experience, and that the subject matter of psychoanalysis largely consists in disentangling the

different faces of what is apparently observed (i.e., in finding flaws with the positivistic model)? But the
larger world view was not significantly changed because transference was assumed to be a transient

www.freepsychotherapybooks.org 84
disturbance (a treatment-activated “neurosis”) that ran its course from symptom to cure. Even the
discovery of countertransference did not significantly affect the world view, because motes in the eye
of the analyst were assumed to be subject to repair by way of the training analysis and occasional

consultations as the need arose. The perfectly analyzed analyst, listening with “evenly hovering”
attention, was the model of the neutral, detached (scientific) observer who was in the perfect position

to see the (physical) field clearly. It was not recognized that even this model of neutrality was perhaps

listening with a bias toward coherence and continuity, not fully aware that the “story” being heard was
only one of many possible ways to understand the patient and his or her associations. Nor was it fully

realized that the meaning in the patient’s associations was not always “out there” but many times was

influenced by the immediate context of the hour and that a comparison of patient’s and analyst’s views
of the treatment might reveal significant differences that were not necessarily the workings of

transference or countertransference. Similar questions could be raised about the status of the past, to
what extent it could be reconstructed in some reliable manner, and to what extent the content of
memories was influenced by the context of the session and by the immediate hopes and fears of the

patient.

The continuous tension between naive realism and the Freudian model led to various kinds of

compensatory strategies. Conceptual terms tended to become more and more ossified, as if the shifting
nature of the subject matter could be held in place by sheer repetition of the explanatory concepts.

Despite Freud’s concession that the metapsychology was to be seen as only a set of temporary
conventions that would be replaced by more appropriate terms as the phenomena became better
understood (Freud, 1915, p. 117), the metapsychology seemed to take on a life of its own. It could even

be argued that philosophical realism played an important role in the choice of such concrete terms as
“structure,” “mechanism,” “splitting,” and “barrier” and that the hoped-for reality which could not be
seen in practice could be found in descriptions about practice. Whatever the reasons, there grew up a

tradition of rewriting the clinical event in the largely mechanistic terms of the theory, giving the
literature a solidity and a tangibility that had been assumed in practice but never found.

www.freepsychotherapybooks.org 85
A similar compensation can be found in Freud’s tendency to posit real events in the past as

causes of the patient’s current symptoms (see Jacobsen & Steele, 1979). It is well known that Freud

first assumed that real seduction was the cause of a later neurosis; but even after he recognized that
the memory was probably false, he continued to introduce real events in his explanatory accounts.

Witnessing his parents’ intercourse was the central event in the Wolf Man’s neurosis; viewing a
monograph in a store window was a significant cause of the Botanical Monograph dream—the hard

stuff of reality was at the root of many symptoms. By always moving the hard facts backward in time,

they could be maintained as explanatory devices even if never actually discovered in the treatment
(see Jacobsen and Steele, 1979, for a fuller discussion of this tendency). Thus, the link to reality was

always assumed, and the patient’s associations were listened to as derivatives or transformations of
significant pieces of the past.

In similar fashion, the unconscious was conceived to be a potentially knowable structure that
had form and content and that impinged on the patient’s behavior in a reliable and accountable

manner. The task of psychoanalysis was to discover its contents and make them available to the

patient; the assumption of a knowable reality applied as much to within as to without. The
transference was equally analyzable, and once significant distortions had been accounted for, the

patient would “see” the analyst as the analyst “really was.” It was never admitted that probably no

amount of analysis could ever accomplish this task.

Overlaying the growing tension between an outmoded realism and the nonneutral analyst was

the conflict between public and private. Freud never felt it necessary to disclose all the facts in
reporting his cases, either because telling too much might risk his authority, because it would

jeopardize the doctor-patient relationship, or because it would not add significantly to his power of

persuasion (see Freud, 1912, p. 114). Within this tradition of privileged withholding, it became

respectable to write about the data instead of making it available; and as this tradition persisted, the
clinical details of the case were overlaid by abstract concepts. No such taboos applied to the

metalanguage—it could be used with impunity—and so it happened that the specific observations of

www.freepsychotherapybooks.org 86
the clinical hour were translated into more general (and in many cases, meaningless) categories. In the
process, the postulates of naive realism could be reaffirmed, and because no one else was present
when patient talked to doctor, no one could say whether or not what was described was really “out

there.”

We can now call on Wittgenstein to make clear what happened next. As the language of

metapsychology became the normal language of psychoanalysis, it became second nature to see the

clinical happenings in terms of the theory. All observations became theory laden and yet were reported
as though they were the pure stuff of observation. Fit between observation and theory was not always

perfect, but because the raw data were never available, the match or mismatch could never be checked.

In this way the metalanguage and its naive realism could be perpetuated indefinitely. In some ways,

the followers of Freud, because they were wearing his blinders, were somewhat worse off than the
founder himself. Language was slowly poisoning observation, and because of the private nature of the

data, no one else could participate in the debate.

We now return to Roy Schafer and to his place within this Zeitgeist. Because of circumstances of

training and experience, he was at odds with the tradition on several counts. He trained at the City
College of New York with Gardner Murphy, a well-known personality theorist, and graduated in 1943;

he then entered a long association with David Rapaport, first at the Menninger Foundation in Topeka,
Kansas, and then at the Austen Riggs Center. He received his Ph.D. in clinical psychology from Clark

University in 1950 and completed formal training in psychoanalysis at the Western New England

Institute for Psychoanalysis in 1959. He has been president of the Western New England Society and
clinical professor of psychiatry at Yale University; and is currently adjunct professor of psychology in

psychiatry at the Cornell University Medical College and training analyst at the Columbia University
Center for Psychoanalytic Training and Research. In 1975 he was appointed the first Sigmund Freud
Memorial Professor at University College in London, and in 1983, he received the American
Psychological Association’s Award for Distinguished Professional Contribution to Knowledge.

Early signs of Schafer’s impatience with the tradition of privileged withholding appear in his

www.freepsychotherapybooks.org 87
books on diagnostic testing, which are notable for their verbatim excerpts from patient protocols
(Rapaport, Gill, & Schafer, 1945-46, Schafer 1948,1954). In these works we have not only diagnostic

impressions of a series of patients but a verbatim record of their responses to the Rorschach test, TAT,

Wechsler-Bellevue Scale, and other diagnostic instruments. Schafer explicitly connects the diagnostic
summary with parts of the protocol, so that the referents for such diagnostic impressions as hysterical

or obsessive character could be found directly in the data. By giving the complete record, Schafer and

his collaborators also make it possible for the reader to develop alternative formulations. Standard
procedure and standard format, one might think, but consider how rarely we discuss alternative

formulations in the clinical literature (Kohut, 1979, is a notable exception) and how we never have
access to the complete data from a complete case.

Concern for the clinical data and for the problems of observation and terminology appear in the
early pages of Aspects of Internalization (Schafer, 1968). The reader of the psychoanalytic literature

may well ask, says Schafer in his introduction, “What does this mean?” He sees the need to introduce

order into the discussion by first sorting out the terms, adopting clear definitions of the critical
phenomena, and, when possible, attempting to talk about these phenomena in plain language

accessible to the professional reader. In efforts that anticipate one of the main themes of his later work,
Schafer takes pains to demystify the standard psychoanalytic formulation and get rid of the implicit

anthropomorphism and demonology of the traditional metapsychology. By trying to bring the clinical
phenomena out of the shadows and into the field of observation, Schafer is once again showing the
respect for the data that characterized his earlier books on diagnostic testing and that would appear in

his attention to observable behavior in his later book on action language (Schafer, 1976). Traditional
metapsychology is seen to be the enemy of observation, not only because it structured the questions to
be asked but also because it shifted the discussion away from the concrete “lurking presences” (in the

case of internalization) to the abstract “cathected object representation.” Some of this shift might be
attributed to a fashionable distrust of Melanie Klein and her too vivid (and hence dubious)
demonology; some might be due to a belief that a proper science should use Latin whenever possible,
so that “cathected object” sounds more respectable than “lurking presence.” Schafer is one of the first

www.freepsychotherapybooks.org 88
psychoanalytic writers since Freud to take the chance of being clear rather than sounding learned, one

of the first to show a concern for language and a willingness to be open and forthright about
experience. Theory is important, but not at the expense of the phenomena, and if these cannot be

reliably described, defined, and contrasted with one another in a systematic manner, the theory will be

a wasted enterprise, superficially impressive but at bottom meaningless.

Aspects of Internalization can be read on two, quite different levels. It is, first, an attempt to
identify the phenomena of internalization, identification, introjection, and incorporation and to

compare and contrast these clinical events. From the clinical descriptions a number of definitions are

generated that lead, in turn, to a clarified theory. The examples are often strikingly specific and

explicitly linked to theory; thus each section of the chapter on identification amplifies one part of the
lengthy definition that is printed at the beginning of the chapter. Even the definitions are arresting, as

in the following example:

An introject is an inner presence with which one feels in a continuous or

intermittent dynamic relationship. The subject conceives of this presence
as a person, a physical or psychological part of a person (e.g., a breast, a
voice, a look, an affect), or a person-like thing or creature. He experiences
it as existing within the confines of his body or mind or both, but not as an
aspect or expression of his subjective self.... The introject is experienced as
capable of exerting a particular influence on the subject’s state and
behavior, and of doing so more or less independently of his conscious
efforts to control it [Schafer, 1968, p. 72].

In order to explain what are often fleeting phenomena, rarely seen for any length of time,

Schafer tries to place them in a more familiar context by beginning with experiences that are relatively
commonplace. The daydream is one such starting point; from here, Schafer goes on to show how it may

often imply a significant shift in reality testing, with the result that the subjective experience is taken as

more real than otherwise. Under these conditions, the introject may come into existence as a piece of
psychic reality. The role of introject is further broadened by using the model of projection. In this

mode, the internalized object is not felt directly, but its influence is mediated by the significant people

www.freepsychotherapybooks.org 89
in one’s life. As this projected role is amplified, the person “out there” disappears and his or her place is
taken by the projected object (as in a paranoid system).

The main argument of Aspects of Internalization is to show how identification, introjection, and

incorporation can each be understood as specific forms of internalization, with each form using a set
number of mechanisms and appearing under certain specified conditions. To carry out this task,

Schafer must necessarily rely more on subjective experience than theory, because the latter tends to be

used inconsistently and often introduces more ambiguity than clarification. In his revised formulation,

Schafer makes clear how the subtypes of internalization can be ordered along a primary-secondary
process continuum, with incorporation being the most primitive, followed by introjection and then

identification. Incorporation refers to the concrete representation of the longed-for object, often in an

oral mode; at times it may take the form of a transitional object inside the head. Introjection is a more
socialized and less regressive form of internalization—an introject, as noted in the definition earlier,

refers to an inner presence that one feels and is influenced by. And finally, identification is the least

regressive of the subtypes and the most abstract. One may identify with one’s teacher even when not
being aware of a conscious presence, and the source of the identification may not be discovered

without a good deal of introspection. Not all incorporations are assembled into introjects and not all
introjects are turned into identifications; nor is the sequence necessarily developmental or phase

specific. There seems to be no need to first compose an introject before going on to form a stable
identification, and many times an introject may appear only when identification begins to break down.
Thus, the theory of internalization lacks the kind of tidiness and order that would lead to specific

developmental or behavioral predictions, and the data are probably more interesting, in their various
manifestations, than any kind of theoretical underpinning. Certainly, the more descriptive parts of
Aspects of Internalization are more arresting than the theoretical conclusions and (at least to my ear)

written with more excitement and urgency.

At a second level, Aspects of Internalization can be read as an outstanding demonstration of

clear clinical description—a sample of exposition that, if successful, would encourage others to follow

www.freepsychotherapybooks.org 90
suit and think twice before using archaic terms or outdated concepts. By bringing the phenomena out
of the shadows of metapsychology and by fashioning a set of contrasting definitions, Schafer is able to
find many overlaps between the unusual and the commonplace. In so doing, he is able to sensitize his

readers to aspects of internalization that they may experience all the time but are probably not able to
label as such. In this branch of psychoanalysis particularly, the data of observation are difficult to

identify because they tend to lie on the edges of awareness and are highly dependent on partially

regressed stages of consciousness. It follows that to ask a patient directly about the presence of an
internalized object is to often cause it to disappear, because the very fact of asking encourages and

reinforces secondary-process modes of function. In contrast to the transitional object of the nursery,

which we see the infant fondle, talk to, and take to bed, the felt presence of a dead father is never seen,
rarely hallucinated, and only referred to indirectly and by implication. It thus becomes doubly

significant, in mapping out this shadowy terrain, that the language of observation be used precisely
and consistently, and it is in this regard that Aspects of Internalization stands head and shoulders above
most of its competitors.

Despite its clear clinical examples and careful use of language, however, Aspects of

Internalization failed to bring about a much-needed revision in the style and terms of psychoanalytic

exposition. The attempt failed in part because the critique of metapsychology was relatively polite and
low-key; more specific criticisms were needed, and they would not appear until Schafer’s next book, A

New Language for Psychoanalysis (1976). In addition, the needed stylistic changes were easily
overlooked. Good exposition becomes transparent precisely because it offers no problem for
comprehension; therefore, the lesson being learned, although doubtless appreciated at some level, may

not be part of the reader’s conscious experience and is thus quickly forgotten. Three years after Schafer
had called attention to the demonology of traditional theory, we were again offered such phrases as:
“the phase-appropriate internalization of those aspects of the oedipal objects that were cathected with

object libido” (Kohut, 1971, p. 41), and “the internalization of the narcissistically invested aspects of
the oedipal and pre-oedipal object takes place according to the same principle” (p. 48). Where is the
patient in these descriptions? Where are the data? What country is being described and who are its

www.freepsychotherapybooks.org 91
inhabitants? Metalanguage had so screened off the data of interest from the reader that once again it

had taken on a reality in its own right. Far from being the temporary scaffolding of Freud, ready to be
dismantled when better models came along, it had acquired permanent status and seemed bent on

edging out the clinical phenomena.

As language became less precise and more abstract, it more than ever began to corrupt

observation and diminish the significance of the data. If we are on the lookout for “narcissistically
invested aspects of the oedipal object,” we will be seeing and understanding much less of the clinical

encounter than if we are on the lookout for lurking presences of the absent father or other concrete

manifestations of the different aspects of internalization. Given the fleeting nature of the data in the

first place, their recognition is just about doomed by fuzzy language and pretentious concepts. The
more rarefied the language, the greater room for argument and controversy.

It may be impossible to speak knowingly of cause and effect, but I suggest that the writings of

Kohut and the advent of Kohutian forms of discourse prompted Schafer to be even more specific about

the data of observation and to shift his focus from inside to outside the head. If the subtleties of the
internalized object were lost in the new language of object relations, then the time seemed ripe to shift

the argument to what could be seen—namely, actions—and to recast psychoanalytic theory in terms of
what could be looked at and pointed to—namely, action language. If the outlines of the introject are
always shifting and its location debatable, then we can bring it outside the head by calling thinking an

instance of action (Schafer, 1976, p. 13). With action language firmly in charge, in Schafer’s (1976)

words, we “shall neither engage in speculation about what is ultimately unutterable in any form nor

build elaborate theories on the basis of unfalsifiable propositions” (p. 10). (Schafer is talking about
theories of mental activity at the beginning of infancy, but the same argument could be applied to a

wide range of elusive phenomena, including the majority of the examples discussed in Aspects of

Internalization.) Actions have the virtue of being more clearly visible and more clearly “out there” than
feelings or thoughts, and there is a certain appeal to bringing all psychoanalytic phenomena out of the

person and into the clear light of day. If we cannot speak with certainty about “where” we feel angry or

www.freepsychotherapybooks.org 92
know exactly what is meant by such expressions as “I am angry up to here,” then there is a temptation
to redefine emotion as action and simply say “he acted angrily.” But a close reading of Aspects of
Internalization makes it clear that the pieces of clinical reporting that ring so true and that carry so

much clinical appeal in Schafer’s earlier books are the very pieces that will be dismantled by a
systematic translation into action language. In return for consensus and agreement, we seem to be in

danger of trading away the very stuff of psychoanalysis. If psychic reality must be recast into action

language to make it knowable, then we may have given away the very stuff of greatest interest to the
practicing analyst. Gone forever—or at least radically transformed—would be the lurking presences,

the vague demonic feelings, fleeting deja vus, the sense of enthrallment to the past, and the awareness

of the uncanny—the full range of subjective reports we have been hearing from patients over the past
100 years.

A positivistic bargain is being struck. If the vague sense of a lurking presence or a sadistic

mother cannot be reliably defined and accounted for in terms of metapsychology—and the failure of

metapsychology on this score seems obvious—and if attempts at clinical description that do justice to
the data are more poetry than science, out of reach of all but the few, then it might still be possible to

improve communication by calling the phenomena something else. If our terms are better chosen,

perhaps some of our descriptive problems might be solved. What we lose with respect to the nuance
and subtlety of observation may be more than offset by an increase in consensus and reliability. This

approach seems laudable; but it betrays a concern for description and control that poses serious
obstacles to its being accomplished, and its positivistic position may represent a fatal flaw. We have
seen how asking questions of certain kinds of fleeting phenomena will cause them to disappear; it

would seem to follow that the traditional subject-object separation cannot be applied to certain kinds
of data and that other methods of study must be devised.

The problem is that the object to be described—for example, the longed-for absent father, the
memory of an early girl friend, or the sense of the analyst as secretly sadistic and vengeful—is not the

traditional object of study that can be set apart from the observer and studied in isolation. It is not an

www.freepsychotherapybooks.org 93
action that can be pointed at, not a thing that can be photographed; rather, it must necessarily be
studied in context when and where we find it. It is this sensitivity to context and to the stream of
experience that Schafer illustrated so well in Aspects of Internalization, and to which he returns in his

most recent work on the concept of narrative (Schafer, 1983). But for a variety of reasons, he preferred
to set it aside in A New Language and shift his focus to observable behavior.

The central theme is sounded in the first chapter (Schafer, 1976). After stating that “it is high

time we stopped using this mixed physiocochemical and evolutionary biological language” (p. 3) of
metapsychology, Schafer proposes the alternative of action language:

We shall regard each psychological process, event, experience, or behavior

as some kind of activity, henceforth to be called action, and shall designate
each action by an active verb stating its nature and by an adverb (or
adverbial locution), when applicable, stating the mode of this action.
Adopting this rule entails that... we shall not use nouns and adjectives to
refer to psychological processes, events, etc.…

…We must understand the word action to include all private psychological
activity that can be made public through gesture and speech, such as
dreaming and the unspoken thinking of everyday life, as well as all initially
public activity, such as ordinary speech and motoric behavior, that has
some goal-directed or symbolic properties.…When speaking of any aspect
of psychological activity or action, we shall no longer refer to location,
movement, direction, sheer quantity, and the like, for these terms are
suitable only for things and thinglike entities.…In order to state
observations in a form suitable for systematic general propositions…we
shall use only the active voice and constructions that clarify activity and
modes of activity [pp. 9-11].

Here are some of Schafer’s (1976) examples: Rather than say “What comes to mind?” the
analyst using action language might say, “What do you think of in this connection?” (p. 148). Rather

than say, “His repression of this dangerous impulse was too weak to prevent it from gaining
consciousness,” the action analyst might say, “By failing to be sufficiently on guard about not doing so,

www.freepsychotherapybooks.org 94
he thought consciously of the action he wished to perform and would have performed had he not

deemed it too dangerous to do so” (p. 206). Instead of saying, “He can’t control his sexual drive,” the

action analyst might say, “He continues to act sexually even though he also wishes he did not do so (or

rebukes himself for doing so)” (pp. 207-208).

Speaking somewhat later in the book in a more general vein, Schafer (1976) argues that his aim

is to eliminate the

unsuitable, confusing, unnecessary and meaningless metaphors and

metaphorical preconceptions that are inherent in Freud’s eclectic
metapsychological language. In this endeavor I shall be building a technical
language using plain English locutions. It is one that should make it
possible to specify in a relatively unambiguous, consistent, parsimonious,
and enlightening way the psychological facts and relations that are of
special interest to psychoanalysts and their analysands” [p. 123].

A New Language for Psychoanalysis is divided into three main sections. In the first, “Preparatory

Studies,” Schafer presents some of the philosophical difficulties with traditional metapsychology and
Freud’s unsatisfactory solutions to what Ryle has called the “ghost in the machine.” Schafer focuses in

particular on the problem of the disappearing person and on the fact that metapsychology has no place

for the “I” or agent. A brief discussion of some alternatives (Hartmann’s adaptive ego, Erikson’s

concept of identity, and Kohut’s narcissistic self) finds them each unsuccessful to some degree; what
Schafer calls the “mover of the mental apparatus” remains clouded behind a screen of theory. Action

language is presented as a possible solution to a long-standing theoretical gap. By using what Schafer

calls the “native tongue of psychoanalysis,” we should be able to catch sight of the disappearing
person.

The second section describes action language, illustrating how it might be applied to a number

of clinical situations and how it clarifies such problems as internalization and resistance and the
understanding of such disclaimed actions as slips of the tongue, motivated forgetting, and so forth. The

www.freepsychotherapybooks.org 95
third section applies action language to emotion by translating noun into verb or adverb. (Instead of

saying, “I am happy about my recent promotion” I might better say, “I view my recent promotion

happily”.) This section presents many examples of how common language is heavily dependent on

metaphor and how metaphor can be misleading and lead to bad theory. Schafer makes clear how it has
invaded metapsychology.

One problem is apparent from the outset. In an effort to divorce himself from the traditional
Freudian metaphor, Schafer must also cut himself off from popular speech and from the way we have

grown up thinking about our body and our feelings. For example, the use of location to express altered

states (as in, “I must have been out of my mind”) is a tradition beginning long before Freud. In an effort

to speak unambiguously about important issues, action language may do quite the opposite and make
them seem strange and foreign because they are being described in unfamiliar language. This

dislocation becomes most apparent when dealing with the lurking presences and other vague

experiences so well described in Aspects of Internalization, because these represent actions only in the
weakest sense of the word; to describe them in action language risks turning them into unfamiliar

specimens.

As Meissner (1979) has argued in his recent critique, metaphor is meant to be taken

metaphorically: “I would have to wonder whether Schafer’s approach to such language is entirely too
literal and fails to take into account the significance of figures of speech.…I am not arguing here that
such propositions cannot be interpreted in the sense that Schafer gives to them.…The issue that I am

addressing…is that such expressions do not necessarily connote that [literal] meaning” (p. 293).

Metaphor can be misleading if taken literally; on the other hand, if taken poetically it can capture an
important truth about ways of thinking and feeling that we all share and on which theory must build.

Metaphor may be particularly useful in at least two contexts: in the generation of new theory where we

need tentative formulation (Freud’s comments on temporary conventions come to mind), and in the

dialogue with the patient, where we are attempting to capture a vague experience for the first time. To
insist on action language when the patient is fumbling for the best expression may often inhibit the

www.freepsychotherapybooks.org 96
discovery process that psychoanalysis tries so hard to foster. To insist on action language while
building theory may unnecessarily restrict the scope of the enterprise by limiting our attention to
phenomena that can be clearly described.

Schafer was particularly impressed by the way in which language in general and the passive
voice in particular can be used in the service of resistance, and one of the most original chapters in A

New Language for Psychoanalysis is titled “Claimed and Disclaimed Action.” Language is easily used to

project ideas of helplessness and disclaimed responsibility, as in “the impulse seized me,” “my
conscience torments me,” “this hour just rushed by,” and “doubts creep into my mind.” In each of these

cases, metaphor becomes defense because the patient is acting as if things just happened to him or her

rather than the patient causing them to happen. But the metaphor can be heard on two levels. To hear

it as a metaphor is to give the patient credit for using it in a figurative sense, saying something like, “my
conscience torments me—so to speak,” which opens the way to analyzing the defense. To hear it as a

literal statement of the patient’s view of life, on the other hand, is to run the risk of challenging the

patient who is following the basic rule and saying what comes to mind. Thus, to treat speech in the
literal way that Schafer suggests is to seriously complicate the analytic relationship by saying to the

patient, in effect, “You must say whatever comes to mind but you will be held responsible for each and

every word.” As I have written elsewhere (Spence, 1982): “To call attention to instances of disclaimed
action would seem to imply to the patient that he is really not free to say whatever comes to mind but

that, in a subtle and all-embracing way, he is being held responsible for his thoughts and—what is
more—being held responsible by the analyst. Thus one could argue that the adoption of action
language may seriously jeopardize the analytic contract” (p. 171).

Now, it is certainly basic to psychoanalysis to assume as Schafer (1976) does, that the patient
“actively brings about that from which he or she neurotically suffers” (p. 145), and some of Schafer’s
most telling anecdotes describe ways in which passive victims are led to see that they have been all the
while secretly arranging their misfortune. But should these accounts of disclaimed responsibility be

analyzed in the traditional manner of gradual interpretation and working through, or by a specific

www.freepsychotherapybooks.org 97
focus on the patient’s words guided by the belief that each psychological event, process, experience, or
behavior is some kind of action? The emphasis on the right and wrong way of saying things (what
Anscombe, 1981, calls “linguistic legislation”) would seem to raise serious questions as to whether

associations can truly be free, whether tentative formulations are open to dispassionate study, and
whether the patient and analyst are collaborating in a mutual enterprise of trust and discovery or one

in which the patient is always put in an adversary position. Even though Schafer has intended his new

language to be a replacement for metapsychology and not a recipe for how to practice psychoanalysis,
it is inevitable that sensitivity to issues of avoiding and claiming responsibility would necessarily have

an effect on treatment (see Spence, 1982).

By focusing on action and activity, on visible over invisible, and on clearly stated rather than

roughly approximated, Schafer inevitably turns from id to ego and, in so doing, raises serious questions
about the central standing of the unconscious. And yet, here is where psychoanalysis begins its quest

and acquires its distinctive character. As Meissner (1979) writes:

If the patient comes to the analysis bearing a burden of unconscious

conflicts and resistances, hidden motives and significances embedded in
his current and past life experience, it is that with which the analyst must
work. If these aspects of the patient’s experience are experienced
somehow passively—granted that they may involve the disclaiming action
that Schafer describes—the analyst must begin by accepting that passivity
and that condition of disclaimed action and engage the patient in a process
which draws him towards a lessening of resistance, an increasing
availability to conscious exploration of unconscious motives, meanings,
and conflicts, and thus gradually lead the analysand in the direction of a
more action-based orientation. In other words, psychoanalytic theory
needs to be a theory of non-action. [p. 306; italics added].

If we follow Meissner and claim that psychoanalysis is a theory (and even more, a practice) of
nonaction, we begin to see why Schafer’s proposals seem to generate such controversy. And it may also

offer a clue to one of the troubling characteristics of metapsychology—the fact that the person

disappears in a field of force and a network of hypothetical structures. We have seen that one of the

www.freepsychotherapybooks.org 98
main goals of A New Language for Psychoanalysis was to make the patient visible again, and it was this
concern that led to the stress on action and responsibility and the concept of human agency. But it may

be that only by making the person inactive and not responsible (as in the classic treatment situation)

can we ever discover the deeper reasons for that individual’s hopes and fears. And it may be that only
by creating a theory which is explicitly not about the person as conscious agent can we begin to

generate a suitable context of explanation.

By putting the stress on the patient as agent, Schafer has necessarily weakened our sense of

psychic reality and its fleeting phenomena. One sense of the loss comes out in comparing Aspects of
Internalization with the chapter on internalization in A New Language for Psychoanalysis. In the former,

psychic reality was described with a dramatic richness of language that seems almost poetic; in the

latter, the descriptions are more prosaic and less familiar. “It is our custom,” writes Schafer in A New
Language for Psychoanalysis (1976), “to speak of introjects as though they were angels and demons

with minds and powers of their own. We speak of them not as an analysand’s construction and

description of experience but as unqualified facts.…We forget…that the introject can have no powers or
motives of its own, and no perceptual and judgmental functions, except as, like a dream figure, it has

these properties archaically ascribed to it by the imagining subject” (p. 163). In other words, the
ascribing should be taken as a form of action, and its products become the responsibility of the patient.

But this renaming tends to decrease the extent to which the analyst can empathize with the patient’s
experience, making the analyst less sensitive to just how haunting the presence may feel. And to say
that the patient is only ascribing these properties does not lessen their impact, just as calling

transference reaction unreal does not make it disappear. Here is an instance in which the sense of an
introject as angel or demon captures an important part of the experience; it represents a piece of
clinical data that we lose by turning to action language. And to the extent that the translation does not

match the patient’s experience, we run the risk of increased misunderstanding and losing touch with
the data.

Schafer hoped that action language would replace metapsychology; we now begin to see

www.freepsychotherapybooks.org 99
reasons why this will not happen. Not only does it fail to capture the richness of the clinical data; it also
fails as an explanation. Although, as we have seen, it is not close enough to inner experience to give a
sense of familiarity and recognition, it is ironically too close to provide a suitable explanation. This

failure comes about because the person as agent represents only the conscious part of the
psychoanalytic domain. To use action terms to generate a general theory is something like trying to

explain what happens inside the atom by studying the psychology of the nuclear freeze movement. The

failure of Schafer’s alternative makes us realize the need for some kind of abstract system that
describes experience but is not cast in the terms of experience, much as the theory of color vision

describes a common happening but is framed in terms of frequencies rather than perceived hues.1

What needs to be kept in mind is Freud’s observation that the explanatory system is only

temporary and will undoubtedly be revised as new facts are discovered. As a provisional model, it is
clearly not meant to be taken literally; it is only the metaphor for the moment and useful only as it

seems to provide explanation. But it must also preserve a certain distance from the phenomena to be
explained, and we now begin to see that the problem of the disappearing patient may have been a sign

that Freud was on the right track. Although terms like “force” and “direction” may seem too crudely

mechanistic, they have the advantage of being psychologically neutral—a key requirement for any

general theory. To frame the model in terms of subjective impressions (as in self psychology) is to rule
out the possibility of making any kind of meaningful discovery about the mind, because it rules out any

investigation in the unconscious.

How then can we summarize the impact of Schafer’s revisionary program? Beginning with the
distaste for metapsychology and its crude physics of force and mechanism, A New Language for
Psychoanalysis held out the hope of returning to the data of behavior and to the “native tongue of

psychoanalysis”—action language. Although at times cumbersome to apply and not suited to

everyone’s tastes, in other contexts it helped us think more carefully about clinical facts, sensitizing us

to certain locutions and letting us see farther into the everyday language of the analytic hour. Certain

kinds of expressions (in particular, the references to disclaimed action) were being heard for almost

www.freepsychotherapybooks.org 100
the first time and in a rather new way. By showing us what new meanings could be uncovered that
were not anticipated by Freud, Schafer paved the way for new ways of reading the text of the hour and
opened the door to new ways of listening.

But there were difficulties with the new language as well, and in many ways, it did not behave
like a long-lost native tongue (see Schafer, 1976, p. 362). To translate anger into “acting angrily” or

resistance into “engaging in actions contrary to analysis while also engaging in analysis itself” (p. 224)

is to complicate rather than clarify, and some would argue that the meaning lost in going to action
language is every bit as great as the meaning lost in going to metapsychology. And the translations are

not always transparent. A New Language for Psychoanalysis tends to read as if observations were there

for the making; we now realize that all observation is theory laden and that Schafer’s native tongue is

no exception. Action language, because it deemphasizes unconscious and passive experience and
emphasizes responsibility and conscious choice, carries significant implications for the process of

treatment. Many of these implications are not explicit, and some of the criticisms of Schafer may stem

from private readings of the words “active” and “passive,” readings that he never intended but that his
program must accommodate.

In certain respects, the most significant impact of A New Language for Psychoanalysis has been

to whet our appetite for a general theory. The difficulties in dealing with the unconscious and with
affect in action language would seem to suggest that some kind of abstract metatheory is a necessary

next step. It is also clear that this metatheory cannot be written in the units of everyday experience.

Since action belongs to a relatively restricted domain of behavior, a good part of our emotional and
unconscious life simply cannot be expressed properly in these terms (the chapters on emotions are the

least convincing of the book). The ground where we choose to build our theory must be equidistant
from both ego and id, from conscious and preconscious, from past and present; and the units of this
theory must lend themselves to translation into clinical concepts (and vice versa) with no significant
loss of meaning.

Even though action language has been found wanting, the discussions around it have opened up

www.freepsychotherapybooks.org 101
central issues that are basic to the future of psychoanalysis. By identifying certain kinds of expressions
that appear in the patient’s language and by showing how they may carry certain implications for the

treatment, A New Language for Psychoanalysis has significantly increased what might be called our

sensitivity to surfaces. By calling attention to the way in which patients use and hide behind language
and by hearing literally (and often for the first time) certain stock expressions of the trade, action

language has increased our ability to listen carefully. In this respect, it belongs to a well-founded

analytic tradition. Schafer’s attention to the data of the consulting room is consistent with his earlier
books on testing and their emphasis on verbatim protocols. The emphasis on language and the text of

the analytic hour puts the focus on units that can be studied, measured, and stored. Even though they
are clearly not the whole story (see Spence, 1981), they are clearly data that cannot be ignored.

What, finally, is the status of Schafer’s “linguistic legislation”? The current interest in how
patients and analysts really speak may have produced a significant and humbling change in our

attitude toward the actual data. Schafer took the position that sloppy language leads to sloppy thinking

(a direct outgrowth of the Wittgenstein school) and that by cleaning up the way we (patients and
analysts) speak about ourselves and our feelings, we can gain greater precision and build better

theory. But it is now becoming apparent that the language contains its own wisdom and that careful
attention to the native tongue of metaphor and common speech may teach us important things about

the clinical encounter, things we can learn in no other way. The close look taken by Dahl and his
colleagues (Dahl, Teller, Moss, & Trujillo, 1978) at the way analysts really speak and the coding scheme
developed by Gill and Hoffman (1982) to analyze the appearance and interpretation of pieces of the

transference are efforts in this direction. As computer procedures come into play and allow us to store
and retrieve vast files of patients’ speech, we may discover regularities that we never knew existed.
Thus, metaphor may not only be used in the service of resistance, as Schafer has pointed out, but subtle

shifts in wording may signal subtle shifts in defense. If analysts are educated about such shifts, they
will be able to deepen their awareness of the meaning of the hour and improve the timing of their
interpretations.

www.freepsychotherapybooks.org 102
In Schafer’s most recent book, The Analytic Attitude (1983), he has turned back to more classical

times. Although there are occasional pieces on action language, the overall tone seems more

conservative and closer to Freud. Even the chapters on narration and the discussion of the patient as a
teller of stories seem to be making manifest what was latent in Freud’s approach. In one section, for

example, Schafer tells us that “Freud used two primary narrative structures, and he often urged that
they be taken as provisional rather than as final truths” (p. 213).

The analytic attitude as seen by Schafer can be characterized as one of empathy and trust, which

generates an atmosphere of safety. There are many technical ways of achieving this atmosphere, some
of which Schafer discusses at length, but the theoretical advantages are also emphasized and clarified.

Only by fostering an atmosphere of safety (see Schafer, 1983, chapter 2) can the analyst create the

conditions for both the identification of resistance and its dissolution; for understanding the patient’s

story in all of its complexity; and for seeing clearly the transference and how it changes over time.
Safety permits the patient to show himself or herself in all aspects—naked and clothed, present and

past, angry and happy. Safety is central because discovery is seen as the key to treatment: “The

appropriate analytic attitude is one of finding out…what the analysis itself will be or be concerned with;
where the principal work will be done;…how this work will best be done;…and how to establish a

termination of the analysis” (Schafer, 1983, p. 21).

Neutrality becomes a central part of the atmosphere of safety. The ideal analyst should be

curious and open to surprise. Schafer (1983) says the analyst should take “nothing for granted

(without being cynical about it) and [remain] ready to revise conjectures or conclusions already
arrived at [and] tolerate ambiguity or incomplete closure over extended periods of time…” (p. 7). The

avoidance of either-or thinking is another aspect of this neutrality and has an obvious relation to the

construction of multiple histories (Schafer, 1983, chapter 13) and to the tolerance of different schools

of treatment (see chapter 17, “On Becoming a Psychoanalyst of One Persuasion or Another”).

What is less emphasized in this picture is the influence of what might be called the analyst’s
world view. Analysts come to their task from a special background of training and experience; as a

www.freepsychotherapybooks.org 103
result, they hear the material from within a certain context. Many descriptive terms have already
acquired specific meanings, and as a result, the analyst will inevitably form images of the significant

figures in the patient’s life—images determined by a turn of phrase that the analyst finds familiar or

influenced by reference to a particular piece of history with which the analyst has personal
associations. Once formed, these images tend to persist, and though they may be sensed as incomplete,

they are less often sensed as wrong, waiting to be corrected. Moreover, correction can never be fully

realized because one of the more effective corrections—a face-to-face meeting with the person in
question—will probably never happen. Although it is certainly true that analysts should remain always

ready to revise their conclusions, Schafer seems to underestimate the difficulties of this task. No one
would disagree with the importance of neutrality and empathy and open-mindedness, but more
attention could be paid to the technical and philosophical problems entailed in this quest.

Schafer’s picture of the neutral analyst tends to overlook the fact that all observations are

theory laden and that certain kinds of material can only be understood by first forming a provisional

model. There seems to be a contradiction between neutrality and forming a provisional hypothesis.
True enough, as Schafer says (1983) the “simplistic, partisan analyst, working in terms of saints and

sinners, victims and victimizers, or good and bad ways to live” (p. 5) is clearly shortchanging the
patient; on the other hand, provisional models are always needed to provide a context for isolated

impressions and to suggest areas that still wait to be discovered. The determining role of the primal
scene is one such model; the possibility of such exposure and its impact on the patient, both immediate
and delayed, is a constant concern of many analysts. In similar fashion, when working with a patient

who is the oldest child they will be sensitized to such events as the birth of the second child and be
constantly on the alert for its derivatives.

The use of provisional models can be witting or unwitting. If it is too much of the second we
may speak of countertransference; If too much of the first, of failure of empathy (as in the cool,

detached analyst who is always forming hypotheses and “testing" them against the “data”). What is less
well understood is that much of psychoanalytic theory is still provisional; that assumptions about

www.freepsychotherapybooks.org 104
primal scene exposure or sibling rivalry represent one class of hypotheses that may not be confirmed
and need to be replaced by others. Thus, one of the common violations of neutrality stems from an
overcommitment to theory and an emphasis on certain parts of the received wisdom.

The issue of alternative explanations is taken up at length in The Analytic Attitude chapter on
multiple histories, and a number of different models are developed and discussed in the subsequent

chapters on “Narration in the Psychoanalytic Dialogue,” “Action and Narration in Psychoanalysis,” and

“The Imprisoned Analysand.” Analysis as journey is one example, as Schafer (1983) makes clear (with
references to the Odyssey, the Divine Comedy, Huckleberry Finn, and Ulysses): “The journey is one of the

world’s great storylines.…We know that in the dreams of analysands all journeys are, among other

things, trips through transference country” (p. 259). Using this model helps the analyst to decode

certain kinds of dream material and to understand the emergence of certain kinds of childhood
memories—travel then may be related in subtle ways to travel now. Another model, developed at

length in Chapter 16, is the model of analysis as prison. Schafer develops with great sensitivity the

positive and negative aspects of this storyline. The happy prison and the safety of closed places may be
seen as an ironic extension of Schafer’s earlier emphasis on the importance of safety in the analytic

attitude; under certain circumstances, the analysis becomes too safe and threatens to become

interminable. The prison model has obvious links with the use of passivity as defense and resistance,
two of the major themes in A New Language for Psychoanalysis. And from another point of view, the

model of the happy prison (safe, but going nowhere) is the complement of the journey of discovery in
which each day brings new adventure and a new outlook.2

Where does the narrative come from? In the last part of the chapter on “The Imprisoned

Analysand,” Schafer (1983) begins to explore this question with the provisional suggestion that the
story is developed jointly by both parties: “By this I do not mean that they have developed it in a happy

collaboration; I mean rather that each has made a contribution, often of different sorts, at different

times, and with different degrees of awareness, reflectiveness, and conflictedness” (p. 278). How does
this mesh with neutrality? It is becoming clear, as Schafer goes on to point out, that the story is not

www.freepsychotherapybooks.org 105
simply being uncovered. This follows from the fact that multiple narratives can be constructed and that
several different models can account for the same pieces of clinical material. Good analysts seem to
work within the hermeneutic circle, using a provisional model (what the European philosopher Hans-

Georg Gadamer would call “fore-understanding”) to build a scaffolding to support the early data, taking
subsequent data to reframe and extend (or dismantle) the scaffolding, and using the revised

framework to see further into the patient’s story and to discover new pieces of information.

Thus, neutrality would seem to consist in the ability to be sensitive to new narratives (new
scaffoldings) as they emerge in the material and as they suggest themselves during the analysis, not in

the absence of models. The analytic attitude becomes a deepened awareness of possible storylines. We

can no longer go back to the myth of the analyst as blank screen who evenly registers all information

by giving equal weight to each new item; if this is neutrality, it is as outmoded as the Monroe Doctrine.
At the same time, as the analyst appears less neutral than we once assumed, it becomess increasingly

urgent that we develop a neutral metatheory—a theory that can handle all possible narratives and

provide a framework for all clinical observations. Thus, the focus shifts from the neutral analyst (an
impossibility) to a neutral theory, and it is in this domain that we may look for Schafer’s contributions

in the years to come.

In coming back to the complexities of the analytic attitude and in trying to go beneath the
surface of the experience, Schafer has returned to the poetic strains of Aspects of Internalization and its

respect for the clinical phenomena. His tone seems more mellow and less didactic. Gone are the

legislative turns of phrase that marred many portions of A New Language for Psychoanalysis, and the
reader feels a greater familiarity with the people and the landscape being described. Ambiguity seems

less an obstacle to understanding (recall the criticism of metaphor in A New Language for
Psychoanalysis) and more a potential source of wisdom (as in the idea of multiple histories). Schafer
seems more willing to take the patient’s story at its own words and to treat it with the same kind of
respect we show a text. The impatience with bad usage or faulty observation that ran through much of

A New Language for Psychoanalysis has been replaced by the respect for the clinical happening that

www.freepsychotherapybooks.org 106
came out so clearly in Aspects of Internalization.

In developing the importance of psychoanalysis as narrative, Schafer (1983) underlines the

importance of the tale and of how it is told; the importance of context and structure over isolated fact;
and the variety of ways in which a life can be presented and understood. “The truth of a psychoanalytic
fact,” he writes, “resides ultimately in the way it fits into the system of interpretation within which it

and its significance have been defined” (p. 277). To emphasize the relational nature of truth is to push

back the Ice Age of Positivism and to argue against the traditional subject-object separation of Big
Science. The patient’s history is no longer an object of study like a bluebird or a molecule, but a

constantly changing story that the patient is writing and rewriting, together with the analyst, inside

and outside the analytic hour. We are just beginning to listen.  

REFERENCES

Anscombe, R. (1981). Referring to the unconscious: A philosophical critique of Schafer’s action

language. International Journal of Psycho-Analysis, 62, 225-241.

Dahl, H., Teller, V., Moss, D., & Trujillo, M. (1978). Countertransference examples of the syntactic
expression of warded-off contents. Psychoanalytic Quarterly, 1,7, 339-363.

Freud, S. (1912). Recommendations to physicians practicing psychoanalysis. Standard Edition, 12, 111-
120.

Freud, S. (1915). Instincts and their vicissitudes. Standard Edition, 14, 117-140.

Gill, M. M., & Hoffman, I. Z. (1982). A method for studying the analysis of aspects of the patient’s
experience of the relationship in psychoanalysis and psychotherapy. Journal of the
American Psychoanalytic Association, 30, 137-167.

Jacobsen, P. E., & Steele, R. S. (1979). From present to past: Freudian archeology. International Review
of Psychoanalysis, 6, 349-362.

www.freepsychotherapybooks.org 107
Kohut, H. (1971). The analysis of the self. New York: International Universities Press.

Kohut, H. (1979). The two analyses of Mr. Z. International Journal of Psycho-Analysis, 60, 3-27.

Meissner, W. W. (1979). Critique of concepts and therapy in the action language approach to
psychoanalysis. International Journal of Psycho-Analysis, 60, 291-310.

Rapaport, D., Gill, M. M., & Schafer, R. (1945-46). Diagnostic psychological testing (2 vols.). Chicago: Year
Book Publishers.

Schafer, R. (1948). The clinical application of psychological tests. New York: International Universities
Press.

Schafer, R. (1954). Psychoanalytic interpretation in Rorschach testing. New York: Grune & Stratton.

Schafer, R. (1968). Aspects of Internalization. New York: International Universities Press.

Schafer, R. (1976). A new language for psychoanalysis. New Haven: Yale University Press.

Schafer, R. (1983). The analytic attitude. New York: Basic Books.

Spence, D. P. (1981). Psychoanalytic competence. International Journal of Psycho-Analysis, 62, 113-124.

Spence, D. P. (1982). On some clinical implications of action language. Journal of the American
Psychoanalytic Association, 30, 169-184.

Notes

1) The problems of focusing on the person as agent are further demonstrated by a look at the object
school of Kohut and his associates. Depending on subjective reports to generate the units
of our theory puts us at the mercy of unreliable witnesses and invisible data; once we
move inside the head, we have given up any hope of consensus or external validation.

2) Not to be overlooked is the model that assumes that the narrative lies in the clinical material,
waiting only to be “discovered.” Freud took some pains to emphasize this model of

www.freepsychotherapybooks.org 108
analyst as archaeologist in order to counter charges of suggestion and influence, and it
has come down to us as part of the received wisdom. One of the implicit themes of The
Analytic Attitude is that this model is probably wrong.

www.freepsychotherapybooks.org 109
4
BENJAMIN B. RUBINSTEIN: CONTRIBUTIONS TO
THE STRUCTURE OF PSYCHOANALYTIC THEORY

MORRIS N. EAGLE, PH.D.

Most of the analysts and theorists included in this volume have attempted to add to
psychoanalytic theory by developing their own clinical and theoretical formulations. This sort of

endeavor is visible and often even produces adherents and disciples. A few theorists contribute by

attempting to clarify the basic structure of psychoanalytic theory. Because they work quietly and do
not often generate the kind of stir that creates followers, it is all too easy to overlook their
contributions. A strength of this volume is that it recognizes the important contributions of one such

theorist, Benjamin B. Rubinstein.

In his writings, Rubinstein is essentially a philosopher of psychoanalysis. But this simple

statement does not really capture the nature and quality of his work. Rubinstein is an analyst with

many years of clinical experience, and his work on the conceptual status of psychoanalysis is written,

so to speak, from the inside. The basic questions he poses are questions that arise in the course of
clinical work (and that most of us slough over and ignore). But what he brings to this probing is a

remarkable and sophisticated philosophical knowledge and style of thinking. After coming to the
United States from Finland, through the efforts of David Rapaport, Rubinstein, as Holt (1967) notes,

www.freepsychotherapybooks.org 110
“made himself into one of the few persons who know as well as Rapaport did the divergent literatures
of psychoanalysis and the philosophy of science” (p. 18). He also brings to his task an intellectual

honesty and conceptual clarity that is unsurpassed by any work in this area. In this paper I will discuss

both Rubinstein’s specific ideas and some general issues which these ideas generate, beginning with a
brief attempt to place Rubinstein’s work in a wider historical and intellectual context.

Broadly speaking, modem efforts to explain human behavior and distinctively human features

(such as consciousness and mentation) have taken one of two philosophical directions. One approach

is to view human beings as nothing but mechanism, as essentially sophisticated machines. A clear and
classical expression of this position is La Mettrie’s (1912) Man a Machine. A more sophisticated and

biological version of this view is Huxley’s epiphenomenalistic view of consciousness, as expressed in

the title of his 1874 paper, “On the Hypothesis that Animals are Automata.” According to this
conception, states of consciousness and presumably other psychological phenomena are no more than

effects of bodily processes. As Huxley stated it: “The mind stands related to the body as the bell of the

clock to the works...” (see Edwards, p. 103). What follows from La Mettrie’s and Huxley’s general
philosophical position is that explanations of human behavior are, in principle, not essentially different

from explanations of physical and chemical phenomena.

An alternative approach is that accounts of human behavior require special explanatory

methods and principles. The neo-Kantian distinction between Geissteswissenschaften and

Naturwissenschaften and the emphasis associated with Dilthey (1961) on Verstehen as the appropriate
method for the study of human phenomena are the prime historical examples of this approach. Recent
emphasis on empathy as the distinctive data-gathering method for psychoanalysis (e.g., Kohut,

1959,1977) and on the so-called clinical theory of psychoanalysis (e.g., Klein, 1976) as well as recent
attempts to conceptualize psychoanalysis as a hermeneutic discipline (e.g., Habermas, 1971, 1979;
Ricoeur, 1970, 1977; Schafer, 1976; see also Grünbaum, 1983, for a superb critique of this point of

view) can be seen as contemporary expressions of Dilthey’s neo-Kantian program.1

In contrast to the above views in which human beings are seen as either nothing but mechanism

www.freepsychotherapybooks.org 111
or immune from laws of nature is recognition that we are from one perspective persons and from
another, organisms. This ontological insight permits Rubinstein to reject a dichotomous either-or

approach to explanations of human behavior. Instead, it leads him to accept the complementarity of

explanation by way of meanings and causes (which parallels the basic complementarity of person and
organism) and to recognize the complex inter dependence between the so-called clinical theory and

metapsychology of psychoanalysis. In recognizing this duality, Rubinstein has preserved one of Freud’s

core insights and one of the primary sources of creative tensions within psychoanalysis (see Holt,
1972, for a discussion of Freud’s two images of humankind).

Rubinstein’s writings and contributions cover a wide range. They include a conceptual analysis

of psychoanalytic ideas such as unconscious mental events and defense; an elucidation of the nature of

clinical inferences in psychoanalysis; the development of a model of mental functioning that is

compatible with both psychoanalytic accounts and neurophysiology; lucid discussions of the mind-

body problem and how it relates to psychoanalytic theory; and a beautiful explication of the nature of

metaphor and related phenomena and their relationship to certain psychoanalytic issues. (Rubinstein’s
1972 paper on metaphor in particular is a wonderful combination of clarity of analysis and exquisite

sensitivity to poetic and literary nuances.) In all these areas, Rubinstein substitutes for casual use of
psychoanalytic concepts careful and detailed examination. For example, the notion of unconscious

mental events is utilized in a casual fashion in the psychoanalytic literature without any apparent
recognition of its ambiguity or the conceptual difficulties it entails. Similarly, psychoanlytic
interpretations and inferences are typically made in the course of clinical work and in case history

descriptions without any systematic attention to the nature of the evidence on which they rest or to
their epistemic and explanatory status. Rubinstein’s rare armamentarium of extensive psychoanalytic
clinical experience, a thorough and deep knowledge of the psychoanalytic literature, and a high degree

of philosophical sophistication permits him to subject such psychoanalytic concepts and methodology
to careful conceptual analysis.

Since Rubinstein’s work is so rich and complex, I can deal only with limited aspects of his work

www.freepsychotherapybooks.org 112
here. I will begin with a brief discussion of his demonstration of the dependence of the clinical theory
for its validation on some form of metapsychology, or extraclinical theory (a term Rubinstein prefers
because it avoids confusion with Freud’s metapsychology). Rubinstein (1967) has shown that what he

refers to as “general clinical hypotheses”—the hypotheses of “partial functional equivalence” or of “the

persistent manifestation potential of unconscious motives” (Rubinstein, 1975, p. 13), for example—

function as axiomatic assumptions in the formulation of specific interpretations and clinical

hypotheses in a particular case. Thus, although we may infer unconscious motives in particular cases,
we can “confirm their presence only if we presuppose the actual occurrence of processes by which the

unconscious motives in question, if in fact present, have been rendered unconscious and being

unconscious, are expressed in various, mostly indirect ways”(Rubinstein, 1980b, p. 13). But, Rubinstein
also (1980a) notes, “the occurrence of these processes cannot be confirmed clinically” (p. 435). It is the

assumption of their occurrence that permits the particular clinical inference. To confirm the existence
of these processes requires the analyst to step out of the clinical context and look to nonclinical,
including neurophysiological, evidence.

Consider another example of the dependence of clinical formulations on some form of

metapsychology. We are justified, Rubinstein observes, in considering parapraxes and symptoms as

motivated and in considering certain behaviors as substitute fulfillments because of the assumptions of
persistent manifestation potential and of partial functional equivalence (including symbolic

equivalence) among different behaviors. Now, there is simply no way one could ever confirm the
hypothesis of persistent manifestation potential of unconscious motives solely on the basis of clinical
data. Such a general assumption clearly requires nonclinical evidence for its confirmation.

This demonstration of the dependence of the clinical inferences and formulations on

extraclinical theory indicates quite clearly the limitations, even the futility, of recent related attempts
to define psychoanalysis solely in terms of its so-called clinical theory (e.g., Home, Klein, 1976; 1966;
Shafer, 1976) and to conceptualize it as a hermeneutic discipline concerned only with interpretation

and meaning.

www.freepsychotherapybooks.org 113
Defining psychoanalysis as a hermeneutic discipline seems to represent, in part, an attempt to

avoid the challenge of how to test and confirm the clinical inferences and interpretations the analyst

regularly employs in clinical work. If psychoanalysis is only a hermeneutic activity, one need merely
view clinical interpretations as “narratives” and “stories.” What Rubinstein has shown, however, is that

these interpretations are not “merely” stories, but are based on extraclinical axiomatic assumptions. If
follows that the validity of these inferences and interpretations ultimately can be tested only if one

steps outside the clinical context. The only self-sufficient clinical theory that can be developed is one

which accepts that its clinical inferences and interpretations will remain untested and unconfirmed.
The conceptualization of psychoanalysis as a hermeneutic discipline, limited only to “narratives,”

“stories,” and other constructions seems to reflect an acceptance of this fate, insofar as it fails to come
to grips with and brushes aside the question of the validity of clinical inferences.

One can attempt to dispense altogether with issues of validity and verdicality by limiting one’s
concerns to therapeutic effectiveness, taking the position that all that one claims for one’s

interpretations is that they provide the patient with a new, more helpful, and more constructive

perspective on life. This position, stated explicitly or implicitly, is increasingly frequent these days. In
its extreme relativism and utter dismissal of issues of validity and truth value, this position seems to

run counter to the central values and outlook that inform Rubinstein’s work. It also runs counter to the

central psychoanalytic tenet that in the final analysis (no double entendre intended), the truth is
liberating. Freud (1917) explicitly stated his belief that only interpretations that “tally with what is

real” will be therapeutic. Although this may or may not be true, the question is central in the

psychoanalytic outlook. When psychoanalysis is defined as a hermeneutic discipline, the question is, so

to speak, legislated out of existence. As I have argued elsewhere (Eagle, 1980), most, if not all, patients
who come for psychoanalytic treatment implicitly and explicitly expect, that they will learn the truth

about themselves, not that they will be provided with “narratives” and “stories,” however helpful they

may be. And I strongly suspect that most psychoanalytically oriented therapists, whatever their

philosophical position, believe that while they are doing clinical work they are helping their patients

learn important truths about themselves rather than simply presenting helpful “stories.” Indeed, I

www.freepsychotherapybooks.org 114
doubt that therapists who believe in presenting “stories” can be maximally effective.

Whatever patients and therapists believe, however, the claim that psychoanalytically inspired
“stories” or “narratives” are therapeutic is (1) simply an assumption, and (2) itself a truth claim—it

asserts that the proposition, “Stories or narratives constructed in the course of psychoanalytic therapy

are therapeutic in such and such ways,” is true. Attempts to validate or confirm this truth claim take

one outside the boundaries of hermeneutics, just as, Rubinstein has shown, attempts to validate or
confirm clinical inferences and interpretations take one outside the clinical theory as commonly

understood. Furthermore, talk about new perspectives and liberating “narratives” takes place without

any reference to systematic and controlled outcome studies that would give substance to at least the

therapeutic claims made for these interpretive narratives.

Common to recent attempts to define psychoanalysis as a hermeneutic discipline, to the claimed

independence of the clinical from the extraclinical theory, and to the failure to seriously consider, let

alone implement, more systematic efforts to gauge outcome of treatment, is an implicit insistence on

the self-sufficiency and autonomy of the clinical enterprise—as if this enterprise could somehow
escape or is immune to issues of accountability on both the epistemological level of validation of

clinical hypotheses and the pragmatic level of effects of treatment. These are disturbing developments,
isolating and solipsistic in their effects. It is as if the response to the difficult and seemingly insoluble
problems of validation of interpretations and clear determination of outcome is to declare them

irrelevant and to aggressively hail the independent legitimacy of the clinical enterprise itself. This

defiant proclamation of self-sufficiency seems to mask an underlying despair of being able to deal

effectively with the complex problems generated by the clinical enterprise. In contrast to this position,
Rubinstein has through the years doggedly attempted to unravel and reveal to us the inherent logic of

clinical inferences and clinical hypotheses and the evidence and assumptions on which they rest (see,

for example, Rubinstein, 1975).

Ironically enough, Rubinstein’s (1975) description and defense of the clinical theory in
psychoanalysis is more systematic and complete than that of those who argue for the self-sufficiency of

www.freepsychotherapybooks.org 115
the clinical theory. He demonstrates that it is at least possible to lend additional credence to both the
general and the specific clinical hypotheses of psychoanalytic theory. In addition, his discussion of

Popper’s falsifiability in the context of confirmation of clinical hypotheses is a gem of lucidity and

simple ingenuity, worth describing briefly. Popper (1962) argues against the scientific respectability of
psychoanalytic theory by maintaining that it is “simply non-testable, irrefutable” (p. 37). According to

Popper, only refutability rather than confirmation are tests of scientificity because “it is easy to obtain

confirmations, or verifications, for nearly every theory—if we look for confirmations” (p. 36).
Rubinstein shows that this argument can be turned into a defense of confirmation in the following

simple and elegant way: The hypothesis (c) “He has an unconscious wish for A,” although not falsified
by the hypothesis (d) “He has an unconscious wish for non-A,” is falsified by the hypothesis (e) “He
does not have an unconscious wish for A.” It seems clear that to falsify hypothesis (c) one would have

to confirm hypothesis (e). But, Rubinstein (1975) notes: “Hypothesis (e) can only be confirmed by an
absence of data confirming hypothesis (c). Accordingly, data confirming hypothesis (c) must be taken
as valid in favor of this hypothesis. Popper’s falsifiability criterion is fulfilled since, as is evident from

the compatibility of hypotheses (c) and (d), the only condition for falsifying hypothesis (c) is the

absence of data confirming it” (p. 46).2

It seems to me that an all too frequent recent response to criticisms of the scientific status of

psychoanalytic theory is to declare that psychoanalysis is to be judged by criteria other than the rules
of evidence and inference characterizing the sciences. Bowlby (1981) sees this response as a reaction

of despair at dealing even adequately with these criticisms. Rubinstein’s response, as the examples
given here demonstrate, is to deal carefully and systematically with such criticisms and to try to make
explicit the kinds of evidence and inference that are critical in the testing of clinical hypotheses.

I will now turn to a concern that, in greater or lesser degree, permeates a good deal of

Rubinstein’s work—the mind-body problem. This problem seems never far from the center of

Rubinstein’s thoughts on psychoanalysis. Consider the themes and issues that have been detailed:
persons and organisms, meanings and causes; clinical and extraclinical theory. All these relate in

www.freepsychotherapybooks.org 116
relatively clear fashion to the mind-body problem. I noted earlier Rubinstein’s pervasive recognition of
the duality of human existence. This should not be misread to mean that Rubinstein takes a dualistic
position on the mind-body problem. On the contrary, he forcefully (and in my view, correctly) rejects

any philosophical position or option which ignores the central fact that we are embodied beings, and
whatever it means to be a person cannot be entirely separated from that embodied status. Rubinstein

rejects not only a metaphysical dualism, which treats mental events as if their ultimate nature were

made up of mental “stuff,” separate and apart from physical matter, but also what can be called a
methodological dualism, which claims autonomy for psychological explanation, whatever its relation

(including one of contradiction) to explanation at the level of neurophysiological functioning. In either

case, Rubinstein rejects the self-sufficiency of mind.3 In his view, a psychological explanation or
account, however clever and ingenious it may be, however intuitively or empathically correct it may

seem, cannot be valid if it contradicts what is known about the principles of neurophysiological
functioning. This will seem self-evident to many, but it is obviously not self-evident to those who take
the position that the formulations and hypotheses of psychoanalytic theory are and should be entirely

derived from the psychoanalytic situation, whatever the logical relationship of these formulations to
other bodies of knowledge.

In rejecting a psychology that implicitly advocates the self-sufficiency of mind and ignores our

embodiment, Rubinstein is being faithful to a core and critically valuable aspect of psychoanalytic
theory. It is Freud’s recognition of the central fact of our embodiment, as expressed in his instinct

theory, that forms the foundation for psychoanalytic theory. Although many of the specifics of Freudian

instinct theory may be deficient or mistaken, what remains valid is Freud’s insistence that our basic
motives and desires as well as our modes of behavior derive from biological imperatives and are

intimately linked to our neurophysiological structure. In rejecting dualism and in keeping in the

forefront the fact of our embodiment, Rubinstein is reminding us of that general insight.

It may seem strange to link Rubinstein to instinct theory. But what I am pointing to is
Rubinstein’s emphasis on our neurophysiological structure as the source of both our motives and the

www.freepsychotherapybooks.org 117
manner in which we go about dealing with these motives. In this sense Rubinstein preserves the
insights that remain valid in Freudian instinct theory; and the rejection of these central insights
characterizes attempts to separate psychoanalysis—either methodologically or substantively—from

the facts of embodiment.4

Rubinstein’s philosophical position on the mind-body problem is expressed in his discussion of

the nature of unconscious mental events. What can it possibly mean, he asks, to speak of unconscious
wishing, wanting, thinking, etc.? According to Rubinstein, unconscious mental events are theoretical

terms that can be described in (1) the language of psychological observables; (2) the language of
neurophysiology or “protoneurophysiology” (as in a disposition for conscious wishing);5 and (3) “as-if ”
mental or phenomenal terms. With regard to the third description, by prefixing the term “unconscious”

to ordinary mental terms such as “wishing,” “desiring,” and “thinking,” we intend to convey the idea

that the person is behaving and acting as if he or she were wishing, desiring, and thinking such and
such, when in fact, in the ordinary sense of these terms, which includes the element of conscious

experience, the person is not so behaving.

A further consideration of how terms such as “desiring” and “thinking” are used in ordinary
discourse helps us make the transition to talking about unconscious desiring and thinking. In ordinary

discourse, to say that one is desiring or thinking X does not necessarily mean that either content X or
the activities of desiring or thinking continually occupy all of one’s conscious experience. There is a

dispositional element to many such psychological terms, by which I mean that someone consciously

desiring X both behaves and is predisposed to behave in certain ways, whether or not, at any given
moment, that person is consciously aware of X or of experiencing desire for X. In thinking, similarly,

when we focus on a problem, for example, we are not necessarily aware of a continual stream of

thoughts or of the uninterrupted experience of thinking. As is well known, one may arrive at a solution
following a period in which one neither consciously experienced any relevant thoughts nor was aware

of thinking. As Rubinstein (1977) and, more recently, Dennett (1978) note, during this period we, as
persons, did not do anything. Rather, our brains did. I would add that we can get some idea of the

www.freepsychotherapybooks.org 118
structure of these brain events by noting the nature of the solution. In describing the solution and the
structure it implies, we often allow ourselves to say that it is as if we engaged in conscious thinking of
such and such a kind.

The point of all this is that even in ordinary discourse, mental terms such as “desiring” and
“thinking” presuppose a more continual neural activity underlying the stochastic and sporadic nature

of conscious experience. This observation was made by Freud and was certainly involved in his general

conclusion that the major part of mental life goes on without awareness. Now, if the ordinary use of
terms such as “desiring” and “thinking” imply neural activity plus a process in which aspects and

portions of that activity are represented in conscious experience, it seems reasonable that unconscious

desiring and thinking, which by definition do not include the element of conscious experience, would

refer to neural activity.

Once having recognized that statements referring to unconscious mental events can be viewed
as “as-if” statements that ultimately refer to neural events, a number of questions immediately arise.

One basic question is whether the conception of unconscious mental events retains the intentionality

(both in Brentano’s [1960] sense and in the ordinary sense of the term) we have in mind when we
speak about mental events. Let me comment here that philosophers are not necessarily entirely in

agreement regarding what is meant by intentionality or the criteria by which a system is judged to be
an intentional one. But for our purposes, we can agree that intentionality refers to such conscious

properties as having purposes and goals, planning, and thinking. Freud’s approach to this issue, which

is entirely consistent with Rubinstein’s, is that the essence of the mental is somatic (neural) processes.
However, Freud (1915b) said, these unconscious mental processes “have abundant points of contact

with conscious mental process.…They can be transformed into, or replaced by, conscious mental
processes, and all the categories which we employ to describe conscious mental acts, such as ideas,
purposes, resolutions and so on, can be applied to them” (p. 168). Hence, Rubinstein (1965) concludes,
for Freud, unconscious mental events are neurophysiological events which are classified as mental on

the two assumptions

www.freepsychotherapybooks.org 119
(a) that observed phenomena resembling the effects of such phenomenal
events as wishing, intending, fantasizing, etc., are in fact the effects of
these neurophysiological events, and

(b) that the latter are in some ways transferable to the particular
neurophysiological events that are correlated with the phenomenal
events, the effects of which their effects resemble [p. 43].

Hence, when we say “Unconsciously, Harry wants to do X,” although strictly speaking we are

referring to a neural event, we generally mean that although Harry does not experience wanting to do

X and will deny wanting to do X, he behaves (here behavior is widely defined to include thoughts,

dreams, slips, and symptoms) as if he wants to do X. Such talk of unconscious mental events is
serviceable and not simply an aberration or anomaly of language, as some philosophers have claimed

(e.g., Field, Aveling, & Laird, 1922), because, as Freud noted, these events have points of contact with

and are describable in terms of conscious mental processes.

We recognize that we can say little regarding the neural events underlying what we describe in

the language of unconscious mental events. What we can do, however, is develop models in a neutral
language that is compatible with both conscious experience and neurophysiological functioning. As we

shall see, Rubinstein attempts to present just such a model. The challenge for any such model is to
accomplish the necessary depersonification of ordinary psychoanalytic statements required by a
scientific rendering and, at the same time, retain the intentionality contained by the ordinary

statements. For example, in an increasingly scientific rendering, a statement such as “Unconsciously,

Harry wants to do X” must be depersonified, but in a manner that will not lose the intentionality that
the original statement contains and that permits the use of “as-if” descriptions. Any depersonified

scheme must reflect, as Rubinstein (1980a) puts it, “not the experience, but what we may regard as the
phenomenological structure of wishing” (p. 438).

In a difficult but provocative paper, Rubinstein (1974) has presented a psychoanalytic

theoretical model of mental functioning which, by virtue of being theoretical, is depersonified, but

www.freepsychotherapybooks.org 120
which nevertheless is intended to be consistent with the phenomenological structure of the activities

of persons. It is also intended to be consistent with, or at least not contradict, what is known about

neurophysiology. The model is presented in terms of classificatory processes and in a neutral language

that is neither neurophysiological nor mentalistic.

One of Rubinstein’s basic intentions is to construct a model in which the kinds of phenomena

that psychoanalysts are interested in, such as motivated behavior and dream symbolism, are generated
and elucidated by the design features of the system. Think of trying to build a machine that is so

designed that it can perceive, recognize, engage in goal-directed behavior, and so on. Such a machine

might yield some insight concerning the formal characteristics necessary to do such things as perceive,

recognize, and so on. In this regard, Rubinstein’s model is in the general tradition of artificial
intelligence and computer simulation. Let me briefly describe the outlines of the model in order to give

some idea of Rubinstein’s attempt to link the psychoanalytic conception of mental functioning to

current scientific thinking.

The model is mainly of motivational processes and the related processes involved in motivated
activity, including perception, recognition, and imagery. Rubinstein’s model of perception is based on
the now commonly accepted central idea that perception is not a passive registration of external

objects, but an active processing of input. This active processing is based on a hierarchically organized
analysis of features. According to this view (e.g., Neisser, 1967), a percept is the result of an active
synthesizing of analyzed features. Thus, the percept orange is achieved by synthesizing the features of

its size, color, texture, smell, etc. And we recognize an orange by classifying it in accord with these

various features. (It can be seen that in this view perception and recognition are closely related
processes). Based on the idea of analyzed features, Rubinstein introduces the concept of object

classifier. A classifier is made up of subclassifiers, each subclassifier corresponding to a different

attribute or feature of the object. Subclassifiers are general properties or features such as “elongated

object,” “two syllables,” “round,” or “begins with the letter s.” Hence, it can be seen that most
subclassifiers will be common to many different object classifiers. It can also be seen that a percept is

www.freepsychotherapybooks.org 121
“constructed” out of subclassifiers or features, much like the title of a book or play is constructed in a
game of charades.

It should be apparent that the notion of a classifier corresponds to the psychoanalytic concept

of object representation. Obviously, a human object classifier will consist of many subclassifier features,
including physical, aesthetic, psychological, and moral attributes. Looking at it this way, one can

imagine the possibility that of a total set of subclassifiers that normally combine in a single percept or

image, particular subsets can become functionally organized, yielding such representations as “good
mother” and “bad mother”. As Rubinstein notes, just as there are object classifications, there are also

self-classifications, which probably interact in various ways with self-standards that we set.

Having elucidated the concept of object classifier, Rubinstein then turns to motivational
processes and introduces the concepts of goal-situation classifier and fulfillment-situation classifier

(which are combined into goal-fulfillment- or GF-situation classifiers) and goal-act disposition. A GF-
situation classifier can be activated from within, which is analogous to the activation of an object

classifier when we think about an object in its absence; from without, as is the case with situations we

refer to as temptations; or spontaneously, as in the case of periodic fluctuations of sexual desire. A GF-
situation classifier is activated by an existing goal situation much the way an object classifier is

activated by the presence of the corresponding object. In both cases, the input is subjected to feature
analysis, which then partly determines whether or not the classifier will be activated.

Rubinstein makes the assumption that once activated, a GF-situation classifier remains active at

least until the motive is fulfilled. What activates a motive is a mismatch between GF-situation classifier
and a particular perception of a situation. Normally, a mismatch will result in instrumental activity
until a match is achieved (which will occur when, during the consummatory act, the situation is

classified as a fulfillment situation). However, a GF-situation fantasy may be activated, particularly if

instrumental action is “judged” not to be feasible. We may note the correspondence between this kind

of fantasy and mental imagery (that is not related to a wish) of an object. In the case of mental imagery,

the classifier activates features in the feature storage (rather than features of perceptual input); while

www.freepsychotherapybooks.org 122
in fantasy, the GF-situation classifier activates corresponding stored GF-situation features.

It should be noted that in this model, although a goal-situation percept can match a goal-

situation classifier, it will not match the goal-fulfillment classifier. The latter is activated by the

activation of the goal-act disposition and the release of the goal act. In ordinary terms, this is
tantamount to saying that although one can experience a situation as an appropriate goal for one’s

motive, one will not experience fulfillment of that motive until the goal act is performed (unless one

posits something like hallucinatory wish fulfillment). Such fulfillment is associated with pleasure and

with a disintegration of the motive structure and its reversion to a mere disposition. There are motives
that do not involve a consummatory act (Rubinstein’s example is a motive such as the desire to be
understood). In such cases, “fulfillment” of the motive is determined entirely by the goal-situation

classifier.

I have given only the briefest sketch of Rubinstein’s model and have omitted quite a number of

details. We can obtain additional insights into the model by seeing how it accounts for certain

phenomena of interest to psychoanalysis. Consider dream symbolism. The basic idea is that goal-
situation (and object) classifiers break up into subclassifiers, with one or more operating

independently to organize an image. For example, if a penis classifier is part of an active goal-situation
classifier, the subclassifier or feature “elongated object” may operate independently and organize an

image of a snake or baseball bat.

How does the model deal with repression? Briefly, certain active goal-and fulfillment-situation

classifiers or a particular set of subclassifiers may match a superego classifier (that is, a classification of

what must not be done, thought about, felt, etc.), which then prevents awareness of the motive as a
motive and also, therefore, of all the subsequent steps that normally accompany awareness of a motive

(such as instrumental action).

I want to remind the reader once again of Rubinstein’s intention to construct a model in which
the design features of the system can yield and account for the kind of motivational and intentional

www.freepsychotherapybooks.org 123
phenomena that are of greatest interest to psychoanalytic theory. Also to be stressed is that the terms

of the model are in a neutral language that is neither mentalistic nor physiological but hopefully

compatible with both. Finally, it is of utmost importance to Rubinstein that the model be not just verbal

description, but falsifiable and discardable if it is not heuristic or is contradicted by the facts.

Returning to the issue of depersonification of explanatory schemes and theoretical models,

Rubinstein (1976b) tells us that in talking about unconscious mental events we extend ordinary
language applicable to persons or, more specifically, to “a sense-of-being-person-doing something” (p.

245). There is no harm in this, as long as we know that we are speaking in this “as-if,” extended

language. Strictly speaking, however, the unobservable and unexperienced activities referred to by

unconscious mental events “are part of our everyday human world in name only” (p. 254). In fact, they
refer to the depersonified natural science world of organisms. This fact tends to arouse in many deep-

seated fears and suspicions toward a scientific enterprise which, in the process of concerning itself

with human behavior, depersonifies it. However, it is important to keep in mind that, as noted earlier,
such depersonification need not and must not eliminate such characteristically human features as

intentionality (in the general sense of the term). Theoretical models need to describe and explain these

features rather than eliminate or ignore them.6 Having said that, however, it is important to note that
theoretical models need not themselves employ the personal language of wants, wishes, and desires.7
As I have argued elsewhere (Eagle, 1980) although wants, wishes, and desires serve an explanatory
function in ordinary discourse, they are themselves phenomena to be explained in a scientific

conception of humankind. One would hardly expect a scientific explanation to limit itself to the

concepts that describe the very phenomena it aims to explain. This is something of what Rubinstein
has in mind when he informs us in a highly condensed fashion that the scientific rendering of

“Unconsciouly, Harry wants to do X” will necessarily involve the depersonification of that statement.

Perhaps the most condensed description of why this is so is Rubinstein’s (1977) reminder that “from a
critical point of view it is illusory to regard a person as the subject-in the sense of being the agent-of an

unconscious activity” (p. 13).8

www.freepsychotherapybooks.org 124
Implicit in Rubinstein’s insistence that the existential referents for unconscious mental

processes are neural events and implicit in Freud’s belief that the essence of the mental is somatic

processes is the seemingly strange idea that neural events themselves (or rather systems of neural
events) possess at least some of the features we normally attribute to and by which we characterize

conscious mental processes. I am not at all certain that one can justifiably speak of, let us say, the
intelligence of neural events, except perhaps in a metaphorical sense. But, at least in a certain sense,

they are intelligent-perhaps in the same sense that computers are intelligent. It has been customary to

think of all physical processes as inherently “blind,” that is, without intelligence or intentionality, and
to locate these latter qualities in the mind and/or the person. However, there are certain perceptual

and cognitive phenomena that imply often elegantly intelligent processes which are not and often
cannot be represented in conscious experience. I will provide some examples.

Consider as the first example the dichotic listening situation in which subjects are presented
with messages simultaneously on two different channels and are instructed to attend to and read aloud

a message on one of these channels. Typically, they can report only gross physical features (for

example, a male voice) from the unattended channel and cannot report the content. However, Lackner
and Garrett (1973) have shown that messages in the unattended channel influence the particular

interpretation given to ambiguous sentences presented in the attended and shadowed channel, even

though subjects could not report what they heard in the former. As Dennett (1978) notes, “the influence
of the unattended channel on the interpretation of the attended signal can be explained only on the

hypothesis that the unattended input is processed all the way to a semantic level, even though the

subjects have no awareness of this—that is, cannot report it.” (p. 211).

As another example, consider an experiment by Lazarus and McCleary (1951) in which subjects

are presented a series of words exposed tachistoscopically for a brief duration and are asked to report

what they see. When the stimulus word is “raped,” many subjects report seeing “rapid.” Their galvanic
skin response (GSR) measurements, however, are of a magnitude associated with emotionally laden

words such as “raped” rather than neutral words such as “rapid.” As in the first example, some aspects

www.freepsychotherapybooks.org 125
of the subject’s response indicate that the stimulus has been processed accurately, even though the
subject is not aware of it and does not report processing the stimulus.

The next two examples are somewhat different from the first two. They focus on phenomena

that reflect the problem-solving nature of perceptual processes which are not and cannot be
represented in conscious experience. The first example is the well-known Ames room, in which the

ceiling and floor are sloped in a manner unobservable to the viewer. A child standing in the corner of

the room where ceiling and floor converge will look markedly taller than an adult standing in a comer
where ceiling and floor diverge. This illusion is irresistible and persists even if the onlooker is told how

the room is constructed. What is perceived seems based on a tacit inference that someone whose head

is very close to the ceiling is obviously taller than someone whose head is not so close. Normally,

ceilings and floors are parallel to each other, and this tacit inference or “rule” will be highly accurate
and serviceable. In the context of the Ames box, that what is immediately perceived and experienced

follows that tacit “rule” rather than what is consciously known. In fact, the immediate experience is, as

noted, irresistible and not changed by one’s conscious knowledge.

The second example in this area has to do with stroboscopic movement. If, let us say, the image
of a triangle flashes at point A and then, after an appropriate interval, at point B, one will experience

the triangle moving from A to B. As Rock (1970) notes, this perceptual experience is based on the tacit
inference “that if an object is now here in this field and, a moment later, it is not there but elsewhere,

then it must have moved” (p. 9). Indeed, Rock reports that the experience of movement can be

eliminated “if, simultaneous with the flashing on of B, A reappears in its original location as well; in
other words, if you flash A then A-B, then B then A-B and so forth, A need not be ‘deduced’ to have

moved to B if it is still where it was a moment ago” (p. 9). The experience of movement can also be
destroyed if a and b appear as two objects being successively uncovered and covered. As Rock notes,
the experience follows the “impeccable logic” that “if the first object is covered over, it has not moved
to location but remained where it is” (p. 9). Evidence such as this leads Rock to conclude that

“perception turns out to be shot through with intelligence” (p. 10) and to support Helmholtz’s (1962)

www.freepsychotherapybooks.org 126
rule that “…objects are always imagined as being present in the field of vision as would have to be
there in order to produce the same impression on the nervous mechanism.” (p. 5).

Finally, consider the seemingly simple phenomenon of experiencing vertigo after getting on an

escalator that is not moving. One infers that the person experiencing such vertigo had unconsciously
“expected” the metal stairs to move. That such unconscious expectations are different from ordinary,

conscious expectations is evidenced by the fact that knowing beforehand that the metal stairs are not

and will not be moving does not eliminate the vertigo. As Polanyi and Prosch (1975) note with regard
to “tacit inferences” in general, such phenomena seem to be relatively immune to adverse evidence. To

say that one unconsciously expected the metal stairs to move is, to Rubinstein’s way of thinking, an “as-

if” use of “expectation,” which does no harm and is certainly useful insofar as it is structurally

analogous with both conscious experience and the neural events for which it is an approximate
description. However, as Rubinstein warns us, to give existential implications to unconscious

expectations is erroneous. In a certain sense, there is no such thing as an unconscious expectation. It

provides only a very approximately and vague linguistic window on certain neural events that
intervene between getting on the stationary metal stairs and experiencing vertigo.

The ontological status of the processes involved in the phenomena described in these examples

is difficult to pinpoint. At least since Helmholtz advanced his concept of “unconscious inference,” there
has been debate regarding the status of such processes. Helmholtz recognized that these processes

have a cognitive, inferencelike property and yet are immediate and automatic and are not represented

in conscious experience. The term “unconscious inference” was intended to capture both aspects of the
process. Helmholtz’s concept fell into disrepute, mainly as a result of the criticism that, by definition,

inferences could not be unconscious and, therefore, the notion of an unconscious inference was an
absurdity. However, the phenomena in our examples, attesting as they do to the inferencelike
processes involved in perception, have led to a revival of the concept of unconscious inference.

It is instructive in this regard to consider the situation in so-called cognitive science. In that

area, descriptions are given of inferred and hypothetical cognitive processes that are neither

www.freepsychotherapybooks.org 127
represented in conscious experience nor tied to specific brain events. Rather, the emphasis is on the
structure of these cognitive processes. Similarly, one can say of the processes represented by

Helmholtz’s “unconscious inference” that they are not in conscious experience, nor can one specify the

neural events to which they refer. What the concept does, however, is to inform us that leading up to
some perceptual experiences are certain inferencelike processes—that is, they function as if they were

making a conscious, logical inference of an if-then kind. Hence, Helmholtz’s concept essentially reveals

the (inferred) structure of certain processes that, at this point, cannot be further specified. One can
interpret them as ontologically neutral. Similarly, Chomsky’s (1965) concept of “deep structures” is

also a structural description that is neither represented in conscious experience nor tied to specific
neural events. It is meant to reveal some important things about the structure of the mind; however, it
is embodied. It seems to me that in Rubinstein’s way of looking at the concept of unconscious activity,

an interpretation cast in the language of unconscious wishes or wants is a structural description that,

despite their differences,9 functions much like Helmholtz’s “unconscious inference” and Chomsky’s
“deep structures.”

It says something like: “Your behavior and associations are patterned as if you wish or want

such and such,” much like a statement of unconscious expectations says that one’s vertigo is as if one
expected the escalator to move. Casting statements about unconscious activity in the ordinary

language of “narratives” about wishes and wants has the dual advantage of not only being potentially

therapeutic, but also permitting one the freedom to describe patterns of behavior with as few
constraints as possible. But Rubinstein’s conceptualization of unconscious activity makes clear that

these “narratives” ultimately have to answer to what is actually the case. That is, they must be

consistent with what we know about the structure of neural processes. This single consideration seems
to me to be a sufficient reason that a psychoanalytic theory that makes use of the concept of

unconscious activity cannot be entirely construed as a hermeneutic discipline. As Rubinstein (1974)

observes, “…no matter how apt an interpretation of a symbol in terms of its meaning, if the processes
by which symbol formation is explained are improbable, we have no alternative but to discard the
interpretation” (p. 105).

www.freepsychotherapybooks.org 128
Keeping Rubinstein’s clarifying comments regarding unconscious activities in mind, it would

seem that the perceptual experiences in our examples are as if we were engaging in logical inferences.

But such inferences or, more accurately, inferencelike processes, cannot be ascribed to a person insofar
as the person is not aware of such activities. Hence, it seems to me that such intelligent, inferencelike

processes must be ascribed to neural events and brain processes.10 To state it generally, intelligence
resides in subpersonal neural processes. I do not pretend to be able to explicate this notion much

further, except to say that such neural processes must have been selected out in the course of evolution

and to point to the work of others who have attempted to develop further this idea of subpersonal
intelligence and intentionality (e.g., Dennett, 1969, 1978).11

It seems to me that the notion of subpersonal intelligence and intentionality is also implicit in
some of Freud’s basic formulations. This is seen in a number of ways. The very basic scheme of

partitioning the personality into id, ego, and superego can be seen as implying subpersonal intelligence

and intentionality. Strictly speaking, insofar as id, ego, and superego are unconscious processes, they
are, ontologically, brain processes ascribable to an organism. However, as we have seen earlier, Freud

(1915b) tells us that unconscious mental processes “have abundant points of contact with conscious
mental processes” and can be described by the categories applicable to conscious processes. Hence, id,
ego, and superego are not simply metaphors of what persons do, as is claimed, for example, by Schafer

(1976) in his “action language,” but are both (1) labels for particular constellations of neural events

and brain processes and, (2) names for classes of wishes and dispositions to behave in certain ways
and to have experiences of certain kinds. In other words, as with Rubinstein’s classification model

discussed earlier, one can think of id, ego, and superego as a neutral language description that will

ideally capture something of the structure of both neurophysiological functioning on the one hand and
behavior and conscious experience on the other.

If unconscious processes are, ontologically speaking, neural in nature and if, as Freud

maintained, such processes constitute the basic psychic reality, then the seemingly peculiar conclusion
one is led to is that psychic reality is neural! (See Nagel, 1974, for a further elaboration of this

www.freepsychotherapybooks.org 129
argument.) This conclusion is not as peculiar as it may seem. For what else can unconscious activities
be but neural events? But they are at the same time mental, insofar as they are characterized by
intelligence and intentionality. For Freud, it should be noted, what defined “mental” was not

phenomenal experience but what I am referring to here as intelligence and intentionality. For Freud,
conscious experience was not the essence of the mental but only a surface and sporadic representation

of an ongoing underlying activity. Hence, when Freud writes that the ultimate and underlying psychic

reality is unconscious, he is essentially saying that the underlying psychic reality is neural. Although
Freud abandoned his attempt to implement in detail this point of view (in the Project for a Scientific

Psychology, 1895), this general conception of psychic reality was never abandoned.

A critical question raised by the psychoanalytic conception of unconscious activity is how an

unconscious want or idea becomes transformed into a conscious want or idea. If, as Rubinstein
maintains, an unconscious want refers essentially a kind of neural activity, how does it ever get

represented in conscious experience? This question has always been central to psychoanalysis. How

does the unconscious become conscious, and how does it get to be represented in personal experience?
I do not pretend to have even the beginnings of an answer to this question. But implicit in

psychoanalytic theory and in some current conceptions is the idea that much of our behavior is guided

by subpersonal intelligent and intentional processes and that only the products of some of these
processes are represented, with varying degrees of distortion, in conscious experience. Conscious

experience can be conceptualized as a selective and constructional rendering of products of underlying

subpersonal processes. If, as Dennett (1978) suggests, there is a subpersonal system that processes
“inner events,” it is the products of such processing that are selectively represented in consciousness

(just as it is the products of visual processing that are selectively represented in perceptual visual
experience). Surely this is implied by Freud’s belief that the major part of mental life goes on outside
awareness.

The opportunities for defense and dissimulation arise in the representation and rendering of

these subpersonal products. In strictly psychoanalytic language, this would be stated largely in terms

www.freepsychotherapybooks.org 130
of the degree to which conscious experience and the ego accurately represent unconscious instinctual
aims.12 If unconscious aims are only metaphorical descriptions of neural activity, the issue becomes
the degree to which conscious experience and the ego accurately represent the subpersonal neural
activity we are really referring to when we talk about unconscious aims.

Another issue that has been central to psychoanalysis is the degree to which subpersonal aims
are integrated into those structures we think of as consciousness and selfhood. Obviously, that which is

not represented in these structures cannot be integrated into them. But it is possible for certain

subpersonal aims to be represented in but not integrated into consciousness, as in the case of an ego-

alien obsessive thought. Indeed, what we mean by a partial failure of repression is that the aim linked
to the obsessive thought is rather clearly represented in consciousness, but in an unintegrated, ego-

alien state—in contrast to a more complex repression in which the aim is only very indirectly

represented or hardly represented at all.

In any case, the point here is that in psychoanalytic theory, a central aspect of personality
integrity concerns not only representation but the successful integration of subpersonal tendencies

and aims into a superordinate, higher-order structure identified as one’s (largely conscious) self. This

central idea is conveyed by the dictum “where id was, there shall ego be,” which can also be translated

as “where the impersonal ‘it’ was, there shall the personal ‘I’ be.” Although the impersonal “it” has been

equated with instinctual aims, it can also be interpreted as referring to all those unconscious
subpersonal tendencies that are not but can become part of the “I,” the personal self. If, however, the

impersonal “it” is essentially neural activity (and it is difficult to see what else it can be), then Freud’s
dictum is tantamount to the assertion that one can claim or reclaim, so to speak, bits of neurology and
transform them into psychology. Or, to put it somewhat differently, the self assimilates bits of the

impersonal and transforms as well as integrates them into the personal, thereby expanding the realm
and domain of the latter.

No wonder the mind-body issue is at the center of both Rubinstein’s writings and of
psychoanalytic theory! It may seem less strange to speak of transforming and integrating bits of

www.freepsychotherapybooks.org 131
neurology into psychology if one takes the perspective that every bit of conscious experience
represents such a transformation. According to the logic of Freud’s conception of psychic activity,

every bit of conscious experience entails making the unconscious conscious. What is distinctive about

the process when it is discussed in the therapeutic context is that active forces (i.e., repression) have
both rendered certain contents unconscious (hence, the concept of the dynamic unconscious) and

interfered with the smooth transformation of unconscious (neural) activity into conscious experience.

The picture of psychoanalysis that emerges from Rubinstein’s (as well as Freud’s) conception of

unconscious mental events is radically different from the current conception of psychoanalysis as
hermeneutics and from the current emphasis on “stories,” “narratives,” and related constructions. As I

have already noted, for Rubinstein these “stories” and “narratives” ultimately depend for their validity

on confirming through nonclinical means general hypotheses regarding our basic structure. Now, from
the point of view of unconscious activity as neural events, to ascribe to the individual unconscious

wishes, wants, and so forth is, in an approximate and metaphoric way, to describe the structure of that

person’s mind, with mind identified as an intentional but nevertheless neural system. Hence, it is not
merely a matter of a “story” or “narrative” that makes sense, but of an account that cannot contradict

what we do know about the structure of mind in general. In other words, implicit in Rubinstein’s view
is the idea that psychoanalytic interpretations regarding unconscious mental events, although cast in

the ordinary language of desires, wants, and actions, are, in some cases, groping descriptions of brain
processes and hence, not only must not contradict what we know about brain processes, but must
actually reflect something about the structure of the latter.

The final issues I want to deal with in this discussion of Rubinstein’s work emerge from

contrasting his formulation with Schafer’s (1976) “action language.” By referring to all mental events,
including unconscious activity, as action, Schafer makes it clear that in his view all the phenomena with
which psychoanalysis deals are to be ascribed to the person. Unconscious motives are to be seen as

disclaimed actions. One consequence of ascribing unconscious activity to the person (rather than
ascribing such activity to the organism) is Schafer’s conclusion that we are all responsible for such

www.freepsychotherapybooks.org 132
activity. (Thus, Schafer’s cites with approval Freud’s [1925] comment that we are responsible for our
dreams). Schafer’s conclusion does, indeed, logically follow from his premise. For, if unconscious
activities are things we do to accomplish particular ends, then they fit the model of action and the

practical syllogism that describes action; and, if these activities are actions, we are responsible for
them. But Rubinstein’s analysis of unconscious activities sensitizes us to such questions as how an

activity, the goal or aim of which we are not consciously aware, can be an action. He points out that

certain motives, particularly unconscious motives, function more as causes propelling activity than as
reasons for action. Schafer does not deal with these issues, but rather attempts to resolve the

conceptual difficulties inherent in the notion of unconscious activity merely through the verbal device

of labeling such activity “action.” As for our responsibility for unconsciously motivated behavior,
Rubinstein’s analysis suggests that this whole issue represents confusion between different universes

of discourse. That is to say, the whole question of responsibility applies to the everyday world of
persons and actions (and the social-legal-ethical contexts it generates), whereas talk about
unconscious wants and desires, insofar as it is a metaphoric description of neural activity, belongs to

the world of organisms.

There is a good deal of Rubinstein’s work that has not been covered here, and what has been

discussed has not had the rigor and details that characterizes Rubinstein’s own work. That, of course, is
inevitable in a chapter of this kind, but I believe that what I have discussed represents some of the

more important and central themes of Rubinstein’s work.

In summarizing some of the main related themes in Rubinstein’s writings, first and foremost is
Rubinstein’s awareness of the dual perspective one can adopt toward human existence—that is, we are

both persons in an everyday human world and organisms in a natural science world. This awareness—
which, I believe, is also central to psychoanalytic theory—permeates much of Rubinstein’s work. It
permits him, for example, to understand the complementarity of meanings and causes rather than
pitting one against the other.

A second theme is Rubinstein’s relentless quest to understand the relationship between the

www.freepsychotherapybooks.org 133
world of persons and the world of organisms and to avoid confusion between the two worlds and
contexts. This quest is reflected in his analysis and clarification of psychoanalytic concepts such as

unconscious mental events and in his writing on the mind-body problem. It is also reflected in

Rubinstein’s attempt to develop a “neutral language” model of mental functioning that will be faithful
to the worlds of both persons and organisms.

A third theme in Rubinstein’s work is his elucidation of the logic and nature of clinical inference

in psychoanalysis. More than any other psychoanalytic theorist, Rubinstein attempts to explicate

clearly and, as I described it earlier, “doggedly” the nature of the evidence and inference rules that
legitimate clinical inferences. His description of how particular clinical hypotheses are confirmed

represents one of the few systematic attempts in this area. Also, his demonstration of the logical

dependence of particular clinical formulations on general clinical hypotheses and the dependence of
the latter on extraclinical sources of evidence represents the most effective argument against an overly

narrow conception of psychoanalysis.

As important as such specific themes and contents, however, is the unrelenting intellectual
honesty, clarity, and rigor of Rubinstein’s thinking. I hope I have given the reader some idea of these

qualities.  

REFERENCES

Abelson, R. (1977). Persons: A study in philosophical psychology. New York: Macmillan.

Ames, A., Jr. (1946). Nature and origin of perception. Hanover, NH: Institute for Associated Research.

Bettelheim, B. (1982, March 1). Reflections: Freud and the soul. New Yorker, 52-93.

Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York: Basic Books.

Bowlby, J. (1981). [Symposium on E. Peterfreund on Information and Systems Theory]. Psychoanalytic

www.freepsychotherapybooks.org 134
Review, 68, 187-190.

Brandt, L. W. (1966). Process or structure? Psychoanalytic Review, 53, 50.

Brentano, F. (1960). The distinction between mental and physical phenomena. (D. B. Terrell, Trans.). In
R. M. Chisholm (Ed.), Realism and the background of phenomenology (pp. 39-61). Glencoe,
IL: Free Press.

Chomsky, N. (1965). Aspects of the theory of syntax. Cambridge, MA: MIT Press.

Dennett, D. C. (1969). Content and Consciousness. London: Routledge & Kegan Paul.

Dennett, D. C. (1978). Toward a cognitive theory of consciousness. In C. W. Savage (Ed.), Minnesota

Studies in the Philosophy of Science: Vol. 9. Perception and Cognition: Issues in the
foundations of psychology (pp. 210-228). Minneapolis: University of Minnesota Press.

Dilthey, W. (1961). Meaning in history. (H. P. Rickman, Ed.) London: Allen & Unwin.

Eagle, M. N. (1980). Psychoanalytic interpretations: Veridicality and therapeutic effectiveness. Noús,

14, 405-425.

Eagle, M. N. (1980a). A critical examination of motivational explanation in psychoanalysis.

Psychoanalysis and Contemporary Thought, 3, 329-380. (Also in L. Laudan, Ed., 1983.
Mind and Medicine. Berkeley and Los Angeles: University of California Press, pp. 311-
353).

Eagle, M. N. (1980b). Interests as object relations. Psychoanalysis & Contemporary Thought, 4, 527-565.

Eagle, M. N. (1984). Recent Developments in Psychoanalytic Theory: A Critical valuation. New York:
McGraw-Hill.

Fairbairn, W. R. D. (1952). Psychoanalytic Studies of the Personality. London, Tavistock.

Field, G. C., Aveling, F., & Laird, J. A. (1922). Symposium on Is the conception of the unconscious of value
in psychology? Mind, 31, 413-442.

www.freepsychotherapybooks.org 135
Flew, A., (1949). Psychoanalytic explanation. Analysis, 10, 8-15.

Freud, S. (1895). Project for a scientific psychology. Standard Edition, 1, 295-397.

Freud, S. (1915a). Instincts and their vicissitudes. Standard Edition, 14, 117-140.

Freud, S. (1915b). The unconscious. Standard Edition, 14.

Freud, S. (1917). Introductory lectures on psychoanalysis. Standard Edition, 16,448-463.

Freud, S. (1925). Some additional notes on dream interpretation as a whole. Standard Edition, 19, 25-
30.

Gedo, J. K. (1979). Beyond Interpretation. New York: International Universities Press.

Grünbaum, A. (1984). Foundations of psychoanalysis: A philosophical critique. Berkeley: University of

California Press.

Habermas, J. (1971). Knowledge and Human Interests (J. J. Shapiro, Trans.). Boston: Beacon Press.

Habermas, J. (1979). Communication and the Evolution of Society (T. McCarthy, Trans.). Boston: Beacon
Press.

Helmholtz, H. Von (1962). Treatise on Physiological Optics (Vol. 3; Translated from the third German
edition). J. P. C. Southall, (Ed. & Trans.). New York: Dover.

Holt, R. R. (1967). [Editor’s introduction.] In R. R. Holt (Ed.), Motives and thoughts: Psychoanalytic
essays in honor of David Rapaport. Psychological Issues, 5 (2/3, Monograph 18/19), 18-
19.

Holt, R. R. (1972). Freud’s mechanistic and humanistic images of man. In R. R. Holt and E. Peterfreund
(Eds.), Psychoanalysis and Contemporary Science. New York: Macmillan.

Home, H. J. (1966). The concept of mind. International Journal of Psycho-Analysis, 47, 42-49.

Huxley, T. H. (1967). In P. Edwards (Ed.), Encyclopedia of Philosophy, Vol. 3 (pp. 102-103). New York:

www.freepsychotherapybooks.org 136
Macmillan. (Original work published 1893-94).

Klein, G. S. (1976). Psychoanalytic theory: An exploration of essentials. New York: International

Universities Press.

Kohut, H. (1954). Introspection, empathy aand psychoanalysis. Journal of the American Psychoanalytic
Association, 7, 459-483.

Kohut, H. (1971). The restoration of the self. New York: International Universities Press.

Lackner, J. R., & Garrett, M. (1973). Resolving ambiguity: Effects of biasing context in the unattended
ear. Cognition, 1, 359-372.

La Mettrie, J. (1912). Man a machine. La Salle, IL: Open Court Publishing Co.

Lazarus, R. S., & McCleary, R. A. (1951). Autonomic discrimination without awareness: A study of
subception. Psychological Review, 58, 113-122.

Louch, A. R. (1966). Explanation and human action. Oxford: Blackwell.

Nagel, T. (1974). Freud’s anthropomorphism. In R. Wollheim (Ed.), Freud: A collection of critical essays
(pp. 11-24). New York: Anchor Books.

Neisser, U. (1967). Cognitive psychology. New York: Appleton-Century-Crofts.

Polanyi, M., & Prosch, H. (1975). Meaning. Chicago: University of Chicago Press.

Popper, K. R. (1962). Conjectures and refutations. New York: Harper, 1968.

Ricoeur, P. (1970). Freud and philosophy: An essay in interpretation. New Haven: Yale University Press.

Ricoeur, P. (1977). The question of proof in Freud’s psychoanalytic writings. Journal of the American
Psychoanalytic Association, 25, 835-871.

Rock, I. (1970). Perception from the standpoint of psychology. In Perception and its disorders, Vol. 68
(pp. 1-11). Association for Research in Nervous and Mental Disease.

www.freepsychotherapybooks.org 137
Rubinstein, B. B. (1965). Psychoanalytic theory and the mind-body problem. In N. S. Greenfield & W. C.
Lewis (Eds.), Psychoanalysis and current biological thought. Madison: University of
Wisconsin Press.

Rubinstein, B. B. (1967). Explanation and mere descriptions: A metascientific examination of certain

aspects of the psychoanalytic theory of motivation. In R. R. Holt (Ed.), Motives and
thoughts: Psychoanalytic essays in honor of David Rapaport. Psychological Issues, 5, 2/3
Monograph 18/19, 20-77.

Rubinstein, B. B. (1972). On metaphor and related phenomena. Psychoanalysis & Contemporary Science,
1, 70-108.

Rubinstein, B. B. (1974). On the role of classificatory processes in mental functioning: Aspects of a

psychoanalytic theoretical model. In L. Goldberger & V. H. Rosen (Eds.), Psychoanalysis
and contemporary science, Vol. 3 (pp. 3-57). New York: International Universities Press.

Rubinstein, B. B. (1975). On the clinical psychoanalytic theory and its role in the inference and
confirmation of particular clinical hypotheses. Psychoanalysis & Contemporary Science, 4,
3-57.

Rubinstein, B. B. (1976a). Hope, fear, wish, expectation, and fantasy: A semantic-phenomenological and
extraclinical theoretical study. In T. Shapiro (Ed.), Psychoanalysis and contemporary
science, Vol. 5 (pp. 3-60). New York: International Universities Press.

Rubinstein, B. B. (1976b). On the possibility of a strictly clinical psychoanalytic theory: An essay in the
philosophy of psychoanalysis. In M. M. Gill & P. S. Holzman (Eds.), Psychology versus
metapsychology: Psychoanalytic essays in memory of G. S. Klein (pp. 229-264). New York:
International Universities Press.

Rubinstein, B. B. (1977). On the concept of a person and of an organism. In R. Stern, L. S. Horowitz, and
J. Lynes (Eds.), Science and psychotherapy (pp. 1-17). New York: Haven.

Rubinstein, B. B. (1978). Psychoanalytic hypotheses and the problem of their confirmation. In K. D.

Irani, L. Horowitz, & G. Myers (Eds.), Pathology and consciousness (pp. 35-50). New York:
Haven.

Rubinstein, B. B. (1980a). On the psychoanalytic theory of unconscious motivation and the problem of

www.freepsychotherapybooks.org 138
its confirmation. Noús, 14, 427-442.

Rubinstein, B. B. (1980b). On the psychoanalytic theory of unconscious motivation and the problem of
its confirmation. Psychoanalysis & Contemporary Thought, 3, 3-20.

Rubinstein, B. B. (1980c). The problem of confirmation in clinical psychoanalysis. Journal of the

American Psychoanalytic Association, 28, 397-416.

Rubinstein, B. B. (1980d). [Review of The self and its brain: An argument for interactionism, by K. R.
Popper and J. C. Eccles.] Journal of the American Psychoanalytic Association, 28, 210-219.

Rubinstein, B. B. (1981, January 27). Person, organism, & self: Their worlds and their psychoanalytically
relevant relationships. Paper presented at a meeting of the New York Psychoanalytic
Society, New York.

Schafer, R. (1976). A new language for psychoanalysis. New Haven: Yale University Press.

Sellars, W. (1963). Science, Perception, and Reality. London: Routledge Kegan Paul.

Taylor, R. (1964). The Explanation of Behavior. London: Routledge and Kegan Paul.

Winnicott, D. W. (1949). Mind and its relation to the psycho-soma. In D. W. Winnicott Collected Papers:
Though Pediatrics to Psychoanalysis. New York: Basic Books.

Notes

1) 'Within philosophy, certain formulations concerning the nature of human action and its claimed lack
of susceptibility to causal explanation also seem to me to be contemporary expressions
of this program (see for example, the work of Abelson, 1977; Louch, 1966; Taylor, 1964).
One also sees in Schafer’s (1976) work-in the very notion of “action language”-the
influence of these philosophers of action on a conception of psychoanalysis.

2) The letters of Rubinstein’s passage have been changed to conform to my example.

3) It seems to me that the recent cluster of formulations including the hermeneutic vision of

www.freepsychotherapybooks.org 139
psychoanalysis, the autonomy of the clinical theory, and the exclusive emphasis on the
psychoanalytic situation reveals an underlying attitude that implicitly proclaims the
autonomy of the mental and that denies our embodied, material nature. This attitude, in
part propelled by a reaction against the purported dehumanizing influence of the
scientific Weltanschauung, characterizes many recent intellectual developments,
particularly in the social sciences. Ironically, although the failure to include and do
justice to such essential psychological considerations as subjective experience and
intentionality in an explanatory system may be dehumanizing, it is equally
dehumanizing, though perhaps in a less obvious way, to fail to include and do justice to
our embodied status. We certainly recognize in our clinical thinking that the isolation
and separation of mind from body is alienating and dehumanizing. For example,
Winnicottt (1954) points to the role of excessive mentation and the separation of mind
from what he refers to as the psycho-soma in schizoid conditions.

4) I am not suggesting, as do some defenders of Freudian instinct theory, that all divergences from and
criticisms of that theory are based on a rejection of the biological—of our embodiment.
Indeed, some of these criticisms entail an expansion of the instinctual. For example, Bow
toy's (1969) rejection of what he calls Freud’s “secondary drive” theory of the infant-
mother relationship is based on the positing of an independent instinctual attachment
system. And Fairbaim’s (1952) dictum that “libido” is object seeking” can be construed as
positing an inborn response to objects. (See Eagle, 1981, for a further discussion of these
ideas.)

5) A disposition to behave (or think or feel) in a particular set of ways can be taken as the
manifestation of a neural structure.

6) As Sellars (1963) puts it, “…to complete the scientific image we need to enrich it not with more ways
of saying what is the case, but with the language of community and individual intentions,
so that by construing the actions we intend to do and circumstances in which we intend
to do them in scientific terms, we directly relate the world as conceived by scientific
inquiry to our purposes and make it our world and no longer an alien appendage to the
world in which we do our living” (p. 40).

7) Indeed, even if such terms are used in a theoretical model, they will have meanings different from
the ones they ordinarily have, as Rubinstein has shown is the case with unconscious
wants, wishes, and desires. It is also possible, as Chomsky (1965) notes, that in giving a
physical, depersonified explanation for such mental phenomena as wants, wishes, and

www.freepsychotherapybooks.org 140
desires “the very concept of “physical explanation’ will no doubt be extended to
incorporate whatever is discovered in this [mental] domain, exactly as it was extended
to accommodate gravitational and electromagnetic force, massless particles, and
numerous other entities and processes that would have offended the common sense of
earlier generations” (pp. 83-84).

8) That one is not the agent of an unconscious activity or, more accurately, some variation of this
insight, is undoubtedly one of the important considerations that lies behind Freud’s
division of the personality into id and ego. If one goes back to the original German terms
this becomes clearer (see Bettelheim, 1982; Brandt, 1966). Thus, Das Es or “the it”
(rather than the id) obviously represents those aspects of the personality that are not
experienced as agent but rather as impersonal happenings; Das Ich or “the I” (rather
than the ego) clearly is meant to include those aspects of the personality that one
experiences as personal agent. Although the concept of ego came to include more than
this-unconscious defensive activities, for example—the fact remains that in Freud’s
tripartite division of the personality, that which is experienced as personal agent belongs
to the ego. Freud’s equation of id with instinct and his difficulty in deciding whether it
was to be defined psychologically, biologically, or somewhere between the two (see
Freud, 1915b, pp. 111-116) reflects the fact that Freud’s id-ego division is, in part, body-
mind distinction (see Eagle, 1984). As is the case with the concept of unconscious
activity, in the concept of id, one is not the subject—in the sense of being the agent—of id
strivings. And yet, also as in the case of unconscious activity, id strivings are nevertheless
intentional and purposive. Because they are intentional, we want to attribute them to an
agent. As Flew (1949) points out, in our habitual style of thinking we are accustomed to
identify intentional and purposive with conscious and voluntary, not with unconscious,
impersonal, and peremptorily involuntary. We are used to thinking of intentional activity
as doings carried out by personal agents. However, as Dennett (1978) notes, subpersonal
systems can be intentional systems. It seems to me that this point is implied in Freud’s
attribution of motivational aims to subpersonal structural components of the
personality.

9) An essential difference is that whereas unconscious wishes or wants can become conscious,
Helmholtz’s “unconscious inference” and Chomsky’s “deep structures” cannot, almost by
definition.

10) It should be clear that not all the implications I draw from Rubinstein’s formulations would
necessarily be shared by Rubinstein himself.

www.freepsychotherapybooks.org 141
11) In the examples of perceptual phenomena, a subpersonal system such as the visual system has
what Dennett (1978), borrowing from computer language, calls “computational access”
to certain stimuli. What the person has access to, continuing with computer language, is
some of the computational products of the visual system’s processing. The latter are
represented in conscious experience, whereas neither the stimuli to which the visual
system has “computational access” nor the visual system’s processing are so
represented. One may also speculate, as Dennett does, that just as the visual system has
access to certain stimuli, there is very likely an “affect” system within the person that has
access to certain “inner” events (for example, hormonal secretion and hypothalamic
stimulation). Continuing with the analogy, just as the products of the processing by the
visual system are consciously experienced percepts, so the products of the hypothetical
“affect” system are experienced as wants and desires. Finally, just as the individual
“constructs” percepts when the product of visual processing is unclear, so one
“constructs” reasons, desires, motives when the product of the “affects” system
processing is unclear.
It should be noted that although I mention subpersonal intelligence and intentionality,
the problems presented by each are not necessarily equivalent. Thus, the essence of
certain machines is that they are intelligent, as the very term “artificial intelligence”
indicates.

12) It is interesting and consistent with what I have been proposing that Gedo (1979) states as an
important goal of psychoanalytic treatment the raising of biological aims and needs to
the level of conscious awareness.

www.freepsychotherapybooks.org 142
5
EMANUEL PETERFREUND: THE INFORMATION
REVOLUTION

STANLEY R. PALOMBO, M.D.

When Emanuel Peterfreund’s Information, Systems, and Psychoanalysis appeared in 1971, it

posed a new and unusual challenge to traditional psychoanalytic beliefs. Peterfreund’s work was not

simply a development of ideas that were already competing within the ideological arena marked out by

Freud’s discovery of unconscious mental activity. More significantly, it presented a new framework of
ideas within which the unique achievements of traditional psychoanalysis could be integrated with the
profound conceptual changes currently taking place throughout the natural and biological sciences.

The effects of these changes are still only beginning to be felt, but they have already produced a

picture of the universe quite different from that of Freud’s time, a picture in which information has

replaced energy as the central unifying concept. My primary objective in this essay will be to trace the

significance of this changing world picture for psychoanalysis. Information, Systems, and Psychoanalysis

has a central role in this inquiry. It raised many of the basic questions that must be answered if
psychoanalysis is to maintain its position at the forefront of scientific thought.

Rubinstein (1975, 1980), whose investigations of psychoanalysis as a revolutionary episode in

www.freepsychotherapybooks.org 143
the history of science have cast a bright light on the conceptual problems inherited by psychoanalytic

theory from the prepsychoanalytic past (1975,1980), says in his preface to Information, Systems, and

Psychoanalysis: “We are here on an adventurous journey, into what, from the viewpoint of most

analysts, must appear as strange and exotic territory. But it is a journey that must be undertaken. The
alternative is a standstill, as a consequence of which current metapsychology will most likely become

increasingly alienated from science generally and hence scientifically irrelevant” (p. 6). And, in a recent
symposium on the significance of Peterfreund’s work, Bowlby (1981) notes:

The material of psychoanalysis, it is sometimes contended, is not a kind

that can be dealt with by means of conventional scientific procedures: it
needs special procedures of its own. An alternative reaction is to search
the current scientific scene to discover whether any of the more recent
concepts and theories that have been developed can be harnessed to
provide a model for psychoanalysis better fitted to its subject matter. This
is what Emanuel Peterfreund has done. (p. 187)

Reppen (1981), in his introduction to the same symposium, remarks: “It is curious that
Peterfreund in his updating of Freud delivers another narcissistic blow to man’s old view of himself as

central in the universe. To Freud’s earlier observation that man is not master in his own house must
now be added the notion that man may be merely an automaton-one must hasten to add, perhaps to

soften the injury, an incredibly complicated one” (p. 159). Reppen seems to be suggesting that this

narcissistic injury was responsible in large part for the “considerable neglect” from which

Peterfreund’s work has suffered. This neglect has taken place, as Reppen notes, despite Peterfreund’s
training and origins in the mainstream of psychoanalysis. Peterfreund, who attended the City College

of New York and the University of Chicago Medical School, trained at the New York Psychoanalytic

Institute. He is an associate clinical professor of psychiatry at Mount Sinai Medical School as well as a
member of the American Psychoanalytic Association.

The resistance to Peterfreund’s revolutionary contribution to psychoanalytic theory illustrates

the difficulties encountered by a scientific community when its investment in the past becomes an

www.freepsychotherapybooks.org 144
obstacle to further advancement. Information, Systems, and Psychoanalysis goes directly to the heart of
the problem in the older theory. It provides a point of view, first of all, from which the conceptual

inconsistencies of traditional metapsychology can be clearly inspected. It makes the cumbersome

improvisations required to circumvent these inconsistencies visible for what they are.

As in the psychoanalytic process itself, the diagnosis is the beginning of the cure. A different

kind of theory was needed, a theory at once simpler in its essentials and more advanced in its powers

of implication, the kind of theory Copernicus offered to the tradition-mired astronomers. While many

analysts were wondering whether the desperate remedy to the problem of theoretical obsolescence
was to cut the remaining ties between psychoanalysis and the rest of the science, Peterfreund was

showing that better science was the only real choice.

SCIENCE AND PSYCHOANALYSIS

Freud was fascinated by the emotional shock effects produced by sudden, radical changes in the
scientific world view. He saw the massive resistance to Copernicus and Darwin as evidence that their

discoveries had undermined a collective fantasy of human centrality and mastery. He believed that

psychoanalysis was meeting the same massive resistance because his discovery of the unconscious had
undermined that fantasy even further (Freud, 1914).

Freud showed how tentative is our control over our own minds and how much of what we

ordinarily consider to be within our conscious control is better thought of as belonging to something

external to our self-awareness, a “psychic apparatus” functioning outside our consciousness and our

capabilities for rational decision making. But his imagery for representing the psychological opposition
between what he called the “I” and the “it” was little more than a metaphorical letting loose in the

human mind of the purely physical forces that had been tamed during the nineteenth century in the

factory and the laboratory.

www.freepsychotherapybooks.org 145
Natural science in Freud’s time was dominated by the notion of energy. The conceptual

vocabulary from which Freud created his metapsychology was formed by the great advances in the

physical sciences during his own lifetime. The discovery of the various forms and manifestations of
physical energy, their interchangeability, and the conservation of quantities through transformations

from one form of energy to another, made it appear that the physical universe had been completely
understood. The human mind seemed to stand outside this rush of physical transformations, as an

interested but uninvolved observer.

Part of the shock effect of psychoanalysis resulted from its refutation of the myth of progress
engendered by the advances in the physical sciences. This idea had been grasped by many as a

replacement for the outmoded religious mythology that put humankind at the center of creation. But

another aspect of the shock was its appropriation of the vocabulary of the physical sciences to reach its

pessimistic conclusions about the power of the human mind. Freud’s success in turning the myth of
scientific progress against itself seemed to finish off whatever was left of collective human narcissism.

Nevertheless, the psychoanalysts who followed Freud developed a myth of their own, a myth

that became an obstacle to further theoretical changes. If the discovery of the unconscious meant the

ultimate deflation of human vanity and self-deception, then no further surprises about the nature of

the mind could come from the other sciences. Any claim to that effect would have to be treated as a
denial of the importance of the unconscious. The psychoanalysis of Freud’s time would become the

permanent basis for “a general psychology,” even though its conceptual scheme had been inherited

from nineteenth-century physics.

But, as Information, Systems, and Psychoanalysis repeatedly points out, science in our own time

has moved in a direction that makes nineteenth-century physics increasingly irrelevant to psychology.
Contemporary science is primarily concerned not with forces but with structures and procedures. Its
subject matter is the accumulation of patterned information in complex systems, biological and

otherwise. An organism is no longer thought of by biologists as a collection of chemical reactions, but

as a hierarchy of organizational structures. Within this conceptual framework, the human mind takes

www.freepsychotherapybooks.org 146
its place as a system like others, differing in the degree of its complexity but not in its possession of
unique attributes or qualities.

Despite Freud’s repeated minimizing of the role of rational thought in determining human

behavior, his model of the mind in conflict requires the presence of a rational human agency—the ego
—striving through intelligent procedures to dominate the naturally occurring chaos of instinctual

forces (Freud, 1923). Freud’s attempt to derive the structure of the ego from an evolutionary process

guided only by the clashing of these unstructured natural forces was brilliantly conceived but doomed

to failure from the beginning. Without a scientific conception that included information and structure
as essential features of all natural process, it could not succeed.

To complete Freud’s project for understanding the origin and development of the ego as a

sequence of natural events, it is necessary to see that the natural world includes not only the clash of

unstructured forces but, even more important, a hierarchy of procedures for conserving and
transforming information as well. Taking this approach, we are drawn inevitably to the idea that the

large-scale intelligent procedures used by the human mind to do its work in the real world must be
integrated systems of smaller and smaller intelligent subprocedures. These subprocedures, in turn,

must exist independently in relatively simple nonhuman systems—in the genetic mechanisms of the
living cell, for instance, and in intelligent computer programs.

Although psychoanalysis seems in retrospect to have been the first of the information sciences

(Pribram & Gill, 1976), Freud’s energic metaphor for the world of nature did not allow him to
anticipate either the shocking realization that we share our sapience with microorganisms and
machines or the freedom from anthropomorphic misconceptions that follows from the realization. By

introducing psychoanalysis to the higher level of generalization made possible by concepts of

information processing, Peterfreund restored the discovery of the unconscious to its proper place in
the continuing sequence of disillusionments that must accompany the progress of science. Freud’s

momentous contributions were relieved of the burden of incredibility assumed by all final revelations.

www.freepsychotherapybooks.org 147
THE PRIVILEGED EGO

The realization that the executive ego is also an “it” has not yet penetrated very deeply into the
psychoanalytic consciousness. Peterfreund showed that the privileged ego, exempt from the

constraints that apply to all other natural systems, has been a refuge for psychoanalysts from the
seriousness of Freud’s scientific goals. Many analysts appear to believe that it would be
“dehumanizing” to venture even a single step beyond the limit of Freud’s personal achievement in

unmasking the mechanical element in human mental life. Some have even insisted that subjective

emotion, the most complex of integrative experiences, receives its due as an influence on human life
only when it is represented with the poetic simplicity of a thunderstorm or a tidal wave. (The science

of our grandparents’ generation always seems soothingly humanistic when compared with our own.)

Information, Systems, and Psychoanalysis met this resistance head-on. Part 1, “A Critique of

Current Psychoanalytic Theory,” made the privileged ego the special target of its criticism. To the
analyst who already thinks of the ego as an organizational concept rather than as the experiencing self,

Peterfreund’s proposal to remove this familiar term from the lexicon of psychoanalytic theory entirely

may seem rather bewildering. But, despite the emphasis placed on this concept by the psychoanalytic
ego psychologists, confusion on this point is still widespread in the psychoanalytic community.
(Hartmann’s [1950,1952] attempts to integrate organizational concepts into a theoretical framework

derived from the concept of energy could never be fully convincing, for reasons already discussed
here.)

Peterfreund’s proposal was intended to focus attention on the inconsistency that results when

the ego is exempted from the chaotic imagery with which Freud depicted the rest of the psychic
apparatus. By disregarding the role of the ego as an organizational structure in ego psychology,
Peterfreund was deliberately sharpening the contrast between the inconsistencies of the older theory

and the rigor promised by the new. As a tactic in the reform of psychoanalytic theory, this move may

have misled many of the analysts he was trying to reach. As a statement that the structural attributes of
the ego are not derivable from the energic axioms of the metapsychology, however, it has its own

www.freepsychotherapybooks.org 148
internal logic.

THE INFORMATION FRAMEWORK

Part 2 of Information, Systems, and Psychoanalysis (Peterfreund, 1971) “Basic Information-

Systems Concepts,” outlined the new frame of reference within which Peterfreund was locating

psychoanalysis. Here, with the collaboration of Jacob Schwartz, a computer scientist, he presented a
technically rigorous view of information

[as] having to do with…patterns of physical events or the relationship

between patterns of events. A pattern of one physical form can be
transduced into a pattern of another physical form, and the latter in turn
can be transduced into a pattern of still another physical form. What
remains the same in this sequence is the information; it is the common
factor in the sequence of changing patterns [p. 115].

What will seem strange to the psychoanalyst in this view is its neutrality with respect to the

origin and meaning of the patterns being transduced or transmitted. Information is not necessarily

“about” anything. It doesn’t have to be a “message” from a “transmitter” to a “receiver.” The motivation,

if any, of the agents concerned, if any, is a separate problem to be taken up at another structural level.

When a tree falls in the forest, a pattern of compression waves radiates through the surrounding
atmosphere. This pattern constitutes information. Whether it falls on the ear of an organism, and
whether that organism can interpret the information as the sound of a tree falling, are separate

questions entirely. Peterfreund thus begins with the fundamental distinction in information theory. It
separates the physical traces of events, the “evidence,” from any possible interpretation of their
meaning or significance. The sound of a tree falling is a function not only of the pattern of air waves

radiating from the tree, but of the information contained in the ear and brain of the listening organism
as well.

How can this distinction be useful to the psychoanalyst? The analyst is concerned precisely with

www.freepsychotherapybooks.org 149
questions about the highest levels of organization, questions about motivation and meaning. That the

sound of middle C is heard when the air is vibrating at so many cycles per second is hardly relevant to

the experience of the opera lover. In contrast to Peterfreund’s position, Rosenblatt and Thickstun

(1978) would restrict the use of the term “information” to the coded record of the physical events
within the listening and interpreting organism or machine. This, they believe, would overcome the

trivialization that might occur if “every nonrandom phenomenon in the observable universe” were
considered to be information. But I think Peterfreund’s point that the more general definition has

greater power is a valid one, especially when psychopathology is concerned. The internally coded

record of a physical event must in some essential respect be isomorphic with the actual event. For a
particular listener or processor, how “nonrandom phenomena” are recognized as being both

nonrandom and relevant to the listener’s or processor’s interests and needs, is still an important

empirical question.

Peterfreund’s formulation provides the useful reminder that in every hierarchical system, all
constraints that apply at a lower level of the system also apply at all higher levels. We can substitute a

patient relating a fantasy for the tree falling in the forest. What is heard by the listener is a function of

the information contained in the listener’s ear and brain as much as in the words and the tone of the
speaker, but the listener must begin by responding to what is actually there in the patient’s
communication.

If the analyst hears what he or she considers to be evidence of a repressed infantile wish, the

analyst’s judgment must be tempered by specific information about the patient’s state of mind at the

moment and at crucial moments in the past, by general information about human development and the
psychic mechanisms of repression and symptom formation, and, finally, by information about the

analyst’s own state of mind in the present and throughout his past. This information must all be

internally consistent and it must all fit together to form the optimal interpretation.

Peterfreund points out that at any moment in a typical analysis much of this information is

either unavailable or unverifiable. A major function of psychoanalytic theory in the clinical situation is

www.freepsychotherapybooks.org 150
to provide hypotheses to fill in temporarily for the missing information. When these hypotheses are
themselves consistent they can be helpful to the analyst in organizing the information actually

available and in identifying specific questions that still need to be answered. But if the analyst’s

hypotheses contain internal contradictions, they will necessarily produce distortions in what the
analyst hears.

Because every theoretical formulation is the product of its own historical development, it will

always be subject to further modification as new information becomes available. Information theory

suggests a number of ways to minimize the consequences of having to work with a fallible theory. One
is to be on the alert for inconsistencies between levels in the hierarchy of theories that supports the

theory in question. This means that although biological and psychological theory cannot “explain” the

particular phenomena with which the psychoanalyst is concerned, no hypothesis of psychoanalysis can
be allowed to contradict what is known at the time about biological and psychological processes.

Perhaps the most dramatic example of a contradiction in the hierarchy of theories underlying

psychoanalysis is the one that resulted from the discovery by Aserinsky and Kleitman (1953) that
dreaming sleep occurs in a constantly repeated pattern of 10 to 20 minute periods occurring at regular

90-minute cycles throughout the night, regardless of the content of the dreams. This laboratory finding
renders untenable the traditional psychoanalytic view that dreams are caused by the eruption of

repressed impulses from the unconscious (Freud, 1900). It takes nothing away from the clinical

observation that repressed wishes are expressed in the content of dreams, of course. But it does
undercut the entire theoretical structure built on the assumption that impulses are capable of
achieving expression without the cooperation of the executive apparatus.

In the case of dreaming, the executive apparatus is creating and delimiting the opportunities for

repressed wishes to be expressed as dream contents. This implies that the expression of the repressed
wishes is not the result of a self-initiated drive for discharge but part of an adaptive process for

evaluating the urgency of the impulses being aroused by current life experience (Palombo, 1978,

1980). A consistent psychoanalytic theory will have to take these nonpsychoanalytic facts into account.

www.freepsychotherapybooks.org 151
This example illustrates the general point that every higher-level theory has embedded within it a host
of lower-level theoretical assumptions. For this reason, psychoanalytic theory cannot be skimmed off

the top of the human sciences and treated as if it were completely independent.

Without an explicit awareness of lower-level assumptions, it is difficult to specify what would

constitute reliable evidence for or against a prediction made by a higher-level theory. Observations and

predictions must each be formulated at the same level of precision to be useful in testing the validity of

a prediction. For the psychoanalyst trying to match global theoretical conceptualizations with

fragmented samples of the patient’s intrapsychic experience, this can be critical. Information theory
can be of considerable help here, because it requires the theorist to be clear about relationships

between hierarchical levels and component subsystems.

The result is an opportunity to subdivide a problem as often as necessary for its components to

match the scale of the phenomena being observed. John Clippinger’s brilliant computer simulation of a
patient’s production in psychoanalytic therapy provides a dramatic demonstration of this method at

work (1977). The simulation begins with a repressed sexual wish uncovered toward the end of a
session. Five interacting structures transform this input by generating a formal expression for the

wish, giving it a linguistic form, censoring it, revising it to conform to the censorship, and reintegrating
the censored version with aspects of the original wish.

The output of the simulation is a passage that almost exactly matches the transcript of an earlier

interaction during the hour in which the patient’s conflict was expressed in what seems like a random
and aimless digression. Of special importance is the network of connections among the five internal
structures of the simulation that Clippinger calls Leibnitz, Calvin, Machiavelli, Cicero and Freud. Each of

these structures has the power to interrupt and modify the output produced by some but not all of the

others. The original wish passes through each of these structures many times. By dividing the
processing among these interacting components, Clippinger was able to master the complexity of the

patient’s internal production of the text.

www.freepsychotherapybooks.org 152
Peterfreund gives priority to what can be directly monitored in the therapeutic situation—the

feedback loops that regulate the interaction of the patient and the analyst. The importance of feedback

as an error-correcting procedure was recognized by Shannon and Weaver (1949) as early as 1942.
Monitoring the differences between the current situation and the desired outcome was shown to be an

essential feature of any complex problem-solving system by Newell, Shaw, and Simon (1957). Miller,
Galanter, and Pribram (1960) applied this principle as a general tool for analyzing the behavior of

organisms. They showed that every action performed by an organism presupposes a preexisting goal

and a plan for reaching that goal. After an action has been completed, its success in reaching the
designated goal is evaluated by the organism. Before any subsequent action is to be taken, errors are

identified and the plan modified to reduce them. The stream of behavior produced by the system is
therefore the integration of many repetitive cycles of planning, acting, evaluating, and correcting.

Analysts who understand only a part of what they need to know about a patient can add to their
understanding by breaking down the patient’s stream of behavior, identifying the patient’s moment-to-

moment goals, reconstructing the patient’s plans for achieving them, and noting how the patient

modifies the plans when they fail. To do this, analysts must continually test their own theoretical
formulations for their success in helping to identify the patient's goals and to reconstruct his or her

plans. From the analysis of these feedback loops, larger structures can be discovered. For example, it is

quite likely that the patient has his or her own set of more general plans for modifying unsuccessful
moment-to-moment plans. These more general plans may be either adaptive or defensive. If adaptive,

they will enhance the flow of information through the patient-analyst system. If defensive, they will

constrict the flow of information. The same may be said for analysts’ procedures for dealing with

discrepancies between their theoretical formulations and the patient’s actual behavior. Most of this
monitoring and processing takes place outside the analyst’s direct awareness. It is usually referred to

in noncognitive terms, as intuition, identification or empathy.

Peterfreund’s argument suggests that there is nothing to lose and everything to be gained in

making these procedures explicit. His new book, The Process of Psychoanalytic Therapy (1983), shows

www.freepsychotherapybooks.org 153
how ideas derived from information theory can make a critical difference in the technique of
psychoanalysis. This important practical issue will be considered, along with this new work, later in
this chapter.

THE SELF-INITIATING IMPULSE

The conceptual distinctions of information theory lead to significant theoretical differences with

traditional psychoanalytic metapsychology. As I mentioned earlier, the idea of a self-initiating impulse

has been radically undermined by the findings of the sleep laboratory. But this idea is also
incompatible with the information theory point of view on very general grounds, as well as with the

principle of psychic determinism, emphasized by Freud as a major discovery of psychoanalysis. Within

the information-processing framework, the Freudian “impulse” is actually a compound formed by

matching competing demands for the gratification of a need with competing plans for achieving a
desired gratification. At any moment, priorities must be assigned to current needs and then an optimal

plan chosen from the many possible plans available. The choice of a plan will depend on many

determining factors. If the demand is an urgent one, for example, the corresponding action will very
likely follow a preplanned routine designed to be set in motion on extremely short notice. A
preplanned action of this kind will necessarily be simple, direct, nonspecific, and inflexible. These are

the characteristics used in traditional psychoanalytic theory to support the notion that “impulses” are
self-initiating, peremptory, and indifferent to the particular channel for discharge open to them at the
moment. This, in turn, is taken to justify the radical separation of “impulses” from other forms of

mental activity.

That neurotic patients see their impulses as alien objects breaking into their minds from outside
is evidence that the mechanism of repression is at work, nothing more. Since the objective of

psychoanalytic treatment is to overcome patients’ needs for such dramatic misrepresentation of their
own mental contents, it is surprising to find many psychoanalysts feeling that a scientific theory of the
neuroses should adopt this subjective misperception.

www.freepsychotherapybooks.org 154
A misunderstanding that comes up again and again in my conversations with other analysts

about this issue is the belief that terms like “planning,” “decision making,” “goal seeking” and “problem

solving” are anthropomorphisms inappropriate for describing the simplest expressions of biological
and emotional need. When I point out that these operations can be carried out at any level of

complexity and that very simple computer programs act in ways that can only be described in these
terms, they tell me that the human mind is not logical like the computer. I asked one rather thoughtful

senior colleague if she believed that the human mind is more like a pot of boiling water. After a

moment’s reflection, she nodded her head and, in all seriousness, said yes.

We are talking here about programming structures built up from conditional statements that

take the following form: “If X is true, carry out the next instruction, Y; otherwise jump to instruction Z.”

The logical result of this procedure would be exactly the same as in the case of an “impulse” seeking

discharge through one (preferred) channel but moving on to another if it finds that the first one is not
accessible. The structure of the discharge channels and their gates is no less a logical structure than

that of the computer program. The difference is not in the logic, but in the relationship between the

logical elements and the activity of the system as a whole. In the computer program, as in the simplest
organism, the logical structure is incorporated into the process that initiates, regulates, and terminates

the activity of the system. For example, the X in the statement, “if X is true, do Y,” is not usually a value

fixed before the execution of the program, but rather is the result of a computation determined while
the program is actually running. Interlocking feedback loops give the system the potential for

combining simple logical structures to form more complex ones. The logical structure is flexible, active,

and self-modifying.

In contrast, the logical structure that determines the discharge pathway of the “impulse” in

traditional psychoanalytic theory is rigid, passive, and inert. Because the impetus for the act of

discharge comes entirely from the impulse, the logical structure has no motivation to respond to the
passage of the impulse or to modify itself as a result. It is simply not an interested party to the

transaction. For this reason, it is often represented metaphorically as a hydraulic system of rigid

www.freepsychotherapybooks.org 155
channels and solid barriers.

It is difficult to imagine either how or why such a system would evolve into an executive ego

capable of adapting itself to a complex external environment. “Reality” is supposedly the agent of
change here, but, to my knowledge, neither a mechanism nor a source of motivation has ever been
proposed through which such a system might be capable of organizing itself to interact with the

outside world. This conceptual poverty is the price psychoanalysis has been willing to pay for a

selfinitiated “impulse” that operates outside the adaptive information-processing structure of the
organism in which it resides.

HIERARCHICAL STRUCTURE

Part 3 of Peterfreund’s (1971) Information, Systems, and Psychoanalysis, attempts to show how
complex hierarchical structures that evolve naturally from simpler information-processing structures

can give a comprehensive account of the subject matter of psychoanalysis with a significant gain in

coherence. A critical issue is whether the systems of structures traditionally classified as “id” and “ego”
can be distinguished through their relationship or lack of relationship to the outside world.

Peterfreund argues persuasively that a motivational structure, as the id is considered to be,

must be able to direct its activity toward actual opportunities for gratification and not merely to rigid

“discharge channels” (see Rosenblatt & Thickstun, 1977, 1978, for a fuller development of this theme).

An immature and vulnerable organism cannot afford the luxury of self-initiating impulses lacking an

adaptive function. The human infant is a little different in this regard from its phylogenetic ancestors.
With the protection and support of its parents, it can afford, temporarily, the fantasy of an autonomous

impulse life. But even if the infant could actually dispense with an adaptively functioning psychic

apparatus in the earliest weeks or months of life, there is no possible scenario through which even a

temporarily nonadaptive psychic apparatus could have survived the evolutionary struggle.

www.freepsychotherapybooks.org 156
For similar reasons, there is no possibility that an adaptive ego could evolve ontogenetically

from a primitive psychic apparatus that itself lacks the capacity to adapt. The Freudian id is a concept

that ruptures the evolutionary sequence just at the point of its crucial transition from slow-motion
information processing in the genetic mechanisms to highspeed information processing in the brain.

Interposing a state of chaos between these intimately interactive stages of evolution is mythological
thinking, supported, like all mythological thinking, by out-of-date science. It makes little difference that

the out-of-date science in this instance is only a century old.

Peterfreund’s proposals for a unified conceptual framework represent an important advance

over the “continuum” of structures extending from id to ego suggested in 1963 by Gill. It replaces the

one dimensional continuum with a multiply branching hierarchy, in which id functions are

distinguished from ego functions by their relative simplicity and more direct relationship to biological

events, but not by a lack of adaptive significance.

FEELING AND FUNCTION

The psychic apparatus in Peterfreund’s theory is driven not by subjective feeling states, but by

adaptive decision making. As we know, important decisions made in pathological states may have

serious maladaptive effects. Peterfreund points out that this is often due to a deficiency in the quality
or appropriateness of the information on which the decisions are based, because of repression and

other information-degrading defensive operations. Alternatively, it may be due to developmental

defects caused by failures of feedback at crucial stages of structure building. The subjective experience

of an intruding impulse is a mental representation of the faulty outcome of a decision-making

procedure. It is not an actual perception of the psychic apparatus at work.

Since this is a point that is difficult for many people to grasp, I think it is worth elaborating. A
frequent complaint about information theory is that it does not “account for” the subjective experience
of feeling or the motivating effects of feeling states. As we noted in the example of the falling tree, two

www.freepsychotherapybooks.org 157
very different kinds of theory are required to understand the nature of the information generated by
an event and the interpretation of that information by a living observer. This is no less so when the

event and the observation take place within a single person.

Although a systematic method of interpretation may produce substantial benefits (as Freud’s
system of dream interpretation does), it does not constitute a scientific theory if it explains only the

subjective interpretation of events by the human mind and not the events themselves. Freud was

aware of the importance of this issue when he tried to supplement his method of dream interpretation

with a theory of dream construction. Now that many of the original assumptions of that theory have
been refuted in the sleep laboratory, it is possible to see more clearly that the interpretive method is

largely independent of it.

Maintaining this distinction can be helpful. It means that (1) the interpretive method will not be

any less valuable if the old theory of dream construction loses its persuasiveness, but (2) the
interpretive method is very likely to be improved if it is supported by a more accurate understanding

of the psychological events underlying the subjective states of dreaming. “Psychological” in this context
refers to the vast amount of information processing that normally takes place outside conscious

awareness, even during sleep.

The sensory impressions experienced by the dreamer are something quite distinct from this
underlying process. So, too, are the dreamer’s affective states. Freud’s psychic energy theory was a

response to his realization that neurotic patients acted as if they were “feeling” something they were
not subjectively aware of. The supposed transformations of psychic energy represented the unknown
events underlying the otherwise inappropriate actions. The idea that these actions are the expression

of “unconscious affects” is contrary to the spirit of Freud’s attempt to support his observations with a

noncircular scientific explanation. Information theory is a more rigorous approach to the underlying
events that Freud was trying to reach.

Peterfreund shows that the traditional treatment of feelings in psychoanalytic discourse (not

www.freepsychotherapybooks.org 158
necessarily Freud’s) is dualistic and inconsistent. The complaint that information theory is too

complicated to explain the directness and simplicity of instinctual impulses is contradicted by the

objection that it is not complicated enough to explain the subtlety and discriminative capacity of

higher-level feeling states usually associated with the ego. These include, among many others, esthetic
judgment, creative inspiration, and empathic identification.

These higher-level feeling states are usually described in two different and mutually exclusive
ways, often at the same time. In the more traditional description, higher-level feeling states are

considered to be cognitive processes that are simply accompanied by painful or pleasurable affective

charges of varying intensity. Here the complaint that information theory is inadequate to explain these

states is irrelevant from the beginning, because for this model there is no structural relationship of any
kind between thought and feeling. Feeling is either painful or pleasurable. Any element of

differentiation belongs to the cognitive rather than the affective order.

In the more current psychoanalytic approach, higher-level feeling states are derived from

lower-level states through a process of internal differentiation and maturation, under the guidance of
the developing ego. This idea is more consistent with the data of child observation accumulated over
many decades. But the crucial point is that this process of differentiation and maturation cannot be

described without the concepts of information theory. A simple thing cannot evolve into a complex one
except through a reorganization of its original substance. Information theory is the science of
organization. A simple feeling, like the pleasure of sex, and a complex feeling, like the mature love of a

sexual partner, are somehow made of the same stuff, differently arranged and organized. That

observation was and is still the underpinning of Freud’s therapeutic method.

Both these descriptions of higher-level feeling states lead, if thought through independently, to
information theory. The failure of one well-known attempt to circumvent this conclusion can be helpful

in understanding why. Hartmann (1952) tried to derive the development of psychic structure from the

process Freud called “neutralization.” His idea seems to have been that when sexual and aggressive

energies are mixed in the right proportions, their “active principles” react with each other to form a

www.freepsychotherapybooks.org 159
stable product. (We will overlook for the moment the absence of a mechanism for determining “the
right proportions” or for regulating the process of mixing, whatever that is taken to be.)

The analogy is clearly with the chemistry of acids and bases. When solutions of an acid and a

base are mixed, their ionic components, initially distributed at random in the solution, combine and
precipitate out to form a crystalline structure lacking the corrosive properties of the original reagents.

The salt formed in this way becomes the metaphor for the ego. But the structure of the salt is simply an

endless repetition of a simple geometric form. It is an arrangement with no capacity to change in

response to events or to incorporate any new information from its environment into its own structure.
The “growth” of the crystal has nothing whatever to do with the maturation of the ego. As a metaphor

for human development it is completely lifeless.

But a tolerance for dead metaphor as a substitute for missing theory is not uncommon in the

psychoanalytic world. For some, a theory need not be any more than a recognizable word picture. Its
purpose is to resemble the mind, rather than to explain it. Like my friend who thought that a pot of

boiling water is a meaningful representation of the mind, they believe that a muddled theory is needed
to do justice to the muddle of motivations contained in the unconscious.

THE PERSISTENCE OF THE PRIMITIVE

This brings me to an important area in which Peterfreund’s thinking needs to be supplemented

by a further application of information theory. This is where he tries, unsuccessfully in my opinion, to

deal with an important set of observations that motivates much of the dualism in psychoanalytic

thought. One might call this issue “the persistence of the primitive.”

As Peterfreund sketches the hierarchy of psychic functions, he stresses the dimension of

complexity almost to the exclusion of other differences that may exist between lower-and higher-level

functions. The picture he presents is one in which simpler functions appear to lose their individual
identities as they are incorporated into or evolve into the more complex. In information-processing

www.freepsychotherapybooks.org 160
language, the levels of the hierarchy are “tightly coupled” (Pattee, 1973). An example in nature is the

multicellular organism. This is a hierarchy in which the smallest units, the cells, combine to form the

tissues; the tissues join to make organs; and, finally, the organs interact to constitute the complete

organism. Only at the level of the organism as a whole is there anything that can be called an
independent unit.

This might appear at first glance to be a natural model for the psychic apparatus, functioning as
it does as the control system for an organism. Schafer (1976), for example, makes a point of insisting

that only the hierarchical level of the whole person be acknowledged by the psychoanalyst. But in

taking this position, Schafer disregards another of Freud’s (1911) major discoveries, that at least two

levels of the psychic apparatus—the primary and secondary processes—are, in functional terms, only
“loosely coupled.” Hierarchical levels that are loosely coupled function independently of one another.

In the large-scale organization of matter, for example, stars and galaxies are very loosely coupled.

Emergent properties often appear when a higher level is only loosely coupled with those below it, as
when molecular properties emerge from atomic interaction or linguistic behavior from hominid

intelligence. Living systems are loosely coupled with their physical environments, although tightly

coupled within themselves.

Interesting questions arise when we try to determine the conditions under which tight
couplings seem to change to loose couplings and vice versa. The origin of life is one of these, as is the
separation of individual galaxies from the primordial mass of matter and energy. A possible definition

of psychoanalysis might be “the study of the psychic conditions in which the coupling of primary and

secondary processes changes from tight to loose and back again.”

Peterfreund prepares us for the view of the primary process as a loosely coupled level of
psychic organization when he speaks of it as an information-processing activity that takes place at a

lower level of complexity than waking thought. For him, the critical question is the membership of the

primary process in the hierarchy of adaptive functions. He tries to derive the functional properties of

primary process activity from the features it has in common with more complex cognitive activities

www.freepsychotherapybooks.org 161
that have clear-cut information-processing functions.

This demonstration is persuasive, but it fails to answer a question that has drawn some public

criticism to Peterfreund’s work. This, once again, is the question of the persistence of the primitive.

Why, if the primary process is simply a lower level of psychic functioning, does it take on a life of its
own, both in dreaming and in other mental states, where it appears at times to intervene in the normal

processes of waking thought? Why, under these conditions, is it only loosely coupled with the higher-

level activities into which one might expect it to be absorbed? (We are putting aside for the moment

the observation that the primary process is always at work behind the scene of waking consciousness,
supplying memories and correspondences not accessible through the normal channels of logical or
narrative thought. Under ordinary circumstances of waking life the primary process does function as if

it were tightly coupled to the higher levels of mental activity.)

How are we to explain those occasions, most notably dreaming, in which the primary process

appears to be very loosely coupled, if at all, with more advanced forms of cognitive activity? There is a

simple and straightforward information-processing explanation. The primary process has its own
cognitive function that is separate from, although necessary to, the functioning of higher-level

processes. This explanation implies that the adaptive goals of primary process activity can and must be
achieved independently of whatever further use the secondary process may make of them.

When we observe the primary process working to accomplish its own adaptive goals, as in
dreaming, it is only loosely coupled to higher processes. When we observe the products of the primary

process being utilized directly in the pursuit of goals of a higher order, as they are, for example, in the

creative process, the two levels of mental activity appear to be tightly coupled. The “products” of the
primary process I refer to are the uniquely individual associative links that combine to form the

treelike structure of human long-term memory. These links connect the isolated elements of our

experience across a range of contexts much wider than their original historical relationships. They
provide the raw materials for all forms of reasoning by analogy, from simple problem solving to

inspired acts of the creative imagination.

www.freepsychotherapybooks.org 162
In dreaming we find the primary process doing its normal adaptive task of matching new

experience with related experience of the past. The dream image is a composite of past and present

events, a test, as in Galton’s photographic method, of their “family resemblance.” (Freud (1900)
described how Galton had superimposed photographs of family members to find their common

features. He suggested that the mechanism of condensation in dreaming is doing the same with events
and experiences.) The process of dreaming is physiologically isolated from waking thought so that the

full resources of the sensory projection mechanisms can be used for this task. (Palombo, 1976, 1978).

Loose coupling of the primary process is also characteristic of the neuroses. But the explanation
for it in this case is not the same as it is in dreaming. In neurotic symptom formation, the uncoupling of

primary and secondary processes is an artifact, the result of pathological defensive operations

motivated by anxiety. It was one of Freud’s (1894) earliest discoveries that this uncoupling of

consciousness does not result, as intended by the defenses, in the exclusion of primary process input
from the behavioral control mechanisms. Instead, the primary process input influences behavior

directly, without passing through the normal sorting and filtering by higher-level cognitive processes.

This capacity for independent action is strong evidence that behavioral control did not pass

automatically from the lower to the higher structures as the cognitive hierarchy evolved. The higher-

level structures must be something much more like coordinating mechanisms than structures of direct
control. Control actually remains distributed at all levels, perhaps most tenaciously at the lowest. It is

the conscious illusion of control that makes neurotic patients vulnerable to sabotage by the products of

their repressed and unintegrated primary process activity.

THEORY IN PRACTICE

How is the theoretical difference between the self-initiating impulse and loosely coupled lower-
level information-processing structures applicable to the psychoanalytic treatment process? The

traditional theory addresses itself to two kinds of therapeutic events, the release of dammed up

www.freepsychotherapybooks.org 163
psychic energy at the primary process level and the acquisition of insight at the secondary process
level. Through the insight that comes from having the “unconscious made conscious,” the released

energy is said to be redirected into more adaptive discharge channels.

The problem with this model is that it fails to account for the building of new psychic structure
during an analysis. It rests on the assumption that the therapeutic effect of psychoanalysis results

exclusively from the removal of defensive barriers to the utilization of already existing structure.

Developmental theory and object relations theory have moved far beyond this view of psychoanalytic

therapy, but they have not supplied a rigorous theoretical alternative to it. Despite its enormous
promise for psychoanalysis, for example, Piaget’s information-processing approach to development

has not yet been successfully assimilated by object relations theory. But even without new theory, the

empirical evidence gathered in the analyst’s office shows very little correlation between patients’
conscious insight and the therapeutic benefit of analytic work.

It has become the custom in the psychoanalytic world to speak of analytic treatment as an

integrative process that may become conscious to the patient in varying degrees. And it is customary to
speak of the integration of the more primitive aspects of the patient’s mental life into the larger

structure of his or her ego. How this happens in the day-to-day work of the analysis remains a mystery
for the traditional theoretical model, which does not provide the mechanisms for this transformation.

Most particularly, it does not allow for the active participation of the primary process in the work of

integration.

From the information-processing point of view, the treatment process is a series of

coordinations or couplings that bring lower-level functions isolated by the defenses into a more

collaborative relationship with higher-level functions within the hierarchical structure of the ego. The

primary process is not merely a passive partner in this work, a source of energy to fuel the organizing
activity of the ego. It supplies vital information about current needs and about the accumulated record

of past events in which similar needs were acted on with varyingly successful outcomes.

www.freepsychotherapybooks.org 164
As the analysis proceeds, defensively isolated associative structures are restored to functioning

through the reopening of blocked connecting pathways. At the same time, new pathways are

constructed and integrated into a reorganized set of more efficient higher-level structures, as required
by the particular circumstances at each point in the patient’s development. A vital part of every

analysis is the discovery of these requirements. Nothing of the complexity of interaction among the
multiple components of the patient’s psychic apparatus is captured by (or comprehensible to) a theory

that views all change as the simple rechanneling of impulses.

Peterfreund’s long chapter on the treatment process in Information, Systems, and Psychoanalysis
(1971) appears near the end of the book, but was actually written first. It describes the treatment

process from a more intuitive position derived from an examination of the analyst’s empathic

identification with the patient as a feedback process that governs the progress of treatment. In a series

of later papers (1973,1975a, 1975b, 1978, 1980; Peterfreund & Franceschini, 1973), particularly in
“How Does the Analyst Listen? On Models and Strategies in the Psychoanalytic Process,” Peterfreund

(1975a) refined and expanded this application of information theory to the treatment situation.

A HEURISTIC APPROACH TO PSYCHOANALYTIC TREATMENT

These later ideas have been brought together in The Process of Psychoanalytic Therapy
(Peterfreund, 1983). As in the earlier works, Peterfreund’s emphasis is on the role of feedback

processes in the moment-to-moment interaction between patient and analyst, rather than on the long-

term buildup of psychic structure within the patient. This is a return to the problem of technique that

originally motivated his interest in information theory. He mentions his concern that his efforts to
promote the assimilation of a comprehensive new theoretical system into psychoanalytic thought may

have diverted attention from his more concrete technical proposals.

The Process of Psychoanalytic Therapy attempts to circumvent this problem by separating the
technical issues, as far as that is possible, from the theoretical. Traditional ideas are criticized in this

www.freepsychotherapybooks.org 165
work not because they are inconsistent or illogical, but because they impede the flow of information
between the patient and the analyst. Although information-processing concepts underlie the technical

approach, information-processing terminology is replaced for the most part by more familiar language.

Questions of the scientific authenticity and historical development of psychoanalysis are relegated to
the remote periphery of the discussion. Everything is subordinated to the single issue of clinical

efficacy.

The result is a profoundly illuminating demonstration of the applicability of information theory

to a central problem of clinical psychoanalysis. The book begins with a discussion of the analyst’s use
of theoretical knowledge in working with a patient. Peterfreund distinguishes between stereotyped

and flexible approaches, which he compares with the “algorithmic” and “heuristic” methods of problem

solving used by intelligent computer programs.

In the algorithmic method, a fixed sequence of procedures is designed that will guarantee the
desired result if followed precisely. This way of doing things works only for very simple problems,

where a limited number of possible outcomes can be evaluated within a reasonable time. In more
complex situations, the problem solver must be able to search the enormous array of possible solutions

by comparing the alternatives at each decision point according to the probable outcomes calculated
from its previous experience with similar situations. It must also be able to back up from a

disadvantageous position when past experience has failed to provide the required solution for the

problem immediately at hand and return to the previous decision point to begin the search once again.

This method is heuristic because it allows the problem solver to find his or her way without
knowing the exact dimensions of the problem in advance. The problem solver is discovering what the

problem is in the process of solving it. (“Heuristic” comes from the Greek verb heurein, to find or

discover, as in “Eureka!”) This is, Peterfreund says, what analysts are required to do. When they
approach a patient’s problem heuristically, they use their own theoretical knowledge to evaluate the

probabilities at the many decision points that must be traversed in the process of discovering the real

nature of the problem. They must be able to judge whether their theoretical expectations have been

www.freepsychotherapybooks.org 166
fulfilled as the process continues and to back up and modify their expectations when they have not.

Analysts who are working stereotypically do not follow these steps. They allow themselves to

think they understand the problem before having had the opportunity to investigate it. They then try to

fit what the patient says in the office into their initial formulations and tend to ignore or misinterpret
whatever fails to fit. Although this might seem like an easy pitfall for any well-meaning analyst to

avoid, Peterfreund shows with examples taken from the psychoanalytic literature and from his own

experience that there are many hidden traps for the unwary. Most important, he shows that the

reductionistic bias of traditional theory encourages the tendencies to clinical stereotyping created by
the paucity and distortion of information with which the analyst must always contend.

Over and over again, it becomes clear how the analytic interchange can be transformed from a

feedback loop into a vicious circle if the analyst allows theoretical expectations to interfere with efforts
to discover what really happened in the mind of the patient during development and current life

situation. Problems can be resolved if the analyst cuts through the circularity of the patient’s defensive

operations and directs the patient’s attention to the fact or feeling missing from the repetitive story he
or she has been telling. Peterfreund reports his successful interventions and his missed opportunities

with equal objectivity, using follow-up inquiries as well as retrospective reconstruction to pinpoint the
critical turn in each case.

The idea of the self-initiated impulse reappears in this context as an obstacle to the therapeutic
process. After Freud’s (1905) disillusionment with his mistaken idea that hysterical patients had been

seduced by their fathers, he began to see the actual events of his patients’ lives (intrapsychic as well as

interpersonal) as of only minor significance. The real sources of the patients’ difficulties were their
dominating instinctual impulses. These impulses could seize on and control any fragment of the

patients’ experience that suited them as a means to their expression or “discharge.”

Patients’ presentations were valuable pictures they provided as they had been those
experiences. The specific details of that experience were somehow relevant to their illness, but could

www.freepsychotherapybooks.org 167
not be identified with their causes. The lack of coherence in the patients’ life stories was evidence of

conflict in dealing with their impulses, but the missing details of the stories were not expected to

explain the nature of the conflicts.

Few analysts (certainly not Freud himself) have tried to model their conduct of analytic

treatment exclusively on this rigid schema. But Peterfreund shows how the idea of the self-initiating

impulse can operate in the background as a justification for denying the analyst’s need to know the
specific details of the patient’s life story. The two kinds of impulse, aggressive and libidinal, are a small

but well-known quantity. It is not unreasonable to try to resolve uncertainties about the meaning of

the patient’s communications by appeal to the most primitive features of instinctual life. The extensive

clinical examples in The Process of Psychoanalytic Therapy show that these assumptions can be fatal to
the progress of an analysis.

WORKING MODELS

To describe his technique for bringing coherence to the patient’s life story, Peterfreund adopts

Bowlby’s conception of “working models.” These are, in Bowlby’s (1969) words, “the internal worlds of
traditional psychoanalytic theory seen in a new perspective” (p. 82). In Peterfreund’s thinking, these
models are like the stored programs used by a computer. They provide not only a representation of

some limited area of experiences, but also a plan of action for operating within that area. Unlike the

vast majority of computer programs currently in operation, however, working models are self-
modifying in the light of further experience.

They are, in other words, component systems in the overall adaptive structure that generates
and regulates the experience and behavior of the person. Many of these components are actually
miniature versions of the entire system, functional representations of the system as a whole. They can

be temporarily modified for the purpose of exploration and experimentation, so that they can perform
what Freud called “trial actions” with minimal risk.

www.freepsychotherapybooks.org 168
None of these miniature representations is complete, of course. They are simulations,

constructed from a variety of simplifying assumptions. For this reason, a great many of them are

required to represent the overall system to itself, including its various modalities of interaction with
the outside world. Conflict between the models is not only possible but inevitable. Leibnitz, Calvin,

Machiavelli, Cicero, and Freud in Clippinger’s (1977) simulation are each working models within the
larger working model of the main program itself.

Peterfreund lists eight major working models employed by the analyst. The first is the analyst’s

knowledge of the world in general, as it operates in normal circumstances. Second is the analyst’s
model of his or her own personal history and the stable elements of his or her own selfrepresentation.

Third is the normal developmental sequence of cognitive and emotional experience. Fourth is the

phenomenology of the analytic process. Fifth is the analyst’s general clinical experience. Sixth is the

analyst’s model of the particular patient as a “total experiencing human being.” The seventh model
includes two theoretical metamodels, one a theory that explains psychopathological mechanisms, the

other a theory that accounts for the therapeutic effect of the analytic process. Finally, there is an eighth

model, a higher-level metamodel that integrates the explanatory concepts generated by all the others.
There are obviously a great many component models at all levels with varying degrees of

independence and interdependence. When the analyst processes the information provided by the

patient, he refers it to each of his own relevant working models. The analyst then begins the sometimes
arduous labor of reconciling inconsistencies that develop among the various models as they are

updated by the new information.

Most of The Process of Psychoanalytic Therapy is devoted to a discussion of clinical cases in

which the reconciliation of these inconsistencies required both flexibility and insight from the analyst.

Peterfreund’s illustrations are themselves models of the therapeutic process at work. It would be a

grave injustice to try to condense them into a few words in a brief essay such as this one. Suffice it so
say that The Process of Psychoanalytic Therapy is probably the best book of its kind available to

teachers and students of psychoanalytic therapy today.

www.freepsychotherapybooks.org 169
Beyond this major work of Peterfreund’s, one can see still another contribution of information

theory to clinical psychoanalysis. This will be a rigorous account of primary process activity as it

relates to the therapeutic process. The patient’s primary process thought is, in fact, the primary source
of information about the patient’s earliest experience in dealing with his or her needs and wishes. The

primary process of the analyst functions in the therapeutic situation by matching his or her internal
models of the patient’s mental life with the derivative representations of that early experience

communicated to the analyst by the patient.

Peterfreund (1983) brings us to the edge of this conception when he says, for example:

All working models are changed by the very information received. They
must be constantly updated, adapted, readapted, checked and rechecked
for consistency—both for internal consistency as well as for consistency
with other models. Such processes are basic aspects of learning and are
apparently in large part associated with the phenomena we call
“consciousness” or “awareness” [p. 83].

The “larger part” not associated with consciousness or awareness has attracted the puzzled
attention of psychoanalysts for a very long time. Peterfreund’s translation of “empathy” into a system
of working models is an important step toward the solution of the puzzle.

CONCLUSION

The movement of history has carried psychoanalysis beyond the limits of Freud’s extraordinary
achievement. Peterfreund’s contributions mark the entrance of psychoanalysis into a new era of
scientific thought. As with all pioneers, he leaves many tasks of exploration and consolidation to be

done. But he has established the broad outlines of a comprehensive new framework within which
traditional psychoanalysis can be safely embedded.

Scientific revolutions, no matter how long postponed, have an inevitability about them. They

www.freepsychotherapybooks.org 170
succeed by sheltering the living tradition within a reconceptualization of greater power and
comprehensiveness. Copernicus’ first concern when he turned the solar system inside out was to save
the phenomena of astronomical observation. Peterfreund’s work has already fulfilled the promise of

information theory to extend the conceptual universe of psychoanalysis while making its day-to-day
observations clearer and more precise.

REFERENCES

Aserinsky, E., & Kleitman, N. (1953). Regularly occurring periods of eye motility and concomitant
phenomena during sleep. Science, 118:273-274.

Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York: Basic Books.

Bowlby, J. (1981). [In J. Reppen (Ed.), Symposium on Emanuel Peterfreund]. Psychoanalytic Review, 68:
187-190.

Clippinger, J. H. (1977). Meaning and Discourse: A computer model of psychoanalytic speech and
cognition. Baltimore: Johns Hopkins University Press.

Freud, S. (1894). The neuro-psychoses of defense. Standard Edition, 3: 45-61.

Freud, S. (1900). The interpretation of dreams. Standard Edition, 4-5.

Freud, S. (1905). Three essays on the theory of sexuality. Standard Edition, 7, 130-243.

Freud, S. (1911). Formulations on the two principles of mental functioning. Standard Edition, 12, 218-
226.

Freud, S. (1914). On the history of the psychoanalytic movement. Standard Edition, 14, 7.

Freud, S. (1923). The ego and the id. Standard Edition, 19, 12-66.

Gill, M. (1963). Topography and systems in psychoanalytic theory. New York: International Universities
Press.

www.freepsychotherapybooks.org 171
Hartmann, H. (1950). Comments on the psychoanalytic theory of the ego. The Psychoanalytic Study of
the Child, 5, 74-96.

Hartmann, H. (1952). The mutual influences in the development of ego and id. The Psychoanalytic Study
of the Child, 7, 9-30.

Miller, G. A., Galanter, E., & Pribram, K. H. (1960). Plans and the structure of behavior. New York: Holt.

Newell, A., Shaw, J. C., & Simon, H. A. (1957). Empirical explorations of the logic theory machine: A case
study of heuristics. Proceedings of the Joint Computer Conference (pp. 218-230).
Washington, DC: Spartan Books.

Palombo, S. R. (1976). The Dream and the Memory Cycle. International Review of Psycho-Analysis, 3:65-
83.

Palombo, S. R. (1978). Dreaming and memory: A new information-processing model. New York: Basic
Books.

Palombo, S. R. (1980). The Cognitive Act in Dream Construction. Journal of the American Academy of
Psychoanalysis, 8, 186-201.

Pattee, H. H. (Ed.). (1973). Hierarchy theory: The challenge of complex systems, New York: George
Braziller.

Peterfreund, E. (1971). Information, Systems, and Psychoanalysis. Psychological Issues, 7 (1/2,

Monograph 25/26). New York: International Universities Press.

Peterfreund, E. (1973). On information-processing models for mental phenomena. International

Journal of Psychoanalysis, 54, 351-357.

Peterfreund, E. (1975a). How does the psychoanalyst listen? On models and strategies in the
psychoanalytic process. In D. P. Spence (Ed.), Psychoanalysis and contemporary science,
(Vol. 4, pp. 59-101). New York: International Universities Press.

Peterfreund, E. (1975b). The need for a new general frame of reference for psychoanalysis.
Psychoanalytic Quarterly, 44, 534-549.

www.freepsychotherapybooks.org 172
Peterfreund, E. (1978). Some critical comments on psychoanalytic conceptalizations of infancy.
International Journal of Psycho-Analysis, 59, 427-441.

Peterfreund, E. (1980). On information and systems models for psychoanalysis. International Review of
Psycho-Analysis, 7, 327-344.

Peterfreund, E. (1983). The process of psychoanalytic therapy. Hillsdale, NJ: The Analytic Press.

Peterfreund, E., & Franceschini, E. (1973). On information, motivation and meaning. In B. B. Rubinstein
(Ed.), Psychoanalysis and contemporary science (Vol. 2). New York: Macmillan.

Pribram, K. H., & Gill, M. (1976). Freud’s “Project” reassessed. New York: Basic Books.

Reppen, J. (Ed.). (1981). Symposium on Emanuel Peterfreund, Psychoanalytic Review, 68, 159-161.

Rosenblatt, A. D., & Thickstun, J. T. (1977). Energy, information and motivation: A revision of
psychoanalytic theory. Journal of the American Psychoanalytic Association, 25, 537-558.

Rosenblatt, A. D., & Thickstun, J. T. (1978). Modem psychoanalytic concepts in a general psychology
(Parts 1 & 2). Psychological Issues, 11 (2/3 Monograph 42/43), New York: International
Universities Press.

Rubinstein, B. B. (1971). Preface. In E. Peterfreund, Information, Systems, and Psychoanalysis.

Psychological Issues, 7(1/2, Monograph 25/26). New York: International Universities
Press.

Rubinstein, B. B. (1975). On the clinical psychoanalytic theory and its role in the inference and
confirmation of particular clinical hypotheses In D. P. Spence (Ed.), Psychoanalysis and
contemporary science (Vol. 4. pp. 3.57). New York: International Universities Press.

Rubinstein, B. B. (1980). The problem of confirmation in clinical psychoanalysis. Journal of the

American Psychoanalytic Association, 28, 397-417.

Schafer, R. (1976). A new language for psychoanalysis. New Haven: Yale University Press.

Shannon, C. E., & Weaver, W. (1949). The mathematical theory of communication. Urbana: University of
Illinois Press.

www.freepsychotherapybooks.org 173
6
MERTON M. GILL: A STUDY IN THEORY
DEVELOPMENT IN PSYCHOANALYSIS

IRWIN Z. HOFFMAN, PH.D.

Merton Max Gill was bom in Chicago in 1914. He was the second of three boys. For business
reasons, the family soon moved to Milwaukee. In high school, Gill was among the top performers in an

oratory club. He was an excellent student and graduated first in his class.

Gill received his Ph.B. (Bachelor of Philosophy) from the University of Chicago in 1934, having
majored in psychology. His interest in psychoanalysis developed quite early, inspired, in part, by a

reading of Freud’s Introductory Lectures. By the time he entered medical school at the University of
Chicago, he was certain that he wanted to become not only a psychiatrist but also a psychoanalyst. Gill

received his M.D. from the University of Chicago in 1938 and went on to do his internship at Michael

Reese Hospital from 1939 to 1941.

In 1941, Gill began his residency at the Menninger Clinic in Topeka, Kansas. Here his
psychoanalytic career was launched under the influence of such notable psychoanalysts as Karl and

William Menninger and Robert Knight. The most important intellectual influence was that of David
Rapaport, the head of the Department of Psychology and subsequently of the Department of Research,

www.freepsychotherapybooks.org 174
with whom there quickly developed a very strong reciprocal bond. Gill became involved initially, along

with Roy Schafer, in Rapaport’s work on diagnostic psychological testing (Rapaport, Gill, and Schafer,

1945-46, 1968). Later, he collaborated closely with Rapaport on the development of psychoanalytic

metapsychology. At Topeka, Gill was also introduced by Margaret Brenman to hypnosis, a technique
that was useful in treating the many war-related cases of traumatic neurosis at that time. This was the

beginning of a 16-year collaborative investigation of hypnosis and related phenomena. Gill also met
George Klein and Robert Holt, among others, at the Menninger Clinic.

After graduating from the Topeka Psychoanalytic Institute in 1947, Gill, along with Rapaport

and Brenman, moved to Stockbridge, Massachusetts, to join Knight who had become the director of the

Austen Riggs Center. At Riggs, between 1948 and 1950, Gill continued his work on hypnosis and on
metapsychology. From 1950 to 1953, Gill was at Yale where he collaborated with Newman and Redlich

in writing The Initial Interview in Psychiatric Practice (1954). While at Yale, he was appointed training

analyst at the Western New England Psychoanalytic Institute.

Gill moved to Berkeley, California, in 1953, where he had a private practice and an appointment
as training analyst at the San Francisco Psychoanalytic Institute. Supported by a grant from the
Foundations Fund for Research in Psychiatry, Gill continued his collaboration with Brenman (Gill &

Brenman, 1959) and with Rapaport (Rapaport & Gill, 1959). He met with Rapaport three or four times
each year to exchange ideas and to go over papers and drafts of chapters for the book they were
writing. Gill also teamed up with Timothy Leary to do research on psychotherapy, an effort that led to a

coding scheme designed to give a comprehensive account of the psychotherapeutic process (Leary &

Gill, 1959). Toward the end of this period in California, Gill collaborated with the neuropsychologist
Karl Pribram in a study of Freud’s Project for a Scientific Psychology. This work was shelved, however,

and was not prepared for publication until many years later (Pribram & Gill, 1976).

Rapaport’s untimely death in 1960 was a great personal loss for Gill. Soon after, Gill completed

the monograph they had begun together (Gill, 1963) and began collecting Rapaport’s papers

(Rapaport, 1967). With George Klein, he also wrote a summary of Rapaport’s contributions (Gill &

www.freepsychotherapybooks.org 175
Klein, 1964). Later he contributed a paper on the primary process to Robert Holt’s collection of essays
in Rapaport’s honor (Gill, 1967).

In 1963, as the recipient of a lifetime Research Career Award from the National Institute of

Mental Health, Gill moved to Brooklyn, where he became Research Professor in Psychiatry at the
Downstate Medical Center of the State University of New York. He also began in earnest to record

psychoanalysis for research purposes, although he had done a good deal of recording previously in his

studies of hypnosis and psychotherapy.

From 1968 to 1971 Gill was a Fellow at the Research Center for Mental Health at New York

University. Here, Gill rejoined Klein and Holt, both of whom further influenced Gill’s thinking on
psychoanalytic metapsychology.

In 1971, soon after Klein’s death, Gill returned to Chicago where he became professor of

psychiatry at the University of Illinois at the Medical Center and a supervising analyst at the Chicago

Institute for Psychoanalysis. With Leo Goldberger, he edited George Klein’s book for publication (Klein,
1976). In 1976, he and Philip Holzman edited a collection of papers in Klein’s memory dealing with the

controversy that surrounded psychoanalytic metapsychology. In this volume, Gill (1976) published his

own full-scale critique of metapsychology, calling into question much of what he himself had written
over the years. In Chicago, changes in Gill’s thinking about the psychoanalytic process were further
stimulated by Samuel Lipton’s ideas, particularly Lipton’s distinction between the personal

relationship in psychoanalysis and technique (Lipton, 1977a) and his close attention to various forms
of resistance to the transference (Lipton, 1977b). Here, Gill (1979,1982) crystallized his own revised
view of psychoanalytic technique and the beginnings of a method for systematically studying its

application (Gill and Hoffman, 1982b).

Among the most important influences on Gill’s thinking were his own experiences as a patient

with several analysts. These experiences left him with a deep sense of the difference that the

personality of the analyst can make in the analytic process, as well as of the differences attributable to

www.freepsychotherapybooks.org 176
varying points of view on technique. Gill’s convictions have been informed and inspired by a wide

range of experiences as an analyst, as an analysand, and as an intimate co-worker with many of the

most seminal psychoanalytic thinkers of our time.

THE SCOPE OF GILL’S CONTRIBUTIONS

Merton Gill’s contributions to psychoanalytic thought encompass a wide range of interrelated

issues that are fundamental to the development of psychoanalysis as an intellectual and professional

discipline. One of the extraordinary things about the corpus of Gill’s work is that it embodies some of
the major tensions in the field, with Gill himself standing among the leading spokespersons on both

sides of a fundamental controversy. Thus, it is not unusual for Gill in 1984 to find himself differing

sharply with someone who cites “Gill, 1954” to buttress his or her own position. Similarly, no critique

of the earlier Gill is more thoroughgoing and unsparing than that which is stated or implied in Gill’s
later contributions.

At the heart of this movement in the history of Gill’s ideas lies the renunciation of

psychoanalytic metapsychology. In this respect, Gill’s intellectual history is closely related to those of
other students and colleagues of David Rapaport who moved away from the perspective of this

extraordinary, charismatic teacher even while continuing to reflect his inspiration in the energy, rigor,

and imagination of their own work (Holt, 1965, 1976; Klein, 1976; Schafer, 1976). Despite the striking

commonalities among the members of this group, Gill’s intellectual metamorphosis is especially
noteworthy because he was probably the closest to Rapaport of his students. It was Gill who took it

upon himself to collect Rapaport’s papers (Rapaport, 1967). It was also Gill who completed the

ambitious theoretical project that began with the landmark paper on the metapsychological points of
view that he and Rapaport wrote together (Rapaport and Gill, 1959) and that culminated with the

publication of Topography and Systems in Psychoanalytic Theory (Gill, 1963), most of which was written
by Gill himself after Rapaport died. With Klein, as noted earlier, Gill also wrote an extraordinary

summary of Rapaport’s contributions (Gill & Klein, 1964). It is a tribute to Gill’s intellectual

www.freepsychotherapybooks.org 177
independence and courage that he, too, finally broke with Rapapport and became one of the most
thoughtful and careful critics of psychoanalytic metapsychology.

In this essay, Gill’s contributions are reviewed in relation to three fundamental tasks or

challenges that are critical for psychoanalysis and that are brought into sharper focus in Gill’s work.
The first is the challenge of determining and describing the nature of psychoanalysis as a discipline. I

include under this heading Gill’s contributions to metapsychology as well as his later critique of

metapsychology and his argument against the natural science framework that psychoanalytic
metapsychology utilizes. To include Gill’s early contributions under this heading is to take a

questionable liberty, since Gill was not raising questions about psychoanalysis as a discipline while he

was immersed in elaborating its theoretical structure within a natural science frame of reference. Only

with hindsight can one argue that seeds of the later critique were sown, paradoxically, by the very
thoroughness of the earlier work. In this respect I am taking my cue from Gill (Reppen, 1982) himself,

who has said of Rapaport: “It was the clarity, brilliance, and persistence with which he pursued the

implications of metapsychological theory that exposed its structure and problems” (p. 169).

The second challenge is that of defining the nature of the psychoanalytic situation itself and the
optimal psychoanalytic technique. What are the distinguishing features of psychoanalysis as compared

with other therapies? How does psychoanalytic theory of technique take account of the fact that the
analyst is a person who inevitably bears a personal relationship to the patient? This concern has been

central for Gill throughout his career, as has a variant of this question, one which lies on the interface of

the metatheoretical and the clinical-theoretical areas, namely, what are the implications of the fact that
psychoanalysis is a discipline in which the human mind is simultaneously the subject and object of

investigation? In his clinical contributions, too, we find a major shift in Gill’s position. Unlike the shift in
his metatheoretical perspective, however, there are relatively clear and direct precursors of Gill’s later
ideas on psychoanalytic treatment in his earlier work.

Finally, we have the challenge of subjecting psychoanalysis as a mode of therapy to some kind of

systematic observation and empirical test, despite the requirements of confidentiality and the

www.freepsychotherapybooks.org 178
enormous complexity of the whole phenomenon. On this matter, Gill has been unwavering throughout
his career, insisting that the propositions of psychoanalysis must be verifiable according to the usual

criteria of science. To reject the notion that psychoanalysis is a natural science, Gill has insisted, is by

no means to reject the notion that it is indeed a science. Or, approaching the matter from the other side,
to espouse the notion that psychoanalysis is a hermeneutic discipline is not to relinquish its scientific

accountability.

It is somewhat artificial to separate Gill’s contributions into these three areas because of the

extent of their interrelationship. This is particularly true with respect to Gill’s later work. Over the
years, Gill’s ideas have developed into an increasingly coherent and internally consistent position. His

metatheory, his clinical theory, and his attitude toward research have developed into a unified

perspective on psychoanalysis. What Gill now has to say about research in the psychoanalytic situation
follows logically from what he has to say about the nature of the psychoanalytic situation itself and

about the essence of psychoanalytic technique. The latter, in turn, bears a close relationship to his

critique of metapsychology and his espousal of a rigorous hermeneutic position for psychoanalysis.
This conceptual integration was absent in Gill’s earlier work; the contributions to metapsychology

were either unrelated to the concurrent clinical contributions or bore a strained relationship to them.
Similarly, the research focus was only partially related to the metapsychological investigations or to

clinical psychoanalysis.

As noted earlier, Gill’s current perspective amounts to a telling critique of his earlier ideas. In
this overview of Gill’s contributions, a bias will be evident in that the earlier work will be considered in
the light of its relationship to later developments in Gill’s thinking. This approach automatically risks

denigrating the intrinsic value of certain earlier positions and contributions because they will appear
either as germinal vis-à-vis what comes later or as unworthy of further development in their own right.
No history is unbiased, however, and this author would be hard pressed to look at Gill’s earlier work in

any other way because I share his later perspective and have had the good fortune to collaborate with
him on some aspects of its development. What follows is a selective review of Gill’s extensive writings,

www.freepsychotherapybooks.org 179
drawing primarily on books and papers that seem to represent culminations or crystallizations of
phases of his work and thought.

CONTRIBUTIONS TO PSYCHOANALYTIC METATHEORY

GILL’S IMMERSION IN METAPSYCHOLOGY

Rapaport and Gill (1959) set out to complete a program that they believed Freud left unfinished,
namely, the spelling out of “that minimal set of assumptions upon which psychoanalytic theory rests”

(p. 1). They group these assumptions under five headings that, they propose, constitute the basic points
of view of psychoanalytic metapsychology. These points of view are at the highest level of abstraction

in the theory (Gill, 1963, p. 153). Presumably, to be complete, a psychoanalytic explanation of any

psychological phenomenon must include reference to all five points of view. According to Rapaport and
Gill, these points of view are the dynamic (having to do with force), the economic (having to do with

energy), the structural (having to do with “abiding psychological configurations”), the genetic (having

to do with origins and development), and the adaptive (having to do with relationship to the
environment). The genetic and adaptive points of view are additions to those explicitly formulated by

Freud. The structural point of view refers specifically to the division of the mental apparatus into the

systems of id, ego, and superego, and replaces the topographic point of view insofar as the latter refers
specifically to the division of the mental apparatus into the systems of unconscious (Ucs.),

preconscious (Pcs.), and conscious (Cs.). Rapaport and Gill (1959) argue that, although he moved in

this direction, “Freud never explicitly replaced the topographic point of view of metapsychology by a
structural one” (p. 2).

The Elucidation of Theoretical Inconsistencies. Gill’s contributions to psychoanalytic

metapsychology bear the stamp of Rapaport’s influence both in style and substance. In Topography and

Systems in Psychoanalytic Theory (1963), Gill’s most extensive metapsychological work (see Ross,
1965, and Spence, 1964, for synopses and reviews), we find a scrupulous attention to Freud’s writings.

www.freepsychotherapybooks.org 180
Often, quotations and page citations on some aspect of the subject are followed first by a highlighting of
internal inconsistencies and then by a creative attempt at integration, including whatever
reformulation seems necessary or useful. This kind of careful exegesis of Freud’s writings, one of

Rapaport’s legacies (Gill & Klein, 1964), invariably underscores the complexity of Freud’s thinking and
the elusiveness of what Freud “really meant” by various terms, such as “ego,” “id,” “primary process,”

“secondary process,” and even “metapsychology” itself. This very elusiveness is a tribute to Freud’s

scientific temperament and his refusal to become comfortable with formulations that are simplistic,
incomplete, or inconsistent with other theoretical propositions and with clinical data.

Apfelbaum (1966) is critical of Gill for implying that the contradictions in Freud are avoidable

and that it is, in principle, possible to develop a more coherent and internally consistent account of the

systems of the mind. He writes:

One aim of Gill’s monograph is to give the coup de grace to the topographic
model, so as finally to settle the issue from what point of view the mental
systems are to be established. Gill finds in Freud’s unwillingness to drop
this model a difficulty of Freud’s rather than a difficulty inherent in the
structural approach itself [p. 467].

In point of fact, however, Apfelbaum does Gill something of a disservice here in that Gill’s

monograph, quite in the spirit of Freud’s writings on the subject, raises as many if not more questions
than it answers about psychic structures.

Indeed, Gill’s discussion shatters any illusion one might wish to maintain that the replacement

of the topographic model by the structural model does away with internal inconsistencies within the

various subsystems of the mind. Gill diligently follows Freud in his attempt to localize various

properties of mental content in one or another subsystem. He examines each of the dimensions with
which Freud was struggling: the relationship of contents to consciousness, the condition of their

energy, their mode of functioning, whether or not they employ neutral energy, and, finally, whether
they are associated with the repressed or the forces of repression. For the sake of scientific elegance, it

www.freepsychotherapybooks.org 181
would have been convenient if unconsciousness, free energy pressing for immediate discharge,

primary process (that is, drive-organized ideas), absence of neutral energy, and contents that are

considered to be repressed could all have been located in one system. Conversely, consciousness,

inhibited or bound energy, the secondary process, neutral energy, and the forces of repression, ideally,
would all be correlated and form a second major system. The fact that the repressing forces—that is,

the defenses—could themselves be unconscious was decisive in leading Freud to partially discard the
topographic model, that is, the criterion of consciousness for defining systems. Instead, Freud chooses

to group together the repressed in the system-id and the repressing forces in the system-ego. However,

as Gill (1963) notes:

Freud’s solution of the difficulties of the topographic system leaves

unresolved a number of issues relating to these difficulties. Even if the
relationship to consciousness is dropped as a criterion of mental systems,
it is still necessary to account for the exceptions to parallelism between the
relationship of contents to consciousness and their mode of organization
and kind of cathexis; and a division of the repressed and repressing into id
and ego fails to account for the similarity between them indicated by the
fact that they are both dynamically unconscious [p. 51].

It is noteworthy that Gill’s extraordinary effort to reconcile these contradictions ends up with

his raising a significant question about the validity of the structural model itself insofar as it connotes a
set of internally consistent, relatively well-demarcated systems of the mind. There seem to be no end
to the “exceptions to parallelism” that are exposed by clinical experience. Perhaps one of the most

important and bold contributions of Gill’s (1963) monograph is the blurring of the distinction between

id and ego:

I favor, then, a definition in which id and ego are conceived of as a

hierarchical continuum of forces and structures existing at all levels of the
hierarchy.

Such a solution argues that Freud’s resolution of the fourth difficulty of the
topographic systems was not a good one, because, by putting force into one

www.freepsychotherapybooks.org 182
system and counterforce into another, it obscured the existence of a
hierarchy of force-counterforce integrations, and while conceptualizing
counterforce in structural terms, did not do the same for force. The
recognition of this hierarchy, furthermore, makes it clear that, on any
particular level of the hierarchy, force and counterforce, despite their
antithesis, show similarities in mode of functioning, energy employed, and
energy regulated [pp. 146-47; italics added].

Gill’s emphasis on continua of types of mental activity throughout might be regarded as a

forerunner of his later holistic approach, which places the whole person at the center of the theory.

This will be discussed further later on. The main point I wish to make here is that one comes away
from Topography and Systems with anything but the sense that the whole notion of systems has been

salvaged and clarified. Indeed, whether it is useful to think at all in terms of discrete psychological

systems, at least in accord with the various criteria that Freud was juggling, seems questionable and is
explicitly challenged by Gill.

The Depreciation of Consciousness. In Topography and Systems, Gill (1963) discusses the

considerations that argue for discarding the topographic perspective as a metapsychological point of
view. Central to his thesis is the idea that “the relationship [of contents] to consciousness can be

subsumed under the five metapsychological points of view” (p. 159). Gill takes pains to emphasize that
to demonstrate that “a topographic point of view in metapsychology is unnecessary” and is not

intended to “belittle the importance of the relation of contents to consciousness and of consciousness

as such.” On the contrary, topographic conceptions retain “an important place in psychoanalysis, both

clinically and theoretically” (p. 148).

However this disclaimer is unconvincing. To say that the topographic status of a mental event,
which encompasses its phenomenological status, can be “subsumed under” (p. 159) the other points of
view, or is “explicable in terms of the more basic hypotheses” (p. 159) associated with them, or can be

“accounted for” (p. 61) in their terms is to denigrate consciousness as a source of explanation in the
theory. Elsewhere, Gill and Klein (1964) indirectly acknowledge as much when, speaking of Rapaport,

www.freepsychotherapybooks.org 183
they state that “he observed that with the replacement of Freud’s topographic systems by the tripartite
model of ego, id, and superego, consciousness was reduced in importance” (p. 493). Applying Gill’s

(1976) own critique of metapsychology, I believe that the idea that topographic considerations are

reducible to the other points of view follows from the mistake of assuming that quasi-
neurophysiological concepts describable in terms of the dimensions of natural science are of a higher

order or are more abstract than psychological concepts. That this is the mistaken assumption

underlying the exclusion of a topographic point of view may be obscured by the fact that the view itself
can be framed largely in natural science terms. However, such terms are applicable primarily to the

preconditions for the emergence of conscious experience, not to the impact of consciousness itself on
the organization of experience and behavior. We can see this clearly if we examine the terms of Gill’s
(1963) discussion of the clinical importance of consciousness (chapter 9) and compare it with the

terms of his argument against the inclusion of the topographic perspective among the basic
metapsychological points of view (chapter 10). In the first discussion, for example, Gill speaks of the
hypothesis that “insight plays a vital role in changing behavior” (p. 151). In the second discussion, Gill

argues that “access to consciousness is determined by competition among external forces, among
internal forces, and between external and internal forces” (p. 155). Applying Gill’s later critique (1976,

1977a), the first of these statements is framed in psychological terms, whereas the second is framed in

quasi-neurophysiological terms. According to Gill in 1976, only the first is relevant to psychoanalytic

theory, but in 1963 the first statement was regarded as subordinate to the second in line with the
assumption that psychological phenomena must be explained by antecedent neurophysiological

conditions.

One is left then with this non sequitur: the preconditions of consciousness can be described in
terms of the economic, structural, dynamic, genetic, and adaptive points of view. Therefore, the

difference that consciousness makes in experience and behavior is subsumable under these points of
view. The rejoinder may be that even the changes that follow from consciousness or, more particularly,
from insight, may be describable in terms of the other points of view. We find such a formulation in the
following statement by Gill (1963): “The sense organ Cs. plays the highest role in the hierarchy of

www.freepsychotherapybooks.org 184
regulations of psychic functioning, increasing the cathexis of contents to which the attention cathexis is

directed, bringing about an advance in synthesis of the contents which excite it, and making possible
the cathecting even of contents which give rise to unpleasure” (p. 158).

The weakness of this argument is transparent, since consciousness is reduced to some sort of

sensory apparatus, and it is not at all clear how a sensory apparatus can “direct” anything. Moreover, to

the extent that it does direct ensuing processes, it is not all evident how this element of control could
be described without reference to consciousness itself, that is by referring only to the interactions

among various other forces. The fact is that the directive properties of the system Cs. carry us

inescapably into the realm of human intention and into the universe of discourse in which intention,

meaning, and self-conscious reflection have their proper place. Among the critics of psychoanalytic
metapsychology, Klein (1976) probably has been the clearest and most emphatic on this issue.

The restoration of consciousness in psychoanalytic theory does not in any way imply a

denigration of the crucial role of unconsciously motivated actions. However, terms like “intention” and

“meaning,” which Gill now believes are the proper terms for psychoanalytic discourse, are, to begin
with, categories of conscious experience. These categories are then attributed to phenomena that lie

outside the realm of conscious experience but that nevertheless act to a significant degree “as if” they
were conscious. As Gill (1977a) has written: “Let it be recalled that Freud insisted that only on the
assumption that unconscious psychological processes must be understood in the same terms as

conscious ones, except for the fact of consciousness itself, could one fill in the gaps in conscious life and

construct a coherent, meaningful psychological continuity” (pp. 585-586).

In the end, Gill himself equivocates about the demotion of the topographic perspective from the

level of formal point of view to the level of clinical theory. He concludes Topography and Systems
(1963) with a telling disclaimer: “It is of course also possible that with some future redefinition or

reclassification of the metapsychological points of view a topographic point of view will be included.

The issue is, after all, one of definition” (p. 159). Nevertheless, it is a measure of the degree to which
Gill uncritically adopted the natural science frame of reference of metapsychology that in his major

www.freepsychotherapybooks.org 185
theoretical contribution to metapsychology he slights the point of view that is most useful clinically
and that is closest to the data of the psychoanalytic situation and of interpersonal experience generally.

Implicit in the holistic “person point of view” that Gill (1983b) was later to adopt as the supraordinate

point of view of psychoanalysis and implicit also in the theory of technique that Gill came to advocate is
a recognition of the fundamental importance of the topographic point of view and of consciousness in

psychoanalytic explanation.

The Depreciation of Object Relations. Another indicator of the depth of Gill’s immersion in

metapsychology was his relative neglect of internal and external objects in his discussion of the
systems of the mind. Freud’s superego provides the basis for a bridge from the mechanistic model in

which the forces of the id are harnessed by the apparatuses of the ego to one in which the person’s

experience is seen as shaped by his or her interactions with others. Yet not only is the person as such
virtually absent from Gill’s account of mental processes in 1963, but so are other persons, which is

merely the other side of the same coin.

The systems of the mind, in the framework of classical metapsychology, house and process
various stimuli from within and from without. Presumably, the stimuli that are associated with

encounters with other human beings, who are perceived eventually as whole persons, are the most
important in determining the quality of experience, behavior, and development. Freud’s concept of the

superego (even though it may derive its power from the forces of the id) theoretically requires

attention to object relations—that is, to the meaning of interpersonal encounters as opposed to

impersonal stimuli as they impinge on the individual. As Apfelbaum (1966) points out, ego psychology
tended to systematically underemphasize the superego precisely because it is not readily accounted

for in a mechanistic model:

The omission of the superego on a level of formal theorizing by Hartmann,

Rapaport and Gill further illustrates the point that the structural approach,
as they have developed it, no longer refers to the study of the
interrelations of id, ego and superego, but to formulations having to do
with “the control of structure over drive.” To put this another way, the

www.freepsychotherapybooks.org 186
structural approach now refers to the construction of a psycho-analytic
model which relies wholly on explanation in terms of energy and
structure. A dynamic conception such as the superego is not congenial to
this model since it cannot be rendered in these terms [pp. 460-461].

Apfelbaum goes on to praise Melanie Klein, Erikson, Zetzel, and Winnicott for their focus on the

superego and the corollary understanding that “the nature of the ego is determined at all times by its

relations with internal and external objects” (p. 461). This view is consistent with Gill’s later critique of
metapsychology, which grows out of a hermeneutic position. This position, for Gill, is inseparable from

an object relations perspective.

The Seeds of the Later Critique. Gill’s metapsychological contributions pull simultaneously
toward the deepest possible immersion in a natural science framework and toward the extrication of
psychoanalytic theory from it as an inappropriate universe of discourse. As counterpoint to Gill’s

depreciation of consciousness and his underemphasis of object relations, we find a surfacing of

fundamental questions that jeopardize the entire way of thinking entailed by psychoanalytic
metapsychology. In the first place, the internal contradictions are so cumbersome and the moves

necessary to resolve them so convoluted and so distant from the data that they allegedly comprehend
that the viability of the whole project seems precarious. In the end, as we have seen, Gill’s proposals

are actually quite radical in that they challenge the validity of existing attempts to define clearly
demarcated subsystems in the mind and argue instead for an emphasis on continua (see also, Gill,
1967). Freud (1923) himself said that “the ego is not sharply separated from the id; its lower portion

merges into it” (p. 24). Gill (1963, p. 141) goes beyond Freud, however, encouraging an almost
complete erosion of the boundaries between the two systems. His position actually foreshadows a
retreat from the notion of a primary source of energy and force having a prepsychological, quasi-

organic basis. The infusion of the id with the properties ordinarily reserved for the ego represents a pull
away from drive theory as conceptualized in traditional metapsychological terms. At the same time, the
infusion of the ego with the motivational properties ordinarily reserved for the id pulls away from the
notion of a rational agency in the mind that has access to the outside world uncontaminated by

www.freepsychotherapybooks.org 187
subjectivity. Thus, although it was clearly not part of his intent, Gill, in 1963, had already laid the
groundwork for the dissolution of the sharp dichotomy of subjectivity and objectivity that
characterized Freud’s epistemology and that so colored his clinical theory. Moreover, Gill’s redefinition

of the id represents a precursor of his later attack on the “energy-discharge point of view” as distinct
from the “person point of view” in psychoanalysis (Gill, 1983b).

THE REPUDIATION OF METAPSYCHOLOGY

Gill’s movement away from metapsychology had to be a painful process, given his closeness to
Rapaport. He did, however, have the support of Schafer and Klein among others. Klein, in particular,

had a strong influence on Gill’s thinking. Gill’s (1976) critique of metapsychology further develops

Klein’s (1973) original notion that psychoanalytic theory is characterized by a mingling of terms from

two universes of discourse, the psychological and the biological, and that the two must be disentangled
before psychoanalytic theory can develop in any useful way.

The reversal of Gill’s position on the value of classical metapsychology for psychoanalysis is

reflected in a dramatic way in the book he wrote with Pribram on Freud’s Project for a Scientific

Psychology (Pribram and Gill, 1976). Here, Pribram and Gill elaborate on metapsychology as a
theoretical model for neuropsychological investigations. In a certain sense, this effort is in keeping

with Gill’s claim that metapsychology is, in fact, in a different universe of discourse than psychoanalytic
psychology. However, it also carries the implication that the development of psychoanalysis itself will
be promoted by investigations that focus upon “brain-behavior-experience interfaces” (p. 168). In a

conclusion that was added around the time of publication, more than ten years after much of the
collaborative work was completed (M. M. Gill, personal communication), there is an unusually candid
statement of sharp disagreement on this issue between the two authors. The book concludes with this

provocative comment: “Where we differ is that Gill feels that psychoanalysis must go its own way and
that means purging it of its natural science metapsychology, while Pribram welcomes psychoanalysis
back into the natural sciences. Pribram doubts that the differing views of the two authors are really, in

www.freepsychotherapybooks.org 188
the long run, incompatible, while Gill finds them irreconcilable” (p. 169).

The format of Gill’s tour de force on metapsychology (Gill, 1976) is once again, in the Rapaport

tradition in that it begins with a detailed examination of Freud’s writings in order to clarify the
implications of Freud’s theoretical propositions. In particular, Gill does psychoanalysis an inestimable
service by documenting Freud’s continuing tendency to gravitate toward neurophysiology despite his

many disclaimers and his acceptance, at times, of psychoanalytic psychology as a science in its own

right. Repeatedly, as Gill shows, Freud betrays an underlying feeling that the phenomena of psychology
must be explained by neurophysiology.

It is important to emphasize that Gill is not rejecting theory as such, including the whole
hierarchy of concepts that characterizes a fully developed theory, ranging from concepts that are close
to the data to those that are more distant and more abstract. This is a common misunderstanding that

goes hand in hand with the idea that the survival of Freud’s metapsychology is equivalent to the
survival of analytic theory itself. Rather, Gill is arguing that concepts having to do with space, force,

energy, and the like are not on a higher level of abstraction than those that are clinically derived and

that are framed in strictly psychological terms. Instead, the former are hypotheses about the
neurophysiological correlates of psychological phenomena. What is wrong here is not only that they

happen to be bad neurophysiology (Holt, 1965), but that they are intended as higher-order concepts
than those of the clinical theory.

In fact, Gill is not even entirely rejecting the metapsychological points of view. He argues, for

example, that although the terms of the economic point of view so consistently reify the notion of
quantities of energy and force that they should be discarded, the other points of view, especially the
structural and dynamic, might be salvageable if reformulated in psychological terms. In fact, Gill (1976)

concludes his critique of metapsychology with a statement that is much milder than the title,
“Metapsychology is Not Psychology,” suggests:

Metapsychological propositions and clinical propositions that are purely

www.freepsychotherapybooks.org 189
psychological must be disentangled and examined on their appropriate
grounds. For this reason, despite the argument that there is no direct
connection between metapsychology and psychology, the present state of
affairs in psychoanalytic theory is such that it makes no sense to say
globally that one accepts or rejects metapsychology [pp. 103-104].

Following Klein and Schafer, Gill insists that not only is the quasi-neurophysiological theory of
metapsychology detrimental to the development of psychoanalysis, but so is any “metatheory” that

implies that psychological phenomena must be explained in terms of mechanisms known from another

universe of discourse. Thus, it is not surprising that Gill rejects the attempt to substitute the model of
the computer and information theory (Peterfreund, 1971) for traditional metapsychology. Information

theory is seductive because it seems, on the surface, to be addressing the problem of meaning itself,

thereby avoiding the pitfall of traditional metapsychology. However, Gill (1977a) claims that the terms
of information theory are either being used in an informal, nontechnical way, in which case they

amount to “no more than a restatement of psychoanalytic propositions in technical sounding terms like

‘feedback’ and “match and mismatch’ ” (p. 591), or else they are being used in a technical sense, which
means they are located in a natural science frame of reference. Once again, in other words, the

assumption is being made that psychoanalytic theory building must subsume the phenomena of self-
conscious human experience under the rubric of an allegedly more general set of phenomena in which

the person as agent is absent. Gill’s point is that the very exclusion of the person identifies information
theory as one that deals with a different universe of discourse than psychoanalysis rather than as one
that is at a higher level of abstraction.

Gill (Reppen, 1982) is arguing for a theory that assumes the existence of the person as “a

unitary human agent conceived of as initiating and in that sense responsible for pursuing humanly
meaningful aims” (p. 179) and that proceeds to identify patterns and regularities among such aims and
the adaptations that accompany them. To a certain extent, especially with respect to the content of

basic human motives, Gill has left open the question of what will evolve from a purely clinical, person-
oriented psychoanalytic theory. Gill (1977a) has been loathe to give up the central importance of
drives in development:

www.freepsychotherapybooks.org 190
The close association in our literature between the concept of
peremptoriness, instinctual drives, and psychic energy apparently leads
many to believe that the abandonment of the concept of psychic energy
amounts to giving up the idea of instinctual drives. That is simply not true.
What is true is that the biological phenomena related to instinctual drive
cannot be directly translated into the realm of psychoanalytic psychology,
but become relevant there only in terms of their meaningfulness [p. 593].

There are indications in Gill’s writings, however, of a questioning of the concept of the primacy

of instinctual drives, even if recast in psychological terms. Thus, for example, he has described as
“fateful” the theoretical step Freud took when he conceptualized conflict between the systems of the

mind in terms of forces seeking expression and those opposed to such expression. Gill (1978) explains

that this step “opened the way to designate a special class of motivations as the ones seeking

expression in contrast to that class of motivations which sought to keep them from expression. The
class seeking expression was referred to as the instinctual impulses and those were in turn related to

bodily needs, in particular sexual” (p. 484). The emergence of the structural theory did nothing to

change this basic distinction between the two types of motivation, Gill continues:

Though [Freud] had thus disposed of the error of assuming that defensive
processes had ready access to consciousness while the processes
defending against did not, he was still left with a class distinction between
processes seeking expression and processes seeking to prevent such
expression.

An alternative scheme would have been to conceptualize the contending

processes as equally striving for expression and to sever the idea of
processes seeking expression from any special relationship to the body,
but such a scheme would have violated his conception of a hierarchy of
psychic processes with the base of the hierarchy constituted by the
somatic “drives” [pp. 484-485].

Although Gill does not explicitly draw the implications, his position certainly is consistent with

the kind of theorizing that George Klein (1976) undertook. Freed of the encumbrance of

www.freepsychotherapybooks.org 191
metapsychology, Klein set forth a revised view of human sexuality and proposed other types of “vital

pleasures” that have a kind of irreducible status.

GILL’S NEW METATHEORY: AN EPISTEMOLOGICAL POSITION

It is true, nevertheless, that Gill himself has refrained from formulating specific notions of the

fundamental motives that organize behavior and experience. Gill has devoted most of his attention to

psychoanalytic theory of technique, complete with lower-level concepts, such as types of

communications by the patient and types of interventions by the analyst (Gill, 1982; Gill & Hoffman,

1982b), and higher-level concepts, such as resistance to awareness of transference, resistance to the

resolution of transference, and propositions about the interrelationships among all of these (Gill, 1979,

1982). At the highest level of abstraction we find a bridge to the new metatheoretical perspective that
Gill has adopted. Although in some of his writings Gill has equated metapsychology with Freud’s

energy discharge model, Gill (1983b) recently stated that he regards as a “cogent objection” the idea

that “any system of thought must have a ‘meta’ organizing principle, whether implicit or explicit” (p.
525). The organizing principle that Gill believes should replace Freud’s basic concept of energy

discharge is “the person point of view.” For Gill, the term person” connotes both the agency of the
subject of analytic investigation and treatment and the subject’s social nature.

What appears to be left out of Gill’s theory of technique are propositions about the content of

the issues that one would expect to be sources of conflict for the individual and that would become the
objects of resistance. Gill’s theory of technique, in this particular sense, is content free. But it is

important to recognize that this absence of attention to content is not merely the reflection of Gill’s

particular area of interest. There is, rather, something intrinsic to Gill’s position that is resistant to
generalizations about the content of human motivation. This feature lies at the heart of what has

evolved as Gill’s epistemology and, at the same time, is the organizing principle at the apex of the

hierarchy of concepts that constitute his theory of technique. This is the principle of perspectivism or
constructivism: The meaning of any emotionally significant experience is actively organized by the

www.freepsychotherapybooks.org 192
person according to a particular perspective that he or she brings to bear in interpreting it. This
position is necessarily skeptical of or actively critical of propositions about universal motives, since
such propositions may imply a transcendence of perspectivism—a revelation, one might say, of the

motivational factors that generate perspectives in the first place. A perspectivist position, by definition,
does not allow for the possibility of such transcendence. This is not to say that perspectivists may not,

for heuristic purposes, posit the existence of certain basic motives, but they would naturally be

skeptical about the applicability of such motives in any particular culture, subculture, individual, or
individual at a certain moment in time.

The definition of perspectivism just given does not refer explicitly to one important feature of

the principle—its social basis. Peoples’ perspectives develop and are sustained or eroded in the

context of their interactions with other persons. In analysis, according to Gill, one never reaches a point
where one discovers something that comes solely from the patient, independent of the influence of

other persons. Instead, one finds specific interactions, out of which certain perspectives emerged that

were to color subsequent interactions. In these interactions, the patient, with the participation of the
other persons involved, constructed an identity, a social world, and a way of living with other people,

which the patient perpetuates in subsequent encounters. This way of being with others is not the only

way available to the patient, although he or she may subjectively experience it as such.

There is one basic human tendency or motive that Gill’s perspectivism can accommodate, and

that is the universal human tendency to make sense of experience in an interpersonal context. The

need for meaning and the need for other people are inextricably intertwined. In his most recent
writings, Gill has gravitated toward theories, such as Bowlby’s, that emphasize human attachment as

the basic motive in the hierarchy of human motives (Gill, 1983b; cf. Eagle, 1981). Gill has shied away
from considering the seeking or construction of meaning as primary motives (cf. Basch, 1977) because
interpersonal human relatedness for Gill is paramount. However, Gill’s epistemology and his emphasis
on human interaction can readily be integrated.

Gill’s theory of technique can be viewed, in effect, as the clinical application of his epistemology.

www.freepsychotherapybooks.org 193
This is not, of course, a reflection of the way in which the theory of technique evolved. On the contrary,
Gill moved from particular clinical experiences, as an analysand and as an analyst, toward a deep

conviction about a way of working with people that he felt would be most conducive to change. If

anything, his epistemology grew out of his clinical theory. Once the epistemology is articulated,
however, it is not difficult to go back and see its reflection in the theory of technique.

Gill sees the psychoanalytic situation as one in which two people interact and continually try to

establish the meaning of that interaction as one of them experiences it. Gill’s focus on the here and now

could be viewed, in part, as an intensive molecular study of the process by which meaning gets
constructed by one human being—the patient—in interaction with another—the analyst. This process

is understood to be liberating precisely because it entails a movement by the patient from an absolute

view of his or her predicament, which is dominated by the neurotic or obstructing transference, to a
perspectivist view, which allows for the realization of latent potentialities. This change is born out of

an emotionally meaningful interpersonal experience in which patient and analyst work together to

extricate themselves from the repetitive patterns that the neurotic transference and
countertransference impose, as if these patterns defined the only ways in which the two participants

could relate. Before giving a fuller account of Gill’s current theory of technique, let us go back and
review the clinical contributions that antedate it and that, in varying degrees, contain the seeds of its

development.

THE CLINICAL CONTRIBUTIONS

As noted earlier, unlike the integral relationship between his current theory of technique and

his current metatheoretical position, there is only a partial connection between Gill’s earlier clinical
contributions and his metapsychological contributions. In discussing this second aspect of Gill’s work, I
will focus on three main areas of clinical contribution: hypnosis, the initial psychiatric interview, and

psychoanalytic technique.

www.freepsychotherapybooks.org 194
STUDIES OF HYPNOSIS AND RELATED STATES

Gill’s research and writing on hypnosis bring together many of his major areas of interest.
Unlike his work on the initial interview (Gill, Newman, & Redlich, 1954), which maintains a strictly

clinical focus throughout, Gill’s discussion of hypnosis includes the ambitious attempt to synthesize
empirical observations and clinical theory, on the one hand, with classical metapsychology, on the
other. Consistent with a value that runs through all of Gill’s professional life, however, the work on

hypnosis was inspired by an interest in developing an approach that could be applied usefully to deal

with a pressing clinical problem—in this instance, that of traumatic neurosis during World War II.

The publication of Hypnosis and Related States (Gill & Brenman, 1959) represented the
culmination of his collaboration with Brenman on a wide range of studies, which involved the
participation of many outstanding clinicians, including Knight, Karl Menninger, and Schafer. Over time,

the authors’ interest in the clinical application of hypnosis evolved into a much broader task, which
was to understand regressive states generally, including those encountered in the usual psychoanalytic

situation.

The entire complex project utilized a combination of methods, including observation of

hypnosis in psychotherapeutic situations, experimental procedures, and even the use of
anthropological data gathered by Bateson and Mead (1942) in their study of trance states in Bali. The

work by Gill and Brenman is a model of clinical research; not only are many methods used and
systematically compared, but the authors are extraordinarily diligent in openly discussing the process
of the research and the thinking that went into each piece of work that they undertook. Hypotheses

and findings are always accompanied by candid discussion of uncontrolled variables affecting the
authors’ sense of confidence in their own hunches and conclusions. Systematic quantitative studies are
supplemented by a wealth of rich clinical material throughout.

The studies of hypnosis include a fascinating oscillation between the poles of the strictly

psychological and the biopsychological. Significantly, and in accord with Gill’s later work, Gill and

www.freepsychotherapybooks.org 195
Brenman (1959) state: “For many years we found ourselves accumulating two apparently independent
bodies of data from our observations of the hypnotic state, but were unable to discern any theoretical

bridge between them. The observations of ‘altered ego function’ and of “transference phenomena’

seemed to us to be in quite separate realms of discourse” (p. xix).

However, in sharp contrast to Gill’s current psychoanalytic focus, which is deliberately confined

to the realm of “transference phenomena” in the broad sense (that is, the realm of the interpersonally

meaningful), Gill and Brenman considered the integration of the two realms of discourse to be of great

importance. Although, to be sure, the subject matter in this instance was hypnosis and not
psychoanalysis or psychoanalytic theory per se, the authors were operating with a psychoanalytic

perspective and the work itself was undertaken in the spirit of a psychoanalytic investigation, as the

subtitle, Psychoanalytic Studies in Regression, makes clear. Thus, it is legitimate to contrast Gill’s
implicit perspective on psychoanalysis as a discipline in this book with his current viewpoint. The

earlier work epitomizes a contribution born out of the view that psychoanalysis can and should be a

general psychology. What makes this possible, Gill and Brenman (1959) argue, is the development of
ego psychology as represented in the work of Hartmann, Kris, Loewenstein, and Rapaport (p. xxi). As a

result of the efforts of these theorists, it is possible, the authors claim, to investigate the effects on the
ego of various kinds of environmental factors, including the presence or absence of various quantities

of “stimulation.” The authors state the “basic theoretical premise of their book” as: “hypnosis is a
particular kind of regressive process which may be initiated either by sensory motor-ideational
deprivation or by the stimulation of an archaic relationship to the hypnotist” (p. xx). It is not that Gill

would now argue that only the meaning of the relationship to the patient is necessary to describe or
explain the phenomenon of hypnosis, and that the effects of stimulus deprivation as such are
unimportant. He would assert, however, that an investigation defined as psychoanalytic would be

confined to and would be designed to maximize what could be understood about that aspect of the
phenomenon having to do with its meaning to the participants.

Another issue that sharply divides Gill’s point of view in his work on hypnosis from his current

www.freepsychotherapybooks.org 196
perspective is the role of regression in the psychoanalytic process. For Gill in the 1950s, there was little
doubt that an induced regression lay at the heart of the psychoanalytic process. His intensive
investigation of hypnosis and related regressive states was undoubtedly fueled in part by the

assumption that anything learned about regression in hypnosis would not only have clinical utility in
itself, but would also further the understanding of psychoanalytic treatment. Gill and Brenman (1959,

pp. 117, 134-135, 329) cite and agree with Macalpine’s (1950) conceptualization of the psychoanalytic

process as a kind of slow-motion hypnotic procedure. Hypnosis also has in common with
psychoanalytic treatment the fact that, ideally, the regression is not a total one (“regression proper”)

but rather a partial one, which “a subsystem of the ego” undergoes in keeping with Kris’ concept of

regression in the service of the ego. This regression is brought about in hypnosis as well as in analysis
by a combination of impersonal factors (such as stimulus deprivation) and interpersonal factors (such

as promotion of a submissive, dependent attitude). Although the regression itself, in either case, is not
spontaneous but induced, the particular form that the regression takes bears the stamp of each
patient’s history and neurotic conflicts. Important points of agreement and disagreement between

Gill’s earlier view of transference and his current view are well illustrated in the following statement
from Hypnosis and Related States:

We know in general that when a “transference interpretation” is made, it

should in fact be an interpretation which shows the patient that his
response is not appropriately geared to the actual behavior of the
therapist, but is in fact an expression of something ancient in himself
which he has brought to the situation. We must now ask ourselves what is
our position in this connection if we introduce a technique (hypnosis)
which implicitly states, “By dint of what I am doing you will find yourself
able to do things you otherwise cannot and unable to do things you
otherwise can.” In short, if we take a position which implies superior
power, how can we ask the patient to analyze the irrational, transference
aspect of his being hypnotizable at all? Yet we have done this, usually in
the face of bitter resistance from our patients, some of whom, as we have
seen, said they would prefer to give up the use of hypnosis entirely rather
than analyze its meaning for them. As one might expect, despite what one
might call the “reality provocation” of inducing hypnosis, it has been

www.freepsychotherapybooks.org 197
possible to tease out the specifically personal projections of each
individual and, on the basis of what the hypnotic relationship seemed to
mean to him, to make use of these in the treatment.

Yet is this qualitatively different from the non-hypnotic standard

psychoanalysis where we ask the patient to lie down while we sit up,
where we arrogate to ourselves the privilege of responding or not as we
see fit, where we ask the patient to let us see him completely though he
cannot see us. and finally where from time to time we tell him what is
“really” going on? Does not all of this too imply that we regard ourselves as
“in charge” of the situation in a uniquely powerful way? Indeed, how
commonly this is the lament of the analysand. Yet, the fact remains that
each analysand reacts in his own way to this “provocation” too, and
reveals his archaic and established patterns of feeling and behavior as
transference phenomena [pp. 369-370].

Clearly, then as now, Gill was concerned about the influence of the analyst’s behavior on the

patient’s experience of the relationship. Indeed, the strained quality of his effort to reconcile the
classical view of transference with his appreciation of the influence that the analyst exerts is striking

and seems to almost beg for the new resolution he was later to achieve. Gill no longer defines
transference, even with its particular idiosyncratic nuances, as divorced from the way in which the

analyst participates in the process. Also, in accord with the perspectivist position he has developed, Gill

no longer divides the patient’s experience into an aspect appropriate to the present circumstances and
one grossly inappropriate to them which comes entirely from the past.

In a major departure from his earlier views, Gill no longer considers the deliberate attempt to

induce a regression to be essential or even desirable in the psychoanalytic situation. Gill now objects,
in principle, to any intentional manipulation of the patient regardless of its purpose. At the core of the
psychoanalytic process is the exploration of the ways in which the patient is assimilating the advertent

and inadvertent influences that the analyst exerts via the analyst’s inevitably significant emotional
participation in the interaction (Gill, 1982, 1983a). For Gill now, a molecular analysis and explication of
the patient’s conscious and preconscious way of constructing and construing the immediate

www.freepsychotherapybooks.org 198
interaction with the analyst has replaced the induction of a state of mind that is allegedly closer to that
of the primary process and the unconscious. Thus, there is no question that Gill has abandoned the

effort to investigate in a direct way, in the psychoanalytic situation or in any other context, those

mental states that appear to be discontinuous with familiar, secondary process modes of thought. The
condensed, often uncanny symbolic richness of the material represented in many of the vignettes in

Hypnosis and Related States seems to be absent from much of the clinical material Gill has published

recently to illustrate his current view of analysis of transference (e.g., Gill & Hoffman, 1982a).

It is clear that Gill has turned away from the mysteries of hypnotic states, dreams, fugue states,
and so on in favor of the more readily accessible nuances of interpersonal interactions. It would be

easy, but also a mistake, to assume that Gill’s course represents a flight from a dangerous and foreign

world to a more familiar and safe one. The fact is that although the content and modes of organization
of thought in the psychoanalytic discourse that Gill now encourages may seem familiar or mundane,

the type of interaction he seeks is rare indeed, and the route toward its achievement is not without its

own special psychological perils. Gill would be inclined now to suspect that being caught up with the
psychodynamic meaning of symbolic material may represent an escape from the greater anxiety

associated with directly confronting what the patient and the analyst are experiencing in their
immediate interaction but which is unformulated or unspoken.

It is important not to leave this area without underscoring the important lines of continuity

between the ideas presented in Gill’s work on hypnosis and his current viewpoint. To begin with, there
is the notion of two universes of discourse, as noted earlier, which remain separate throughout
Hypnosis and Related States despite the authors’ determination to integrate them. Second, there is an

emphasis on the importance of the relationship throughout and on the element of mutuality in the
process. Gill and Brenman (1959) take the position that “hypnosis is at least in part a dovetailing of the
unconscious fantasies of the two people involved, and that strictly speaking one should not speak of

‘the hypnotic state’ but rather of ‘the hypnotic relationship’ ” (pp. 60-61). They spell out the specific
form that this reciprocity of roles takes:

www.freepsychotherapybooks.org 199
From analysis of the two sets of data, on subject and on hypnotist, it
appears to us quite clear that hypnosis is a complex dovetailing
relationship between the two participants wherein the overt role taken by
the one is the covert fantasy of the other. Thus, while the hypnotist is
overtly being the powerful figure, whether as a domineering tyrant or a
boundless source of “supplies” he is covertly on the receiving end of this
power and/or bounty in his fantasy. ...

As we have seen, on the other side of this coin, the hypnotic subject takes
overtly the role of the obedient, super-compliant puppet; covertly he is not
only sharing in the hypnotist’s presumed omnipotence, but is pushing this
in fantasy to the point of the hypnotist’s having to abdicate completely [p.
98].

Here we have just the kind of emotional reciprocity that Gill would now be on the alert for in his
work as an analyst, in which the interpretation of the transference always includes reference to the
patient’s plausible ideas about the analyst’s countertransference response.

Toward the end of the book, Gill and Brenman discuss the reasons why many therapists often

give up the use of hypnosis even as an adjunctive technique. Among the reasons they give is the

growing awareness of the unconscious wish to assume the role of the omnipotent parent or, covertly
and vicariously, of the helpless, regressed child. Whatever personal factors were involved, Gill left

hypnosis behind both as a treatment technique and as an instrument for investigation of psychological

phenomena. Instead of the unabashed exercise of psychological power that hypnosis epitomizes but
which is more subtly represented in standard psychoanalytic technique, Gill has opted for a rigorous,

critical understanding of interpersonal influence in the psychoanalytic situation as a means of

liberating the patient from closed and repetitive patterns of interaction with others.

THE INITIAL PSYCHIATRIC INTERVIEW

During his relatively short stay at Yale between 1950 and 1953, Gill collaborated with Newman
and Redlich on The Initial Interview in Psychiatric Practice (1954), a book that is remarkable for the

www.freepsychotherapybooks.org 200
extent to which it anticipates Gill’s later perspective. In the first place, the book is based on three

transcribed sessions, reflecting Gill’s commitment to recording, which had already taken hold in the

mid-1940s. In the second place, the book is a critique of the medical model as it is generally applied in

psychiatric diagnostic interviewing—a critique that foreshadows Gill’s (1976, 1977b) later repudiation
of both biologistic psychoanalytic theorizing and medically tinged conceptualizations of psychoanalytic

technique that emphasize the analyst’s detachment rather than participation in the process.

The rejection of the medical model in this early work bears some special attention. Gill and his

collaborators object to the prevalent practice of gathering information from the patient under various

headings instead of following the patient in a relatively open-ended fashion. Of special interest, in view

of what comes later, is the authors’ rejection of history taking as the necessary way to arrive at a valid
diagnosis. This position has continuity, of course, with Gill’s later concern that analysts are often

interested in genetic reconstruction at the expense of understanding the patient’s experience of the

relationship in the here and now.

There are, of course, important differences between Gill’s current theory of technique and his
approach to interviewing in 1954. Although some of these may be explained by the fact that the two
types of clinical situations are not fully comparable, I think Gill’s current position regarding technique

does entail principles that he believes can and should be extended to diagnostic interviewing. For Gill,
what is of central importance diagnostically is the way in which a person relates in the here and now,
including the patient’s capacity to reflect upon the meaning of his or her immediate experience with

the interviewer. Other considerations may also be important, but they are secondary. What we do not

yet see in The Initial Interview in Psychiatric Practice is the full emergence of the focus on the
relationship, including the technique of systematically searching for and interpreting disguised

allusions to the transference, even at the beginning of the treatment (Gill & Muslin, 1977).

There are many examples in The Initial Interview in Psychiatric Practice of Gill’s early conviction

that the immediate process should take priority over collection of content. Concerning departures in

interviewing from a “psychiatric copy of medical schedules” Gill and his collaborators (1954) wrote:

www.freepsychotherapybooks.org 201
“Probably the most important [departure] was the psychiatrists’ realization of the significance of the
patient-therapist relationship as the very framework within which the nature and meaning of the

patient’s productions must be understood” (p. 19).

Gill wrote in 1954 as though the mental status exam and other aspects of traditional, medically
oriented interviewing were already passé, but what he had to say then is still quite germane

considering contemporary zeal about ferretting out the biological factors in mental disorders.

Foreshadowing his later sharp distinction between a psychological realm of discourse and a biological

realm, Gill makes clear that the assessment he is talking about is one of psychological (that is,
interpersonal) functioning, not of biological factors. This is not to say that the latter are not relevant to

a complete understanding of the patient’s functioning, but only that assessing psychological factors is

something separate and apart, something requiring the adoption of an attitude that is not compatible
with the type of diagnostic attitude associated with assessment of organic factors. As Gill, Newman, and

Redlich (1954) put it: “In the psychiatric interview the interpersonal relationship is focal. The

psychiatrist must deal adequately with this relationship to insure that the desired communication
between patient and therapist shall take place. Should there be any indication of a somatic disorder,

the patient must be further studied by techniques which are not our concern here” (p. 65; italics added).

One of the reasons these authors gave for the persistence of the “old-fashioned mental status

examination” was the psychological function it served for the interviewer:

The second reason for the “deaf and dumb” quality of the older
examination lies in the psychiatrist’s need to retain his equilibrium by
demonstrating his superiority. The inquisitory technique is used, then, as
an unconscious defense against the threatening content of the experience
of his disturbed patient, and against an emotional relationship with the
patient—particularly against the patient’s emotional demands [p. 23].

Here, again, we find a foreshadowing of Gill’s later emphasis on the inescapable fact that
whatever an analyst’s or therapist’s behavior, it carries meaning that derives from his personal

www.freepsychotherapybooks.org 202
participation in the process. Gill might well say the same today about the function of silence and other

allegedly neutral postures that an analyst may adopt. Similarly, Gill, Newman, and Redlich (1954)

wrote of the inevitability of the reciprocal influence of patient and interviewer: “Reactions of both

doctor and patient will of course change as each meets the reactions of the other in that progressive
redefinition which is the essence of any developing relationship” (p. 66-67).

As a final example of this early conviction, consider the following comment, which so clearly
anticipates Gill’s current emphasis on the here and now in psychoanalysis and his relative deemphasis

of historical reconstruction if undertaken without reference to a reliving in the transference: “It has

naturally occurred to us that we may be trying to push too far the idea of abandoning the collection of

historical data in order to emphasize current interaction. But we are persuaded that doubts about our
technique are caused by our inability more completely to divest ourselves of long-established and

anxiety-reducing habits of professional practice” (p. 412).

It is also of interest to note that of the various influences on the development of their own

orientation, Gill and his co-authors consider Sullivan’s to be the strongest, although they note that
Sullivan does not go as far as they do in giving up adherence to a “relatively formal ‘reconnaissance’
and ‘detailed inquiry’ ” (p. 62). Significantly, Gill has recently immersed himself in Sullivanian literature

and has written about the continuity of his own ideas and those of Sullivan as well as about important
differences between them (Gill, 1983a).

There are many examples in the commentary on the transcribed sessions, the phonograph
recordings of which were also published, in which the authors commend or criticize the interviewers

depending on whether they follow the patients’ leads or retreat defensively to some agenda of their

own. Where the advocated technique and mode of listening depart from Gill’s current view is in the
failure to systematically interpret or even identify disguised allusions to the transference. I believe that

if Gill were to criticize the interviews and the authors’ commentaries on them now, he would point out

that although much emphasis is apparently placed on the interaction, in practice, the approach fails to

follow the patient’s experience of the relationship in a systematic way, one that would require constant

www.freepsychotherapybooks.org 203
attention to disguised allusions to the transference in the patient’s associations (Gill, 1982, 1983a,
1984a; Gill & Hoffman, 1982a, 1982b).

PSYCHOANALYTIC TECHNIQUE

Probably the most carefully elaborated statement on psychoanalytic technique that Gill made

while he was still immersed in psychoanalytic metapsychology and writing on hypnosis is found in the
paper, “Psychoanalysis and Exploratory Psychotherapy” (Gill, 1954), published in an issue of the

Journal of the American Psychoanalytic Association devoted entirely to papers on technique. It is useful

and illuminating to compare Gill’s views as represented in that paper with his current ideas (1979,
1982, 1983a), keeping in mind always that Gill’s point of view in 1954 is probably representative of

much that is still in the mainstream of classical psychoanalytic thought. Gill himself has written a paper
(1984a) that undertakes such a comparative analysis.

Gill’s definition of psychoanalytic technique in the 1954 paper is well known and often cited:
“Psychoanalysis is that technique which, employed by a neutral analyst, results in the development of a

regressive transference neurosis and the ultimate resolution of this neurosis by techniques of
interpretation alone” (p. 775). The definition has three main elements, which Gill goes on to elaborate:

the neutrality of the analyst, the necessity of regression, and the importance of relying on
interpretation alone to resolve the transference neurosis.

Gill’s (1984a) recent comparison of the classical position as he himself formulated it in 1954

and his current view emphasizes the differences between the two. In light of Gill’s own emphasis on

the contrast, a reader of the 1954 paper may be somewhat surprised at the points of continuity
between the ideas Gill had then and those he has now. It is important to recognize, however, that

although Gill may sometimes underestimate the element of consistency in his ideas about technique

and, perhaps even more so, the element of continuity in the kinds of issues that have concerned him,
the ways in which his ideas have changed are very important and substantial. Even the apparent points

www.freepsychotherapybooks.org 204
of agreement pale when considered in their total context. The changes are associated with the
paradigm shift noted earlier, and Gill is justified in regarding them as “radical.”

One apparent point of agreement is the irrelevance of the arrangement—either the frequency of

visits or the use of the couch—to the definition of the technique. On this matter, Gill could hardly be
more emphatic now than he was in 1954 when he labeled as “foolish” and “ridiculous” the tendency to

regard such “outward trappings” as essential (pp. 774-775). However, this position has a very different

meaning in the context of Gill’s current overall point of view than it had then. In 1954 these

“trappings,” although disdained as part of the definition of technique, were nevertheless consistent
with the attempt to “enforce” a regression; an attempt which was a defining feature of psychoanalytic

treatment (pp. 778-779). In 1984 these trappings have lost this connection to the essence of the

technique Gill advocates.

A second apparent point of agreement is the fact that neutrality does not mean the absence of
any emotional involvement on the part of the analyst. Gill already recognized in 1954 that the analyst

was a participant in the process and not just an observer and that there was room in the
psychoanalytic situation for the analyst to feel and even to show a range of emotional responses,

including amusement, irritation, and sadness (p. 780). Moreover, then as now, what separated an
analysis from psychotherapy was not the absence of any interpersonal influence or suggestion in the

process, but the attempt in the long run to “resolve the suggestive influence of the therapist on the

patient” by means of interpretation (p. 790).

However, in 1954 Gill also wrote: “The clearest transference manifestations are those which
occur when the analyst’s behavior is constant, since under these circumstances changing

manifestations in the transference cannot be attributed to an external situation, to some changed

factor in the interpersonal relationship, but the analysand must accept responsibility himself” (p. 781).
Gill would never make such a statement today since he sees the analyst as implicated on a moment-to-

moment basis in the nuances of the transference as they emerge. On the other hand, he would agree

that the aim of analysis includes a heightened appreciation by patients of their share of the

www.freepsychotherapybooks.org 205
responsibility for the quality of the interaction as they experience it. The relationship between these
points of agreement and disagreement might be clarified if we realize that what was a main point with

regard to technique in 1954 becomes a qualification in 1984 and vice versa. Thus, for example, the

1954 position on neutrality might be paraphrased as follows: Although there is always an element of
suggestion in every analysis, the analyst should try to maintain a relatively constant demeanor in order

to be able to demonstrate to patients that the responsibility for their experience of the relationship lies

primarily within themselves.

The 1984 position would have the emphasis reversed, so that the attitude encouraged is quite
different, that is: Although one of the goals of analysis is to enhance patients’ appreciation of their

responsibility for their experiences of themselves and others, and although it is important that the

analyst avoid being so active or intrusive as to prevent this realization from emerging, on a moment-to-
moment basis, the analyst must pay attention to the ways in which he or she is contributing to the

patient’s experience and should include reference to these contributions, as they are plausibly construed

by the patient, in his or her interpretations.

In line with this important difference, Gill’s main recommendation in 1954 to practitioners of

“intensive psychotherapy” is that they be less directive, in order to bring the process closer to an
optimal psychoanalytic one. His principle recommendation in 1984 to the same end is that they

systematically analyze the transference—it being understood, of course, that analyzing the

transference has a different meaning for Gill in 1984 than it had for him in 1954 and than it has for
most classical analysts.

The principal differences between Gill’s position on technique in 1954 and his current one

center on the following issues: (1) the type of influence that the analyst intentionally exercises; (2) the

extent to which the transference itself is understood and interpreted as a plausible construction on the
patient’s part; and (3) the optimal frequency of transference interpretations.

With regard to the first issue, in 1954 Gill believed, following Macalpine (1950), that inducing a

www.freepsychotherapybooks.org 206
regressive transference neurosis was an essential feature of technique. As noted earlier, Gill has

abandoned this view. He no longer considers the achievement of any particular regressive state,

beyond what the patient brings to the analysis, necessary or desirable. The work is no less analytic if

the issues explored reflect high levels of ego functioning than if they are more overtly primitive, and
there is no requirement that they become more primitive for the process to be called an analysis.

Perhaps even more to the point, he is opposed to manipulating patients in a manner that does not in
itself become a subject of analytic investigation. Instead, the analyst ought to openly encourage

patients to explore their experience of the relationship, understanding that this encouragement may

also have repercussions that require exploration.

Consistent with the creative, dialectical nature of Gill’s thinking throughout his career, even
with regard to the matter of regression, the 1954 discussion includes a foreshadowing of his later

views. He argued then that theoretical advances in ego psychology lent greater weight to the

importance of intrasystemic conflicts in the ego that achieve relative autonomy as opposed to the
intersystemic id-ego conflicts from which they derived. He also reminds us that Freud himself (1926, p.

83) raised a question as to whether, in Gill’s (1954) words, “after repression the original impulse

necessarily persists in the unconscious” (p. 794). Gill suggests that the derivative conflicts may “exist in
a form which allows a relatively firm resolution,” particularly when psychotherapy takes on more of
the character of psychoanalysis by being “more intensive and less directive.” Moreover, he argues that

“this may result in a quantitative shift which may not be so completely different from what often

happens in psychoanalysis” (p. 793).

With regard to the second issue, the differences are both subtle and critical. It is clear that Gill
recognized in 1954 that the general phenomenon of regression in analysis was not spontaneous but

rather induced (e.g., pp. 778-779). However, the particular form that this regression took was one that

was relatively free of the analyst’s influence, which Gill (1954) described as “a nonspecific, steady,

unremitting regressive pressure” (p. 780). The analyst, as noted earlier, could put himself or herself in
a position that would enable him or her to show the patient that the particulars of the transference

www.freepsychotherapybooks.org 207
were coming from the patient alone.

In contrast to this view, and in keeping with his perspectivist orientation, Gill now sees the

analyst as implicated in the transference in highly specific ways, since the transference is associated
with continual plausible speculations on the patient’s part about the analyst’s inner state. Thus, the
best transference interpretations generally refer to some way in which the analyst could plausibly be

understood to have contributed to the patient’s experience. To say this is not to abandon leverage for

demonstrating the responsibility of patients for their own experience. Ultimately, the analysis leads to
patients’ heightened awareness of the repetitive patterns of interaction to which they are prone. The

point is that they repeat patterns of interaction, and patients have reason to believe that the analyst’s

inner experience and outward behavior are colored by the pressure they exert to make the

relationship repeat those patterns.

This shift is apt to be confused with the more common emphasis on the “real” influence of the
analyst. This emphasis is often presented as an alternative to the view that the analyst functions only

as a screen for the transference and as a technical instrumentality. As noted earlier, Gill has moved to a

perspectivist view of reality, especially of emotionally significant interpersonal reality. From this point
of view, the patient’s ideas about the analyst are usually neither simply veridical nor simply groundless

fantasy.1 Gill’s views are similar in some respects to those of Racker (1968), Levenson (1972), and
Sandler (1976), among others. What these theorists—nominally, a Kleinian, a Sullivanian, and a

Freudian—have in common is the idea of an inevitable degree of interlocking of transference and

countertransference and a conviction that empathic transference interpretations must take this
interplay into account (Hoffman, 1983).

The third difference noted in Gill’s position on technique has to do with the frequency of
transference interpretations that his current theory seems to encourage. Although Gill has emphasized

that considerations of tact and timing are exceedingly important, and although he has recognized the
importance of allowing the patient to have the initiative and to develop his or her own thoughts, the

overall thrust of his position nevertheless encourages a generally more interactive stance and certainly

www.freepsychotherapybooks.org 208
more frequent interpretations of allusions to the transference than standard technique would
recommend.

To some extent, this emphasis on regular interpretation follows from the fact that Gill is no

longer interested in facilitating a mode of experiencing and communication that is remote from
secondary process thinking. The conversation between analyst and patient is a special kind, to be sure,

but there is no technical principle, such as the induction of regression, that is opposed on a moment-to-

moment basis to the principle of analyzing the transference. In fact, instead of inducing regression by

depriving the patient, Gill now feels that the transference should be “encouraged to expand” by
continually explicating its immediate direct and indirect manifestations (1979, 1982).

A second consideration that is consistent with more frequent interpretations is Gill’s emphasis

on the “ubiquity” of disguised allusions to the transference (1982, pp. 69-79). Gill (1982, p. 80) differs

with the following statement by Freud, (1913) especially with the first part to which Freud gave special
emphasis: “So long as the patient’s communications and ideas run on without obstruction, the theme of

transference should be left untouched. One must wait until the transference, which is the most delicate
of all procedures, has become a resistance” (p. 139). Gill believes that Freud failed to consistently

recognize resistance in indirect references to the transference. Gill’s review of Freud’s ideas about
transference (1982, pp. 139-175) is thorough and illuminating, revealing Freud’s tendency, despite

some important statements to the contrary, to see the analysis of the neurosis as primary and to see

the transference as an obstacle that has to be dealt with when it obstructs the analysis of the neurosis.
Gill (1982) believes, instead, that the neurosis will find its way into the transference in one way or
another, either in transference of wish or defense (p. 32), and that the transference can be usefully

interpreted in a relatively molecular way throughout the analysis. Moreover, Gill is not concerned that
interpretation of transference per se will interfere with the therapeutic alliance (p. 84). On the
contrary, tactful interpretation of transference from the first session on will promote the alliance2
since it addresses issues that are troubling the patient in a very immediate sense but that the patient
resists speaking of or thinking of explicitly for fear that they will not be accepted or understood.

www.freepsychotherapybooks.org 209
Instead of being concerned about managing his own behavior so as to promote a nonspecific
regression on the one hand, and a spontaneous specific transference on the other, Gill’s attention as a
clinician is devoted to identifying the various disguised expressions of transference, such as

displacement and identification. Citing Lipton’s (1977b) discussion, Gill (1982, p. 170) finds that
familiarity with identification, which is less commonly recognized than displacement (or, I might add,

projection) as a vehicle for indirect communication, greatly expands the range of associations in which

it is compelling to infer that there is an implication for the transference.

Despite his encouragement to the analyst to regularly interpret disguised allusions to the

transference. Gill certainly allows latitude for a wide range of frequency, depending on the patient, the

type of material that is coming up, and the style of the analyst. It is a mistake to regard frequency of

interpretation per se as the crux of the difference between Gill’s position and the classical one. Indeed,
it is quite compatible with Gill’s ideas to be critical of overzealous interpretation of transference, a

perversion to which Gill’s theory of technique may be prone but which certainly is not required by it.

More at the core of Gill’s departure from the classical model, in my view, are the changes associated
with the first and second issues I have cited, namely, the opposition to deliberate manipulation and the

understanding and interpretation of transference as a plausible construction, given the inevitability of

the analyst’s personal participation in the process.

All that I have said here pertains to what Gill (1979,1982) calls “interpretation of resistance to

the awareness of transference,” as distinct from “interpretation of resistance to the resolution of

transference.” There is no question that Gill gives priority to the former as a matter of technique,
feeling that a good deal of resolution of transference will follow spontaneously from its explication in

the here and now. The patient will come to recognize that, for example, he or she paid selective
attention to the features of the analyst’s behavior that he or she had previously disavowed entirely.
Also, the patient will spontaneously recall experiences from childhood that will help to show that his or
her perspective has particular historical origins. Gill (1982) is very concerned that genetic

interpretations may be used as a flight from the here and now, but he also recognizes their importance

www.freepsychotherapybooks.org 210
and the importance of other kinds of interpretation for the resolution of the transference:

It is important that the analyst not be tied to some rigid rule that he should
make only transference interpretations. Not only can extratransference
interpretations be useful, but the spontaneity of the analyst’s behavior is
essential for the conduct of an analysis. If an extratransference
interpretation occurs to the analyst as a plausible clarification, he should
make it. At the same time, he should be alert to its possible repercussions
on the transference—but then he should be alert to the repercussions on
the transference of a transference interpretation too.

I conclude that while extra-transference interpretations play a role in

analysis—and extra-transference clarifications certainly must—priority, in
both time and importance, should go to transference interpretations. This
principle may be more readily accepted if I emphasize that attention to
resistance to the awareness of transference should come first and that,
even though priority in interpretation designed to resolve the transference
should go to interpretation within the analytic situation, working through
requires extra-transference, transference, and genetic transference
interpretations [pp. 125-126].

Gill speaks of the person paradigm or point of view, the interpersonal paradigm, and

perspectivism almost interchangeably, because for him each implies the others. Gill’s more specific

ideas about the analytic situation follow directly from these supraordinate concepts. In the old
metapsychology, according to what Gill (1983b) calls the “energy discharge” point of view, the patient
was encouraged to regress in order to arrive at the underlying infantile neurosis that would bear the

stamp of the patient’s bodily urges, relatively independent of environmental influences. In the old
paradigm it was thought that “free association and regression will in time lead to the relatively direct
expression of bodily urges little related to interpersonal interaction, whether with others in the past or

with the therapist in the present” (p. 546). These urges are the decisive factors underlying the
transference and the distortion of reality the transference entails. In other words, the emphasis on the
past is linked with the idea of a somatic drive that precedes and determines interactions with others.

But for Gill, there is nothing unearthed or reconstructed in psychoanalysis that antedates interactions.

www.freepsychotherapybooks.org 211
The present interaction, moreover, is the best place to look for the person’s fixed ways of organizing
interpersonal experience. Transference is redefined as a way of looking at things and as a way of being

with other people—not a distorted way in any simple sense, but a rigid way that cuts off alternative

potentials. The analyst always interprets in the spirit of acknowledging the plausibility of the patient’s
perspective. There is no absolute reality to which the analyst has access but to which the patient is

blind because of the transference. In fact, the patient’s transference perspective may shed light on

some aspect of the analyst’s own participation (overt and covert), which the analyst resists. Neither
participant has a corner on the truth, and yet they try to hammer out an understanding that makes

sense to both of them and that has the feel of cogency. Even as they are doing so, the analyst must take
the lead in turning a critical eye on what they have decided and how they have decided it. That is why
historical exploration cannot get very far without being interrupted by a question as to its purpose

right now. To raise this question does not preclude returning subsequently to reconstructive work.

This continual scrutiny of the relationship is, in the ideal, not an intellectual exercise, but rather

a moving, new interpersonal experience that represents and promotes personal growth (Gill, 1982, pp.
118-120). The new experience associated with the analysis of the transference rests in part on the

analyst’s openness to the possibility that, wittingly or unwittingly, he or she has been the patient’s
accomplice in the perpetuation of the old, fixed patterns of interaction that the transference

represents. At the very moment in which this openness is conveyed to the patient, the analyst stands a
good chance of extricating himself or herself from the role of accomplice. Gill is fond of citing Strachey's
(1934) and Loewald’s (1960) classic papers on the therapeutic action of psychoanalysis in which both

of them state that analysis cures because the analyst offers himself or herself not only as a technical
instrumentality but as an object with whom the patient can have a new kind of experience. What
Strachey and Loewald omit or underemphasize is the element of mutuality in the shaping of the

transference and the countertransference on a moment-to-moment basis.

With psychoanalytic technique redefined in a manner that encourages more active engagement
of the patient and that demands more systematic exploration of the patient’s immediate experience of

www.freepsychotherapybooks.org 212
the interaction, the whole question of analyzability is thrown open. Gill’s views now on assessment of
analyzability parallel those of Freud in calling for a “trial analysis” as the best way to begin. Gill does
not have fixed ideas on who can benefit from a rigorous psychoanalytic approach based on standard

types of classification. He feels that too many patients are written off as “unanalyzable” because they
cannot adapt well to the couch and to a relatively silent and remote analyst who is systematically

trying to induce a regression (1984a). For Gill, this procedure is misguided, and a patient’s refusal or

inability to comply with it could be a sign of strength. Analyzability for Gill has to do with a patient’s
ability to engage in and reflect upon a relationship with an analyst who is emotionally available and

who thinks of himself or herself as a co-participant in the process. Gill (1983a) mocks the standard

view of analyzability, stating that “an analyzable patient is a patient with whom the analyst can
maintain the illusion of neutrality” (p. 213). In fact, there may be relatively healthy patients with

particular temperaments who could not tolerate standard technique but who would respond well to
Gill’s approach. Similarly, there may be some very sick patients who are automatically written off
because of their inability to adapt to the standard psychoanalytic situation but who may be able to

respond relatively well to the more active focus on the here and now that Gill espouses.

RESEARCH IN PSYCHOANALYSIS

As noted earlier, Gill has been unwavering throughout his career on the necessity of systematic

research on the psychoanalytic process, the third aspect of Gill’s work that I will discuss here. He has
never accepted the common psychoanalytic view, which Freud himself promulgated, that the case
study method, however much it has contributed to theory and practice, can obviate the need for a more

rigorous application of scientific methods to the gathering and analysis of psychoanalytic data. Perhaps

Gill’s most important contribution to the development of psychoanalysis as a science has been his

pioneering effort to make the raw data of psychoanalysis available for study by independent observers
through audio-recordings of psychotherapeutic and psychoanalytic sessions. Inspired partly by Carl

Rogers and others of the client-centered school, who made recording and research a central part of
their practice from the start, Gill, along with a few others, began recording psychotherapy at the

www.freepsychotherapybooks.org 213
Menninger Clinic as early as the middle 1940s. The advent of tape recording greatly facilitated this

effort. Over the years, Gill has collected samples from each of a number of tape recorded analyses,
some conducted by himself and some by the relatively few other analysts who were willing to

contribute.

Recording was instrumental in Gill’s research on hypnosis during the 1940s and 1950s—for

example, in a study of spontaneous fluctuations in the depth of the hypnotic state during sessions of
psychotherapy (Brenman, Gill, & Knight, 1952). In this study, ego functioning around the time of the

fluctuation was assessed by having independent judges examine associations surrounding statements

such as “I’m going deeper” or “I’m coming up lighter.” The method itself provided a model that was

later adapted by Luborsky (1967) in his studies of momentary forgetting in psychotherapy. The model
also evolved into the broader “symptom context method” for investigating the appearance of physical

and psychological symptoms during psychotherapy sessions (Luborsky and Auerbach, 1969).

In the 1960s, Gill and his collaborators (Gill, Simon, Fink, Endicott, & Paul, 1968) wrote a

landmark article on recording and psychoanalysis, in which they take up and challenge many of the
common sources of resistance to recording, some personal and some more clinical or theoretical. On

the personal side, for example, there is fear of exposure and criticism. Gill (Reppen, 1982) recognizes
that for some patient-analyst pairs, recording may pose insurmountable difficulties (p. 171). In
general, however, he feels that the fear of exposure and criticism, both on an individual and

institutional level, must be overcome if analysis is to have any hope of growing or even surviving as a

scientific discipline.

Objections on the clinical side to recording for research purposes include concern about

compromising confidentiality as well as about introducing another purpose into the analytic situation
that is extraneous to the analytic work itself and to the immediate interests of the patient. There is no

question, of course, that patients will react in various ways to these circumstances. However, Gill

(Reppen, 1982; Gill et al., 1968) argues, it does not follow that they preclude a successful analysis. Here
Gill’s point of view on technique dovetails with his attitude on recording. The transference does not

www.freepsychotherapybooks.org 214
develop in a vacuum. However the analytic situation is set up, its transference repercussions must be
explored in the spirit of recognizing the element of plausibility in the patient’s view of the situation. In

the first place, the research situation is not so different from other analytic situations in which

confidentiality is compromised or in which the analyst’s self-interest is readily apparent. When, for
example, an institute candidate sees a patient as part of training, confidentiality is not inviolate, and the

patient has cause to feel used. These circumstances are generally understood to be important

complications that need to be explored, but not of such a magnitude that they preclude a successful
analysis. In the second place, even when the patient has reason to believe that confidentiality is strictly

maintained, the situation is likely to have particular meanings that must be investigated. In what might
be regarded by many as the optimal analytic setting, the promise of strict confidentiality, the strong
recommendation that the frequency of sessions be four times per week or more, the use of the couch,

the analyst’s fee, and the analyst’s silence are all very powerful stimuli, which the patient construes in
plausible and yet also personally expressive and, in principle, analyzable ways (Gill, 1984a).

Gill’s theory of technique is also congenial to research on the psychoanalytic process in that it
invites attention to each analytic hour as a unit that has a certain integrity of meaning. In the classical

paradigm, in which the intent is to foster the unfolding of a regressive transference neurosis over a
long period of time, it would be difficult to assess the quality of the analyst’s technique as well as other

variables because the context of each event is so temporally broad and so difficult to know and take
into account. Although Gill is fully aware of the importance of context and knows that the analyst may
be in a position to take it into account more than an external observer who has only a small sample of

the data, Gill’s molecular focus on the analysis of transference in the here and now lends itself to
investigation of smaller and more manageable units of data. One of the fruits of Gill’s commitment to
systematic research has been the development of a coding scheme (Gill & Hoffman, 1982b), which

permits classification of various kinds of patient communications and analyst interventions. The
highlight of the scheme is the delineation of criteria for identifying disguised allusions to the
transference in associations not manifestly about the transference. The research judge cannot claim
that such an allusion has occurred without giving a specific basis for this inference. The basis may have

www.freepsychotherapybooks.org 215
the form of a previous statement by the patient that is explicitly about the relationship, or it may have

the form of some readily recognizable event in the interaction about which neither of the participants
has spoken. These criteria for coding allusions to the transference have clinical utility as well, because

they set up at least partial guidelines to indicate when a transference interpretation might be called for

and when it might not—guidelines that have been vague or lacking in clinical theory.

Gill recognizes that the development of this coding scheme is only a small first step toward a
more comprehensive program of systematic research on the analytic process as he conceives of it. In

the long run, Gill would want to see variables defined and operationalized and methods developed so

that it would be possible to study, on a molecular level, the differential effects of various types of

intervention on the process and, on a molar level, the relative efficacy of one treatment approach or
another.

Gill’s commitment is not to research for the sake of research but to clinical research that deals

with theoretically meaningful variables. It is a commitment to the scientific study of human intention

and meaning, including the interpersonal conditions that promote change and growth. In this sense,
psychoanalysis for Gill is a hermeneutic science, a contradiction in terms for some (Blight, 1981; Eagle,

1980) but for Gill a category that connotes the special combination of values he feels should
characterize psychoanalysis as a discipline.

CONCLUSION

As we have seen, Gill’s point of view has changed radically over the years. His ideas have

continued to evolve. Any attempt to capture the thrust of his position at a given time is unlikely to do
justice either to various important qualifications that he has proposed or to new ideas and revisions of
theory that are still germinal in his thinking. For example, recently Gill (personal communication,

November, 1984) has been wrestling with several questions. Are there, after all, universal conflicts that
play a role in every analysis, and, if so, what are they? As noted earlier, Gill has been considering the

www.freepsychotherapybooks.org 216
conflict between attachment and autonomy as a primary issue of this kind. With regard to theory of
technique, does the elimination of the principle of deliberately inducing a regression leave a gap that

invites, instead of excessive restraint, overzealous interpretation of transference? What provision is

there in the theory, formally speaking, to prevent the reductio ad absurdum that would have the analyst
forever interpreting the transference repercussions of overzealous interpretation? Is it enough to

emphasize the importance of common sense, tact, and timing, or to say that one does not interpret

until one has a compelling sense of a latent transference meaning in the patient’s associations? Or is it
necessary to formulate another principle of technique to balance the principle of analysis of

transference? Perhaps for Gill this principle would be supraordinate to the analysis of transference and
would, in the most general terms, have to do with promoting a certain quality of interpersonal
experience. This experience might be most powerfully served by the analysis of the transference but,

presumably, could also be undermined by it at times. What this quality of experience would be, of
course, needs to be spelled out. It would also have to be located in relation to other conceptualizations
in the literature of the interpersonal experience in analysis, such as Gill’s own concept of the new

experience that accompanies the analysis of the transference, Zetzel’s therapeutic alliance, Winnicott’s
holding environment, Kohut’s self-selfobject tie, and Schafer’s more generic atmosphere of safety.

Gill’s intellectual style is to steadfastly pursue the logical implications of a particular line of

thought without shrinking from their consequences for entrenched tradition. At the same time, in
dialectical relationship with this tendency, his convictions about theory, research, and practice are
united by his readiness to turn a critical eye on his own perspective and to consider other points of

view. Thus, despite the vigor with which he has advocated and defended his position, Gill has also
actively explored the points of convergence and divergence of his own views and those of Gedo (Gill,
1981), Melanie Klein (Gill, 1982, pp. 129-137), Kohut (Reppen, 1982, pp. 183-186), Sullivan (Gill,

1983a), and Langs (Gill, 1984b), among many others. The very fact that he is actively engaged in
dialogue with exponents of these diverse perspectives (as reflected in his publications, speaking
engagements, and extensive correspondence) testifies to the bridge-building role that Gill now
occupies in the field. I believe that such a role is congenial to him because of his disdain for

www.freepsychotherapybooks.org 217
parochialism; he has a deep conviction that psychoanalysis will survive and grow only if exponents of

diverse viewpoints engage each other in an ongoing process of reciprocal, constructive criticism and
ultimately submit their differences to the arbitrating power of systematic clinical research.  

REFERENCES

Apfelbaum, B. (1966). On ego psychology: A critique of the structural approach to psychoanalytic

theory. International Journal of Psycho-Analysis, 47, 451-475.

Basch, M. F. (1977). Developmental psychology and explanatory theory in psychoanalysis. Annual of

Psychoanalysis, 5, 229-263.

Bateson, G., & Mead, M. (1942). Balinese character: A photographic analysis. Special Publications of the
New York Academy of Sciences, 2.

Blight, J. G. (1981). Must psychoanalysis retreat to hermeneutics? Psychoanalytic theory in the light of
Popper’s evolutionary epistemology. Psychoanalysis & Contemporary Thought, 4, 147-
205.

Brenman, M., Gill, M. M., & Knight, R. P. (1952). Spontaneous fluctuations in depth of hypnosis and their
implications for ego-function. International Journal of Psycho-Analysis, 33, 22-34.

Eagle, M. N. (1980). A critical examination of motivational explanation in psychoanalysis.

Psychoanalysis & Contemporary Thought, 3, 329-380.

Eagle, M. N. (1981). Interests as object relations. Psychoanalysis & Contemporary Thought, 4, 527-565.

Freud, S. (1913). On beginning the treatment: Further recommendations on the technique of psycho-
analysis. Standard Edition, 12, 147-156.

Freud, S. (1923). The ego and the id. Standard Edition, 19, 12-66.

Freud, S. (1926). Inhibitions, symptoms and anxiety. Standard Edition, 20, 87-172.

www.freepsychotherapybooks.org 218
Gill, M. M. (1954). Psychoanalysis and exploratory psychotherapy. Journal of the American
Psychoanalytic Association, 2, 771-797.

Gill, M. M. (1963). Topography and systems in psychoanalytic theory. Psychological Issues, 3 (2,
Monograph 10). New York: International Universities Press.

Gill, M. M. (1967). The primary process. In R. Holt (Ed.), Motives and thought: Psychoanalytic essays in
honor of David Rapaport. Psychological Issues, 5 (2/3, Monograph 18/19), 260-298.

Gill, M. M. (1976). Metapsychology is not psychology. In M. M. Gill & P. S. Holzman (Eds.), Psychology
versus metapsychology. Psychoanalytic essays in memory of George S. Klein.
Psychological Issues, 9 (4, Monograph 36), 71-105.

Gill, M. M. (1977a). Psychic energy reconsidered: Discussion. Journal of the American Psychoanalytic
Association, 25, 581-598.

Gill, M. M. (1977b). The two models of the mental health disciplines. Bulletin of the Menninger Clinic, 41,
79-84.

Gill, M. M. (1978). Freud’s concepts of unconsciousness and the unconscious. In A. S. Prangishvilli, A. E.

Sherozia, & F. V. Bassin (Eds.), The Unconscious (Vol. 1). Tbilisi, U.S.S.R.: Metsniereba
Publishing House.

Gill, M. M. (1979). The analysis of the transference. Journal of the American Psychoanalytic Association,
27 (Supplement), 263-288.

Gill, M. M. (1981). The boundaries of psychoanalytic data and technique: A critique of Gedo’s Beyond
Interpretation. Psychoanalytic Inquiry, 1, 205-232.

Gill, M. M. (1982). Analysis of transference: Vol. 1. Theory and technique. Psychological Issues
(Monograph 53).

Gill, M. M. (1983a). The interpersonal paradigm and the degree of the therapist’s involvement.
Contemporary Psychoanalysis, 19, 200-237.

Gill, M. M. (1983b). The point of view of psychoanalysis: Energy discharge or person. Psychoanalysis

www.freepsychotherapybooks.org 219
and Contemporary Thought, 6, 523-551.

Gill, M. M. (1984a). Psychoanalysis and psychotherapy: A revision. International Review of Psycho-

Analysis, 11, 161-179.

Gill, M. M. (1984b). Robert Langs on technique: A critique. In J. Raney (Ed.), Listening and interpreting:
The challenge of the work of Robert Langs. New York: Jason Aronson.

Gill, M. M., & Brenman, M. (1959). Hypnosis and related states: Psychoanalytic studies in regression. New
York: International Universities Press.

Gill, M. M., & Hoffman, I. Z. (1982a). Analysis of transference: Vol. 2. Studies of nine audiorecorded
psychoanalytic sessions. Psychological Issues (Monograph 54).

Gill, M. M., & Hoffman, I. Z. (1982b). A method for studying the analysis of aspects of the patient’s
experience of the relationship in psychoanalysis and psychotherapy. Journal of the
American Psychoanalytic Association, 30, 137-167.

Gill, M. M., & Klein, G. S. (1964). The structuring of drive and reality: David Rapaport’s contributions to
psychoanalysis and psychology. International Journal of Psycho-Analysis, 45, 483-498.

Gill, M. M., & Muslin, H.(1977). Early interpretation of transference. Journal of the American
Psychoanalytic Association, 24, 779-794.

Gill, M. M., Newman, R., & Redlich, F. C.(1954). The initial interview in psychiatric practice. New York:
International Universities Press.

Gill, M. M., Simon, J., Fink, G., Endicott, N. A., & Paul, I. H. (1968). Studies in audiorecorded
psychoanalysis: 1. General considerations. Journal of the American Psychoanalytic
Association, 16, 230-244.

Hoffman, I. Z. (1983). The patient as interpreter of the analyst’s experience. Contemporary

Psychoanalysis, 19, 389-422.

Holt, R. R. (1965). A review of some of Freud’s biological assumptions and their influence on his
theories. In N. S. Greenfield & W. C. Lewis (Eds.), Psychoanalysis and current biological

www.freepsychotherapybooks.org 220
thought (pp. 93-124). Madison: University of Wisconsin Press.

Holt, R. R. (1976). Drive or wish? A reconsideration of the psychoanalytic theory of motivation. In M. M.

Gill & P. S. Holzman (Eds.), Psychology versus metapsychology: Psychoanalytic essays in
memory of George S. Klein. Psychological Issues, 9 (4 Monograph 36).

Klein, G. S. (1973). Two theories or one. Bulletin of the Menninger Clinic, 37, 102-132.

Klein, G. S. (1976). Psychoanalytic theory: An exploration of essentials. New York: International

Universities Press.

Leary, T., & Gill, M. M.(1959). The dimensions and a measure of the process of psychotherapy: A system
for the analysis of the content of clinical evaluations and patient-therapist verbalizations.
In E. A. Rubinstein & M. B. Parloff (Eds.), Research in Psychotherapy. Washington, D.C.:
American Psychological Association.

Levenson, E. (1972). The fallacy of understanding. New York: Basic Books.

Lipton, S. D. (1977a). The advantages of Freud’s technique as shown in his analysis of the Rat Man.
International Journal of Psycho-Analysis, 58, 255-273.

Lipton, S. D. (1977b). Clinical observations on resistance to the transference. International Journal of

Psycho-Analysis, 58, 463-472.

Loewald, H. W. (1960). On the therapeutic action of psychoanalysis. International Journal of Psycho-

Analysis, 41, 16-33.

Luborsky, L. (1967). Momentary forgetting during psychotherapy and psychoanalysis: A theory and
research method. In R. R. Holt (Ed.), Motives and thought: Psychoanalytic essays in honor
of David Rapaport. Psychological Issues, 5 (2/3, Monograph 18/19).

Luborsky, L., & Auerbach, A. H. (1969). The symptom-context method: Quantitative studies of symptom
formation in psychotherapy. Journal of the American Psychoanalytic Association, 17, 68-
99.

Macalpine, I. (1950). The development of the transference. Psychoanalytic Quarterly, 19, 501-539.

www.freepsychotherapybooks.org 221
Peterfreund, E. (1971). Information, systems, and psychoanalysis. Psychological Issues, 7 (1/2,
Monograph 25/26).

Pribram, K.H., & Gill, M. M. (1976). Freud’s “Project” reassessed. New York: Basic Books.

Racker, H. (1968). Transference and countertransference. New York: International Universities Press.

Rapaport, D. (1967). The collected papers of David Rapaport (M. M. Gill, Ed.). New York: Basic Books.

Rapaport, D., & Gill, M. M. (1959). The points of view and assumptions of metapsychology. International
Journal of Psycho-Analysis, 40, 1-10.

Rapaport, D., Gill, M. M., & Schafer, R. (1945-46). Diagnostic psychological testing (2 vols.). Chicago: Year
Book Publishers.

Rapaport, D., Gill, M. M., & Schafer, R. (1968). Diagnostic psychological testing (rev. ed.; R. Holt, Ed.).
New York: International Universities Press.

Reppen, J. (1982). Merton Gill: An interview. Psychoanalytic Review, 69, 167-190.

Ross, N. (1965). [Review of Topography and systems in psychoanalytic theory, by M. M. Gill.]

International Journal of Psycho-Analysis, 46, 254-256.

Sandler, J. (1976). Countertransference and role-responsiveness. International Review of Psycho-

Analysis, 3, 43-47.

Schafer, R. (1976). A new language for psychoanalysis. New Haven: Yale University Press.

Spence, D. P. (1964). [Review of Topography and systems in psychoanalytic theory, by M. M. Gill.]

Journal of Nervous and Mental Disease, 139, 401-403.

Strachey, J. (1934). The nature of the therapeutic action of psychoanalysis. International Journal of
Psycho-Analysis, 15, 127-159.

Notes

www.freepsychotherapybooks.org 222
1) Since Gill’s views have been changing, some of his recent writings show remnants of his earlier,
more traditional stance that are inconsistent with his newer ideas. Thus, for example, in
his recent monograph on technique (1982), Gill sometimes divides the patient’s
experience into transferential and realistic components (e.g., pp. 94-96), although
beginning with chapter 7, he adopts a more consistently perspectivist position.

2) Although Gill sometimes uses the term “alliance,” he agrees with Lipton (1977a) that the concept is
objectionable insofar as it denotes or connotes something that should be deliberately
fostered with special techniques and that is uncontaminated by transference (see Gill,
1982, pp. 96-106; Reppen, 1982, pp. 173-174).

www.freepsychotherapybooks.org 223
7
ROBERT LANGS: THE COMMUNICATIVE
APPROACH

ZVI LOTHANE, M.D.

Expounding is propounding: It is not possible to expound another person’s views without, at the
same time, propounding one’s own. This has been true of expositions of Freud and applies to those,

such as the theorists in this volume, who went beyond Freud.

“Going beyond” is a spatial metaphor, which implies being in one place and then going
somewhere else or toward something else. Psychologically, it means an identification with and a

departure from. Thus, an exposition of thinking beyond Freud requires a preliminary exposition of
Freud. But here, too, both the analytic and the lay expositions of Freud are face to face with the ever-

present question: How to read Freud? For Freud is a protean thinker. Like Proteus, the Greek god of

prophecy who, when consulted, refused to give answers but instead assumed various shapes, so Freud

constantly eludes the attempt to give a definitive reading of his text. He has been claimed by many

domains, from the biological (Sulloway, 1979), to the theological (Homans, 1970). Who can truly
proclaim what Freud really said? Many of those who made such a claim have created the most hair-

raising revisions (Lothane, 1983b). What the Italians say about translations applies here: traduttore,
tradittore (the translator is a traitor).

www.freepsychotherapybooks.org 224
Robert Langs began as a Freudian analyst before he developed his communicative approach.

Therefore, I shall present my reading of Freud and show what he took over from Freud and where he

took off.

Freud the methodologist, rather than Freud the ideologist or metapsychologist, is my focus in
reading Freud (Lothane, 1980, 1981a, 1981b, 1982a, 1983a, 1984b). The methodological focus is
operational: it studies what mind does as against what mind is. From this vantage point, Freud’s was

from the outset a depth psychology, which was dynamic, dialectical, and dualistic—that is, concerned

with the conflict of strivings and actions both between man and man (interpersonally) and within man
himself (intrapersonally).

Freud’s psychology encompasses the following varieties of dynamics and dialectics:

1. The dialectics of the surface versus the depth, of the conscious versus the preconscious, of the
manifest versus the latent, of the remembered versus the forgotten, of the explicit versus

the implicit.

2. The dynamics of defense, originally repression but later including other modes of defense
(also referred to as resistance, both intrapersonally and interpersonally).

3. The dynamics of dream thinking versus waking modes of thought.

4. The dynamics and dialectics of transference versus love and self-love.

5. The dynamics and dialectics of emotion and desire, including sexual desire.

Historically, during the first two decades of his psychoanalytic work, from 1895 to 1915, Freud
was concerned with method and clinical theory. Thereafter he became increasingly concerned with

metapsychology. The gist of the method is given in the Studies on Hysteria (1895d) and The

www.freepsychotherapybooks.org 225
Interpretation of Dreams (1900a). Both the psychological symptom and the dream are seen as similarly
constructed: They show the same correspondence between the way they were caused and the way
they were cured. The memory of a painful (traumatic) or conflictual event in the past is transformed by

the silent (unconscious) operation of defense (repression) into a symptom. The cathartic method (the

technique of hypnosis) creates the conditions of widening of consciousness and the emergence of
memories in pictorial (imagic) forms and in words. Overcoming defense (resistance) facilitates this
process and thus erases the pathogenic sting of memory. To be sick is to reminisce. To be cured is to

recall and erase the record. Similarly, the painfully or conflictually experienced event prior to the
dream—the day residue—evokes a psychological reaction: the latent dream thoughts. Under the
influence of the censor and through the silent (unconscious) operation of dream work, the latent

thoughts are transformed into the manifest content, or the dream as remembered. The cathartic
method is now replaced by the psychoanalytic method. It is a homologue of the hypnotic technique and

of the preconditions for dreaming. The withdrawal of attention from goal-directed pursuits and critical

selection creates an altered state or frame of consciousness and fosters the emergence of pictorial
modes of thought: images of memory, of imagination, of dreams and hallucinations. The combination of

spontaneous free association and of directed free association to the separate elements of the manifest
dream content leads to a retrieval of the day residue and the antecedent thoughts and feelings that

were the reaction to the day residue. The central conception here is that action, whether symptom or
dream, is determined by external reality (Lothane, 1983a). Both the symptom and the dream are a

personal response to and a commentary on an episode of lived reality. Rapaport (1960) called this the

adaptive point of view. I would like to refer to it as the action-reaction conception of the symptom and
the dream.

What is being reemphasized here is that Freud’s depth psychology is also a conflict psychology,

and it is meant to explain inhibitions and distortions of memory and of sense perception. Both the
symptom and the dream are shaped by defense in all its varieties and by the dream work in all its

varieties. In this sense, the symptom is a return of the repressed and the dream a transformation of the

latent content into the manifest content. Both are strange and puzzling manifestations and require a

www.freepsychotherapybooks.org 226
solution. To recall the memory that caused the symptom means to analyze, or dissolve, it; to trace the

latent content from the manifest content is to interpret a dream, or solve it. As Freud (1900a) writes:

The [latent] dream thoughts and the [manifest] dream content are given to
us as two depictions [Darstellungen] of the same content in two different
languages.…The dream content is expressed, so to speak, in a picture
[hieroglyphic] script whose signs have to be translated, one by one, into
the language of the [latent] dream thoughts. We would obviously be led
into error if we were to read these signs according to their picture value
instead of according to what the signs refer to…. [pp. 283-284; author’s
translation, italics added; see also Freud, 1900a, pp. 277-278].

The transformation wrought by the dream work can be undone by the activity which, Freud

(1901) says, is “the counterpart of this [dream] work, which brings a transformation in reverse, which

I already know of analysis-work” (p. 645; author’s translation, italics Freud’s). This original text, the
first consciously registered reaction to trauma, is the cause of the second, edited text. The first text is

now unconscious and replaced by the now conscious second text, which is a derivative, disguised,

displaced, condensed, dramatized, pictorial, or encoded version of the first text. To analyze a dream, or
a symptom, is not to read it cognitively or literally, according to the picture value, but to decode it. Such

decoding can only be accomplished by a recourse to the special dynamics of the psychoanalytic
situation: undoing of repression, fostering of images and memories, and tracing the associative chains
of reference from the signifier (the manifest content) to the signified (the latent content). The

emergence of imagic forms of thought proceeds hand in hand with a shift in the dynamics of

repression. Just as the conditions of sleep and dreaming decrease waking vigilance, undo repression,

and facilitate the emergence of the repressed, so the psychoanalytic situation also fosters just that; to
the extent that waking, conversational give-and-take is decreased, the repressed memories, attitudes,

and expectations have a chance of emerging. In this way the psychoanalytic-therapeutic situation

provides the conditions for a dynamic (undoing of repression) and associative (emergence of images)
unfolding and decoding of the patient’s story.

Both repression (defense) and the dream work are unconscious, or silent, processes. The

www.freepsychotherapybooks.org 227
emphasis is on unconscious as an adjective qualifying the nature of this mental activity, not on the
reified unconscious and its various connotations. The reason for this emphasis is twofold: (1) to

underscore the dynamic-reactive nature of this activity in response to a reality stimulus; and (2) to

hold to the conception of a continuous counterpoint between direct modes of memory and perception
and indirect, or distorted, modes. Freud provides his own emphasis in these words in a footnote added

in 1928 to The Interpretation of Dreams (1900a):

I used at one time to find it extraordinarily difficult to accustom readers to

the distinction between the manifest content of dreams and the latent
dream thoughts.…But now that analysts at least have become reconciled to
replacing the manifest dream by the meaning revealed by its
interpretation, many of them have become guilty of falling into another
confusion.…They seek to find the essence of dreams in their latent content
and in so doing they overlook the distinction between the latent dream
thoughts and the dream work. At bottom dreams are nothing other than a
particular form of thinking made possible by the conditions of the state of
sleep. It is the dream-work which creates that form and it is alone the
essence of dreaming—the explanation of its peculiar nature [pp. 506-507;
italics Freud’s].

The action-reaction paradigm is from the start opposed by another line of thought in Freud: the
role of sexuality. At first, sexuality was considered as an aspect of external reality, and in the form of

seduction it played the role of an external traumatizing event, evoking its proper response. The

overthrow of the seduction theory by Freud went hand in hand with another development: the concept
of the dream as wish fulfillment. These two ideas pave the way for the final conception of the body as
an internal source, distinct from external reality, which generates two kinds of movers of human

action: the drive and the (dream-) wish.

With this new emphasis on action in response to inner sources of stimulation goes a
deemphasis of the environment as a stimulus to action in the service of adaptation.

A most important corollary to this new orientation is the attitude toward the dream and

www.freepsychotherapybooks.org 228
daydream, or fantasy. Fantasy is no longer, via the dream work, a reaction to and commentary on

events in external reality but an internally generated action. This reformulation of the dichotomy of

internal-external creates a new approach to defining paradigms of cause (pathology) and paradigms of

cure (analytic technique). One example is the so-called structural theory and the preponderant
preoccupation with metapsychology. Although a fuller discussion of these is beyond the scope of the

present essay, the central implication of this ideological shift was a tendency among analysts to
embrace scholastic debates about internality, the remote infantile past, hybrid concepts, and theories

removed from the realities we live in.

The internal-external dichotomy rears its head once again with the full blooming of the concept

transference, first defined in Freud’s Studies of Hysteria (1895). Although the notion of internally
generated drives and fantasies only presupposes an object, the transference actualizes it; in the

psychoanalytic-therapeutic situation, the other person is experienced simultaneously as a real and an

imaginary other. Thus the one-person psychology of drives and wish fulfillment becomes the two-
person psychology of the interpersonal realm—of dialogue, communication, and interaction. Of the

many aspects of transference, in addition to its traditional definition as reenactment of the past, two

others are immediately relevant to the present argument: (1) its relation to dreaming (Lothane,
1983a), and (2) its relation to the dichotomies truth/error and reality/delusion. The one-person
psychology and the intrapersonal dynamics of drives and internally derived fantasies had this effect on

the concept transference: They tended to convert it into a monadic instead of a dyadic reality, divorced

from the reciprocal personal influences between the participants in the psychoanalytic dialogue.

These trends have resulted in a curious double standard in the analytic profession. Although the
ruling theories (metapsychology) have been formulated largely in terms of a one-system, one-person

psychology and mechanism, the clinical practice has been rolling along in the context of interpersonal

relations, conversation, and interaction. On the one hand, the one-system orientation created its

dogmatics (Hartmann) and schismatics (Schafer) and bitter theological warfare within the
psychoanalytic movement. On the other hand, the interpersonal approach has remained bereft of a

www.freepsychotherapybooks.org 229
systematic theory. Against this background, we can now proceed to examine the contribution of Robert
Langs. I met Langs in 1980 as a result of having published a review of his The Listening Process (Langs,
1978a; Lothane, 1980). I later joined the faculty of the Lenox Hill Hospital Psychotherapy Program, of

which Langs is the founder and director. In the exposition that follows, however, I have limited myself
to the use of published material so that my assertions can be checked against verifiable sources and

debated accordingly.

Robert Joseph Langs graduated from the Chicago Medical School in 1953. He later became a
psychiatrist and graduated from Downstate Psychoanalytic Institute (now the New York University

Institute). He joined the faculty there and was enrolled as member of the American Psychoanalytic

Association. By 1971 he had become active in the practice of psychotherapy and psychoanalysis in

clinical research and was on the staff of the Long Island Jewish and Hillside Hospitals. He had
published clinical and research papers (Langs, 1978b). His first major psychoanalytic paper, “Day

Residues, Recall Residues and Dreams: Reality and the Psyche,” appeared in 1971. It contains the germ

of his future views and “proved to be a fateful beginning” (Langs, 1978, p. 6).

In this paper Langs rediscovered external reality and its relevance for fantasy life, past and
present. The clinical fact that led to this rediscovery was that the day residue—that is, events in

external reality—was crucial to the understanding of the dream. As shown earlier, this was Freud’s
own perennial insight, which was replaced by formulations about the varieties of intrapsychic movers

(the id, “the” unconscious, and unconscious fantasies). Toward the end of the paper, Langs(1971)

argues for a

reassessment of Freud’s thinking regarding infantile seduction. In essence,

we can see that Freud was actually correct in both of his formulations
regarding the role of reality in the formation of neurosis: real seductions
do occur on many levels, while unconscious fantasies are also constantly
being created and revised from both experiencing and imagining.
Together, interacting, creating a totality, they lead to the anxieties and
conflicts out of which neurosis develops [p. 521].

www.freepsychotherapybooks.org 230
In this, Langs anticipated the recent surge of interest in seeing the original seduction theory

reinstated (Klein & Tribich, 1979; Lothane, 1983a; Musson, 1984; Swales, 1982).

This then is Langs’ fundamental idea, the foundation on which the Langsian approach rests: The

day residue is the stimulus to which the dream is a response. If for the day residue we substitute the
psychoanalyst, the analyst’s actions and conduct, and the way they affect the patient, we obtain the gist
of Langs’ method. Every sequence of the psychoanalytic session shows an adaptive context, that is, the

action of the analyst, the reality trigger, and the patient’s double-layered reaction. This reaction has its

manifest content and its latent content, what Langs (1978b) termed the “specific unconscious fantasies
and memories contained in derivative and disguised form in the manifest material” (p. 10). This

seminal idea has subsequently led Langs to a number of extrapolations, which are both an extension of

Freud’s method and a departure from it. We shall have a closer look at these issues later.

The next stage in the evolution of Langs’ ideas is seen in the two volumes of The Technique of

Psychoanalytic Psychotherapy (1973, 1974) as well as in a clinical paper, “A Psychoanalytic Study of

Material from Patients in Psychotherapy,” (1972). The two volumes of The Technique of Psychoanalytic

Psychotherapy constitute a textbook that reflects the best in the classical psychoanalytic tradition. They

show Langs to be a seasoned psychoanalytic clinician who writes lucidly and persuasively.

Continuing the line of thought about day residues and dreams, Langs places central emphasis

on

human adaptation in neurotogenesis.…Functioning, responding, and

adapting are set off by environmental alterations…[the] environmental
stimulus may, in general, be positive and supportive or negative and
traumatic. Most crucial for the development of neurotic disturbance are
the intrapsychic responses to traumatic stimuli. It is these major, currently
disruptive stimuli which have the potential to set off inappropriate or
maladaptive (neurotic) responses that I have identified as the primary
adaptive task [pp. 281-282].

www.freepsychotherapybooks.org 231
This is in the spirit of Freud in the Studies on Hysteria. The traumatic reaction is the paradigm of

disease, and identifying the trauma is the cure. Such identification is barred by the patient’s

defensiveness, which must be analyzed first. But the traumas are not limited to intercurrent reality
events in the patient’s extra-analytic life. A major event may be the previous session, “the therapist’s

interventions or lack of them, which may have traumatized the patient and evoked responses in him”
(p. 284).

The extension of the traumatizing event to include the actual behavior of the therapist is the

beginning of the specific Langsian emphasis. Iatrogenic trauma, injury caused by the doctor’s actions, is
a medical commonplace. Speaking of the dangers of hypnosis, Freud (1895) states: “Where I caused

damage, the reason lay elsewhere and deeper.” (p. 266). Many analysts have acknowledged the

potential of the therapist to cause harm by countertransference. But no analyst before Langs has

defined the therapist as an ever-present traumatizing agent, and none has made this point of view into
a system, as Langs has. This topic will be discussed further.

In this textbook, Langs also develops the other methodological idea of Freud’s first two decades,

the idea of the manifest and latent content.

This cardinal concept of dream psychology is conjoined by Langs with the idea of the adaptive

task. In this way, a new methodological tool has been created for getting hold of the meaning of the

patient’s communications.

To understand a communication in context is different from understanding it in isolation. Freud

addressed this issue squarely in The Interpretation of Dreams, where he contrasted the reading of a
dream according to a universal symbol key with reading it in reference to a specific day residue and a

specific decoding or tracing of the manifest dream thoughts to their latent antecedents. With the
growth of the assorted psychoanalytic causal doctrines and ideologies, analysts have developed a
fondness for stock formulas and clichés, used in the manner of what Freud called the “Egyptian dream
book.” Notions like castration anxiety, penis envy, the Oedipus complex, identity, and separation

www.freepsychotherapybooks.org 232
became the stock-in-trade of what Sandor Feldman (1958) called “blanket” interpretations. Otto

Isakower (1968, 1971) warned, similarly, against the habit of diagnosing set patterns and trends in a

given sequence of an analytic session rather than getting the drift of the actual mental images and their

role in the communication. Like many others, Langs was faced with the sterility of the analytic cliches
and chestnuts and went in search of the truth of the given moment in the lived experience.

Freud’s idea of the manifest and latent was not limited, however, to the transformation wrought
by the dream work. His depth psychology also addressed the issue of honesty versus hypocrisy in

human communication, the difference between what is said and what is intended or meant. The

content aspect of “meaning” has had a greater hold on both the popular and the professional

imagination than the intent aspect. Although alive to the importance of intention, Langs has followed
established habit in using the shorthand “unconscious fantasy” to refer to the deeper, implied, indirect,

concealed intentions in communication. The manifest content is seen as a hidden, disguised, allusive

reference to the direct idea or intention that lies latent in the manifest material and manifests itself as a
derivative of the antecedent direct idea. Thus, any piece of material may be read not naively and at face

value, but as a derivative pointing to deeper-lying truths. This seminal Freudian idea underwent some

transformations in Langs’ writings, which will be examined later.

It should be sufficient to note at this point that the use of the term “unconscious” is liable to
certain pitfalls related to the tendency among analysts to reify the concept of the unconscious. Another
tendency is, as in Freud’s caution quoted earlier, to sacrifice the manifest content to the latent content,

as if the manifest were second hand goods to be bypassed on the way to the latent content. Here Langs

(1973) notes that “manifest content screens or conceals, but also reveals some of what lies beneath it”
(p. 296). To deny the manifest would imply a wholesale repudiation of all art. For what is art but the

giving of artful, or derivative, expression, that is, in various guises and disguises, to what can be more

simply and directly expressed?

Two central conclusions emerge from this for Langs: (1) what to listen for in the material of the

hour; and (2) the order of priorities in such listening. Regarding the first, every hour revolves around

www.freepsychotherapybooks.org 233
two contexts: the therapeutic context, that is, the manifestations of the patient’s psychopathology as
reported, and the adaptive context, as defined earlier. Both these are listened to on both the manifest

and latent, conscious and unconscious levels. The highest priorities in listening in preparation for

intervening, however, Langs concludes are indeed these “reactions to errors by the therapist and acute
symptomatic crises” (p. 364). The other aspects, in order of decreasing priority, are “disturbances in

the therapeutic alliance arising from sources other than the therapist’s errors; other resistances;

current intrapsychic conflicts and unconscious fantasies related to them; the genetic basis for the
patient’s reactions to the therapist and for his present symptoms and inner conflicts; reality issues and

problems” (p. 364).

This exposition contains the essence of Langs’ thought. It is on the one hand firmly rooted in the

classical psychoanalytic tradition, and it marks a departure from it, on the other. The point of
departure is the transition from a one-system, intrapersonal conception to a two-system, interpersonal

or interactional conception. It should be noted that Langs has not completely given up the

intrapersonal habit of conceptualizing. His clinical theories of symptom formation and the role of
memory and fantasy are traditionally intrapersonal (intrapsychic). At the same time, his interpersonal

formulations differ in this respect from those encountered in the classical literature: The delineation of
the adaptive context has from the beginning led Langs to a consideration of the analyst, or therapist, as

an ever-present traumatogenic agent. This one idea has been driven by him relentlessly to its logical
limit. Let us examine this more closely.

The adaptive context emerges as the final common strand that gathers into itself all of the
following threads: (1) how one listens to the session, how one discovers the sense, the point, the central

message of any given session; (2) the reciprocal action of the patient and therapist upon each other,
each manifesting a response to the other; (3) the two-layered derivative structure of each
communication: the conscious/unconscious, latent/manifest, explicit/implicit (subsequently defined

as truth/lie) levels of each utterance.

A convenient mid-point in the evolution of Langs’ ideas is his book, The Listening Process

www.freepsychotherapybooks.org 234
(1978a), which I have discussed elsewhere at some length (Lothane, 1980). In it he spells out clearly
the interweaving strands of the adaptive context.

The classical view stressed the thematic content of the patient’s utterances in the

psychoanalytic-therapeutic situation. The themes in the patient’s narrative were related to the
patient’s memories, past and present-day realities. Initially, in the Studies on Hysteria, the analyst

understood—that is, interpreted—these themes the same way a reader understands or interprets a

told or printed story: by becoming aware of its meanings, messages, references. But in the Studies on

Hysteria, Freud had already become aware of a story within a story, a drama within a drama: the
emergence of transferences, or the effect of the patient-doctor interaction on the story as told. Thus,

whereas at first memory was subjected to the same dispassionate scrutiny in the therapeutic session

as was the histologic section under the microscope, it soon became evident that such scrutiny had to be
tempered by clarifying the personal equation. Thus, the evidential status of the seduction stories was

reevaluated as an attempt on the part of the patients to fake such stories in order to have a personal

effect on the listener. Even with the recognition of the distorting potential of the here and now on the
there and then, however, the latter was still viewed as a result of the intrapersonal dynamics of the

patient, with the therapist remaining the dispassionate observer and interpreter of the patient’s inner
drama as remembered and enacted in the analytic situation.

As I have argued elsewhere (Lothane, 1983a), Freud replaced the trauma and dream paradigms

of the symptom with the concept of intrapsychic dynamics of instinctual drives and defenses and
intrapsychic determinism, as consistent with a one-system psychology. However, the fact remained
that the analyst was not only a naturalistic observer from above of the goings on within the patient, a

diagnostician of symptom complexes and mechanisms of defense; the analyst was also a participant
observer. This placed an insoluble strain on the one-system conception. For whereas it takes one
person to remember or to dream, it takes two to talk. Speaking and listening are in their very nature

interpersonal and interactional. The basic one-system orientation persisted with the emergence of the
concept of transference. The analyst in the transference was seen as an inert screen onto which dreams

www.freepsychotherapybooks.org 235
and memories were projected. And such projecting does exist. This mode of functioning of the
psychoanalytic interaction is still valid as an instrument for the clarification of the there and then. But
it is insufficient for the understanding of the here and now, the actual goings on in the living doctor-

patient encounter.

This emphasis on the here and now and the present-day relationships as opposed to the there

and then and past relationships first began with Ferenczi (see Lothane, 1983a). It was followed

independently by Sullivan (see Lothane, 1984), Wilhelm Reich (1949), Szasz (1961), and now, among
others, by Langs. In this connection, however, it should be appreciated that Freud’s concept of the

hysterical symptom was interpersonal from the start; the symptom was a statement with meaning that

was intended for another person and thus could be decoded by an observer or listener.

It is the theories that were either intrapersonal or interpersonal, not the phenomena in

question. Similarly, the doctor-patient relationship was viewed as a personal relationship from the
very start, in the Studies on Hysteria. It is only the politics and ethics of this personal relationship that

were not spelled out till some 20 years later in Freud’s (1912-15) papers on technique. Thus the often-

touted achievement of the so-called object relations theorists is not a finding but refinding of a truth
already present in Freud.

To return to the main thread of this exposition, the gist of Langs’ communicative approach (the
latest designation of what was formerly called the adaptational or interactional approach) is these two
ideas: the adaptive context (Freud’s trauma paradigm) and derivative communication (Freud’s dream

language paradigm). Having explained the idea of the adaptive context, let us now turn to the concept
of derivative communication.

The notion of derivative is in Freud. He defined derivative in the context of intrapersonal

dynamics; Langs has redefined it in the context of the dialogue, in a specific way. Freud used the idea of

derivative to refer to something observable that was seen as arising or formed from something else

and prior to it. The notion of derivative is basic to Freud’s method of determining causes and origins of

www.freepsychotherapybooks.org 236
phenomena. In linguistics a word derives from an earlier word. In chemistry one compound is a

derivative of another. Freud (1915) used the word Abkommling (literally, offspring), to state, for

example, that “repression proper concerns psychical derivatives [psychische Abkommlinge] of the

repressed [instinctual drive] representation, or such trains of thought which, arising elsewhere,
become related to it through association” (p. 250; author’s translation, second italics added).

Symptoms and dreams are psychological derivatives of trains of thought that are hidden, repressed, or
warded off. Freud’s psychology thus necessarily implies a surface and a depth. His depth psychology is

based on the manifest and latent dichotomy.

The concept of derivative thus implies two basic judgments: a judgment about cause and

origins, and a judgment about what is primary and what is secondary. A formulation about the
dynamics of a case history, or of a sequence in a session, will thus entail a discussion of all sorts of

derivatives.

The idea of the derivative is another centerpiece in Langs’ communicative method. It is locked

into the idea of an adaptive context and the two are an indissoluble whole. Langs (1978a) made the
following distinctions:

[In] clarifying the types of communication from the patient and the ways
in which the analyst could organize and conceptualize the material...on the
first level, a patient’s associations could be organized around their
manifest contents. This approach, which is essentially nonanalytic since it
totally rejects all notions of unconscious process and content, confines
itself to the surface of the patient’s communications.

On the second level, the analyst organizes material from the patient by
attending to the manifest associations, isolating various segments of this
material and imputing to each a specific unconscious meaning; I term
these inferences Type One derivatives. Here the manifest content is
addressed in relative isolation and the latent content—the unconscious
communication—is determined by the recognition of obvious
displacements, the use of symbols, the intuitive understanding of

www.freepsychotherapybooks.org 237
underlying meanings and a knowledge of a given patient’s communicative
idiom.

A third level of organizing the material from the patient is feasible through
the use of an adaptive context as the dynamic organizer of the patient’s
associations; this yields Type Two derivatives. The model here is that of the
day residue and the manifest dream, the latent content of which is fully
comprehended only with the knowledge of the dream’s precipitant and the
related associations....

Each adaptive context itself has both manifest and latent meanings. ... A
true understanding of the nature of an adaptive stimulus and of the
responses it evokes (associations and behaviors) is founded on the self-
knowledge of the analyst—his sensitivity to the conscious, and especially,
unconscious meanings and projections conveyed in his verbal
interventions, silences, and efforts to manage the frame.

Type Two derivatives, then, are always viewed dynamically and as

responses to adaptive stimuli. As a rule, they imply that virtually all of the
communications from the patient must, on this level, be appended or
related to the analytic interaction—those representing perceptions and
introjections as well as fantasies and distortions. At this level, many
seemingly divergent and relatively indecipherable associations accrue
significance in the light of the recognized adaptive content [pp. 562-563].

“The efforts to manage the frame” are the most important doings and sayings of the therapist
and the most telling impingements on the patient. This is the crux of Langs’ interactionist emphasis.

Consequently, the realm of the ground rules of the therapeutic situation is viewed as the prime arena
of interaction. It comes to this: Any action by the therapist or the patient intended either to make or

break the “frame,” or the ground rules, will create a most important reality stimulus. The reaction to

this stimulus is the adaptive context, expressed in an encoded form, or in Type Two derivatives.

The above defines a model of disease and a model of cure. Neurosis is an interactional or
communicative creation, and its treatment, or resolution, is also interactional and communicative. The

www.freepsychotherapybooks.org 238
treatment, is a series of ever-evolving interactions and communications, but all betray a basic pattern
of action and reaction. At all times the stimulus emanating from one person produces both a conscious
or manifest level and an unconscious or latent level of reaction in the other person. The conscious or

direct message is only the misleading surface of the communication. The true and valid level is the
latent, encoded, derivative, embedded, or hidden message. The manifest message has to be decoded,

unmasked, driven from its hiding place of disguise and exposed to yield the hidden message. All is in
the interaction and in the here and now, and therapy means decoding the latent meanings of this

interaction. All else is secondary. The most mature expression of this idea is given in Langs’ latest

systematic exposition of the communicative approach, Psychotherapy: A Basic Text (1982), and in a
recent paper (1981).

It is essential to appreciate Langs’ insistence on derivative, or encoded, communication, and the

distinction between manifest content, Type One derivatives, and Type Two derivatives and
formulations (Langs, 1981,1982). In this distinction, Langs remains rooted in the classical analytic

tradition but develops a new emphasis. The traditional way is to view neurosis as confined to one

person who relives the memory of his past in the form of symptoms, dreams, and daydreams. Langs

(1981), following Freud, refers to this as the unconscious fantasy constellation. The traditionally
oriented therapist will treat these constellations as self-contained products and apply to them, in
Langs’ (1981) words, to the

familiar avenues of affective cognitive insight, through which the nature

and effects of the unconscious fantasy constellations are interpreted to the
patient, who then affectively understands them and works through them.
This procedure, on all evidence clinically prevalent today, frees the
patient’s ego for growth and the development of relatively flexible and
adaptive resources with which to cope with and resolve intrapsychic
conflicts, and to modify pathological aspects of the unconscious fantasy
constellations.

The one-person emphasis is characterized by Langs as resting on two misleading and

interrelated approaches: (1) concentrating on the manifest content of the patient’s consciously

www.freepsychotherapybooks.org 239
expressed thoughts and the manifest themes; and (2) formulating Type One derivatives. The manifest
content approach takes the patient’s statement at face value and the analytic relationship is addressed

on its surface only (Langs, 1981). Type One derivatives, related to such conscious thoughts and themes,

are the traditional dynamic and genetic formulations applied to such material. But this is tantamount
to throwing the book at the patient. Such Type One derivatives exist in a vacuum, and they do not

become useful until activated in response to a stimulus from the therapist. The manifest content and

Type One derivative approach, furthermore, implies that “the burden of pathological inputs is placed
almost entirely on the patient and the sources of his seemingly distorted communications are seen to

reside exclusively with his own unconscious fantasy constellations” (Langs, 1981). Langs will not deny
that the manifest unconscious fantasy constellation is in itself a derivative. But here is the crucial point
of departure: Since Langs (1981) goes for interaction, he chooses to emphasize interactional, or Type

two, derivatives, over all else: “This lays the foundation for the second avenue of symptom resolution
[which] involves the object relationship between patient and analyst, the nature of their unconscious
communicative transactions and projective and introjective identifications of each.”

Once again, Langs invokes Freud’s notion that the manifest content of the dream is a disguised

edition of the latent content and of perception in external reality. In the context of the interaction, the
patient’s reaction to the therapist’s impact upon him or her is not expressed directly, but in a

derivative, that is, disguised and allusive, manner. In ordinary social intercourse, hypocrisy is more
common than honesty. Like the king’s jester, the patient disguises his or her true reaction to the
therapist. Like Pinel, who came to the rescue of the insane, Freud went out to rehabilitate the worth of

the opinions of neurotics. Similarly, Langs makes a case for the patient’s correct and astute perceptions
of the therapist’s mistakes, foibles, lies, evasions, and abuses. The rigid therapist, like the authoritarian
parent, may think he or she is above criticism and be quick to attribute the patient’s complaints to

“transference,” “sickness,” or “acting out.” The correspondingly cowed patient might talk in allusions,
or, as Langs puts it, in derivatives. Langs (1981), however, generalizes an indirectness to every patient
and every interaction:

In essence, every association and behavior by the patient is analyzed in the

www.freepsychotherapybooks.org 240
light of the stimulus or adaptive context that provoked it. Extensive
empirical evidence suggests that these precipitants are almost without
exception the silences and interventions of the analyst. ... All other stimuli,
whether from within the patient himself or from traumatic outside
relationships, are seen as secondary adaptive contexts and are, as a rule,
linked to primary adaptive contexts within the therapeutic experience. On
this level, the patient’s material is given specific organization and meaning
in the here-and-now as derivatives that must be understood in the light of
the stimulating adaptive context, a concept modelled on Freud’s
conception of the day residue for the dream.…Listening at this level
consistently addresses all manifest associations as derivatives of
unconscious contents and processes, a term of both fantasies and
perceptions [italics added]…[of] the extensive pathological
communications contained in the therapist’s and analyst’s erroneous
interventions and mismanagements of the framework. With remarkable
consistency, patients unconsciously perceive and introject the implications
of these errors. Similarly, when the analyst intervenes properly,
representations of a positive introject and Type Two derivative validation
ensue.…Making use of Bion’s discussion of lies, liars, and the thinker
(1970) we might advance the flowing postulate: truth as it pertains to the
patient’s neurosis within the psychoanalytic situation can be identified
only by taking into account the unconscious communicative interaction
between patient and analyst as this relates to the manifestations of that
neurosis on the one hand, and to the central adaptive contexts for both
patient and analyst on the other (Langs, 1980a, b). Truth must include a
recognition of introjective and projective processes, transference and non-
transference, countertransference and non-countertransference, and the
valid and disturbed functioning of both participants. Any formulation
which excludes any aspect of this totality, or which makes use of one part
of the total picture as a means of denying or excluding the rest, should be
viewed as a barrier to the truth. On this basis it becomes possible to
distinguish truth therapy from lie therapy, and to develop a
conceptualization of distinctive modes of symptom alleviation.

Langs (1982) also notes: “The distinction between Type One and Type Two derivative listening

shows the need for a basic revision in the nature of psychoanalytic listening in the direction of adaptive

www.freepsychotherapybooks.org 241
context formulations.” This claim makes the current milestone of the fruition of a seminal idea. Starting

with a reaffirmation of the importance of reality, via clinical investigations of the role of day residues in
dreams, Langs found the importance of the reality impact of the analyst. The focus on the actions of the

analyst upon the patient, on the here and now interaction between the participants in the analytic

encounter, then shaped two major areas of concern: (1) concern with the content and form of the
communication, inspired by dream psychology, such that the patient’s stream of consciousness is read

for covert allusions to his or her thoughts and feelings about the therapist, even though the patient is
overtly talking about his or her present and past life; and (2) a concern with the ground rules—the

ethical norms governing the therapist’s professional conduct.

Langs’ emphasis has brought about a transvaluation of the traditional objects of analytic
exploration. The patient’s life, past and present, life’s events and crises, and the time-hallowed

transference are all viewed as secondary to the here and now and as merely a vehicle for the patient’s
reactions to the therapist’s impact upon him or her.

Langs’ innovation is evidently of great heuristic usefulness. It raises our consciousness to the
actual and real inputs of the therapist and shakes our complacency about them. At the same time, in

spite of Lang’s repeated claims, it does not achieve a definitive degree of certainty about the intent of a
given communication. In a given moment, who is the patient really talking about—the patient or the
therapist? Entertaining a silent hypothesis about the intent, not merely content, of a communication

can only lead to presumptive, not conclusive, inferences. We are dealing with interpreting matters of

degree, intensity, accent. Such matters are in the realm of opinion. As such, they become open to debate

and create debate. Their ultimate validation is subject to developments in time—to the judgment of
history.

It is useful to invoke at this point the dialectic of content and form. This dialectic, and the

varying emphasis on now one, now the other aspect of the content-form unity, has been in evidence

from the very beginning of psychoanalysis. At first the idea predominated—the what (content) of the
communication; for example, hysteria was defined as ideogenic, an idea persisting in time. Later Freud

www.freepsychotherapybooks.org 242
discovered the how (form) of the communication—the mood of dreaming, latent and manifest content,
and free association. As dream psychology and interest in content waned, form came more and more to

the fore, first as the emphasis on manners of disguise and encoding, later as transference and especially

transference-resistance. This focus on resistance then led to two further developments: a shift of
interest from communication to conduct, from the what to the how, and from the how to the what-for—

that is, a shift from content to intent (Lothane, 1983a). Consequently, there was a greater stress on the

contract aspect of the conduct of the two people, on their interactions in the here and now rather than
on events in the there and then, and on the ethical norms governing this conduct. This development

was traversed by Freud in his movement from his works on hysteria and dreams to the 1912-15
papers on technique. Along this path Freud gave his attention to two basic sorts of form: (1) the depth-
surface, latent-manifest, straight-encoded forms of communication and conduct revealed by dream

psychology; and (2) the honest-dishonest, cooperative-resistant, love-hate, gratification-abstinence

forms of conduct reflected in the observance of the analytic contract. Langs has traced a similar course.
On the one hand, he defined what to listen to in the communication (identifying the adaptive context)

and how to listen for it (decoding derivatives). On the other hand, he has defined the ground rules (the
frame). Over the decades, analysts have debated and battled about both these aspects of the analytic

encounter.

Having given what is hoped is a balanced critical exposition of Langs’ views, I shall now proceed
to quote two reactions to Langs in the literature. To date, the orthodox analytic establishment has
ignored Langs totally. To this, the reaction of Leo Stone, is both a unique and instructive exception.

Searles has always been a maverick himself, viewed with suspicion by the orthodox.

In two dialogues of Langs with Leo Stone (Langs & Stone, 1980) and Harold Searles (Langs &
Searles, 1980), there are expressed many interesting opinions, agreements, and disagreements. In both
dialogues the disagreements are not so much in the realm of the frame of communication but in the

realm of conduct.

Stone comes across as a pillar of the orthodox analytic establishment, a man both humane and

www.freepsychotherapybooks.org 243
urbane, who believes that the situation is more important than the rules. The Sabbath is made for
human, not human for the Sabbath. He admits to having been influenced by Ferenczi (via his first

analyst) and espouses an approach to the patient marked by common sense, justice, reasonableness,

flexibility, and a modicum of gratification in the relationship. In this dialogue, Langs espouses a
fundamentalist position on ground rules and their rigid application. He takes Stone to task for giving a

patient an extra session when the patient requested pills to calm his anxiety. The patient’s subsequent

dream of the pills slightly chewed up and accompanied by an image of two worms was read by Langs
as indicating the patient’s view of the extra hour as “a dangerous contaminated gratification” (Langs &

Stone, 1980, p. 173). Stone, pressed by Langs to concede the point, defends himself by seeing Langs’
position as “Calvinist” and prohibitionist.

This brief vignette highlights the perennial problem of interpreting a record of a live text,
especially when its author is not around. The interpreter can never be certain about the exact referents

of an author’s content or intent; the interpreter can only offer a plausible hypothesis. In this case,

furthermore, the debate is not so much about the dream’s meaning as a proposition as about the dream
as the patient’s judgment of the usefulness or helpfulnesss of the analyst’s conduct. As such, it is less a

matter of logic and more a matter of love. But the canon of love differs essentially from the canon of
logic. Yes and no, true and false have different implications in love; they mean acceptance or rejection,

like or dislike, preservation or annihilation. In logic it is possible to achieve certainty a priori, before
the fact; in love, in fortunate circumstances, certainty comes a posteriori, after the fact. The truth of
love is tested in time.

Stone feels that Langs views the frame as sacrosanct, as a bed of Procrustes, as too rigid.

Certainly, an important difference in background surfaces in the dialogue: Stone speaks mainly of his
own experiences with patients, Langs of his reactions to accounts of cases by students and residents,
who, in Langs’ view, are both prone to error and vulnerable to countertransference. Such facts are

important to remember in order to understand the positions espoused. Langs also makes reference to
his experience as an analytic candidate. Both Langs and Stone concur about the “conditions of training

www.freepsychotherapybooks.org 244
analysis [as] a disastrous fact of our training…a gross modification of the analytic situation. Gross!” (pp.
18-19). Langs also expresses the view that “all analytic research is an effort to complete the unfinished
business of one’s personal analysis. The gift is to do it in a creative way” (p. 17).

The unfinished business of men in analysis with men is often the father-son relationship, a
problem for Freud and his followers and for many analysts and their analysands-students ever since. It

is endemic to the profession. The father-son dilemma is in evidence here, too. Langs’ efforts to educate

Stone to see the unconscious implications of his consciously well-intentioned behavior are met with
Stone’s temperamental query, “Are you ‘wild analyzing’ me?” (p. 286).

Perhaps the most interesting exchange between Langs and Stone is about the relationship of
reality to fantasy. Invoking the patient’s true and valid unconscious perceptiveness as manifested by
the patient’s introduction of a modification in the frame, Langs proceeds to interpret—that is, to

translate—the manifest as a derivative communication:

I would argue that the patient has actually perceived, unconsciously,

kernels of truth regarding unconscious motives within yourself for
deviating. These would be communicated indirectly, as a rule, in what he is
saying, and his response would not be totally distorted. Granted that the
therapist is not consciously involved in homosexual fantasies about the
patient, granted that he does not have conscious sexual wishes for his wife,
nonetheless, I think the patient would be entitled to feel that there are
some unresolved, unconscious, homosexual and seductive problems
within the therapist and that they were expressed through the acceptance
of the modification in the frame [pp. 284-285].

Trying such a formulation out for size on himself, Stone cannot hold back his sense of outrage: “I
think you are absolutely wrong here, due to the fact that unconscious fantasy is given preeminence and

predominance in life that is utterly unrealistic” (p. 285). Stone here and elsewhere feels that Langs is
ignoring reality. Langs’ surprise is just as poignant as Stone’s outrage: “It is odd to hear you imply that I
ignore reality when I am actually stressing it—realities of which the analyst may be unaware. You are

www.freepsychotherapybooks.org 245
addressing manifest reality; I acknowledge its presence and add latent reality, if I may use the phrase,

as well” (p. 284).

This exchange amounts to a reversal of roles. The orthodox Freudian analyst professes a
commonsense faith in the external, consensually validated reality of overt action as prior to an internal,

intrapsychic reality of a hypostatized unconscious fantasy. Langs the innovator is ultraorthodox in his

faith in an intrapsychic reality as a valid criterion for judging external reality. This contradicts his other
emphasis on day residues. The dream is a reaction to an event, not prior to an event. The dream or

daydream is not an unconscious fantasy—it is an outcropping into consciousness of unconsciously

transformed other thoughts, prior in time, which can be recalled.

But who is the proper judge of the validity of memory or of the validity of imputed motives—the

person who remembers and who avows motives, or another person with a vested interest, who listens
to the story? How can the debate between Langs and Stone be settled to satisfy the requirements of
scientific, or other, proof?

From Freud on, analysts have been tempted to consider themselves experts in the unconscious,

implying a special perceptiveness about other people’s hidden motives. This expertise betrays a hidden

authoritarianism stemming from the reality of social, economic, or other status. As authoritarian as
Freud was in his politics, he was egalitarian about “the unconscious.” In “the unconscious” we all covet,

lust, and murder, but in real life a father can say to a son: “Do as I say, don’t do as I do.” Langs has again

created a transvaluation. Whereas the traditional analyst is an expert on the analysand’s unconscious,
the innovation is to set up the patient as an expert adjudicating the analyst’s unconscious, or hidden,

motives. No wonder Stone was outraged at such a revolutionary turning of the tables.

This is also related to Langs’ pervasive skepticism toward direct and truthful human

communication. He will not take a straight yes or no for an answer. Yet, although the concept of
encoded, derivative communication squares with the human capacity to conceal truth and practice
duplicity through the use of language, it does not follow that direct communication does not exist.

www.freepsychotherapybooks.org 246
Langs has converted a potentiality into an actuality, a probability into a certainty—a consequence of

taking the notion of “the unconscious” too literally. Furthermore, the suspicion of direct

communication, if pushed too far, can be as disabling as the disregard for indirect, or derivative,

communication. Both modes of communication need to be subjected to the test of truth.

The dialogue between Langs and Searles (1980) creates a different atmosphere. Two

circumstances make for an immediate affinity between them: Searles’ political status as a maverick in
relation to the analytic establishment; and his ideological approach (inspired by the teachings of

Sullivan) that psychopathology is an interactional product, that the patient cures the doctor, and that

the doctor may himself be disturbed, or have a “psychotic core.” Searles has been known for years as a

therapist marked by originality, probity, courage, and bluntness in his dealings with some of the most
severely disturbed patients at Chestnut Lodge, Maryland, and for his numerous imaginative

contributions to the literature.

Searles is sympathetic to Langs’ position on the frame, the concept of the adaptive context and

derivative communication, and the approach of monitoring one’s behavior toward the patient rather
than chalking problems up to the patient’s transference.

The two men also share many private sentiments about the injustices of the analytic
establishment. Langs describes poignantly his dissatisfaction with his training analyst: “I think I will be

forever ungrateful and angry about the modifications in the framework of my analysis and its lasting

effects on me. And then I have to have a perspective. Such deviations have been and still are a reflection
of a shared blind spot” (Langs & Searles, 1980, p. 93). He also described his break with his institute and

society:

One of the very positive things about my alienation from my colleagues

and Institute is that it helped me to resolve a good piece of—I’ll never
resolve it entirely—but a good piece of my largely inappropriate need for
their approval, for their sanction, for their love, which had been among the
conscious motives for my work.…These needs are reflected in my

www.freepsychotherapybooks.org 247
technique books (Langs, 1973, 1974), which I wrote with my teachers at
the Institute in mind. At the time, I believed that what I wrote was true,
and I was already establishing my independence by working in ways
regarding which they openly disapproved [p. 99].

The break between Langs and the group has been complete. (A number of Langs’ former
teachers and peers, whom I approached recently for reactions to his work, declined to get involved.)

Langs’ isolation has even led him to “keep asking myself, Am I trying to be a martyr? Am I inviting all of
this condemnation? And I have absolutely decided that this is not martyrdom or masochism, but a love
of truth—yes, a dedication to fathom the truth regardless of personal cost” (p. 86). But Langs still

worries:

...On one level, I really feel that I have freed myself in many ways, but I
don’t mean to imply that it’s not still a great concern. In fact, one of the
things that disturbs me most at this time—in all honesty—is that I am still
preoccupied with just that very area. How much is my work being
accepted? When will I have my day? When will they regret it? When and
how will it all be resolved? There is something I haven’t worked through. I
know it, I am working on it. Still, I think that in terms of what I am writing
and creating now, I have become far more free of those shackles than I had
been before, in a very positive sense. I didn’t mean to imply, though, that it
doesn’t remain a kind of hurt and almost a damned obsession [p. 102].

These personal statements illustrate the ubiquitous connection between the man and his

creation in matters belonging to the sphere of thought and action. The personal equation has even
penetrated such a priori, impersonal disciplines as physics and astronomy. How much more important
is the personal element in a profession like therapy or in a discipline like psychoanalysis.

Scientific consensus, doctrinal compliance, and group loyalty are forever a vexing problem for

analysts. It is possible to be a lonely investigator in the laboratory, but a psychotherapist cannot

survive in isolation. He needs a group and a public. Freud rightly described himself as the leader of the

psychoanalytic movement. He also created the paradigm of the drama of the innovator and future

www.freepsychotherapybooks.org 248
leader—the initial experience of the revelation of truth, the revolt against an establishment, the

gathering of faithful disciples and the appearance of schismatics, the spread of the message and the

creation of a wide following, the institutionalization of the group as an organized body, and the

conversion of revealed truth into dogma.

The story of Langs’ “schism” has not been published. It was not a heresy but a manifestation of

individuality and a quest for independence. The problem is with the group, which cannot accommodate
an ideological variant in its midst, and with the individual, who craves the approval of the group but

will not sell his originality short. The docile stay and the naughty go away. As Langs, inspired by

Winnicott (1949), avows: “Whenever an analyst writes, it is an effort to complete his own analysis” (p.

48). And, he should have added, it is to resolve the business of relating to the group, to teachers,
students, and patients as well.

Give a dog a bad name and hang him, as the saying goes; such name-calling is a strategy for
maintaining group cohesion. The epithet “wild” (Chessick, 1981) is one of the mildest of those thrown

at Langs in a number of reviews of his books. On the other hand, a follower of Langs qualified the
absence of serious debate on the communicative approach in the psychoanalytic literature as
narcissistic defensiveness (Raney, 1983). Cursing enemies, excommunicating heretics, and pinning

psycho-pathological labels on opponents—the varieties of name-calling. In this case, the establishment

chose silence. Obliteration is a fate worse than excommunication.

But we are dealing with careers, not curses. Langs has gone on to create a career, to win friends
and influence people, in imitation of Freud’s example. Freud’s motto in The Interpretation of Dreams

(1900b), “Flectere si nequeo superos, Acheronta movebo” (If I cannot bend the upper gods, I shall move

the underworld) had a political correlate: Since he could not conquer the Viennese academic
establishment, he went directly to the public and created a world movement. He understood the

sociopolitics of groups and of ideologies (Lothane, 1983a). He chose to express it in the terminology of

the sciences. Like Freud, Langs has shifted from the career of therapist to a career of teacher, author,

lecturer, leader, and reformer. Freud and others published their cases; Langs has decided to refrain

www.freepsychotherapybooks.org 249
from writing about his cases. His clinical examples are from the caseloads of students in supervision.
This sociopolitical fact deserves some consideration, because it has a bearing on the evolution of his

theoretical emphases. What Einstein said about physicists applies even more so to analysts: “If you

want to find out anything from theoretical physicists about the methods they use, I advise you to stick
closely to one principle: Don’t listen to their words, fix your attention on their deeds” (quoted in Szasz,

1961, p. 2).

In Langs’ case, he has largely taught residents and young therapists. These are practitioners in

institutional settings or beginning in private practice, therapists who are relatively inexperienced and
insecure. People in institutional, as compared to entrepreneurial, settings often treat individuals who

lack social or economic independence. Both therapist and patient are at the mercy of the system; they

are not free to choose what they want or to decide policies. This situation is not unlike the fate of the
training analyst and the candidate in an analytic institute. Only the most powerful training analysts in a

system can call their own shots. Others, along with their trainees, are subject to scrutiny and pressure.

The institutional frame is a compromised one from the word go.

It is thus understandable that Langs, dealing with interactions in such settings, should have

placed such a great deal of stress on issues of frame and developed a method so heavily focused on the
therapist. To be sure, increased self-awareness and responsibility is a moral duty of every practitioner,

institutional or entrepreneurial. Free entrepreneurs are beholden only to their conscience and to

society at large, whereas those within the system are beholden to their supervisors, an obligation that
becomes a third-party infringement in the patient-therapist relationship. Ultimately, the patient bears
the consequences of the supervisor-supervisee struggle. This struggle is often irrelevant to the patient

in the system, it is relevant only to the needs of the other two players, and the patient is used as a pawn
in their game. From the outset, furthermore, it has to be decided whose agent the supervisor is going to
be, the patient’s or the therapist’s. Langs (1979) defines his position unequivocally: “The supervisor’s

commitment must be primarily to the patient in therapy and only secondary to the trainee; physicianly
responsibilities precede all else in any type of therapeutic situation. Supervisory interventions for

www.freepsychotherapybooks.org 250
which the supervisee may be unprepared are thus at times indicated, in the interests of securing for
the patient a sound therapy situation” (p. 324). Such an advocacy of the patient, a third party to the
teacher-student relationship, can result in the interests of the student being sacrificed (see Lothane,

1984b).

Langs has commented in print on his analyst but not on his supervisors at the institute or how

they affected him personally and the analyses of his patients. The amount of pressure to which he may

have been subjected can only be surmised from its reverberations in the dialogue with Searles (Langs
& Searles, 1980), where Searles expresses the following reaction:

I feel you are going to destroy me. You are starting to put the squeeze on
me. It is similar to what you did with some of those poor bastardly
therapists in the “Bipersonal Field” (Langs, 1976a) and, my God, I dread it
and I cringe and I can’t supply those answers. ...I have told many audiences
that, in my work with nonschizophrenic patients, at one or another
juncture, relatively infrequently, I express feelings with an explicitness
which is relatively commonplace in my work with schizophrenic patients;
but what determines my timing of my doing so I cannot, I can’t possibly
say (p. 124).

Searles juxtaposes his freely flowing, intuitive style with Langs’, who says of himself: “I am a

stickler on methodology” (p. 125). Searles is also critical of Langs’ stance as teacher: “…I doubt very
much that you realize how pulverizingly critical and condemnatory you are being, at least verbally, to
the therapist.…Nonverbally you’re much less unkind than your words would indicate” (p. 131).

Speaking doctor to doctor, Searles offers the following advice: “I would recommend to you that, in your
work with the therapists in your seminar, you utilize something of the same allusive subtlety that you
recommend they utilize in their work with their patients, as regards any implied acknowledgement of

the therapist’s psychopathology as it becomes revealed by their work with the patients” (p. 138). Langs
concedes the point and notes in self-defense: “And I do use discretion and modulation.…It’s a dilemma.
It is not me that disturbs the supervisee, but the patient; I am trying to be open and helpful” (p. 139).

www.freepsychotherapybooks.org 251
Langs’ teachers at the institute may have also sincerely felt that they were critical of their

student for his own good, or for the good of the patients they thought they were protecting. It is easier

to achieve consensus in medicine, where the target of treatment is the disease, not the patient. In the
field of psychological treatment there is the perennial conflict of personal vested interests. There is

also the inflated narcissism of minor differences. An interpretation may be brilliant, but also off by a
hair’s breadth; it is a matter of personal taste. But in medicine as elsewhere the dilemma has always

been: whose interests come first, the patient’s or the doctor’s?

Can beggars be choosers? The wave of consumerism that has changed the nature of the practice
of medicine and psychiatry may soon sweep through the schools that teach psychoanalysis and

psychotherapy. Students will claim that the teacher should be their advocate primarily and the

patient’s secondarily. Physicians have traditionally stuck together. Psychoanalysts have persecuted

peers and students in the interests of their own power and in the defense of their own orthodoxy. The
principle of the adaptive context will have to be applied to the teacher-student relationship. What is

the teacher’s impact on the student? What is the teacher’s hidden agenda? How truthful or deceitful is

the teacher being with the student, how exploitative?

Coming from a different direction than Stone, Searles (Langs & Searles, 1980) disagreed with

Langs’ skepticism about the ability of patients consciously to “tell me when something is quite off the
mark” (p. 98). He also found his own analyst’s self-revelations “very helpful, very helpful. It would have

been intolerably impersonal without them. A lot of it was pretty impersonal anyway; but there was

enough leaven of a person there to make it reassuring. It was very useful” (p. 42). For Langs this is
anathema. Yet, this “leaven of a person” is the leaven of love in human relation. Without it there is no

relationship. Since psychological treatment is a personal relationship, since the person is the

instrumentality of that treatment, it cannot be and grow without this leaven.

Thus posited, the problem of technique can be examined in a new perspective: what is the right

technique, what is right love, and how do the two relate to each other in the enterprise called
psychotherapy? Freud began with the phenomenon of hysteria, stumbled on the phenomenon of love,

www.freepsychotherapybooks.org 252
and invented transference in an attempt to bring law and order into love. The analysis of transference
became his definitive conception of the right technique. He did not often treat of love, but did on

occasion treat with love. Ferenczi, by contrast, emphasized love.

Freud’s abandonment of the seduction theory had momentous consequences for the
development of psychoanalytic theory and practice (Lothane, 1983b). What would have happened if he

had not abjured the seduction theory? He would have been like Ferenczi, who remained true to the

traumatic conception of neurosis and the neurosogenic effect of cruel parents on their children.

Ferenczi also advocated, according to Szasz (1965), the “abandoning of transference-analysis and,
indeed, analysis of any kind in favor of dwelling sympathetically on the patient’s past disappointments

and making heroic efforts to undo them.” Ferenczi is thus the father of the here-and-now wave: of

Horney’s and Sullivan’s emphasis on the present over the past; of Wilhelm Reich’s character analysis;
of Franz Alexander’s corrective emotional experience; of Merton Gill and Robert Langs.

Langs is between Freud and Ferenczi. He does not treat of love directly, except, in the manner of

Freud, by default. Love comes to you indirectly when you do things right, when you apply the right
technique, when you say the right words, when you express the right ideas. Direct love is as impossible

as direct consciousness of it. In his stress on the ideogenic nature of the symptom versus its affective
side, on the pathogenic nature of unconscious fantasy, on the negative value of failed communication,

Langs is like Freud. Like Freud, he also stresses interpretation. But as Freud (1933) himself saw:

The associations to the dream are not yet the latent dream thoughts.…An
association often comes to a stop precisely before the genuine dream
thought. At that point we intervene on our own; we fill in the hints, draw
undeniable conclusions, and give explicit utterance to what the patient has
only touched on in his associations. This sounds as though we allowed our
ingenuity and caprice to play with the material put at our disposal by the
dreamer and as though we misused it in order to interpret into his
utterances what cannot be interpreted from them [p. 12; italics Freud’s].

This is the perennial problem of interpretation: How do we know whether we are interpreting

www.freepsychotherapybooks.org 253
from or into? Are we not dignifying the analyst’s thoughts, the analyst’s associations, by the pretentious

title of interpretations? How do we know whether patients are alluding to the analyst or talking about

themselves? We do not know for sure. But if interpreting is nothing more than entertaining options, it

is of service in making further discoveries, subject to the judgment of history. The danger lies in
claiming premature validity for such interpretations. As the expert in the unconscious, rather than as

an observer of reality, the analyst may be tempted to engage in a kind of imperialism toward the
patient or student.

In Langs’ primary focus on the interaction, in the short shrift he gives to the notion of the

transference neurosis (“for me that is a denial-based myth” [Langs & Searles, 1980, p. 55]) pointing to

the traumatogenic behavior of the therapist, in his views on the seduction theory, Langs is more like
Ferenczi than Freud. Is it technique or love? Technology or personology? Idea or feeling? Content or

form? The choices between these pseudopolarities make up the body and soul of psychoanalysis, past

and present. Langs’ effort is a challenging link in this historical chain.  

REFERENCES

Chessick, P. (1981). Critique: The wild supervisor. American Journal of Psychotherapy, 35, 445-448.

Feldman, S. S. (1958). Blanket interpretations. Psychoanalytic Quarterly, 27, 205-216.

Freud, S. (1895). Studies on hysteria. S.E., 2, p. 12.

Freud, S. (1900a). The Interpretation of Dreams. S.E., 4-5.

Freud, S. (1900b). Die Traumdeutung. Gesammelte Werke (Vols. 2-3). London: Imago, 1942.

Freud, S. (1901). Uber den Traum. Gesammelte Werke (Vols. 2-3). London: Imago, 1942.

Freud, S. (1912). Recommendations to physicians practising psycho-analysis. S.E., 12.

www.freepsychotherapybooks.org 254
Freud, S. (1913). On beginning the treatment. S.E., 12.

Freud, S. (1914). Recollecting, repeating, and working through. S.E., 12.

Freud, S. (1915a). Observations on transference-love. S.E., 12.

Freud, S. (1915). Die Verdrangung. Gesammelte Werke (Vols. 2-3). London: Imago, 1942.

Freud, S. (1933). New Introductory Lectures on Psycho-Analysis. S.E., 22, 12.

Homans, P. (1970). Theology after Freud. New York: Irvington.

Isakower, O. (1968, 1971). Personal communications.

Klein, M., & Tribich, D. (1979). On Freud’s blindness. Colloquim, 2, 52-59.

Langs, R. (1971). Day residues, recall residues, and dreams: Reality and the psyche. Journal of the
American Psychoanalytic Association, 19, 499-523.

Langs, R. (1972). A psychoanalytic study of material from patients in psychotherapy. In R. Langs,

Technique in translation (p. 71-111). New York: Jason Aronson, 1978.

Langs, R. (1973). The technique of psychoanalytic psychotherapy (Vol. 1). New York: Jason Aronson.

Langs, R. (1974). The technique of psychoanalytic psychotherapy (Vol. 2). New York: Jason Aronson.

Langs, R. (1976). The bipersonal field. New York: Jason Aronson.

Langs, R. (1978a). The listening process. New York: Jason Aronson.

Langs, R. (1978b). Technique in transition. New York: Jason Aronson.

Langs, R. (1979). The supervisory experience. New York: Jason Aronson.

Langs, R. (1980a). Interactions: The realm of transference and countertransference. New York: Jason
Aronson.

www.freepsychotherapybooks.org 255
Langs, R. (1980b). Truth therapy, lie therapy. International Journal of Psychoanalytic Psychotherapy, 8,
3-34.

Langs, R. (1981). Modes of cure in psychoanalysis and psychoanalytic psychotherapy. International

Journal of Psycho-Analysis, 62, 199-214.

Langs, R. (1982). Psychotherapy: A basic text. New York: Jason Aronson.

Langs, R., & Searles, H. (1980). Intrapsychic and interpersonal dimensions of treatment. New York: Jason
Aronson.

Langs, R., & Stone, L. (1980). The therapeutic environment and its setting. New York: Jason Aronson.

Lothane, Z. (1980). The art of listening: A critique of Robert Langs. [Review of The Listening Process.]
Psychoanalytic Review, 67, 353-364.

Lothane, Z. (1981a). A perspective on Freud and psychoanalysis. [Review of Freud, Biologist of the
Mind, by F. J. Sulloway.] Psychoanalytic Review, 68, 348-361.

Lothane, Z. (1981b). Listening with the third ear as an instrument in psychoanalysis: The contributions
of Reik and Isakower. Psychoanalytic Review, 68, 487-503.

Lothane, Z. (1982). The psychopathology of hallucinations: A methodological analysis. British Journal of

Medical Psychology, 55, 335-348.

Lothane, Z. (1983a). Reality, dream, and trauma. Contemporary Psychoanalysis, 19, 423-443.

Lothane, Z. (1983b). Cultist phenomena in psychoanalysis. In D. Halpem (Ed.), Psychodynamic

perspectives on religion, cult and sect. Boston: John Wright/PSG.

Lothane, Z. (1984). Teaching the psychoanalytic method: Procedure and process. In L. Caligor, P. M.
Bromberg, & J. Meltzer (Eds.), Clinical perspectives on the supervision of psychoanalysis
and psychotherapy. New York: Plenum.

Masson, J. M. (1984). The assault on truth. New York: Farrar, Straus & Giroux.

www.freepsychotherapybooks.org 256
Raney, J. O. (1984). Narcissistic defensiveness and the communicative approach, pp. 465-490. In J. O.
Raney (Ed.), Listening and interpreting. New York: Jason Aronson.

Rapaport, D. (1960). The structure of psychoanalytic theory. Psychological Issues, 2, (2, Monograph 6).

Reich, W. (1949). Character analysis. New York: Noonday Press.

Sulloway, F. J. (1979). Freud: Biologist of the mind. New York: Basic Books.

Swales, P. (1982). Freud, Johann Weier and the status of seduction: The role of the witch in the
conception of fantasy. The Sciences, 22, 21-25.

Szasz, T. (1961). The myth of mental illness. New York: Harper.

Szasz, T. (1965). The ethics of psychoanalysis. New York: Basic Books.

Winnicott, D. W. (1949). Hate in the counter-transference. International Journal of Psycho-Analysis, 30,

69-75.

www.freepsychotherapybooks.org 257
8
HEINZ KOHUT: BEYOND THE PLEASURE
PRINCIPLE, CONTRIBUTIONS TO
PSYCHOANALYSIS

HYMAN L. MUSLIN, M.D.

Heinz Kohut, the founder of the psychology of the self, died on October 8, 1981. He had come a

long way in developing a theory of the mind which, starting with a relatively modest addition to the

psychoanalytic structural model of the mind, evolved into a totally unique approach to the problems of
modern humanity. Kohut, like Freud was a conquistador in many areas of people’s reactions to their

surrounds. In some areas, his contributions were well worked out, for example in his systematized

work on the crucial developmental issues that lead to either a cohesive self or a self vulnerable to
fragmentation. Kohut’s contributions to psychopathology have also been neatly systematized. Perhaps

his views on the theory of cure in analysis, including his systematization of the transferences and their

role in analytic cure, represent the most compelling of his contributions.

Kohut’s insistence on prolonged empathic immersion into the experience of the patient—away

from external behaviors and preformed theories including theories of self psychology—is perhaps, of

all his contributions the central one. Other aspects of the theory and practise of the psychology of the
self are in need of further elaboration and research. It remains a truism, however, that Kohut’s

www.freepsychotherapybooks.org 258
discoveries and formulations have been a major force in liberating the field of psychoanalysis from the
shackles of insistence on attention to the vicissitudes of the drives and their defenses, especially the

insistence on the oedipal complex as the inevitable pathogenic force for human beings.

Heinz Kohut came to the University of Chicago from Vienna as a neurologist after World War II.
He then began his training in psychiatry and psychoanalysis, although he had been analyzed in Vienna

with August Aichorn. Shortly after his graduation from the Chicago Institute for Psychoanalysis, he

joined the staff at the institute and began his lifelong career in teaching and research in psychoanalysis.

He also continued his affiliation with the University of Chicago as a professor of psychiatry and
lectured there regularly. Kohut’s active participation in the local and national psychoanalytic

community culminated in his election to the presidency of the American Psychoanalytic Association in

1966.

Prior to his first formal paper on narcissistic issues, Kohut wrote on a broad range of
psychoanalytic topics, including empathy and introspection (1959) and psychological reactions to

music. In 1966, with the paper “Forms and Transformation of Narcissism,” Kohut began his total
involvement with the understanding of the self, its development, anatomy, and psychopathology, and

the treatment of the disorders of the self. From 1966 to 1977, Kohut focused on the self as a structure
within the ego—the self that can be recognized within the classical psychoanalytic structural model. He

delineated special psychopathological disorders of the self, as differentiated from the so-called

structural neuroses in which the etiologic variables deal with drives and their conflicts—the
transference neuroses commonly based on the inadequate resolution of the oedipal complex. Kohut
emphasized that the self and its charge of energy, narcissism, should be recognized without bias as an

important entity, separate from the other aspects of the intrapsychic world. Thus, although object
relationships are important, narcissistic interests are of special value in the psyche and, in fact,
narcissism has its own line of development. The next stage in Kohut’s thinking was to delineate the

treatment issues of the disorders of the self, particularly the special transferences he discovered in
people suffering with treatable self disorders.

www.freepsychotherapybooks.org 259
In his 1977 work, The Restoration of the Self, Kohut delineated the self as the “center of the

psychological universe” and the maintenance of its cohesion as the essential ingredient of mental

health. What Kohut terms “Tragic Man”—the individual preoccupied with gaining succor for his or her
depleted self—here replaces “Guilty Man”—the individual preoccupied with the avoidance of oedipal

guilt—as the central problem in Western civilization. The search for esteem, from early life through
death, through the medium of the self-selfobject dyad, replaces anxiety as the central feature of

humankind. Adequate esteem leads to a life of joy, not a life based on the taming of drives. In fact, as

Kohut teaches, drives and their vicissitudes emerge as a central feature of the individual only when the
self breaks down and these drives are recognized as disintegration products.

Kohut’s contributions to psychoanalysis were in the area of the self and its vicissitudes, what he

referred to as “…the phenomena that lie within the area ‘beyond the pleasure principle’ ” (Kohut in

Ornstein, 1978 p. 752). Starting from his initial investigations pertaining to the empathic investigation
of the self (Kohut, 1959), Kohut staked out his arena of concentration as being centered on what he

called “Tragic Man”, the conceptualization of the individual as blocked in his attempt to achieve self-

realization. This version of man is at great distance from Freud’s version of the individual in conflict
over his or her pleasure-seeking drives, the so-called Guilty Man (Kohut, 1971).

Kohut’s investigations into the inner mental life of human beings ultimately encompassed a
theory of the developing self. This became his model of the mind, a theory of psychopathology, a new

approach to the therapies of self disorders and neuroses, and a new version of the essence of the

outcome of psychoanalytic therapies. Although Kohut’s contributions extended into every facet of
psychology and offer new explanations for the distresses of the modem individual, Kohut repeatedly

emphasized that he placed the psychology of the self in the mainstream of psychoanalysis and that he

wished to maintain “the continuity of psychoanalysis” (Kohut 1977, p. 172). He certainly added a new

emphasis, however, by insisting that “...psychoanalysis is a psychology of complex mental states which
with the aid of the perservering introspective—empathic immersion of the observer into the inner

mental life of man, gathers its data in order to explain them” (Kohut 1977, p. 302). For Kohut, then, it is

www.freepsychotherapybooks.org 260
not transference and resistance but empathy that defines the essence of psychoanalysis. Scientific
empathy, as the indispensable tool of the investigator of the inner mental life of humanity at once
defines the field of observations and allows for an adaptation of theories and explanations in

accordance with the data obtained by empathic cognition.

Although an appreciation of Kohut’s contributions to depth psychology must encompass his

theories of the development of the self and his views on the pathologies of the self and the

psychoanalytic treatment of these disorders, it is also important to recognize in Kohut’s work his
conviction that the psychology of the self had important relevance to fields outside of mental illness

and health. As Kohut remarked in his exchange of letters with Erich Heller (Heller & Kohut, 1978):

Whatever their limitations and shortcomings, I know not only that the
psychology of the self explains more meaningfully certain areas of man’s
psychological experiences in mental illness and health than previous
scientific approaches but also that its formulations can be more relevantly
applied outside the field of normal and abnormal psychology. The
explanations of the psychology of the self are in particular able to
encompass the significance of man’s scientific, religious, philosophical and
artistic activities [p. 449-450].

It is my intention in this essay to offer a view of Kohut’s notions of the developing self, the

pathogenesis of self disorders and the treatment of the developing self, emphases of self psychology.
Readers must answer for themselves the question asked by some critics of self psychology: Can an
approach be called “psychoanalytic” if it does not subscribe to the primacy of the drives and especially

the Oedipus complex in neurogenesis? Can self psychology be regarded as offering a psychoanalytic
view if it considers the outcome of psychoanalysis as essentially an impetus to the development of a
stunted self rather than—as classical psychoanalysis would say—the resolution of transferences

centered on the oedipal conflicts of incest and parricide?

If one accepts the Kohutian definition of psychoanalysis as a depth psychology whose areas are

limited only by the limitations of empathic cognitions, self psychology is in the mainstream of

www.freepsychotherapybooks.org 261
psychoanalysis. On the other hand, if psychoanalysis is to be considered as a conflict psychology of

drives versus the restraining and taming forces, self psychology is not an addition to the theory of

psychoanalysis. Its views would then constitute a new school of psychology. The goals of self-

psychology analyses are reached when patients are enabled to seek out and invest appropriate self
objects for the sustenance of their now cohesive selves. This statement of the end point of an analysis

conducted to rehabilitate the self focuses immediately on the significant differences between self
psychology and classical psychoanalysis. Classical psychoanalysis is concerned with the resolution of

conflicts that are purported to be the instigators of the symptoms of neurotic distress. Other end points

of a classical analysis are reached when the consciousness of ego is expanded through insight, when
the patiient’s drives are tamed, and when the Oedipus complex is resolved, with its attendant features

of castration anxiety and excess guilt diminished. Finally, the patient at the end of a classical analysis is

understood to have moved from dependency to autonomy and from narcissism to object love. Thus,
classical analysis holds that an analysis is complete when the symptoms of the pathogenic conflicts are

ameliorated, especially the castration anxiety and the hypertrophied guilt, and when the pathogenic

complexes have become conscious, especially the persisting conflicts centering on the Oedipus
complex, which has been reenacted and become the central focus of the transference drama. The

analyst and the patient, in their constant preoccupation with the manifestations of the archaic oedipal
phenomena, work on bringing to consciousness the buried pathogenic fears (Freud 1917a).

Self psychology holds that an analysis is complete when the self, formerly underdeveloped

through fixations on archaic self-selfobject relationships, is provided with a therapeutic atmosphere in

order to complete its development. The patient will be cured when his or her self is cohesive, when he

or she has achieved sufficient structure from the development-enhancing psychoanalysis to reveal the

activities emanating from a firm self. As Kohut (1977) stated: “Within the framework of the psychology
of the self, we define mental health not only as freedom from the neurotic symptoms and inhibitions

that interfere with the functions of a mental apparatus involved in loving and working, but also as the

capacity of a firm self to avail itself of the talents and skills at an individual’s disposal, enabling him to
love and work successfully” (p. 284).

www.freepsychotherapybooks.org 262
Thus, from the outset, the classical position concerns itself with the fate of the drives—their

conflicts, their resolutions through recreation in the transference and their subsequent working

through, especially of oedipal conflicts and oedipal transferences. Self psychology is concerned with
the integrity of the self. If the self is healthy, drives are not experienced as isolated phenomena and no

pathological conflicts involving drives would then ensue. The analyst, in this view, is occupied with
rehabilitating a self that has become fixated for its sustenance on archaic measures that are

unsuccessful. The result is an enfeebled self, unable to engage in life with vigor in a goal-directed

fashion. The analysis in this view is also concerned with the establishment of transferences and
interpretation, but they are directed at promoting the development of the self. In self-psychology

analysis, the essence of the cure lies in the establishment and resolution (“re-solution”) of selfobject
transferences, each of which replicates an archaic selfobject relationship that has resulted in a fixation
of developmental strivings. The modal psychoanalytic regression reactivates the pathogenic selfobject

transference at the point where the self object functions required for development of the self were
deficient. As will be discussed later in detail, the patient’s stunted self now resumes development of the
particular functions that were inadequately internalized through the failures of the selfobjects in the

surround. The transferences that are established reflect the analysand’s fixations on the point in
psychological time when development ceased, ushering in, for the patient, the never-ending search to
resurrect that particular selfobject from whom the patient tries again and again to obtain the necessary

mirroring or power merging and make it into his or her own. Once the pathognomonic transferences
are established, the mirroring or other selfobject functions are initiated. The patient begins to resume
the development of the missing or defective self functions through transmuting internalizations. The

process of internalization is set in motion through the optimal frustration of analysis which ultimately
intensifies the imagos of the analyst’s selfobject functions to the point of causing a permanent addition
of selfobject functioning to adhere to the patient’s self, thus, for example, eventuating in a movement

from admiration of the mirroring selfobject to self-admiration.

We now turn to the significant aspect of the development of the self and the development and

maintenance of the self-selfobject relationships, which self psychology holds are at the core of

www.freepsychotherapybooks.org 263
psychological life.

THE DEVELOPMENT OF THE SELF

In Kohut’s (1977) view, the self is the center of the psychological universe, by which he meant

that people can only be understood in terms of their experiences—their inner mental life—not their
behavior (Kohut, 1959). It follows from this that any genuine investigation of man must be through the
medium of empathy—vicarious introspection—which therefore defines and restricts the observational

field of psychological understanding. Kohut’s last statement about empathy was that it is to be
understood as the capacity to think oneself into the inner life of another person (personal
communication, 1981).

As we will see in more detail, for Kohut (1979):

The self is the core of our personality. It has various constituents which we
acquire in the interplay with those persons in our earliest childhood
environment whom we experience as selfobjects. A firm self, resulting
from optimal interactions between the child and his selfobjects is made up
of the three major constituents: (1) one pole from which emanates the
basic strivings for power and success; (2) another pole that harbors the
basic idealized goals; and (3) an intermediate area of basic talents and
skills that are activated by the tension arc that establishes itself between
ambitions and ideals [p. 11].

Moreover, the self experience has a line of development as separate from the experience of
single body parts and single functions. As Kohut (1974) comments: “The child’s self experience arises

separately, increasing in importance as it develops next to and more and more above his experience of
body parts and single functions. And finally, the child reaches a stage in which the progressively tamed
experience of single parts and functions has become related to the total experience of a cohesive self—
the parts in other words do not build up the self, they become built into it” (p. 749).

www.freepsychotherapybooks.org 264
The complete self is a supraordinate structure, which functions not only as the receiver of

impressions derived from the environment but as the center of action. It is experienced as continuous

in space and time, as a cohesive entity. The so-called bipolar self can be further identified in terms of its
major constituents: the poles of ideals and ambitions and the intermediate area of talents and skills.

These poles of the self come into their final form through interaction with the significant persons in
infancy and childhood who serve as the instigators of these self functions.

The development of the pole of ambitions is initiated as a result of special activities of the

parent, who functions as an admirer, approver, or echoer of the unfolding self and thus offers to the
child an experience of unquestioning confirmation of the child’s worth. From the point of view of the

child, this parent is experienced as an entity over whom the child has total control—much as one

controls various parts of one’s body—thus the designation “selfobject,” or in this case, the “mirroring

selfobject.” These early relationships are experienced as fusions or mergers—or, psychologically

speaking, immersions—into the body and mind of the caretaking selfobject. Establishing the archaic

selfselfobject mirroring dyad is crucial for psychological life. For structure-building to take place,

however, the self-aggrandizing mirror functions must be interiorized or internalized—actually added to

the contents of the self—so that self-esteem, an intrapsychic function, replaces selfobject mirroring, an

interpersonal activity. In Kohut’s view, internalization of selfobject mirroring functions takes place

along the lines first articulated by Freud (1917b) in Mourning and Melancholia, in which the mourner’s
unique reaction to loss—internalization of significant aspects of the departed person—is seen as a

ubiquitous reaction to separation.

At about the same time in an infant’s development as the establishment of the mirroring

self/selfobject, the second major influence on self development occurs—the establishment of the

idealizing parental imago selfobject. Whereas the mirroring selfobjects respond to and confirm the

infant’s grandiosity, the idealized parent imago are figures whom the child looks up to and merges with
as an imago of calmness, soothing, perfection and thus a source of strength. One other early

selfselfobject experience is ordinarily present in the child’s ontogeny. This is the experience of what

www.freepsychotherapybooks.org 265
Kohut (1977) called the alter ego—a twinship merger in which the child experiences the parental self
as essentially the same as the child’s own. This essential sameness is instrumental in enhancing the
child’s skills and unfolding the child’s talents.

The next phase of the child’s development is significant in the formation of the cohesive self.
This is the internalization of the self objects’ functions of initiating and promoting esteem, so that what

was a feature of the self-selfobject relationship now becomes a set of self functions. Kohut describes

the interiorization of these functions as occurring in two steps: (1) optimal frustration and (2)
transmuting internalization. Optimal frustration refers to the unavoidable disappointments in child

rearing, so that the child does not obtain the instant feedback that he or she may be demanding. These

unavoidable delays, absences, and misappreciations are not protracted or in any way traumatic—thus

they are optimal frustrations. They promote the internalization of the mirroring or other selfobject
functions, so that the mirroring selfobject’s approval is attached, so to say, to the child’s self as a

permanent source of nurturance (Kohut, 1971). Over time, the sequence of optimal frustrations leading

to transmuting internalization creates a cohesive self. This structure is bipolar in its psychological
shape, the archaic grandiosity transformed into the pole of ambition, and the internalized archaic

idealizations transformed into the pole of ideals. In this early self, which can now be labeled the

nuclear self, the pole of ambitions strives to live up to the pole of ideals through the talents and skills of
the self. In fact, in the adult, the cohesion of the self is maintained through the tension arc created by

the pole of ambitions striving to live up to the ideals through the exertions of the talents and skills in
what Kohut (1977) called a program of action: “With the term tension arc,…I am referring to the
abiding flow of actual psychological activity that establishes itself between the two poles of the self; i.e.,

a person’s basic pursuits towards which he is driven by his ambitions and led by his ideals” (p. 180).

The bipolar self now experienced by the child as continuous in time and discrete in space
maintains its cohesiveness—its resistance to breakup (fragmentation)—through two sources of self
cement. One is the pool of endogenous stores of self support derived from the internalized functions of

selfobjects to maintain self-esteem. The other is the continuing need for selfobjects throughout life.

www.freepsychotherapybooks.org 266
Kohut found that self-selfobject relationships form the essence of psychological life from birth

to death. The nature of this relationship, however, changes over time and in functioning. The earliest

self-selfobject contacts, as previously noted, are actually merging types of relationships. They instill in
the child, after optimal frustration, the supplies of esteem. From the archaic selfobject relationships,

there is a developmental line of self-selfobject encounters to what is called the mature selfobject
relationships. These offer an experience of empathic resonance—the admiration of a colleague through

which the adult self can experience a revival of the memory traces of the archaic self object’s mirroring

or calming and soothing, and in this manner restore disequilibrium due to a temporary flagging of
one’s esteem. Throughout the individual’s development, the self requires selfobject refueling to

maintain its integrity. At times, these selfobject encounters will approach the approving, admiring,
calming, merging interactions of the archaic self-selfobject fusions.

Thus, in the so-called anal stage of development, the child’s need for the mirroring responses of
the selfobject parent are necessary for the child’s toilet-training accomplishments to be given value. In

the oedipal phase of development, the child’s selfobject requirement of the parents are that they

respond to his or her increased assertiveness in the sexual and other spheres with admiration and
pride at the vigor and creativeness displayed. The selfobject encounters in these early stages of

development, although not of the earlier, archaic types, still continue to provide supportive

experiences that will be interiorized and serve to enhance the achievement of the youngster in his or
her development. The adolescent’s need for the mirroring selfobject parent to give credence to his or

her creative activities is well known, as is the intensity of the adolescent’s need for intimate contact

with an idealized selfobject. In both these instances, internalization of selfobject functioning is again

affected. In later life, necessary refueling of one’s worth is provided through mature self object
encounters and the phenomenon of empathic resonance. In the senium for example, mirroring of one’s

achievements, of one’s courage in the face of death is necessary.

In sum, the self is maintained in a cohesive manner through the strength of its constituents, the

firm sense of assertiveness, the intact sense of one’s values serving as a compass through life, and the

www.freepsychotherapybooks.org 267
ability to exert one’s skills and talents in the pursuit of one’s programs of actions, from writing a
speech to caring for the disabled. Selves differ considerably in the relative weakness or strengths of
their constituents. There are selves that are firm or enfeebled, resistant to fragmentation (cohesive) or

highly vulnerable to losses of worth and thus prone to fragmentation. Charismatic selves are firm in
the pole of assertiveness, whereas messianic selves are extremely leadership oriented. Some selves are

mirror hungry, while others are chronically searching for a leader (Kohut & Wolf, 1978).

PATHOLOGY OF SELF OR SELF-DISORDERS

The position of self psychology with regard to psychopathology is that all forms of
psychopathology are ultimately derived from defects in the overall structure of the self or from

distortions of the self. Both of these are due to disturbances of self-selfobject relationships in

childhood. Self psychology further asserts, in contrast to classical analysis, that conflicts in the object-
instinctual realm—the realm of object love and object hate, in particular the set of conflicts called the

Oedipus complex—are not the cause of psychopathology, but its results.

As we previously have seen, in adult life as well as in childhood, the cohesiveness or harmony or
fragmentation of the self—whether it is enfeebled, distorted, or firm—is a result of the success or

failure of the archaic self-selfobject relationships. A failure in the self-selfobject relationships in

childhood or adult life leads to the painful experience of fragmentation. Fragmentation, in the view of
self psychology, is the central pathologic experience of breakdown of the self. It is ushered in by a

massive loss of self-esteem, followed immediately by the advent of the global anxiety referred to as
“disintegration anxiety.” Directly after the advent of disintegration anxiety, the self is experienced as
losing its cohesiveness, with the usual experience of splitting or fragmentation of the self functions and

self perception, including reality testing, memory, and orientation in space and time. There is also loss
of the intact experience of self observing; the various experiences of the different organs previously

coalesced together in the intact experience of the total bodyself are now experienced as separate and

become focuses for enhanced attention and even preoccupation (hypochondria). In addition, the

www.freepsychotherapybooks.org 268
patient is in the throes of a separation reaction, with its attendant features of loss of vigor, esteem, and
meaning in life. Finally, a failure in a self-selfobject encounter will commonly lead to a unique rage
reaction. This so-called narcissistic rage reaction represents the reaction to the loss of control of the

selfobject. The individual will vent destructiveness on anyone in the immediate surround (Kohut,
1971).

A self-selfobject failure in childhood has different consequences from a self-selfobject failure in

adult life. In adult life, the cohesive self has continuing mature selfobject encounters, which are of value
in maintaining continuing support to the self through empathic resonance—that is, by supplying

mirroring or firmness to add to the cohesiveness of the self. A failed self object encounter in an adult

with a cohesive self will ordinarily lead to a transitory fragmentation, with hypochondria, loss of

esteem, temporary interference in mentation, and so forth.

In childhood, a failed self-selfobject relationship is of a different order. A massive or chronic

failure during the phases of childhood when the self is unfolding may result in a fragmentation that will

eventually be resolved—that is, the self will reconstitute itself and the fragmentation will subside—but

the self will now have permanent alterations. The overall experience of the self will be that of a self
chronically low in energy, a self depleted of vigor without evidence of the experience of joy. This self

will react strongly to criticism and failures by becoming more withdrawn or, at times, caught up in the
explosion of a narcissistic rage reaction. Depending on the specific type of selfobject failures, the

resultant self distortion may be that of a self weakened in the pole of assertiveness, in the pole of

ideals, or in the area of talents and skills. These defects will of course lead to the absence of formulated
programs of action in life, for example, of educational, athletic, or musical pursuits.

The overall result of such self-selfobject failures may be a self that experiences life as empty and

that is constantly in the throes of loneliness. Despite this loneliness and a desire for human encounters,
this self may be quite resistant to such encounters and may maintain a conscious attitude of

haughtiness and isolation. At times, this self may attempt to gain support for self-esteem through a

variety of activities designed to lessen the chronic emptiness such as compulsive homo- or

www.freepsychotherapybooks.org 269
heterosexuality, addiction to compounds to provide calming experiences, or compulsive episodes of
stealing.

At other times, selfobject failures in childhood eventuate in what appears to be a syndrome of

neurosis. These reactions occur when, after a failed self object encounter in a particular phase of
childhood, the child becomes preoccupied with the drive or developmental task specific to the phase,

ultimately leading to a fixation on that drive or developmental task and leaving the child permanently

preoccupied with the fears of that phase in life, which were never allayed. Thus, an oedipal fixation or

an anal fixation represents a failed self-selfobject relationship in the corresponding developmental era
of childhood. The secondary elaborations of the breakdown of the self during those times in childhood

when developmental tasks need to be mastered involve an exaggerated focus on the drive currently of

concern and defenses elaborated in an attempt to ameliorate or repress the exaggerated drive
fragments. When the cohesive self breaks down or becomes fragmented, in response to a self-selfobject

rupture, it may take one of several pathways. The fragmented self may maintain a state of chronic

fragmentation (protracted fragmentation disorders, borderline personalities); the fragmented self may
repair itself without evidence of the previous state of breakdown (episodic fragmentation); the

fragmented self may reequilibrate itself with newly developed defenses against selfobject bonds
(narcissistic personality disorders); or the fragmented self may focus on the drives that are salient in

the current developmental phase or have been activated as a manifestation of a regressive reaction
(neurotic syndromes) and may secondarily develop defenses against the egress of the specifically
elaborated drives (Kohut, 1971, 1977).

EPISODIC FRAGMENTATION DISORDERS

Reactions to a breakdown in self-selfobject bonds are, of course, ubiquitous, since self-selfobject

bonds and failures are ubiquitous. As has been described, selfobject involvements range from archaic
self-selfobject ties that continue over time to so-called mature selfobject encounters. In adults, the need
to enter into an archaic self-selfobject bond is limited to instances in which the self is subjected to

www.freepsychotherapybooks.org 270
psychological trauma requiring a temporary merging relationship. These are, of course, instances in
which the self is suddenly devoid of narcissistic supplies and is in need of the experience of the fusion
with a mirroring selfobject or a revered leader. Archaic self-selfobject bonds always serve to invest the

self with the experience of worth, of strength, of calming and soothing. In childhood, these experiences
give the self the requisite strength of cohesion; in adulthood, when entered into temporarily in reaction

to the stress of dissolution, they effect a repair to a fragmenting self. Mature selfobject encounters are

entered into when the self is in need of a temporary enhancement of esteem, that is, in a situation of
esteem-deficiency such as is the innumerable states of self-doubt individuals experience. In the mature

selfobject encounter, the self’s experience of the self object is in actuality not that of an object fused

with one’s self and under one’s control; rather the self has a reactivation of the early self-selfobject
mergers and experiences a state of esteem enhancement, thus effecting a repair of the self’s cohesion.

Seen in this way, much of adult interactional life consists of mature self object encounters with others
who function temporarily to repair a flagging self-esteem or symbolic encounters with music or
literature in which the self is uplifted or invigorated.

Thus, episodic fragmentations or near fragmentations or simple instances of loss of esteem or

threatened loss of worth are part of one’s modal reactions to a complex world of victories, near misses,

and failures. In a more or less cohesive self, the repair in most instances will be effected by entering
into a mature self-selfobject encounter. In those instances where the demands for cohesion are intense,

the previously cohesive self will fragment, albeit temporarily and seek out an archaic self-selfobject
encounter in which a merger will be effected. For example, in the case of a person who has just been
informed that his or her longstanding state of weakness is due to a malignancy in the colon, the

psychological reactions are frequently the self experience of fragmentation. This distress, one hopes,
will be followed by the self-selfobject merger effected with a trusted caretaker or relative. In such
situations, if empathic caretakers recognize the manifestations of the fragmentation and respond

appropriately with a dose of mirroring or allow themselves to become the target for idealization, the
fragmentation experience will be short-lived.

www.freepsychotherapybooks.org 271
Self-fragmentation Resulting in Neurotic Syndromes. In the view of self psychology, drives come

into focus when the self is fragmenting-thus the statement that drives are disintegration products of a

fragmenting self (Kohut, 1977). In this light, consider the self of the oedipal-phase child and the
selfobject needs of his or her emerging phase-specific assertiveness, including the child’s sexual

assertiveness of a homoerotic and heteroerotic nature (with hostility toward the parent of the opposite
sex). If the selfobject supports are missing or inadequate and the child experiences the parents’

withdrawal or rejection during this important phase in development, the self depletion will result in a

fragmented self. Thus, in some instances, the result will be not an eruption of undirected narcissistic
rage, but an egress of animus unleashed when a selfobject has failed in its functions—a preoccupation

with the drives derailed from the now-fragmented self. In the ordinary functions of the self, the drives
are a vital part of the self, seeking and maintaining contact with the world, including the world of
selfobjects. In a fragmented self, the drives are now in a free state and clearly visible since they are not

bound up with the functions of the cohesive self.

The unleashed phase-specific drives of the oedipal child whose self is now in a fragmented

condition will eventuate in repetitive experiences of anxiety, centering on tissue destruction—the so-
called castration anxiety, with its attendant features of anxiety dreams of mutilation—and the buildup

of irrational guilt. If, however, the child in the oedipal phase becomes the recipient of helpful selfobject

supports, he or she will emerge from this normal phase of development with heteroerotic and
homoerotic strivings and a minimum of guilt and castration anxiety.

Thus, in contradistinction to classical psychoanalysis, self psychology does not regard the
oedipal phase as “the pivotal point regarding the fate of the self that it is with regard to the formation

of the psychic apparatus” (Kohut 1977, p. 240). The so-called neurotic syndromes, which in classical

psychoanalysis emerge from the predetermined unfolding of the drives coming into intense conflict

with ego defenses and superego, are conceptualized in self psychology as only one of the possible
outcomes of a self in fragmentation. Self psychology holds that if the self is intact, there will be no

preoccupation with the drives in an isolated fashion. Thus, from the viewpoint of the self psychologist,

www.freepsychotherapybooks.org 272
although an oedipal phase of development is ubiquitous, if there is an adequate set of selfobjects, the
child emerges with a firming up of assertiveness, now more adequately controlled, and a firming up of
the gender experience. Conversely, if there has been a selfobject failure to the modal egress of

assertiveness in an oedipal youngster, the derailed (unattached) instinctual drives will emerge as
naked lust and hostility.

The Narcissistic Personality and Behavior Disorders. When self-selfobject failures during the

phase of the early development of the self are protracted, they result in a variety of self disorders.
These are the narcissistic personality disorders and their acting-out varieties, the narcissistic behavior

disorders. They ordinarily result from the failure of the functioning of the mirroring selfobject and the

inability of the idealized parent to compensate for the primary selfobject failure (Kohut 1977). The

cohesiveness of the resultant total self is defective, and both poles of the self are inadequately filled.
This self is vulnerable to fragmentation, especially in relation to further losses of esteem from its

milieu. The self experience is commonly a reflection of the diminutive poles of assertiveness and ideals

—that is, emptiness and/or loneliness. However, the needs of the self for mirroring or leadership are
commonly defended against by attitudes of haughtiness and superciliousness, reflecting anxiety about

allowing any further self object encounters to transpire. Another common experience in persons with

these disorders is to become immersed in transitory relationships in which an archaic self-selfobject

dyad is formed and then rejected, ordinarily out of a mixture of anticipated psychic pain and

disappointment because the relationship cannot offer them the longed-for childhood gratification.
Fragmentation states commonly lead to intense loss of esteem—the so-called empty depression,
without prominent guilt.

Other common features of the fragmentation states are the experience of disintegration anxiety
—an anxiety state marked by panicky feelings, dissociations, and end-of-the world sensations—
followed by mentational dysfunctioning (memory loss, reality-testing deficits, loss of synthesizing, and
derailing of associations), and hypochondriasis. Hypochondriasis in fragmentation states reflects the

state of the “unglued” self. Although the ordinary experience of a single organ or anatomical part is

www.freepsychotherapybooks.org 273
minimal in a cohesive self, when the self is fragmentating, a particular organ percept in the self that is
now functionally split off from the rest of the self may suddenly be experienced in a highly charged
fashion. A patient in the middle of a fragmentation reaction may complain of unusual body feelings and

localize it to an awareness that her or his face, nose, or abdomen is now experienced quite differently.
It may seem too large or too prominent. These experiences reflect the body percepts becoming split off

and, for the first time, prominent in the patient’s awareness. Patients with narcissistic personality

disorders at times exhibit behavior that expresses their reactions to insult or their needs for claiming
and soothing or mirroring. These narcissistic behavior disorders encompass the behavior of the

compulsive homosexual, the addict, and delinquents who steal as a symbolic expression of the self

need for a gift from the selfobject. Those addicts who experience the compound and the effects of the
compound as an aid to calming and soothing are clearly demonstrating and gratifying archaic self

needs, as are those homosexuals who feel mirrored in frantically sought out episodes of fellatio.
Patients who suffer with narcissistic personality disorders do not experience protracted fragmentation
states. Their fragmentation is transitory, and they ordinarily seek relief in complaining of their

experience of isolation and inability to form and maintain human relationships.

In sum, patients with these self disorders have had failures in their self-selfobject relationships

early in life. In effect, their self development is fixated, and thus they continue—albeit, unconsciously—
to effect repeated archaic self-selfobject bonds. This is to no avail, however, since they will shortly

reject these relationships. The failure of adequate internalization of the self in these patients leads to
their vulnerability to fragmentation states. This is resolved in these patients by the self’s capacity to
erect firm defenses against the egress of its desires for empathic understanding and gratification.

PROTRACTED FRAGMENTATION STATES

Patients with borderline disorders and psychoses of all kinds, demonstrate not only a
heightened vulnerability to self fragmentation but a protracted quality to their fragmentation. When a
so-called borderline patient develops a fragmentation state, which is followed by reality-testing loss

www.freepsychotherapybooks.org 274
(psychosis), derailing, and other symptoms of an acute psychotic decompensation, these pathological
states may persist for a long time. Moreover, these patients do not have an adequate capacity to form a
therapeutic self-selfobject dyad based on an alliance of effort to appreciate their inner mental life.

These patients commonly experience an absence of as-if transference phenomena. They commonly
develop a transference psychosis, insisting that the therapist feels this or that and now wishes to cause

the patient harm. To repeat, chronic, protracted fragmentation disorders represent the end point of a

massive failure in the selfobjects in these people’s lives. Due to the failed selfobject functioning, these
people cannot form alliances to investigate themselves because they do not have adequately developed

functions of self observation (Kohut, 1977).

In summary, the central teaching of Kohut on the psychopathological syndromes is that all

forms of psychopathology are due to disturbances of self-selfobject relationships, which result in

structural defects in the self and render that self vulnerable to fragmentation and its vicissitudes.

Whereas Freud’s model of the mind—the model of structural theory—led to erupting instinctual

derivatives coming into conflict with the superego and ego and leading to new defenses (neurotic
symptoms), Kohut teaches that one must empathize with a self that is fragmented due to a current

deficit of cohesiveness brought about by loss of esteem from whatever source. The model of classical

psychoanalysis holds that psychopathological syndromes begins with a psyche in conflict and
therefore in a state of anxiety. Should this conflict become protracted, the initial signal anxiety will

intensify to massive anxiety and there will be a neurotic breakdown. Directly after this event, the
psyche develops new symptoms and the offending drive is rerepressed, the psyche becoming once
again calm (Freud, 1926). The Kohutian model, in contrast, focuses on the self in fragmentation as the

initial manifestation of psychic disequilibrium, which may lead to an episodic fragmentation; a chronic
fragmentation; the syndrome of repression of the self’s needs, defended by attitudes of haughtiness
and superciliousness; or the neurosis that represents the psyche focused on the drives, which are

disintegration products of the fragmenting self.

Classical psychoanalysis holds that the Oedipus complex and its resolution or lack of same are

www.freepsychotherapybooks.org 275
the central instigators of neurosis and character disorders. Kohut’s view, as has been described, is that
if the parents function as supporters of their children’s assertiveness, there will be no castration
anxiety over “malignant” drives. The oedipal phase will end without castration anxiety or guilt if the

parents function as adequate caretakers.

HOW DOES SELF PSYCHOLOGY ANALYSIS CURE?

Self-psychology analysis cures by acquisition of structure in the patient’s self. Since a major

tenet of self psychology, (one could say “finding” rather than “tenet”) is that psychopathology most

often reflects deficits in the self, the major thrust of the curative process is to be of aid in reinitiating
the development of the self that has been fixated and retarded in its growth. As we have discussed,

deficits in the self are seen as outcome products of a failed self-selfobject relationship of childhood,

amounting to a deficit in the self structure (self function) that was inadequately internalized. The cure
in self-psychology analysis is to develop—that is, to acquire, additional structures within the self. In

classical psychoanalysis, in contrast, the cure is to ultimately resolve the fixation of the oedipal

complex through the medium of the unfolding of the transference neurosis (Freud, 1917a). In the work

of classical analysis, the material of the sessions is focused on the myriad manifestations of the oedipal
fixations directed to, for, and against the analyst. The result of the interpretative work is to make the
patient aware—and thus free the patient—of the fixations emanating from the oedipal drama. The

result will be the acquisition of an expanded conflict-free sphere, the expanding of consciousness
(“Where id was ego shall be”) and the reduction of castration anxiety and the symptoms (new
defenses) evoked by anxiety (Freud, 1926).

Self-psychology analysis, like all psychoanalyses, involves the elaboration of transference

phenomena in the analytic work focused on the selfobject transferences and the previously thwarted
developmental needs of the self. Patients who are analyzable (those who, while possessing deficits in

their selves, have the capacity to form and develop stable alliances with their therapists) will have a

spontaneous unfolding of their strivings for structure in the form of a specific self-selfobject

www.freepsychotherapybooks.org 276
transference. These transferences, which reflect the stalled development of the self in relating to a
selfobject, encompass the specific functions that have not been internalized in the self of the analysand.
They represent the stalled developmental needs of the self for confirming, admiration, and echoing

(mirror transference) or the self’s needs for firm ideals, calming, and guidance (idealized parent imago
transference). The spontaneously unfolding transferences represent needs for the development of

structure—not, as in classical analysis, the reliving (in fantasy) of the ancient oedipal strivings that

requires explication. The end point of self-psychology analysis is the internalization of the analyst and
his or her functions. These become metabolized into self structures that perform the now internalized

functions of mirroring and other functions of the selfobject prior to internalizing.

The analysis can be said to begin with the establishment of the basic self-selfobject transference

in which the patient’s self is sustained (Kohut, 1968). In the course of the analysis, the basic selfobject
transference is disrupted time and again by optimal failures of the analyst, akin to the optimal

frustrations of the archaic self-selfobject relationships of childhood. After suitable awareness and

interpretations of the analysand’s retreat and regression (with manifestation of the reinstitution of
archaic selfobject relationships), the basic selfobject transference will be reestablished. However, the

optimal frustration sets into motion the transmuting internalization of the imago of the selfobject

analyst and his or her mirroring or idealized parent function, thus leading to the acquisition of self
structure. The process of analysis can never proceed without experiences that the analysand perceives

as empathic failures. In this category of events one can place unavoidable interruptions (weekends and
vacations) and the analyst’s incorrect interpretations. These frustrations, if nontraumatic, will lead to
interiorizations of the analyst’s essential or basic positively enhancing selfobject functions, especially if

the analyst, after grasping the analysand’s distress or retreat into archaic preanalytic object-related
behaviors, attempts to focus on the experienced rebuff. This latter process, involving empathic
understanding of the analysand’s experience, amounts to a transference interpretation in which the

analyst demonstrates that the patient’s self is held in high regard in the analytic relationship, in sharp
contrast to relationships with the unempathic archaic selfobjects of the patient’s past. The optimal
frustrations that the patient experiences extend to the analyst’s interpretations, since these, too, are

www.freepsychotherapybooks.org 277
not mirroring actions but are only words. The analyst cannot perform mirroring actions as he or she

interprets, an action that serves only to clarify and illuminate.

To summarize, in the normal flow of the analysis, the curative process is a matter of the

analysand’s previously stunted self acquiring selfobject functions through internalization of the

analyst’s selfobject functions. This comes about, as does any building of self structure, through a hiatus

in the relationship (optimal failure), which serves to energize the imago of the selfobject analyst and
his or her functions. These then become absorbed into the self as the self’s mirroring or other

functions. Another way of understanding the structure building that comes out of analysis is to

remember that analysis implies regression, so that the analyst and analysand are locked into a

regressive transference. In an archaic selfobject relationship such as is found in childhood, the archaic
selfobject is the source of regard. In the analytic transferences, the patient enters into a reactivation of

the previously thwarted needs for structure so as to infuse the self with esteem and vigor. The
analysand’s experience of the analyst, the new selfobject, is as if the patient is once again in contact

with a giver of the gifts of worth and value to the self (Kohut, 1977).

THE COURSE OF ANALYSIS

An overview of the course of a self-psychology analysis approximates the process found in a

classical psychoanalysis. There are two phases of treatment to be considered.

1. The Defense Transference. This is, of course, the unconscious position of adjustment the

analysand takes in reexperiencing the analyst as a parent figure. The conforming experience of the
analysand, in the service of maintaining the archaic self-selfobject ties, serves secondarily as resistance

to the new selfobject bond in analysis. Its major purpose is to protect the analysand from the

possibility of recurring disappointment at the hands of unempathic selfobjects. Thus, the genuine
needs of the analysand are repudiated so as to avoid psychic pain.

The defense transference, in the view of classical analysis, is effective in maintaining the

www.freepsychotherapybooks.org 278
repressed instinctual derivatives of oedipal previously buried yearnings for self-structure—for

example, the wish to experience self worth through the confirmatory, admiring attitudes and actions of

the mirroring selfobject—the analyst is called on to express his or her understanding of what the

analysand is experiencing. In this manner, the analysand’s transference strivings are “accepted” by the
analyst, indicating that the analyst is mindful that a period of time, sometimes a long period, must

elapse to allow the transference to unfold without challenge. Premature challenges to these
transference strivings may be taken by the analysand as rejections of these very strivings, thus

repeating the actual childhood milieu in which these self needs went underground, resulting in a

deficient self. Some patients require more or less protracted periods of understanding. In any case, the
analyst must be mindful that to understand these selfobject strivings without interpretation is at times

of crucial importance in the curative process of a self-psychology analysis. Understanding, which is not

simply acceptance, emphasizes that the outcome of a self-psychology analysis is the eventual growth of
the patient’s self through internalization of the analyst’s selfobject ministrations.

The next phase of the analysis centers on the explaining or interpreting function of the analyst.

The analytic work done in this phase of treatment deals with interpretations of the repressed strivings

that ultimately will bring the patient into investing the analyst with attributes of one or other of the
parental roles. If the interpretations are successful, the transference neurosis will now emerge. In a
self-psychology analysis, however, the defense transference is in the service of maintaining out of the

patient’s awareness, the strivings of the self for mirroring and/or the firm ideals, leadership, or

calming of the idealized parent. The analytic work done in this phase is directed at providing an
environment that the analysand experiences as safe and where the analyst, if necessary, can interpret

the defense of haughtiness or isolation against the emergence of the feared wishes for selfobject

support.

2. The Basic Selfobject Transference. Kohut (1978) stated: “The discovery of the selfobject

transferences forms the basis of my whole work concerning narcissism and the self” (p. 20). This dyad
of patient and analyst reactivates the self needs of the analysand that had remained, as a result of

www.freepsychotherapybooks.org 279
faulty interactions in early life, disavowed or in a state of repression. Once the analysand enters an
idealizing or mirror transference, the self achieves a state of cohesiveness. The analyst’s activity in this
phase consists of two sets of attitudes and behaviors, understanding and explaining (Kohut, 1977).

Once the patient begins to establish the analyst as the selfobject to whom he or she can reveal
the previously buried yearnings for self-structure—for example, the wish to experience self worth

through the confirmatory, admiring attitudes and actions of the mirroring self-object—the analyst is

called on to express his or her understanding of what the analysand is experiencing. In this manner, the
analysand’s transference strivings are “accepted” by the analyst, indicating that the analyst is mindful

that a period of time, sometimes a long period, must elapse to allow the transference to unfold without

challenge. Premature challenges to these transference strivings may be taken by the analysand as

rejections of these very strivings, thus repeating the actual childhood milieu in which these self needs
went underground, resulting in a deficient self. Some patients require more or less protracted periods

of understanding. In any case, the analyst must be mindful that to understand these selfobject strivings

without interpretation is at times of crucial importance in the curative process of a self-psychology

analysis. Understanding, which is not simply acceptance, emphasizes that the outcome of a self

psychology analysis is the eventual growth of the patient’s self through internalization of the analyst’s

self object ministrations.

The next phase of the analysis centers on the explaining or interpreting unavoidable

interruption of the steady state of the basic selfobject transference. As the analysis proceeds, with the

analysand now revealing his or her specific self needs in the selfobject transference that has
spontaneously unfolded, an equilibrium is reached, a cohesive self state. This equilibrium, of course, is

dependent on the presence of the selfobject, in the same manner in which infants experience
equilibrium in the presence of their selfobjects. Only after internalization of the necessary mirroring
and other functions performed by the selfobject is the self complete. These functions, as already
described, become interiorized in the self directly after a failure of the selfobject to either empathically

appreciate or respond to a self need—the notion of optimal frustration. Similarly, the analysand

www.freepsychotherapybooks.org 280
immersed in a selfobject transference onto the analyst experiences absences or unavoidable empathic
failures by the analyst as if he or she once again has not been responded to. The patient is once again
with the archaic selfobjects of the past, whose failures were not optimal but fixating because they were

too protracted, too intense—in short, traumatic failures. The analyst’s task here is to help the
analysand recognize his or her experience in temporarily identifying the analyst with the childhood

disappointers. Thus, the explaining (the interpretations) of the analyst in a self psychology analysis is

necessary to reveal what might be called the transference distortions that have interfered with the
structure building in analysis.

Explaining or interpreting is necessary to illuminate not just the dynamics of the transference

interactions but also its genetic roots. As the analyst explains (interprets) to the patient the dynamic

and genetic explanations of the patient’s thwarted needs and the reconstruction of the failed self-
selfobject dyads in childhood, the analyst is offering to the patient an appreciation of the patient’s past.

This will be of service to the patient’s empathic grasp of himself or herself and will be of help both in

the subsequent working-through phase of the analysis and later when the analysis is terminated.
Moreover, when the analyst is explaining, he or she is becoming more objective with the patient, in a

sense, replacing the experience of merger with the experience of resonance. This reflects progress in

the development of the self, from reliance on merger to the use of the empathic closeness of the analyst
selfobject.

After the transference distortion is made clear, the analysand is enabled to experience the

unavoidable interruptions, empathic misunderstandings, and other mistakes of the analyst as

frustration—but optimal frustration. This experience results in the phenomena of internalization of

function, the so-called transmuting internalization. As has been previously explained, transmuting
internalization refers to the intrapsychic process in which the functions such as mirroring that were
previously performed by an outside agency (the selfobject) are now experienced as imbricated or
intertwined in one’s self. The imago of the self object’s functions after an empathic failure takes on

greater intensity. These functions now exert their specific action in response to a specific intrapsychic

www.freepsychotherapybooks.org 281
signal—loss of self-esteem. These signals of need, which formerly were communicated to the selfobject
or were responded to by the selfobject without overt communication, now evoke the intrapsychic
functioning, so that the self is now in a cohesive state without the minute-to-minute presence of a

selfobject that was formerly required. When the entire self is filled out, so that the poles of ideals and
ambitions and functioning skills and talents are operational, a nuclear self exists that can initiate what

Kohut (1977) called “programs of action” (p. 180).

Finally, the termination stage of a self-psychology analysis is arrived at when the patient
experiences sufficient cohesiveness of his or her self as that the patient and analyst believe that further

analysis will not result in further additions to the patient’s self structures and that further insights will

not be beneficial. At the termination stage, the patient’s self will, ideally, be sufficiently strengthened to

have a greatly enhanced resistance to fragmentation as well as an overall decrease in the experience of
the self as lacking assertiveness or firm ideals. Thus, whether the analysis focuses on the patient’s

primary self trauma and its subsequent imbalances or on the patient’s compensatory attempts to gain

self balance, the outcome of the analysis is that the patient has now developed a cohesive self that can
now seek out and invest in mature selfobjects for the necessary support in times of need (Kohut,

1977). In sum, in Kohut’s view, the aim of a psychoanalytic cure is to firmly establish the patient’s

capacity to form mature, empathically directed, self-selfobject bonds so that mature self-selfobject
encounters take the place of the bondage that had previously enslaved the self to the archaic

selfobjects.

APPLICATIONS OF SELF PSYCHOLOGY

Kohut hoped that self psychology would have applications in the field of history and social

sciences and that the psychology of the self could contribute wider meanings than the views of classical
psychoanalysis in literature and the arts. The central contribution of self psychology to an

understanding of humanity, its history, arts, and place in the universe comes from the acceptance of

the empathic outlook in life. As Kohut (1973) stated:"…it (the empathic outlook) constitutes the very

www.freepsychotherapybooks.org 282
matrix of man’s psychological survival” (p. 360). Kohut’s (1975) description of empathy was
summarized in three propositions:

(1) Empathy, the recognition of the self in the other, is an indispensable

tool of observation, without which vast areas of human life, including
man’s behavior in the social field, remain unintelligible. (2) Empathy, the
expansion of the self to include the other, constitutes a powerful
psychological bond between individuals which-more perhaps than even
love, the expression and sublimation of the sexual drive-counteracts man’s
destructiveness against his fellows. And (3) empathy, the accepting,
confirming and understanding human echo evoked by the self is a
psychological nutriment without which human life as we know and
cherish it could not be sustained” [p. 361].

Thus, Kohut believed that the contribution of self psychology to the understanding of people

through scientific empathy added to the values and ideals of humanity, indeed, served to support the
very survival of humankind. Kohut (1971, 1973, 1977) stressed over and over again that the central

problem of humanity in the Western world is the child who is understimulated, not responded to, and

lacking leaders, who becomes the empty, isolated adult, still in search of approval or a target for
idealization—in short, Kohut’s Tragic Man. Kohut pointed to a major change in the structure of families

from Freud’s time, when children had closer ties to their families and the environment was

experienced as close and even sexually overstimulating, leading to the type of conflict and
psychopathology that Freud described. In the families of today, in Kohut’s (1977) view, under

stimulation is rampant, leading to attempts at “erotic stimulation in order to relieve loneliness, in order

to fill an emotional void” (p. 271). Thus, Kohut as social critic, as humanist, striving to appreciate (i.e.,
diagnose) the essential difficulties in humanity’s quest for survival in the modern era, discerned that

our greatest need is to be in an environment in which we can be singled out, appreciated, uplifted by
invigorating leaders, and not be lost as a note in the underground. In short, we need not to be relegated

to the state of anomie.

LITERATURE AND THE PSYCHOLOGY OF THE SELF

www.freepsychotherapybooks.org 283
Kohut often pointed out that the great modern artists were the first to respond to the shifting

problems of the modern individual. Thus, in the works of Ezra Pound, Eugene O’Neill, and Franz Kafka

in the literary field, the emphasis on the breakup of the self and the striving to restore the self of
fragmentation documented and even anticipated the dominant psychological problem of this era. In

Kohut’s view, Kafka’s K is the everyman of our time, as he tries to get close to the great ones in power
(The Castle) or dies a death without meaning (The Trial) or, as Gregor Samsa in The Metmorphosis, lives

like a cockroach without being responded to by his family. Kohut often quoted three lines from

O’Neill’s play, The Great God Brown as an example of man’s longing to restore his self: “Man is born
broken. He lives by mending. The grace of God is glue” (see Kohut, 1977, p. 287).

The findings of self psychology are of great value to the student of applied psychoanalysis in

literature. Kohut’s emphasis on the empathic immersion into the self experience of the other is

especially important in the appreciation of the great figures in literature. One cannot begin to assess
the tragic downfall of the Ajax of Sophocles without immersing oneself in the self of the great military

hero who has become a ludicrous spectacle after destroying sheep whom he thought were his

enemies’, Menelaus and Agamemnon. Consider the self of the aging monarch, Shakespeare’s King Lear,
whose prized daughter has refused his request for self-sustenance as he is about to pass on the baton

of command and retire. Lear’s experience of outrage must be experienced through empathic

immersion into his particular self needs. And again, to gain a heightened regard for the issues with
which Hamlet struggles, one must be able to read empathically into the self of the prince recently

separated from his dead father, confronted with his newly married mother, and denied his ascension to

the throne of Denmark. Once readers have been enabled to sink empathically into the literary figures

presented by the author, they are able to appreciate the self state of the protagonists.

Another set of ideas from self psychology of great service in literary appreciation, is notion of

the self-selfobject bond and its disruptions, which may lead to the experience of fragmentation and its
vicissitudes, including disintegration anxiety, depletion of self-esteem, hypochondriasis, narcissistic

rage, and loss of mentational functions such as reality-testing, synthesizing, and memory. Armed with

www.freepsychotherapybooks.org 284
this methodological approach of empathy and the notions of self psychology, the reader can approach
each of Shakespeare’s tragedies, for example, and illuminate the concerns, and failures of each of the
protagonists and the reparative selfobject functions that each is seeking. Thus, Hamlet can be seen as

responding to the losses he incurred with the reactions stemming from a depleted self and the
interaction congruent with that self state. Lear, as previously stated, has had to suffer the loss of his

major selfobject, Cordelia, and reveals his tragic fragmentation in the tempest. Othello is understood

from the outset, in the view of self psychology, as experiencing concern over the attractiveness of his
black self to his young, Causasian wife, and thus is vulnerable to Iago’s sadistic innuendoes over her

loyalty. Macbeth can be recognized as a man who has lost his selfobject, without whom he falters, and

is compelled to seek surrogate selfobjects, the witches. They too fail him in giving self-support, and he
dies. Thus, the findings and views of self psychology added a needed dimension to the appreciation of

literature that parallel its contribution to the study of the individual in the clinical encounter.

Self Psychology and Music. Kohut expressed the conviction that the great artists, including the

great modern composers, reflected in their art the great psychological problem of our era—the
situation emanating from the endangered self (Kohut, 1977). One gains a unique contribution to the

appreciations of music from the application of self psychology. The experience of music in its function

as a selfobject are part of almost everyone’s life. We may recall the uniquely calming, soothing
experiences of listening to music. For some, these experiences are provided by the Missa Solemnis of

Beethoven or the Mass in B Minor of Bach. For others or at different times, it is a modern popular
singer or instrumentalist or a popular musician of an earlier era. Music, in those who respond to it, can
be felt as an invigorating experience that may cause a quickening of the self and lead to programs of

action. It is, of course, common to seek out music in which one finds an essential likeness—“music to
match one’s mood,” as the expression goes—a twinship type of phenomenon. When one needs
company to share one’s inner mental life, one seeks a particular type or form of music, and one may

seek a certain type of music or performer to merge with in order to shore up a flagging or enfeebled
self.

www.freepsychotherapybooks.org 285
Music can be said to perform selfobject functions as a result of its being linked with memories of

archaic selfobjects of childhood and their self-sustaining qualities. The sounds of an admiring

mirroring selfobject are experienced in musical expressions by the individual self as recapturing the
memories of that blissful union. Similarly, in those to whom music and the state of their selves coexist,

music can be experienced as a phenomenon akin to a twinship merger. In its ability to calm or evoke
action, music performs functions similar to those of the idealized parent imago. The experience of

becoming immersed in robust musical expression is also part of the feeling of being with a leader.

Along the same lines, the experience of listening to music that is spontaneously creative, such as
improvised jazz, or music that is systematically creative, such as the compositions of Arnold

Schoenberg or Alban Berg, may allow the listener to identify with the musicians’ or the composer’s
assertiveness and thus enhance the listener’s self state.

At times, the musical message or tenderness or vigor may be direct, without complex
orchestration, or it may have complex counterpoint or harmony. It may be experienced as too direct in

its impact—too simplistic—or as totally acceptable. Thus, Tchaikovsky’s Sixth Symphony may be

experienced as maudlin, not subtle or beautiful. Some listeners, who lack resistance to direct
communications of gentleness, may appreciate without restraint the operas of Puccini, whereas others

with resistance to direct mirroring messages find it prosaic. Thus, music may serve a variety of self-

object functions in these who can respond to it.

THE SELF IN HISTORY

Kohut (1974b) believed that “History and psychoanalysis should be the most important
sciences of the future. They are important because humanity has reached a point in which populations

will sooner or later have to become stabilized.…If humans are to survive in a way that has any

similarity to what we have prized up till now as being the essence of human life, the narcissistic

motivations, I believe, must come into the ascendancy” (p. 775). Kohut believed that the insights of self

psychology would be helpful to historians in understanding the formation, maintenance, and

www.freepsychotherapybooks.org 286
disruptive processes of groups.

Kohut described the notion of the group self as analogous to the individual’s self. Thus, a nuclear

group self would include the central ambitions and the ideals that characterize the group in its
ordinary operations. To appreciate a group in operation, one would study the economic and social
circumstances that influenced its formation and the specific psychological conditions that evoke

fragmentation or cohesion, including the need for a particular type of leadership. Kohut observed that

groups are held together not only by their shared ego ideal as Freud (1921) maintained, but also by a
shared group self—that is, by shared assertiveness (Kohut, 1972). The group’s integrity may be

disturbed by destruction of the group values or damage to the group outlets for maintaining its

prestige—for example, by an economic depression or military losses. Such imbalances in the group’s

esteem—similar to an individual’s loss of self esteem—may lead to fragmentation of the group. The
ensuing manifestation of narcissistic rage (acute or chronic) may involve the entire group in acts of

vengeance against outside forces who are structured as oppressors (Kohut, 1972).

An important source for maintaining the integrity of any particular group is the leader needed

or chosen by the group in various situations, especially in situations of impending fragmentation.

Kohut identified two types of leaders. In the first type, the messianic leader or personality, there has

been a fusion between the self and the pole of ideals, so that messianic leaders experience themselves
as being in possession of total rectitude. These personalities set themselves up as the perfect leader, a

god, worthy of reverence. Such a leader was Adolf Hitler, who effected repair to the German group self

in its experience of ineptitude after World War I. The second type, the charismatic leader, has become
one with his or her pole of assertiveness and thus experiences and exudes certitude and omnipotence.

Winston Churchill was such a charismatic leader, needed by the British people during the crisis of
confidence of World War II and abandoned when the need for an omnipotent selfobject was at its end
(Kohut, 1976). Thus, the messianic or charismatic leader, who steps in to effect repair to the group self
experiencing a common defect in assertiveness or sharing a common need for an idealized leader, is

then experienced as the selfobject of the group self.

www.freepsychotherapybooks.org 287
SUMMARY

The centerpiece of Kohut’s work is the self and the self-selfobject dyad in the study of historical
characters and literature as well as in the study of the developing person and the distressed patient

petitioning for relief of his or her loneliness. Kohut never lost sight of his central finding, his

anagnorisis that it is the experiences of man—the self—that is crucial to appreciate, not the drives nor
the conflicts of man. From his seminal paper on empathy and introspection (Kohut, 1959) to his final

works on the curative processes in psychoanalysis, Kohut taught that man must be understood

through empathy, the royal road to the appreciation of the inner life. Kohut’s works on the
development of the self, on the archaic and mature self-selfobject dyads, on the theory of

psychopathology and on the theories of cure in psychoanalysis are significant contributions to

psychoanalysis and in my view will continue to exert an impact on the field of psychoanalysis. Will
Kohut’s views and findings be amalgamated into the mainstream of psychoanalytic theory and

practice? This is a question for the future generations of psychoanalysts and one that Kohut would

have welcomed, as he stated: “A worshipful attitude toward established explanatory systems—toward

the polished accuracy of their definitions and the flawless consistency of their theories—becomes

confining in the history of science—as do, indeed, man’s analogous commitments in all of human

history. Ideals are guides, not gods. If they become gods, they stifle man’s playful creativeness; they

impede the activities of the sector of the human spirit that points most meaningfully into the future”
(1977, p. 312).

And further: “My deepest wish, however, is that my work—in amplification or emendation, in

acceptance and even in rejection—will contribute to motivate the rising generation of psychoanalysts

to pursue the path opened by the pioneers of yesterday, a path that will lead us further into the

immense territory of that aspect of reality that can be investigated through scientifically disciplined
introspection and empathy” (p. 312).

In this I have a sense of certitude: Heinz Kohut as theoretician, as practitioner, as humanist and

as a man will never be forgotten.  

www.freepsychotherapybooks.org 288
REFERENCES

Freud, S. (1917a). Introductory lectures on psychoanalysis. Standard Edition, 16, 243-496.

Freud, S. (1917b). Mourning and melanchoia. Standard Edition, 14, 237-259.

Freud, S. (1921). Group psychology and the analysis of the ego. Standard Edition, 18, 65-144.

Freud, S. (1926). Inhibitions, symptoms and anxiety. Standard Edition, 20, 75-174.

Heller, E., & Kohut, H. (1978). Psychoanalysis and literature. Critical Inquiry, 1,449-450.

Kohut, H. (1959). Introspection, empathy and psychoanalysis. Journal of the American Psychoanalytic
Association, 7, 459-483.

Kohut, H. (1966). Forms and transformations of narcissism. Journal of the American Psychoanalytic
Association, 14, 243-273.

Kohut, H. (1968). The psychoanalytic treatment of narcissistic personality disorders: Outline of a

systematic approach. The Psychoanalytic Study of the Child, 28, 86-114.

Kohut, H. (1971). Analysis of the self. New York: International Universities Press.

Kohut, H. (1972). Thoughts on narcissism and narcissistic rage. The Psychoanalytic Study of the Child,
27, 360-400.

Kohut, H. (1974a). Remarks about the formation of the self. In P. Ornstein (Ed.), The search for the self
(pp. 737-771). New York: International Universities Press, 1978.

Kohut, H. (1974b). The self in history. In P. Ornstein (Ed.), The search for the self (pp. 771-783). New
York: International Universities Press, 1978.

Kohut, H. (1976). Creativeness, charisma and group psychology. In J. E. Gedo & G. H. Pollock (Eds.),
Freud: The fusion of science and humanism (pp. 379-425). Psychological Issues, 9, (2/3,
Monograph 34/35).

www.freepsychotherapybooks.org 289
Kohut, H. (1977). The restoration of the self. New York: International Universities Press.

Kohut, H. (1979, June). “Four basic definitions of self psychology.” Paper presented to the Workshop on
Self Psychology. Chicago, IL.

Kohut, H., & Levarie, S. (1950). On the enjoyment of listening to music. Psychoanalytic Quarterly, 19, 64-
87.

Kohut, H., & Wolf, E. (1978). The disorders of the self and their treatment. International Journal of
Psychoanalysis, 59, 413-425.

www.freepsychotherapybooks.org 290
9
MARGARET S. MAHLER: SYMBIOSIS AND
SEPARATION-INDIVIDUATION

ANNI BERGMAN, PH.D.
STEVEN ELLMAN, PH.D.

In this chapter we will trace the development of Margaret Mahler’s research, clinical works and
theoretical conceptualizations. Although Mahler’s concepts have always been firmly grounded in either

clinical or naturalistic observations, it is interesting and in keeping with a book whose theme traces

developments beyond Freud, to first look briefly at Freud’s concepts of early development and to

compare these concepts with Mahler’s pioneering work. Although Freud did not do observational
studies, he at times wrote about early development, and in our opinion this aspect of his work has been

somewhat neglected.

It would also be of some importance and interest to compare the theoretical statements of
various writers about early development. Certainly such a comparison might include Jacobson,
Hartmann, Winnicott and perhaps other authors from the British object relations school. Mahler’s

work would make this comparison particularly interesting since she and her co-workers have provided
both theoretical conceptions and empirical observations about early development.

www.freepsychotherapybooks.org 291
Freud was frequently concerned with how the infant began to learn about the external world. In

The Interpretation of Dreams (1900), he presents his well-known views of the infant, at first primarily

or only concerned with pleasure and later, through deprivation, coming to know about the external
world. This conception of how the infant turns from its primary concern (pleasure or tension

reduction) to secondary concerns (the outside world) is based heavily on a tension-regulation model.
Freud’s later views, which are contained in large part in his papers on narcissism and his

metapsychological papers (1914, 1915), are less centered on a tension-regulation or reduction model.

In these and other papers, Freud put forth the guidelines of an interesting theory of early development,
but in this chapter we can only sketch out some of his ideas. Freud sees the early mental development

of the infant and child as taking place along three polarities—pleasure-pain, subject-object, and active-
passive—an idea that has a developmental unfolding. In early life, pleasure and pain predominate, and
Freud maintains that for the infant or child (we do not know the age range to which Freud referred),

the external world is at first primarily a matter of indifference. This corresponds to Freud’s notion of
primary narcissism, in which satisfaction or pleasure, from the infant’s perspective, is autoerotic. At
this time, the external world is not cathected with interest (in a general sense) and is indifferent for

purposes of satisfaction (Freud, 1914). Interestingly, although Mahler uses different terminology, her
autistic phase bears striking resemblance to this Freudian phase.

At Freud’s next step in development, we run into something of a paradox. Freud (1915)
postulates that as the infant continues to experience the external world, “it acquires objects from the

external world, and, in spite of everything, it cannot avoid feeling internal instinctual stimuli for a time

as unpleasurable” (p. 135). As the infant builds up perceptions of (primarily internal) stimuli as

unpleasurable and (primarily external) stimuli as pleasurable, it takes into itself (or introjects) the
pleasurable stimuli and casts out (or projects) the unpleasurable stimuli. At this point, Freud (1915)

maintains that “the original ‘reality-ego’ which distinguished internal and external by means of sound

object criterion changes into a purified pleasure-ego” (p. 136). This pleasure ego has divided the world

into all that is pleasurable, which is equated with itself (“ego subject,” in Freud’s terms), and all that is

unpleasurable, which is equated with the external world. One can attempt to equate this idea of the

www.freepsychotherapybooks.org 292
purified pleasure ego with some of Mahler’s findings and formulations, but for the purposes of this

chapter we wish to make several related points about the concepts Freud puts forth.

First, Freud pointed out that development of certain reality-ego functions may be nonmonotic.

Thus, the infant at an age prior to the purified pleasure ego is considered by Freud to be, in some ways,

in better contact with reality than when the pleasure ego is formed. We believe this line of reasoning is

consistent with several of Freud’s concepts at this time (Freud, 1915), but the main point we wish to
dwell on is that at a time when the infant, according to Freud, is indifferent to the external world, it can

still develop a rudimentary reality ego. Thus, Freud saw nothing incompatible with postulating a stage

of primary narcissism in which pleasure is seen as passive, internal, and autoerotic, and yet at the same

time certain types of “learning” can take place. The question for Freud was not whether the infant
could correctly perceive certain aspects of reality but, rather, whether or how the object was viewed in

terms of the infant’s pleasurable and unpleasurable experiences. This is quite a different question than
whether the infant can learn to respond during its first weeks or days of life.

As a second general point, Freud (1914, 1915) begins at about this time to make use of what
today are frequently called projective-introjective mechanisms. These concepts are, of course, used

frequently by Mahler as well as many others, but it is of interest to see the way she has both expanded
and particularized the use of these concepts.

As a third related point, we wish to emphasize how during this era Freud stresses both the

gradual nature of being able to know the pleasure-giving object as a separate entity and, even more

important, the very gradual nature of the development of object love. Freud (1915, 1917) discusses
aspects of the development of object love, but of course Mahler is able to delineate with much greater

precision concepts such as libidinal object constancy on the pathway to object love. As we will see,
Mahler’s concepts and observations in many ways begin to fulfill the promissory notes that Freud left

us in his many brilliant papers.

In this brief introduction we have touched on a few of the concepts that Freud introduced that

www.freepsychotherapybooks.org 293
bear some relationship to Mahler’s work. We could, of course, make a much fuller comparison, but our

intention is only to point out the relationship and set the stage to show how Mahler has built on and

yet gone beyond what Freud could have even anticipated. In a chapter devoted to a historical

recounting of the theorists who bear some important relation to Mahler’s work, one would also have to
include at least aspects of the work of Hartmann, Kris, and Lowenstein and large parts of Jacobson’s

work. Both Mahler (1979) and Kernberg (1980) have emphasized in different ways the importance of
Jacobson’s developmental concepts. Many other influential authors could be named, of course, but in

our opinion, Spitz and Anna Freud’s pioneering empirical studies were, in general, an inspiration to

psychoanalytic researchers in many ways, particularly in demonstrating that theoretical concepts

could be shown to have important empirical consequences.

Although all the authors mentioned have a variety of similarities (and differences) with respect

to Mahler’s work, in Loewald’s (1979) words: “Her clear emphasis on the fundamental importance in

early development and continuing throughout life, of differentiation and separation from an
encompassing psychical matrix…have had a remarkable impact on current analytic understanding of

children and adults.” Although Freud implied the “dual unit” or dyad, Mahler makes it the beginning

and most important part of her observational and theoretical field.

We shall discuss the work of Margaret Mahler in three parts: (1) her early papers, including her
work on infantile psychosis; (2) her research project on separation-individuation and her theory of
subphases resulting in beginning self and object constancy; and (3) applications of separation-

individuation theory to psychoanalytic theory and treatment.

EARLY PAPERS

Mahler began her career as a pediatrician and director of a well-baby clinic in Vienna. The
interests she developed at the outset of her professional life have remained important throughout her

career. Probably the most important of these has been her interest in the mother and baby as a dyad,

www.freepsychotherapybooks.org 294
or, as she later referred to it, as a dual unity within one common boundary, a symbiotic pair. Beginning
with her first paper delivered in this country, entitled “Pseudoimbecility: A Magic Cap of Invisibility”

(Mahler, 1942), presented in 1940 to the Psychoanalytic Institute of New York, she demonstrated her

interest in the pre-oedipal era, in motility, and in the affecto-motor communication between mother
and child.

Between child and mother there exists from the beginning a close
phylogenetic bond which is unique and much more exclusive than
communication by words or thoughts; it is an interrelationship through
the medium of affective expressions…The interrelation between the
unconscious of the mother and the reception of stimulation of the sense
organs of the baby is the prototype for a way of communication between
child and adult which is not confined within the limited sphere of
language, (p. 4)

In her psychoanalytic work, Mahler began to treat several children suffering from childhood
psychosis. This culminated in her eventual formulation of the autistic and symbiotic types of childhood

psychosis (Mahler, 1952). She also became interested in determining how normal infants attain a sense
of separate identity in the caretaking presence of their mothers. Examination of Mahler’s papers of that

period (those that preceded the beginning of observational research) reveals how closely connected in
her thinking were the phases of early normal development and the consideration of extreme

pathology. Mahler is essentially a psychoanalyst and a clinician, and her early papers are filled with

clinical vignettes from the many severely disturbed children whom she treated as a child analyst. Yet
her thinking about pathology never overshadowed her interest in normal mental life and her
conviction about the importance of the early mother-child relationship.

In an early paper (Mahler, Ross, & DeFries, 1949), Mahler was already dealing with the child’s

problem around the waning of omnipotence.

The child gradually realizes that its power is waning. It has not only to
renounce essential gratification, but must in addition lose its sense of

www.freepsychotherapybooks.org 295
omnipotence. The language of violent affect is rendered useless as a means
of communication with the parents, and the child has to renounce them in
favor of speech…It seems as if these affective outbursts at the age of 2 to 3
years are struggling attempts in the child to maintain the archaic common
ground so familiar to it: the intensely pleasurable affective rapport with
the parents in the child’s affective domination of them. This attempt is
destined, like the Oedipal strivings, to fail from the danger of loss of love
and fear of castration.

Direct affective attacks failing, the child searches for other means to regain
entrance to the Garden of Eden. This coincides in time with beginning to
walk and the process of taking in impressions of the outside world with all
the senses, acquiring knowledge and testing reality. The child utilizes
these newly gained discoveries, to share them with mother and father, and
thus restore a common ground with them. The expressions of
enchantment and affection, which the parents give so abundantly at the
first presentations of such fact finding, bring the child a temporary
restoration of the old affective and a new intellectual co-experience with
the parents.

This quotation already contains descriptions of behaviors that later, during the observational

study of separation and individuation, become incorporated into the careful delineation of the
subphases.

Mahler’s papers on child psychosis contain many references to her view on normal

development. In 1952 she stated:

The intrauterine, parasite-host relationship within the mother organism

must be replaced in the postnatal period by the infant’s being enveloped,
as it were, in the extrauterine matrix of the mother’s nursing care, a kind of
social symbiosis....

The turning from predominantly proprioceptive awareness to increased

sensory awareness of the outer world occurs through the medium of
affective rapport with the mother. The baby’s libido position thus proceeds

www.freepsychotherapybooks.org 296
from the stage of fetal narcissism to primary body narcissism, a stage in
which representation of the mother’s body plays a large part....

To understand the dynamics in infantile psychosis, observation and study

of the most important transitory step in the adaption to reality is
necessary; namely, that step in the development of the sense of reality in
which the mother is gradually left outside the omnipotent orbit of the self.
This step is preliminary to, and perhaps alternates with, the process of
endowing the mother with object-libidinal cathexis. The toddler gradually
delimits his own individual entity from the primal mother-infant symbiotic
unit. He separates his own self (and his mental representation) from that
of the mother. This stage in ego development is a very vulnerable one,
particularly in children in whose early life the somatopsychic symbiosis
has been pathological, (pp. 132-134)

Mahler’s interest and views on childhood psychosis and normal development were still closely
intertwined at this point in her work. These remarks on early development occur in the same paper in

which she outlines her views of autistic and symbiotic childhood psychosis. She describes primary

autistic psychosis as a syndrome in which the mother, as representative of the outside world, seems
never to have been perceived emotionally by the infant. The mother, therefore, remains a part object,

seemingly devoid of specific cathexis and not distinguished from inanimate objects. These, according to
Mahler, are infants with an inherently defective tension-regulating apparatus, which probably cannot

be adequately complemented by even the most competent mothers. The inherent ego deficiency of

these infants predisposes them from the very beginning to remain alienated from reality. Mahler

(1952) states:

It would seem that autism is the basic defense attitude of these infants, for
whom the beacon of emotional orientation in the outer world—the mother
as primary love object—is nonexistent. Early infantile autism develops, I
believe, because the infantile personality, devoid of emotional ties to the
person of the mother, is unable to cope with external stimuli and inner
excitations, which threaten from both sides his very existence as an entity.
(p. 145)

www.freepsychotherapybooks.org 297
Mahler contrasts the autistic psychosis with the symbiotic infantile psychosis. Symbiotic

psychosis often goes unnoticed during the first 2 or 3 years of the child’s life. It becomes evident at a

point in development when the phase-specific demands include realization of separateness.

The mechanisms which are characteristic in the symbiotic infantile

psychosis are the introjective, projective mechanisms and their psychotic
elaboration…These mechanisms aim at a restoration of the symbiotic
parasitic delusion of oneness with the mother and thus are the diametric
opposites of the function of autism.…It seems that the symbiotic psychosis
candidates are characterized by an abnormally low tolerance for
frustration, and later by a more or less evident lack of emotional
separation or differentiation from the mother. Reactions set in…at those
points of the physiological and psychological maturation process at which
separateness from the mother must be perceived and faced…agitated,
catatoniclike temper tantrums and panic-stricken behavior dominate the
picture; these are followed by bizarrely distorted reality testing and
hallucinatory attempts at restitution. The aim is restoration and
perpetuation of the delusional omnipotence phase of the mother-infant
fusion of earliest times—a period at which the mother was an ever-ready
extension of the self, at the service and command of “His Majesty, the
Baby.” (pp. 145-6)

THE SEPARATION-INDIVIDUATION PROCESS

Mahler’s observational research study of normal mother-child pairs began in 1959, the findings

of which have been described in the second volume of The Selected Papers of Margaret S. Mahler (1979)
and in The Psychological Birth of the Human Infant (Mahler, Pine, & Bergman, 1975). This research was

prompted by the following questions: How do normal infants, during the first three years of life attain

intrapsychic self and object representations? How do they move out of the state of dual unity or

symbiosis, during which they are not aware of themselves as separate, and achieve awareness of self as
separate from other? How do they attain a measure of libidinal self and object constancy? The

hypothesis of the study was that the human infant begins life in a state of complete dependence on the

www.freepsychotherapybooks.org 298
mothering one and in a state of nondifferentiation, or dual unity. The infant then undergoes a gradual
process of differentiation or hatching out, which results in intrapsychic structures of self and object.
The goal of the study was to learn about the process by which the first level of identity is achieved.

A setting was created in which mothers could interact freely with their infants. This was a large
playroom with many appropriate toys, divided by a low, fencelike partition from the mothers’ section.

There mothers could sit comfortably and chat while watching their children, who were in a stimulating

and safe environment. Participant observers were present at all times, mingling freely with mothers
and children while maintaining a friendly yet neutral atmosphere. The participant observers later

wrote down their observations in detail, and discussions took place in staff and research meetings,

where observers and investigators met at least once but more often twice a week. The research thus

created did not take place in an experimental artificial setting but in a very natural one—an indoor
playground, as Mahler called it, where mothers were in charge of their children.

The observations of the participant observers were checked by regular nonparticipant

observations conducted through a one-way mirror. Nonparticipant observers wrote down what they

saw at the time, and thus could obtain greater objectivity and detail than participant observers.
Participant observers, however, knew the mothers and children; their observations were more

impressionistic and subjective, but, it was thought, more in tune with the affective tone of the mother-
child pairs. The mother-child pairs were observed 3 to 4 times a week for 2½ hour-long sessions over a

period of 2½ years. The frequency and length of sessions provided a large data base from which it was

possible to obtain an intimate and detailed knowledge of each mother-child pair and the development
of their relationship.

In addition to participant and nonparticipant observations, mother-child pairs were regularly

filmed. All mothers were interviewed by senior staff members once a week. These clinical interviews
provided information about the family’s life at home. They also gave the mothers the opportunity to

talk about any aspect of themselves or their children that they chose to discuss. Fathers were

interviewed several times a year, and home visits were conducted regularly, especially during vacation

www.freepsychotherapybooks.org 299
periods.

Several aspects of the study were of special importance. One, as noted, was the frequency with

which observations were undertaken. This provided for a measure of objectivity, since a judgment

made one day could be corrected the next. Another essential aspect of the research design was that it
combined data from longitudinal and cross-sectional perspectives. Each mother-child pair was

observed from the time the child was about 6 months old to 3 years. At the same time, there were

always several mother-child pairs being observed simultaneously. Thus, children of any given age

could be compared both with each other and with himself or herself over a time period.

Another essential aspect of the study was that, although observational in method, it was guided

by psychoanalytic concepts. We believe that there was a good deal of carry-over from the way

psychoanalysts make inferences in the psychoanalytic setting to the way the observers used inferences
in these observational studies. As Mahler has put it, in these studies the psychoanalytic eye was guided

by the observations themselves, as in the psychoanalytic situation the psychoanalytic ear is led by the

analysand’s free associations. Thus, this research study relied heavily on the psychoanalytic acumen
and empathy of the observers and investigators, who were psychoanalysts. It rested on the meaning

and coherence that emerged out of many multifaceted daily observations. In the psychoanalytic
situation, analyst and analysand together create the psychoanalytic life history. In the study of

separation-individuation, the observers created the life history of the unfolding mother-child
relationship and the unfolding sense of self of the infant.

THE SUBPHASES

It was the comparative nature of the cross-sectional aspect of the study that eventually led to

the delineation of the subphases of the separation-individuation process. For example, in the first
group of children observed, a 1-year old girl was seen to explore the room freely, climbing a lot. At first
it seemed surprising that her mother sat calmly, staying in contact with the girl over a distance and

www.freepsychotherapybooks.org 300
directing her to avoid dangerous situations. It was thought at first that perhaps this mother-child pair
did not like physical contact. However, over time, after observing more mother-child pairs with infants

around 1 year of age, it became clear that this kind of exploration with relatively limited physical

contact between mother and child was characteristic of this particular age. This eventually came to be
termed the “practicing subphase.” In another example, a 16-month-old boy seemed to be anxiously

clinging to his mother. It was not difficult to understand this in terms of the particular mother-child

relationship, since the mother had shown considerable ambivalence about her baby after he was born.
But, again, after watching more mother-child pairs with children of that age, it became clear that

greater concern about mother’s whereabouts was a typical phenomenon of the toddler.

The subphases were delineated quite early in the study. However, the intensive study of each

mother-child pair made it possible to observe and study the individual variations within the regularity
of subphase specificity. Such variations involved the timing, intensity, quality, and mood that

characterized each particular mother-child pair. The subphases will be described in the following

pages. This description takes into account some of the more recent findings of infant researchers which
have contributed to and enriched Mahler’s original conceptualizations.

FROM 0-6 MONTHS

Since Mahler undertook her research project on the normal separation-individuation process, a
great deal of research has been done with infants and their caretakers for example that of Brazelton

(1974, 1981); Sander (1976); and Stern (1971, 1974, 1982). This research has shown that neonates
are more active and discriminating, more responsive to outside stimuli, than had ever been thought. It
has even been shown that they are capable of performing complex tasks. In other words, our view of

the infant has been revolutionized. Mahler (personal communication) has reconsidered and rethought
her earlier formulations and has agreed that the word “autistic” does not well describe what we now
know about the neonate.

www.freepsychotherapybooks.org 301
A more recent formulation of what Mahler originally called the autistic phase is that it is the

time during which newborns have the task of adjusting to extrauterine existence, of finding their own

niche in the external world. They have to achieve physiological homeostasis, that is, adequate inner
regulation in synchrony with the vocal and gestural rhythms of their caregiver. Each infant elicits his or

her own mother’s caregiving, and the mother responds with coenesthetic empathy to the needs of a
particular infant. She is enabled to do so by reaching the state described by Winnicott (1956) as

primary maternal preoccupation. Bergman (1982) has attempted to show from the mother’s side how

this particular empathic state is at times reached easily and smoothly and at other times with great
difficulty.

The symbiotic phase, which is reached at around 2 months of age, is of great importance for

separation-individuation theory, since on it rests the idea of a gradual hatching out, a psychological

birth. The findings of contemporary infant research here pointed to the importance of distinguishing
the regressed merger experience of pathology from the attunement and reciprocity of the normal

symbiotic phase. Pine (1981) has hypothesized that what could be referred to as normal merging

occurs during certain brief periods of high drive arousal. Bergman and Chernack (1982), in a paper
dealing with preverbal communication, have shown how, during the symbiotic phase, differentiation

and merging go hand in hand.

Observers agree that attunement, mutual empathy, or communion between mother and infant

are at their height in the period from 2 to 5 months of age. Empathy is not possible without the ability

to freely evoke states of loss of self, while maintaining the ability to regain a state of full awareness.
The same happens in the creative process. Where does such ability come from? We believe that the

blissfulness of the symbiotic stage, which is still longed for in later life, provides us with a reservoir of

self-other experiences, which in normal development are pleasurable and creative.

McDevitt (1981) has elucidated the symbiotic phase from a more cognitive perspective. He

states that by age 2 to 3 months, the infant (1) both anticipates and initiates the pleasure provided by
interaction with the mother; (2) develops a sense of confidence and basic trust in the caregiver and in

www.freepsychotherapybooks.org 302
his or her own initiative; and (3) responds by smiling and direct eye contact. The work of infant
researchers has made us more aware of the capacity of the infant not only to initiate contact but also to

control it through gaze and gaze aversion. Thus, the infant’s sense of self during the symbiotic phase is

fed by experiences that, even at that early period, may be experienced as “his or her own,” especially if
the caregiving environment is responsive to the infant’s more subtle signals and signs. The sense of self

also receives important nutrients from the pleasure and attunement the infant experiences with the

mother. Thus, from early on, there may be two strands to the infant’s experience of self: self-alone and
self-with-other. These should then be the forerunners or beginnings of separation-individuation. To

separate, there must first be self-other and separate-self experiences. Sander (1976) has described
these early experiences of self as being alone in the presence of someone, in Winnicott’s sense. Thus,
the symbiotic phase is the bedrock of libidinal attachment and intimacy on the one hand, and

beginnings of self-alone experiences on the other. Even during the early months, for example, infants
show individual preferences for color, for certain tunes, and for varying amounts of stimulation.

SUBPHASE I-DIFFERENTIATION

The subphase of differentiation begins at the height of symbiosis, when the baby begins more

active and persistent visual and tactile exploration of the surroundings. The baby begins to perceive
things at a greater distance and typically scans the environment, checking back to the mother
regularly. This eminently important process of shifting attention cathexis to the outside is what has

also been called the hatching process. The fully hatched baby, around the age of 9 to 10 months, is
alert, can easily grasp what he or she wants, sits up freely (Resch, 1979), and is characterized by a
general brightening of mood. The differentiation subphase is also the time when unpleasure at the

stranger and even anxiety can begin (Emde, Gaensbauer, & Harmon, 1976). The baby also shows
unpleasure and sometimes cries when left by the mother, but is usually comforted fairly easily by a
nonintrusive mother substitute.

Pushing away from mother and exploration of the environment are quite characteristic of the

www.freepsychotherapybooks.org 303
differentiation subphase. During this time, the child explores, both visually and tactilely, the faces of
individuals other than the mother. The infant is also particularly attracted by appendages that can be
removed, such as eyeglasses, beads, or a pencil in the pocket. All these explorations of both the animate

and the inanimate, of that which can be removed and held by the infant and that which clearly is part of
the other, are important ingredients of the ongoing process of self-object differentiation.

THE PRACTICING SUBPHASE

The practicing subphase begins when the now hatched baby begins to be capable of
independent locomotion. The early practicing period comprises the time of crawling, standing up, and

coasting, whereas the practicing period proper begins with the mastery of upright locomotion. If we
can think of symbiosis as the first blissful stage in human development, the stage of pleasure in
mutuality and recognition and exploration of the mother, we can think of the practicing subphase as

the second blissful period. The mastery of locomotion, at first crawling and then walking, brings with it
an enormous increment of energy and pleasure. The ability to go after and get what one wants by one’s

own efforts, is an immense source of pleasure and satisfaction. Whereas, during differentiation, babies

often cry when their mother or even others walk away from them, beginning locomotion counteracts
the sense of helplessness. This is a period of rapid development, especially of locomotor and

manipulative abilities. The narcissistic investment in the body and in mastery and exploration brings
about a temporary lessening in the investment in the mother, who can now be taken for granted. This

slight lessening of investment in the mother also protects the baby from a full realization of his or her

separateness. The mother is simply assumed to be there unless she is absent for any length of time.
More protracted separation changes the practicing infant’s mood of elation to one of lowkeyedness, a

temporary lowering of mood which is understood to be caused by the need to hold on to the image of
the mother.

Toddlers’ expanding locomotor capacities widen their world; there is more to see, more to hear,

and more to touch. Along with increasing awareness of the outside world goes the more highly

integrated and differentiated knowledge of the body self, as the infant gains increasing mastery over

www.freepsychotherapybooks.org 304
body functions which become more and more intentional and goal directed. Finally, standing and
eventually walking provide a whole new perspective of the world and add further to the small toddler’s

sense of elation and exuberance. Another important characteristic of this period is the relative

hardiness of the infant, who is quite oblivious to the knocks and falls that are, of course, daily
occurrences.

THE RAPPROCHEMENT SUBPHASE

The expansiveness and omnipotence characteristic of the practicing subphase wane as the

toddler increasingly comes face to face with the feeling of separateness caused by frustrations that
occur as explorations are curtailed by obstacles in the real world. The child also has to face the fact that

mother is not always automatically at hand to smooth the way for his explorations. Indeed, there are
times when she curtails them in the interest of protecting the child’s safety. The infant’s former relative

obliviousness of the mother is now replaced by active approaches to her.

This rapprochement subphase is again conceptualized in two parts—early rapprochement and

the rapprochement crisis. During early rapprochement, the generally good mood of the practicing
period still prevails as the toddlers attempt to bridge the gap that they are now beginning to perceive

between themselves and their mother. Toddlers begin to want to share everything with their mother;
most characteristically, they will bring things and put them in their mother’s lap, but they will also seek
out her active participation in their activities. The availability of the mother during this particular

period is of great importance, but even under the most optimal conditions, the maturational spurt of
toddlers’ cognitive development makes them realize their separateness and relative helplessness.
Toddlers, during rapprochement, wish to be autonomous and find all hindrances to their autonomy

extremely disturbing, whether emanating from their own activities, from curtailment by adults, or
from their inability to do what they would like.

The child’s recognition of his or her separateness and limitation threatens his or her sense of

www.freepsychotherapybooks.org 305
omnipotence, which is still very closely connected with the child’s self-esteem. In addition, toddlers
have to come to terms with the fact that their mother’s wishes and their own by no means always

coincide. Toddlers still believe in the omnipotence of their parents and become very angry and

sometimes desperate if the parents cannot do for them what they want. “He thinks we can do
everything,” a mother of a rapprochement-age toddler said recently. Some weeks later, the same

mother said, with great relief: “He’s beginning to accept that somebody or something can be gone and

that I cannot do anything about it.” For example, that morning, when the cereal he had wanted was
gone, her son agreed to eat a piece of bread and butter rather than insisting or crying for more cereal.

While wanting to be independent and autonomous, rapprochement toddlers also often want to

control the whereabouts of their mother and want her to partake in all their activities. Anxious clinging

or daring darting away, hoping to be caught up and brought back by the mother, are typical behaviors.
The toddler at this age does not easily tolerate the mother’s attention being elsewhere and is typically

quite demanding.

In the course of the rapprochement subphase, the child begins to have a separate mental self.
Beginning language and symbolic functioning are very important in bringing a resolution of the

rapprochement crisis. Being able to know and name others and eventually being able to know and
name oneself are important indicators of internal processes that take place at that time. The child

begins to know “mine” (Bergman, 1980), but “mine” at that time can express a wish or demand as well

as a fact. “Mine” is a precursor to naming oneself or using the personal pronoun.

If development goes reasonably well and the mother is reasonably available to the toddler, the
rapprochement crisis is eventually resolved by way of identification and internalization. Successful

resolution of the rapprochement crisis by no means always takes place, however. A badly resolved

rapprochement crisis leads to intense ambivalence and splitting of the object world into good and bad.
The maternal representation may be internalized as an unassimilated bad introject. McDevitt and

Mahler (1980) cite four conditions that would lead to poor resolution of the rapprochement crisis: (1)

the love object is disappointing and unavailable or excessively unreliable and intrusive; (2) the child

www.freepsychotherapybooks.org 306
experiences the realization of his or her helplessness too abruptly and too painfully, resulting in a too
sudden deflation of the child’s sense of omnipotence; (3) there has been an excess of trauma; and (4)

the child experiences to an unusual degree the narcissistic hurt of the preoedipal castration reaction

which accompanies the discovery of the anatomical difference. Under such conditions, rapprochement-
type behaviors persist rather than giving way. Such behaviors include excessive separation anxiety,

depressive mood, passivity or demandingness, coerciveness, possessiveness, envy, and temper

tantrums.

ON THE WAY TO OBJECT CONSTANCY

The fourth and final subphase of separation-individuation is called “on the way to self and

object constancy” and is recognized as being open-ended. In the context of separation-individuation

theory, self and libidinal object constancy (the achievement of this final subphase) is not seen as a fixed

fact, but rather as an ongoing, lifelong process. Nevertheless, a child who has successfully resolved the

rapprochement crisis has made an important qualitative change that is quite unmistakable to
observers.

Self-constancy develops along with object constancy. In the fourth subphase, the toddler’s sense

of self includes actions as well as perceptions and feelings. The toddler beings to like to be admired for
what he or she can do. Earlier, doing and achieving mastery were enough. Now, the participation of the
“other” is an important ingredient in the pleasure of mastery. The qualitative change that comes with

the resolution of the rapprochement crisis is comparable to the qualitative change that comes when
hatching is accomplished.

Hatching, which means living in the outside world while taking the mother for granted, resolves

the crisis of differentiation when the infant, for the first time, becomes exceedingly sensitive to

separation from the mother. The infant needs to take the mother for granted-that is, to stay
omnipotently at one with her, while at the same time, turning to the outside world with curiosity,

www.freepsychotherapybooks.org 307
pleasure, and eagerness. The rapprochement crisis is the second crisis of separation. To bring it to a
satisfactory resolution, the child has to achieve a degree of internalization, which allows the lessening

and eventual relinquishment of omnipotent control. The development of the symbolic function is

intimately connected with the lessening of omnipotent control, as it allows the senior toddler to live
out and practice in play some of the wishes and fears that arise from the conflict over autonomy and

the need or wish to still be “at one” with the powerful, good mother.

SUMMARY

The delineation of the subphases of the separation-individuation process describes the

psychological birth of the human infant. Out of the union or attunement of symbiosis with the mother,

the infant grows to an increasing awareness of separateness and develops his or her own unique
characteristics, in part inborn, in part the result of the intimate interaction between the infant and his

or her love objects, the parents. The infant also grows from a stage in which the object is only dimly

perceived as outside and separate, toward the attainment of a unique attachment to the love object; the
infant grows further, toward the stage of loving in which a positive image can be maintained even in

the face of anger and frustration and in which the capacity for concern for the other takes the place of

the demand for omnipotent control.

Each overlapping stage paves the way for the next. Thus, the solid and pleasurable period of
symbiosis means that the child will be more prepared for the stage of differentiation to follow and will

meet the stranger or strangeness of the outside world with greater confidence and less anxiety.
Similarly, a rich practicing subphase which affords ample opportunity for exploring the outside world
while remaining in contact over distance with a supportive and admiring caregiver will provide the

child with a reservoir of resources with which to withstand the onslaughts of the crisis of
rapprochement.

The task for the parent changes as the separation-individuation process progresses. During

www.freepsychotherapybooks.org 308
practicing, the parent has to be able to follow the cue of the child who now requires more space in
which to try out his or her burgeoning abilities. It is during the period of rapprochement that it

becomes more difficult for the mother to remain emotionally available, as the child who has appeared

more autonomous during practicing now returns to the mother often with conflicting and unfulfillable
demands. Nevertheless, parents who can be playful and patient during the rapprochement period will

help the child toward more favorable resolutions during the period on the way to object constancy.

While each subphase paves the way for the next, each subphase also contains a potential for

repair if optimal conditions have not prevailed in the preceding period. Each subphase is also separate
and discrete (Mahler, Pine, & Bergman, 1975), with its own rewards as well as its own tasks. The little

child’s personality is pliable and patterns are not fixed, leaving a great deal of room for adaptation. For

example, a particular child whose symbiotic phase had been colored by his mother’s depression during
that period seemed to differentiate rather late. It seemed at first like a possible danger signal. It later

seemed, however, that this child had found a way of making up for what he had missed by remaining in

the symbiotic orbit for a longer time by emerging into the outside world only slowly, as he became
ready to do so. Since this particular mother could respond much better to the active child of

separation-individuation, he began to catch up and developed well as time went on.

The theory of separation-individuation is a dynamic developmental theory. It leaves room for

progression and regression as well as for the back-and-forth movement between needs for closeness

and attachment and needs for exploration and disengagement.

An important result of the study of separation-individuation is the enrichment of knowledge on

several topics which, although already familiar to psychoanalysts and developmentalists, were further

illuminated during the years of the research. We would like to mention a few contributions that have

dealt with psychoanalytic concepts from a developmental perspective. In an important paper on the
“Development of Basic Moods,” Mahler (1966) considers the tendency to depressive moods in women

and ties it to conflicts arising during the rapprochement subphase. Furer (1967) writes about

developmental aspects of the superego. He considers “identification with the comforter” as a

www.freepsychotherapybooks.org 309
forerunner of the superego and feels that this identification with the active mother “increases the
child’s capacity to bind its aggression and thus helps bring about the required reaction formation.” In

an examination of the relationship between adaptation and defense, Mahler and McDevitt (1968) say:

The child’s experiences over the course of time, on the basis of his drive
and ego endowment, lead to more or less successful adaptation. His
adaptive style contributes to his character traits, as do his defense
behaviors. We have observed in our research the process by which these
behaviors gradually become internalized as more or less successful
defense mechanisms. (p. 100)

McDevitt (1982) traces the emergence of hostile aggression in the course of the separation-
individuation process. Bergman (1982) describes the development of the girl during separation-

individuation, with implications for later development.

We would like to mention some other important issues that have been elucidated by the

developmental point of view and by the detailed scrutiny of our day-to-day observations in the study of
separation-individuation. The first of these is stranger anxiety.

It was Spitz (1957) who first drew attention to stranger anxiety and considered it the second

organizer at the age of 7 months. This phenomenon has attracted a great deal of attention since Spitz
first described it, and the separation-individuation study has contributed to a more detailed
understanding of a variety of phenomena subsumed under the concept of stranger reactions. Stranger

anxiety is the most visible of a large array of phenomena with which an infant indicates increasing
recognition of mother as unique as well as interest and curiosity in the world beyond mother. Thus, we
prefer the term “stranger reactions” rather than “stranger anxiety.” Stranger reactions can include a

variety of affects, ranging from interest and curiosity to wariness and finally anxiety and distress.
Stranger reactions can even be directed, at a certain age, to the mother or father if they look different
from the way they usually do. Early in the study, we heard about a little boy who, at the age of 4
months, cried when he saw his mother wearing a shower cap. We recently heard of a little girl, age 5

www.freepsychotherapybooks.org 310
months, who was quite concerned when she saw her father after he had shaved his beard and

mustache. These are early indications that the child is beginning to form an inner image, which is

disturbed if what the child sees is suddenly very different from what she or he expects. It seems to us

that the timing, the kind, and intensity of the stranger reaction is intimately connected with the
mother-child relationship. For example, we recently saw a little girl who showed a marked stranger

reaction, even anxiety, at the unusually early age of 3 to 4 months. She was the daughter of a young
mother from a foreign country who had not yet learned the language very well. This young woman had

been quite depressed after she married an American man and came to live in this country. After the

birth of her daughter, her mood improved and she developed an extremely close symbiotic
relationship with the girl. Mother and daughter seemed rather insulated from the rest of the world in

which they lived. Thus, it seemed very interesting that this particular little girl showed such early

stranger reaction and reacted to outsiders not with curiosity or interest, but with displeasure. When
she was seen again at the age of 6 months, she was still rather wary but willing to engage in play with a

stranger as long as her mother stayed close by.

“Customs inspection” is a term that was coined during the separation-individuation study to

describe another type of stranger reaction. This is the way in which the child in the period of
differentiation, around 7 to 10 months, will examine the faces of strangers, both visually and tactilely,
with great interest and absorption. Not all children feel free to engage in this activity with the same

amount of intensity and interest, but most will show some interest in the stranger and wish to touch

and explore parts of the stranger’s face or at least such appendages as beads or eyeglasses.

Yet another kind of stranger reaction was recently observed in a little boy during the
differentiation subphase. This little boy seemed to enjoy attracting the interest of strangers, and he had

learned that when he shouted, most people would look around and smile at him. Thus, in strange

places, he would often shout at strangers and then show great pleasure when they paid attention to

him. Separation-individuation studies have shown us that the outside world is not just a threat to the
unique mother-child relationship, but it is also often a source of great excitement and pleasure.

www.freepsychotherapybooks.org 311
It has long been known that separation from their mother is often painful to children during the

first 2 to 3 years of life. Once again, the study of separation-individuation has given us a developmental

view of such separation reactions. It has shown us that sensitivity to separation is very different during
the different subphases of the separation-individuation process. Of course, each child’s sensitivity to

separations will also be determined by the mother-child relationship and by the way the mother
handles such separations. Regardless of these individual differences, however, we were able to see a

developmental line of separation reactions (McDevitt, 1980b). The period of the differentiation

subphase is a time when most infants first show active protest or distress at separation. This seems to
be when they are on the verge of being able to move independently themselves and are trying to do so,

but cannot do so yet. It is at this time that they seem to perceive their mother walking away from them
and often cry. Most infants at that time accept substitutes without too much difficulty, but the period of
differentiation is a sensitive one. It is as if the infant’s capacities of discrimination are ahead of his or

her capacity to act. The infant is acutely aware that when mother walks away he or she is not yet able
to follow her or call to her. However, the infant has a beginning image of the mother and begins to look
at the door through which she might have left. It is also often comforting to the infant to be taken to a

window. The child seems, at this time, to have a vague feeling that mother is out there. Thus, going to
the window and observing the world in which she is somehow known to be seems to ameliorate the
feeling of helplessness or entrapment that might otherwise be present. One mother who was especially

sensitive observed that her little boy, at a somewhat older age, would wait by the door in the late
afternoon, thereby indicating to her that he was waiting for his father to come home.

By the time they reach the practicing period from about 8 to 16 months, children are quite

aware that their mother might leave and may protest her leaving as soon as she prepares to do so. On
the other hand, their newly found ability to crawl, and later to walk, seems to compensate to some

extent. No longer are they so dependent on a mother substitute for comfort. Children are now more

able to do things for and by themselves that are enjoyable and exciting. They can also attempt to follow

the mother. They can go to the door through which she left. They can be more actively engaged with

substitute caretakers in the mother’s absence.

www.freepsychotherapybooks.org 312
Nevertheless, during the early practicing period, from about 8 to 13 months, it was observed

that infants tended to become much less active when their mother was out of the room. Pleasure and

cathexis in the outside world was definitely reduced, and infants began to withdraw into a state called
“low-keyedness” (Mahler & McDevitt, 1968). Low-keyedness was conceptualized as a state of holding

on to the image of the absent mother by reducing activity and stimulation from the outside. This
withdrawal and low-keyedness can be quite dramatic. Equally dramatic is the way in which the child at

this age will immediately come back to life as soon as he or she is reunited with mother.

It is during the period of practicing that the invisible bond with mother is at its height, and the
infant seems to feel as if she were at one with him or her, even while at a distance. Infants at this age

characteristically will play at a distance from their mother but periodically look at her and check back,

apparently receiving sustenance from the visual contact. Absence of the mother at this age, if it is too

prolonged, and if no adequate substitute caretaker is available, disrupts too suddenly the illusion of
oneness with the mother and thus disrupts the elation that is so characteristic of the practicing

subphase. It may also lead the child to become restless and search for the mother or to get into

dangerous or precipitous situations, probably with the hope of being rescued by her.

The increased sense of separateness during the period of rapprochement brings with it a sense

of vulnerability, loneliness, and often helplessness. Thus, most children become much more sensitive
to separation. Toddlers of the rapprochement subphase are often constantly preoccupied with their

mother’s whereabouts. They insist on following her through the door and will protest vigorously when

separated. Phenomena such as shadowing and darting away have been described as characteristic of
toddlers during this period. They can be quite insistent on their mother’s exclusive attention and, if it is

not easily available, attempt to get this attention by clinging and coercion. Substitutes are no longer as

easily accepted, and often familiar substitute caretakers, even fathers, are angrily rejected when the

mother is desired. A kind of splitting often occurs in which the absent mother is longed for and the
present caretaker is rejected. Beginning feelings of ambivalence are directed toward the mother, who

is often seen as interfering with the child’s budding autonomy. Thus, the mother is split into the good

www.freepsychotherapybooks.org 313
absent mother who is longed for and the bad present one who is rejected. At the time of reunion, the
mother who returns is no longer necessarily experienced as the pleasurable, life-giving force that she
was during practicing. Instead, when she returns she is sometimes avoided. The child veers away and

seems angry instead of smiling at the mother’s return, and it takes considerable time for a pleasurable
reunion to be effected.

The beginning abilities for symbolic play and language help the toddler withstand separation

from the mother. It is only with the advent of the fourth subphase however, on the way to object
constancy, that mother’s absence can truly be accepted and the child can be content for longer periods

of time without her. By then, the child can understand quite well where mother or father is when they

are not with him or her and can pleasurably anticipate their return. Symbolic play and imitation are

important tools for the mastery of separations. These can be played out endlessly by children of
different ages, beginning with the simple peek-a-boo of the young infant.

It is important to remember that Mahler’s study of separation-individuation was designed to

study the emergence of separateness, not the reaction to separation. Children were studied in the

caretaking presence of their mothers. Yet, even in this setting, mothers would leave the room for brief
periods for their interviews, providing some insight into the developmental reactions to separation

from mother.

DISCUSSION

In the beginning of this essay, we briefly mentioned some of Freud’s ideas about early
development. His ideas about introjective and projective mechanisms were an early attempt to

conceptualize how the infant starts to distinguish self and nonself on the basis of other than “reality-

ego” considerations. We have attempted to convey a number of the pathways that Mahler and her co-

workers have taken to elucidate this and many other related issues. Clearly, Mahler agrees with
Freud’s contention that the infant and child can normally develop structures on the basis of factors

www.freepsychotherapybooks.org 314
other than those that Freud referred to as reality-ego related. Mahler’s conceptualization of libidinal
object constancy and the phases of development that lead up to libidinal object constancy are clearly

instances of factors that are not simply based on the reality ego.

We deliberately have not used the term “cognitive” in contrast to “emotional” factors, since we
believe this type of dichotomy is, for the most part, not a useful one in early development. One might

say, for example, that Mahler and Piaget both refer to a series of cognitive structures developed by the

infant or child, but to some extent they are talking about different types of cognitive structures.

Moreover, for Mahler the intermesh of the infant’s and mother’s affective states is often a reliable
indicator or predictor of how the infant’s structures will develop. Thus, Mahler maintains (as did

Freud) that cognitive structures that develop in relation to the self (selves) and important object

representations follow different developmental lines than other cognitive structures such as these
described by Piaget. If this is the case, it raises questions about the relationship between observations

and theoretical concepts from a psychoanalytic perspective, on the one hand, and infant experiments,

observations, and theoretical concepts of researchers from other perspectives (such as those of
cognitive and learning theorists), on the other.

To be more concrete, let us take the example, cited earlier, in which Mahler recently altered her
concept of the autistic phase because of current infant research. Clearly, contemporary studies have

been striking in pointing out the early perceptual and response capabilities of the infant. Moreover, a

number of psychoanalysts, such as Stern, have pointed out that these studies contradict aspects of
Mahler’s and Freud’s thought. Even though the autistic phase is not a central concept to Freud (nor, for
that matter is primary narcissism), the examination of this issue might elucidate some of the

difficulties in comparing findings that are couched in psychoanalytic terms with findings from other
theoretical points of view. This examination might also touch on some of the difficulties of formulating
psychoanalytic concepts.

Freud’s notion of primary narcissism can be interpreted in several ways, but one narrow

interpretation of Freud (or Mahler) is that he was referring primarily to the building of rudimentary

www.freepsychotherapybooks.org 315
representations of self and object by the infant. During primary narcissism, the infant is not concerned
with the object as object—not because the infant cannot discriminate the object, but rather because no

accumulation of experience (in normal development) has occurred that leads the infant to anticipate a

consistent or long period of frustration of primary gratification. If such an accumulation of experience

occurs very early, Freud implicitly predicts traumatic results. Freud (1915) states that even though

“the ego is autoerotic [and] it has no need of the external world, in consequences of experience.…it

acquires objects from that world and, in spite of everything, it cannot avoid feeling internal instinctual
stimuli…as unpleasurable.” The purified pleasure ego, then, develops as a response to unpleasurable

stimuli, and although it brings the infant closer to the object, it also causes some distortion in the
infant’s rudimentary sense of reality.

Freud is here making a unique type of prediction, a prediction that should differentiate to some
extent his theoretical position from other positions. There are, of course, difficulties in testing these

ideas. We have little idea of the time periods that Freud is postulating. It may be difficult to find ways to

measure the infant’s postulated split of the world into all good (inside) and all bad (outside). Freud is
silent about factors that might influence (retard or advance) the development of the purified pleasure

ego, nor does he tell us in detail about factors that might continue the purified pleasure ego longer than
developmentally appropriate or that might lead to the dissipation of the structure earlier than might be

desirable. In short, Freud tells us very little that would enable us to develop a testable theory from his
writings. It is obviously difficult, therefore, to compare his account of development with other
accounts. We believe, however, that even in Freud’s sparse writings on early development, there are

ideas rich enough that if one rigorously applied his assumptions and tried several time sequences, it
might be possible to empirically test his conceptualizations.

Why, one might ask, have we in our summary of Mahler’s contributions reviewed some
relatively obscure sections of Freud’s writings? We have done so in part because Freud’s writings are

finalized and in some ways are a simpler version of an early developmental schema than are other
psychoanalytic theories. In the main, however, we wish to give a brief illustration of both the

www.freepsychotherapybooks.org 316
difficulties and the potential of even such a seemingly “discarded” (Lichtenberg, 1982) part of Freud’s
writings as his metapsychological papers. In our opinion, Mahler’s pioneering work has some of the
same difficulties but clearly much more potential because of the richness of the observations and

concepts of her work.

Let us go, therefore, to Mahler’s conceptualization of the autistic phase, which we roughly

equated with Freud’s ideas of primary narcissism and autoeroticism. This concept is one of the few

aspects of Mahler’s writings that has been actively disputed. In addition, it is clearly not a central
concept for her (her research has not included this phase of development), and therefore might prove

useful as an illustration of the richness of her ideas.

In our opinion, the concept of the autistic phase has been translated as a phase without stable
representations, or an “objectless” period. This, of course, is one possible translation, but one not

necessarily in keeping with Mahler’s ideas or with Jacobson’s notion of the psychophysiological self,
which Mahler has utilized. The key to this notion is the definition of an “objectless” period. If one

means a period where no stable perceptions or memories are retained, then the first month of life is

probably not an objectless state. However, Mahler, Freud, and Jacobson all describe the state of the
infant in this period with respect to gratifying and aversive experiences. They maintain that the infant

is interested not in the object, but in the gratification or maintenance of homeostasis or in something
other than the object itself. The fact that the infant possesses aspects of the rudimentary ego does not

alter this concept. What, then, is the contradiction with other research? For if one means by

“objectless” state an infant whose main interest is in gratification and who is not motivated or
interested in the object, (we are of course simplifying), then there is no contradiction.

Part of the difficulty, then, may lie in the manner in which the concepts are stated. Or perhaps it

is more accurate to say that the difficulty lies in the fact that the concepts are incompletely stated. Here
Mahler has not gone beyond Freud, and all the questions we previously asked about Freud’s ideas can

be appropriately applied to Mahler. We believe, however, that all the conditions are present for

separation-individuation concepts to be put in the form of a theory that can both do justice to the

www.freepsychotherapybooks.org 317
richness of psychoanalytic concepts and at the same time be empirically rigorous.

The line of thought and research that Mahler has pursued in her separation-individuation

research is probably the outstanding example in psychoanalysis of how concepts have guided research

and, in turn, have themselves been enriched and expanded by the research. Given the outstanding
quality and amount of this work, however, we might briefly summarize what we believe are some of

the difficulties in this conceptual field. Difficulty in knowing how to conceptually coordinate

separation-individuation and other aspects of psychoanalysis may in part be an empirical question. At

this point, however, it is hard to know how to coordinate concepts such as psychosexual stages, drives,
or other aspects of psychological structure in ego psychology. For example, one might ask if drive is a
concept that is compatible with the separation-individuation theoretical framework and if so, does a

concept like drive add to this framework? How does one think of psychosexual factors in relationship

to processes of differentiation or individuation? Many questions such as these can be asked, and it is
not a criticism but rather a comment about psychoanalytic thought that there are few substantive

attempts to logically order and coordinate these concepts. Only if this is done can firm empirical

consequences be derived from a theoretical position.

A similar point can be made about separation-individuation concepts even outside the context
of the more general psychoanalytic concepts. It is difficult to know the logical status of certain

concepts. That is, it is hard to know which concepts are absolutely essential and which are more
peripheral. It is also difficult to know how to translate certain concepts into ideas that have firm
empirical consequences. For example, there are many examples in the research of children who

deviated from what would seem to be expected theoretical norms, but the delimiting conditions were

not often given in generalizable statements. A substantial elucidation of these difficulties is beyond the

scope of this chapter, however.

We have attempted to give one example of how some of Mahler’s and Freud’s less-developed
ideas and psychoanalytic explanations may be powerful if stated in more specific terms. When that

occurs, we may see that even the concept of the autistic phase has a good deal of explanatory power.

www.freepsychotherapybooks.org 318
ACKNOWLEDGMENTS

The authors wish to thank David Pollens for his assistance in the preparation of this article.  

REFERENCES

Bergman, A. (1980). Ours, yours, mine. In R. F. Lax, S. Bach, & J. A. Burland (Eds.), Rapprochement: The
critical subphase of separation-individuation (pp. 199-216). New York: Aronson.

Bergman, A. (1982). Considerations about the development of the girl during the separation-
individuation process. In D. Mendell (Ed.), Early female development: Current
psychoanalytic views (pp. 61-80). New York: Spectrum Publications.

Bergman, A., & Chemack, M. (1982). From command to request: The development of language in the
treatment of symbiotic psychotic child. International Journal of Psychoanalytic
Psychotherapy, 9, 583-602.

Brazelton, T. B. (1974). The origins of reciprocity: The early mother-infant interaction. In M. Lewis & L.
A. Rosenblum (Eds.), The effect of the infant on its caregiver (pp. 49-76). New York:
Wiley.

Brazelton, T. B. (1981). Neonatal assessment. In S. I. Greenspan & G. H. Pollock (Eds.), The course of life:
Psychoanalytic contributions toward understanding human development (Vol. 1) (pp. 203-
233). Washington, D.C.: U.S. Government Printing Office.

Emde, R., Gaensbauer, T., & Harmon, R. (1976). Emotional expression in infancy: A biobehavioral study.
Psychological Issues, 10 (1 Monogaph 37).

Freud, S. (1900). The interpretation of dreams. Standard Edition, 5, 339-621.

Freud, S. (1914). On narcissism: An introduction. Standard Edition, 14, 73-102.

Freud, S. (1915). Instincts and their vicissitudes. Standard Edition, 14, 117-140.

Freud, S. (1917). Mourning and melancholia. Standard Edition, 14, 243-258.

www.freepsychotherapybooks.org 319
Furer, M. (1967). Some developmental aspects of the superego. International Journal of Psycho-Analysis,
48, 277-280.

Kernberg, O. (1980). Internal world and external reality. New York: Aronson.

Lichtenberg, J. D. (1982). Reflections of the first year of life. Psychoanalytic Inquiry, 1, 695-729.

Loewald, H. W. (1984). Review of The selected papers of Margaret S. Mahler. Journal of the American
Psychoanalytic Association, 32, 165-175.

Mahler, M. S. (1942). Pseudoimbecility: A magic cap of invisibility. In The selected papers of Margaret S.
Mahler, vol. 1 (pp. 3-16). New York: Jason Aronson.

Mahler, M. S. (1952). On child psychosis and schizophrenia: Autistic and symbiotic infantile psychoses.
In The selected papers of Margaret S. Mahler, vol. 1 (pp. 131-153). New York: Jason
Aronson.

Mahler, M. S. (1961). On sadness and grief in infancy and childhood: Loss and restoration of the
symbiotic love object. In The selected papers of Margaret S. Mahler, vol. 1 (pp. 261-279).
New York: Jason Aronson.

Mahler, M. S. (1966). Notes on the development of basic moods: The depressive affect. In The selected
papers of Margarets. Mahler, vol. 2 (pp. 59-75). New York: Jason Aronson.

Mahler, M. S. (1975). On the current status of the infantile neurosis. In The selected papers of Margaret
S. Mahler, vol. 2 (pp. 189-193). New York: Jason Aronson.

Mahler, M. S. (1979). The selected papers of Margaret S. Mahler (Vols. 1, 2, & 3). New York: Jason
Aronson.

Mahler, M. S. (1981). Aggression in the service of separation-individuation: Case study of a mother-

daughter relationship. Psychoanalytic Quarterly, 50, 625-638.

Mahler, M. S., & McDevitt, J. B. (1968). Observations on adaptation and defense in statu nascendi. In The
selected papers of Margaret S. Mahler (Vol. 2) (pp. 99-112). New York: Jason Aronson,
1979.

www.freepsychotherapybooks.org 320
Mahler, M. S., Pine, F., & Bergman, A. (1975). The psychological birth of the human infant. New York:
Basic Books.

Mahler, M. S., Ross, J. R. & DeFries, Z. (1949). Clinical studies in benign and malignant cases of
childhood psychosis (schizophrenia-like). American Journal of Orthopsychiatry, 19, 295-
305.

McDevitt, J. B. (1980a). The emergence of hostile aggression and its defensive and adaptive modifications
during the separation-individuation process. Brill Memorial Lecture, prescribed at the
New York Psychoanalytic Society, New York.

McDevitt, J. B. (1980b). The role of internalization in the development of object relations during the
separation-individuation phase. In R. L. Lax, S. Bach, & J. A. Burland (Eds.),
Rapprochement: The critical subphase of separation-individuation (pp. 135-149). New
York: Aronson.

McDevitt, J. B. and Mahler, M. S. (1980). Object constancy, individuality and internalization. In S. I.

Greenspan and G. H. Pollock (Eds.), The course of life: Vol. 1. Infancy and early childhood
(pp. 407-424). Washington, D.C.: U.S. Government Printing Office.

Pine, F. (1978). On the expansion of the affect array: A developmental description. In R. F. Lax, S. Bach,
and J. A. Burland (Eds.), Rapprochement (pp. 217-233). New York: Aronson, 1980.

Pine, F. (1981). In the beginning: Contributions to a psychoanalytic developmental psychology.

International Review of Psychoanalysis, 8, 15-33.

Resch, R. C. (1979). Hatching in the human infant as the beginning of separation-individuation: What it
is and what it looks like. The Psychoanalytic Study of the Child, 34, 421-441.

Sander, L. W. (1976). Issues in early mother-child interactions. In E. N. Rexford, L. W. Sander, & T.

Shapiro (Eds.), Infant psychiatry: A new synthesis (pp. 127-147). New Haven: Yale
University Press.

Settlage, C. F. (1974). Danger signals in the separation-individuation process. In D. Bergsma (Ed.), The
infant at risk. (March of Dimes National Florida Symposium, Vol. 10). unpaginated

www.freepsychotherapybooks.org 321
Spitz, R. (1957). No and yes: On the genesis of human communication. New York: International
Universities Press.

Stern, D. (1971). A micro-analysis of mother-infant interaction. Journal of the American Academy of

Child Psychiatry, 13, 501-517.

Stern, D. (1974). The goal and structure of mother-infant play. Journal of the American Academy of Child
Psychiatry, 13, 402-421.

Stern, D. (1982). Implications of infancy research for clinical theory and practice. Paper presented at
the 13th Annual Margaret S. Mahler Symposium, Philadelphia.

Winnicott, D. (1956). Primary maternal preoccupation. In Collected papers: Through Pediatrics to

Psycho-analysis (1958, pp. 300-305). London: Tavistock Publications.

www.freepsychotherapybooks.org 322
10
OTTO KERNBERG: PSYCHOANALYSIS AND OBJECT
RELATIONS THEORY; THE BEGINNINGS OF AN
INTEGRATIVE APPROACH

MONICA CARSKY, PH.D.

STEVEN ELLMAN, PH.D.

It is a difficult task to attempt to summarize and critique Otto Kernberg’s psychoanalytic

contributions, for he has presented the most systematic and wide-sweeping clinical and theoretical

statements of the last decade, perhaps even since Freud. His work touches on many if not most of the

topics that have been of interest to contemporary analysts. In addition, he has been instrumental in
introducing many topics to the American psychoanalytic community. Even reviewers who have been
sharply critical of Kernberg, such as Calef and Weinshel (1979), have stated that “no other single

colleague has been so instrumental in confronting American psychoanalysts with Kleinian concepts
and theories” (pp. 470-471). Clearly, this is damning Kernberg with faint praise, since much of the
American psychoanalytic community is in opposition to many aspects of Melanie Klein’s theoretical

contributions. Although there is no question that Kernberg has been strongly influenced by Kleinian
concepts, however, there is also no question that he is attempting to integrate many different parts of
what is called the British object relations school, as well as aspects of Freudian thought, ego

www.freepsychotherapybooks.org 323
psychology, and different strands of research in neurophysiology and physiological psychology. This
list is by no means complete. Kernberg is strongly interested in research in affect, for example, whether
from psychoanalysis, physiology, or academic psychology.

Given that we are dealing with a theoretical integration of such large proportions, one that
blends the familiar and unfamiliar, it is not surprising that a number of critics have pointed out various

difficulties in Kernberg’s theoretical attempts. Before we try to evaluate Kernberg’s writings, it is

important to put our critical stance into an appropriate historical perspective. In our opinion, there is
no psychoanalytic theorist whose theory would stand up to some of the criticism that has been

directed at Kernberg. Psychoanalysis has yet to produce a full theory as defined by philosophers of

science such as Nagel (1961) or Popper (1962). Leaving aside philosophical conceptions of theory, it is

clear to most students of Freud’s or Hartmann’s writings that many concepts remain without clear
definition and are not well integrated into a theoretical structure. If we are to evaluate Kernberg

reasonably, it must be within contemporary psychoanalytic standards. In addition, much of

contemporary criticism in psychoanalysis is not based on either logical or empirical grounds but rather
is often simply or mostly a reflection of the critics’ values. We will attempt to evaluate Kernberg’s

contributions in terms of both his stated aims and our view of the state of contemporary

psychoanalytic theory.

Some of our views of contemporary theory have been stated elsewhere (Ellman & Moskowitz,

1980; Moskowitz & Ellman, unpublished manuscript), but for this introduction we will briefly restate

them. We believe that in many of the social sciences it is difficult to state clearly how the different
aspects of a theory are organized. Thus, at times, it may not be clear what are the central assumptions

of a given theory, as opposed to assumptions or statements that are more peripheral. Frequently, the
coordinating logic (see Nagel, 1961) of a theory is also unclear, so that it is hard to know what
assumptions should be combined to predict or to explain a given event. Most often, however, the
phenomena to be explained are relatively clear and are at least somewhat separated from the theory

itself. Psychoanalytic theory shares some of these difficulties, and it is not clear at times what

www.freepsychotherapybooks.org 324
phenomena some psychoanalytic theories are addressing.

Fortunately, Kernberg usually indicates clearly what phenomena he is trying to explain. In our

discussion, we will initially introduce the clinical phenomena that have been the main impetus for
Kernberg’s theorizing, and from that point on will go back and forth between Kernberg’s theorizing
and the clinical phenomena or observations he wishes to explain. It will be clear as we proceed that the

observations and the theory become more and more intertwined. Nevertheless, we think that

Kernberg is attempting to explain important clinical phenomena. In fact, this is a major reason for his
present importance in psychoanalysis. We do not believe that Kernberg’s clinical observations are

simply or mainly an artifact of his theorizing.

It may be hard to see in the present context why it would be necessary to mention any of
Freud’s writings. From our point of view, however, there is a somewhat neglected aspect of Freud’s

work that is particularly germane to most object relations theorists. (We are obviously including
Kernberg as an object relations theorist.) This is most clearly seen in Freud’s metapsychological

papers, where he frequently presents his views on early development. Certainly Freud’s (1915) view of

the developmental phase that he termed the purified pleasure ego has been included in one form or
another in the work of a number of contemporary authors (Kohut, 1966, 1971; Mahler, 1968; Mahler,

Pine, & Bergman, 1975). In these writings, Freud deals with what he termed the origins of the three
polarities of the mind and sets the stage for the differentiation of types of identification processes. That

is, Freud (1914, 1917) began to conceptualize the process of introjection or early identifications with

more specific and developmentally later types of identifications. Melanie Klein, in many ways,
expanded on this phase of Freud’s work, as well as Freud’s theory of instinct1 as stated in Beyond the
Pleasure Principle (1920). To greatly oversimplify Klein’s work, one may say that she was the first
psychoanalytic theorist to attempt to integrate object relations and instinctual points of view. Kernberg

has clearly stated that he is also attempting to unite drive and object relations points of view. In

addition, he is attempting to stay within a general ego psychological framework, so that the
psychoanalytic conception of drive that Kernberg is utilizing arises from Hartmann’s emendations and

www.freepsychotherapybooks.org 325
clarifications of Freudian theory. Thus, in a later era, Kernberg is attempting to integrate important
aspects of the British object relations and ego psychological points of view. Given this broad statement,
we should point out that many different and at times divergent points of view are encompassed both

between and within these two so-called points of view. Kernberg is both eclectic and selective, but he is
trying to utilize the concepts that he regards as essential to each position.

Kernberg has attempted to combine at least four elements from either an ego psychological or

object relations perspective. These are the following:

1. Structure. Although Kernberg has moved the concepts such as self-representation, self-image,

and so forth into a more central focus, he has retained Freud’s tripartite structure of ego, id, and
superego. As we will see, with most of these concepts Kernberg utilizes object relations theorizing to a
greater extent when dealing with questions of structuralization early in childhood development and

utilizes the tripartite model in later childhood development, particularly in the oedipal period.

2. Defense. Although defense is certainly a part of structure, the concept of defense is important
enough in Kernberg’s writings to warrant special notice. By and large, what Kernberg calls low-level

defenses are those that have been discussed by object relation theorists (such as splitting and
projective identification), whereas most of Kernberg’s high-level defenses (such as repression and

isolation) stem from Freudian and ego psychological theorists. In Kernberg’s conceptualizations, the
main defenses utilized are thus an important indication of the general state of an individual’s
psychological structure.

3. Development. Kernberg has attempted to integrate the concept of object relations phases
(schizoid, depressive phase) with the concept of psychosexual development and Mahler’s (1968;

Mahler, Pine, & Bergman, 1975) developmental findings. Here again, one may see the viewpoint of

Mahler and object relations being used more extensively in considering preoedipal development, while

Kernberg utilizes ego psychological concepts in considering oedipal development.

www.freepsychotherapybooks.org 326
4. Instinct or drive. This is a concept that Kernberg has consistently maintained in his theorizing.

Since some object relations theorists, such as Fairbairn or Guntrip, have explicitly rejected Freud’s,

Hartmann’s, or Klein’s concept of drive, Kernberg is not combining two points of view but rather
including this aspect of Freudian ego psychological theorizing in his theoretical framework. In fact,

with respect to the concept of drive, Kernberg attempts to integrate segments of modem
neurophysiology and neuropsychology with a psychoanalytic concept of drive. It should be pointed out

that Kernberg’s conceptualization of emotion and affect are particularly important in his theorizing,

and for Kernberg (1976, 1980c, 1982h) these concepts to some extent replace drive as a motivational
concept.

The main focus of Kernberg’s theorizing is the type of patients that the British school (Balint,

1968; Fairbairn, 1952; Guntrip, 1968, 1971; Winnicott, 1965,1975; and others) has been describing for

the last 30 to 40 years. Kernberg has grouped these patients and maintains that many of these other
theorists wrote about people who manifest borderline pathology.

In describing Kernberg’s work, we will first note the clinical observations which Kernberg’s

theory seeks to explain. Then we will present summaries of his contributions in five areas: (1)

development; (2) psychoanalytic classification of character pathology, including the borderline

diagnosis; (3) treatment implications, derived from the developmental theory and diagnostic system,
including the rationale for various treatment recommendations as well as Kernberg’s view of

countertransference and the therapeutic stance; (4) groups and institutions, including issues in

hospital treatment; and (5) a theory of drives and affects. Finally, we will comment on Kernberg’s
critics and will ourselves critically review what we consider to be major elements in Kernberg’s

contributions to psychoanalytic theory.

SUMMARY OF KERNBERG’S WORK

THE CLINICAL OBSERVATIONS

www.freepsychotherapybooks.org 327
Kernberg’s clinical observation of “borderline adults” has been one of the factors that has led

him to expand “traditional” psychoanalytic theory. He observes that this type of person can often

maintain rapidly fluctuating, contradictory ego states. These ego states can be manifested as rapidly
changing, intense transference reactions (from idealization and love to intense hatred and rage) and

can also be seen in initial clinical contacts. Kernberg (1967, 1975a, 1980c) has inferred that these
contradictory ego states are actively separated or split and that a person who shows splitting cannot

reconcile these contradictory states. In fact, if someone else points out the person’s contradictory

attitudes, states, or actions, the person would always manifest anxiety. A person’s reaction to such an
intervention is an important diagnostic indicator to Kernberg (1976, 1981f).

Kernberg feels that when analysts and therapists do not recognize that splitting is taking place

they may fail to understand what is happening in a therapeutic situation. He notes a tendency for

alternating transference states to remain static when therapy is viewed over a long period of time. The
analyst sometimes takes one of the positive states to be the manifestation of a good working alliance

or, alternately, might feel that a patient’s rageful attacks may represent an important breakthrough, in

which the patient may become aware of and begin to understand these “primitive impulses.” Kernberg
believes that often no intrapsychic change is taking place. Instead, the patient simply alternates

presentation of these states. Often, the patient uses the tolerant atmosphere of therapy to derive

greater gratification of (in particular) his or her aggressive impulses than would be allowed elsewhere.
Or, as in the Menninger study of the effects of supportive psychotherapy (where there was little

transference interpretation, and signs of latent negative transference, especially, were

unacknowledged), the patient-therapist relationship is shallow or mechanical (Kernberg et al, 1972).

Kernberg points out that according to traditional observations of this patient population

(Federn, 1947; Frosch, 1960; Knight, 1954; Schmideberg, 1947; Zetzel, 1971) they tend to lose the

ability to test reality adequately in the context of the psychotherapy (transference psychoses), to act
out severely, and to consciously experience primary process material while apparently lacking a

capacity for introspection and insight.

www.freepsychotherapybooks.org 328
The primitive, early reactions of borderline patients to their therapists seem to be not only

preoedipal in content, but also less organized than neurotic transference. Kernberg (1976) concludes

that the work of various object relations theorists (Klein, Fairbairn, Guntrip, etc.) described these
reactions most accurately as recreations of early actual or fantasied object relationships—as “the

pathologically fixed remnants of the normal processes of early introjection” (p. 25).

These observations and conclusions led Kernberg to propose both a developmental model to

account for borderline pathology and technical innovations for the psychotherapy of borderline

conditions.

THE DEVELOPMENTAL MODEL

Kernberg’s (1966, 1975a, 1976) developmental model is organized around the internalization
of object relationships, a process he takes to be crucial in the formation of psychic structures. He posits

three types of internalization or, in his terms, three different identification systems. Each process

results in a psychic structure, which is named accordingly (introjects, identifications, and ego identity).
Thus, the process of introjection results in an introject, and so forth. As Kernberg (1976) described it:

“All processes of internalization consist of three basic components: (a) object-images or object-

representations; (b) self-images or self-representations; and (c) drive derivatives or dispositions to

specific affective states” (p. 26). Psychic organization takes place at two levels. In the earlier and more

basic organization, splitting is the main defense mechanism; during these periods, self-object-affect (S-

O-A) units with opposite affective tones are unintegrated, either as a passive consequence of lack of

maturity or as active process (splitting). In the more advanced level of organization, repression is the
main defense utilized. Ego and superego development and integration can be assessed by the degree to

which repression and its associated higher-level defenses have succeeded the more primitive condition

(Kernberg, 1976).

Kernberg follows Melanie Klein (1946) in taking introjection to play an important role in the

www.freepsychotherapybooks.org 329
early development of the ego. However, he suggests that it is a mechanism based on primary
autonomous functions of perception and memory, rejecting Klein’s views of the importance of very

early oral incorporative fantasies. We will, at a later point, describe Kernberg’s more detailed account

of the relationship between his model and findings in cognition, perception, and neurophysiological
processes.

Kernberg (1976) defines introjection as “the reproduction and fixation of an interaction with

the environment by means of an organized cluster of memory traces” (p. 29) with the S-O-A

components. For him those components are “(i) the image of an object, (ii) the image of the self in
interaction with the object, and (iii) the affective coloring of both the object-image and the self-image

under the influence of the drive representative present at the time of the interaction” (p. 29).

Introjection goes beyond the primary apparatuses because it entails complex organization of the
results of perception and of memory traces, in which perception of the external world is linked to

perception of subjective experience. Although the earliest introjections do not clearly differentiate self

and object images, a dyadic element is present.

The affective tone of the introjection is important because the various S-O-A introjections are

gradually sorted and organized by affective valence. Kernberg (1976) writes “Introjections taking place
under the positive valence of libidinal instinctual gratification, as in loving mother-child contact, tend to

fuse and become organized in what has been frequently called “the good internal object.’ Introjections

taking place under the negative valence of aggressive drive derivatives tend to fuse with similar
negative valence introjections and become organized in ‘the bad internal objects” (p. 30).

Kernberg sees affect in the first months of life as particularly important. Its “irradiating” effect

on introjects (which may include perceived self and object representations) is such that the resulting

perceptual constellations differ most according to their associated affective states. Affect states, then,
are the manner in which introjects of opposite valence are kept apart, since the immature psyche is

unable to integrate different temporal experiences and opposite affective experiences.

www.freepsychotherapybooks.org 330
Although Kernberg stresses the importance of affect in building up separated S-O-A units, his

account of developmental stages parallels that of Mahler (1968; Mahler, Pine, & Bergman, 1975). His

stages may be summarized as follows:

Stage 1. This is the stage of normal autism, or primary undifferentiation in the first month of life,
before the “good,” combined selfobject constellation develops through positive experiences. Pathology
at this stage would mean that this undifferentiated image would not develop, and a normal symbiotic

relationship with the mother would not take place, being replaced by autistic psychosis.

Stage 2. This stage, normal “symbiosis,” from the third or fourth to the sixth or ninth month,
consists of the consolidation of an undifferentiated, “good” self-object representation, and corresponds

to the periods of Mahler’s symbiotic phase and differentiation subphase. Even when self- and object
images begin to be separated-still within the umbrella of libidinally organized S-O-A units-they are

weakly delineated, and, Kernberg (1976) says, there is a “persisting tendency for defensive regressive

refusion of ‘good’ self and object images when severe trauma or frustration determine pathological
development of this stage” (p. 60). Fixation at, or regression to, this self-object dedifferentiation and

loss of ego boundaries is typical of childhood symbiotic psychosis (Mahler, 1968), most types of adult

schizophrenia (Jacobson, 1954), and depressive psychoses (Jacobson, 1966).

Stage 3. In this stage, self- and object representations are clearly differentiated, within both the

core “good” self-object and core “bad” self-object. Self-images from one positively experienced S-O-A
unit are linked with those from other positively valenced S-O-A units, with parallel joining of object

representations. With the increasing complexity of the resulting representations, this process

“contributes to the differentiation of self and other and to definition of ego boundaries” (Kernberg,

1976, p. 30). This stage corresponds to Mahler’s separation-individuation phase (excluding the
differentiation subphase), and lasts from 6 to 9 months of age through 18 to 36 months. Object
constancy (Hartmann, 1964) and stable ego boundaries should be achieved, but relationships are still

with part objects. Integration of self- and object representations occurs only at the close of this stage.
Kernberg follows Mahler in suggesting that borderline pathology follows from fixation and/or

www.freepsychotherapybooks.org 331
regression to this phase of internalized object relationships.

This is the stage in which active separation (the defense of splitting) between good self-images

and bad self-images and between good object and bad object images occurs. In patients with

borderline pathology, the combined S-O-A units of opposing valence persist in an unintegrated fashion,
and are not replaced or accompanied by higher-level developments. Kernberg (1976) maintains that

when opposing S-O-A units are initially introjected, they are kept apart to avoid the anxiety associated

with the negative valences “from being generalized throughout the ego,” and to “protect the integration

of positive introjections into a positive ego core” (p. 36). However, defensive splitting represents a
later development, in which the opposing S-O-A units are actively separated.

Kernberg suggests that the ego comes into being at the point when introjections are used

defensively. This is a state in which the “good internal objects” (mostly undifferentiated self- and object
representations with a positive valence) along with the “good external objects” (positively experienced

aspects of reality), form the purified pleasure ego, while the negative S-O-A units are viewed as outside.

“Good” self-images and “good” object images begin to be separated. Slightly later, all these “units”
become more elaborate, and the differentiation between “good internal objects” and “good external

objects” occurs. Now the defense of projection can be utilized across a relatively clear boundary, so
that the array of “bad external objects” includes some that are “bad” via projection of introjections that

had a negative valence. This clear utilization of projection is an important development of Stage 3.
Correspondingly, the defensive use of active splitting decreases over time, and the individual
successfully traverses to Stage 4.

Although we have been focusing on the building up of S-O-A units and the unfolding of defensive
processes, the second internalization process, identification, also begins to be used in Stage 3. This is a

higher-level form of introjection, which includes the role aspects of the interpersonal interactions and

hence requires some development of perceptual and cognitive abilities so that socially recognized
functions can be conceptualized by the child (Kernberg, 1976, p. 31). The affective components of such

internalizations are also more advanced and differentiated than those associated with introjections.

www.freepsychotherapybooks.org 332
The view of the self is likewise more differentiated, so that it is possible to view the object taking a role
with respect to the self. Identifications, like introjection, contribute to the formation of psychic

structure and yet may also be used for defensive purposes. Identification continues as a process

throughout life at different levels of ego integration, and its results are more subtle and better
integrated when the ego is more integrated and splitting mechanisms are not used (Kernberg, 1976, p.

77). In psychotic identifications, where self- and object images are pathologically refused,2
identifications are distorted by the projection of primitive superego forerunners or repressed drive
derivatives onto the object, so that the internalized object relation is altered in the direction of “all

good” or “all bad” introjections. When pathological identifications occur at a more integrated level, they

result in pathological character traits.

Stage 4. In Stage 4, contradictory self- and other representations are integrated into percepts of

the self and others that more accurately reflect complex experiences of the self and other persons.

Failure to achieve this integration results in “identity diffusion.” In this stage, repression appears as a
defense, and ego, superego, and id are also differentiated. This period begins toward the end of the

third year of life and continues through the oedipal period. Pathology from this stage is that of patients

with neuroses or “higher-level” character pathology (hysterical, obsessive-compulsive, and depressive-

masochistic characters).

Narcissistic personality disorders may also result from abnormal development during this
stage, when instead of integration of self and object, there is, in Kernberg’s (1976) words, “(1) a

pathological condensation of real self, ideal self, and ideal object structures; (2) repression and/or

dissociation of ‘bad’ self-representations; (3) generalized devaluation of object representations; and

(4) blurring of normal ego-superego boundaries” (p. 68; see also Kernberg, 1982f). This results in a

grandiose self, which is separated from negatively valenced S-O-A experiences in a splitting process
more typical of Stage 3.

Kernberg interprets Stage 4 as representing the achievement of what Klein (1948a,b) termed

the “depressive position,” in which, because of the new, more complex view of others as the objects of

www.freepsychotherapybooks.org 333
both hatred and love, both guilt and concern begin to appear. Representations of an ideal self and ideal
object develop as wishes to counteract the increasingly accurate awareness of reality. Repression,

which prevents the irruption into consciousness of various drive derivatives, separates id from ego

during this stage, and the id becomes more organized. Hence, in neurotic or other higher-level
psychopathology, one does not readily see primary process or direct expression of drives.

Integration of the superego as an independent intrapsychic structure takes place in Stage 4. This

has two aspects: the condensation of ideal self- and object images into the basis of the ego ideal, and

the integration of this with the sadistically determined superego forerunners. These superego
forerunners are what Kernberg (1976) terms the “fantastically hostile, highly unrealistic object images

reflecting ‘expelled,’ projected and reintrojected ‘bad’ self-object representations…and reflecting

primitive efforts of the infant to protect the good relationship with the idealized mother by turning the
aggressively invested images of her (fused with the respective self-images) against himself” (p. 71).

With integration come decreases in projection and in the fantastically hostile and unreal nature of the

superego elements.

Ego identity, the third process in the internalization of object relations, begins to occur in Stage

4. Ego identity is “the overall organization of identifications and introjections under the guiding
principle of the synthetic function of the ego” (Kernberg, 1976, p. 72). This refers to the organization of

a self-concept and of deeper, more realistic concepts of others.

Stage 5. Consolidation of superego and ego integration takes place in Stage 5, and ego identity
continues to evolve. The individual is able to learn from experience, and “an integrated self, a stable
world of integrated, internalized object representations, and a realistic self-knowledge reinforce one

another^ (Kernberg, 1976, p. 73). Representations of a social and cultural world are included. The

internal world gives increasing meaning to present interactions and provides support for the
individual in times of crisis. The individual has the capacity to discriminate subtle aspects of him-or

herself and of others and develops “depersonified” attitudes and values with increasing capacity to

communicate views and experiences in a way that others can understand. These capacities are absent

www.freepsychotherapybooks.org 334
in pathological conditions organized at earlier stages; the most striking example is the narcissistic
personality, who cannot convey more than a shallow sense of who he or she is or who the other is in an

interaction (Kernberg, 1976, p. 73). Although intimate connections among drives, affects, object

relations, and cognitive and other ego functions are implied throughout Kernberg’s model, these form a
particularly complex and dense matrix in the successful outcome of Stage 5—the healthy personality.

THE PSYCHOANALYTIC CLASSIFICATION OF CHARACTER PATHOLOGY

Kernberg’s (1980a) model of psychopathology is primarily a conflict model; constitutional

deficits may contribute to the intensity of certain conflicts and hence render the development of

pathological ego structures or character traits more likely. For example, an infant constitutionally

endowed with an intense aggressive drive may project more aggressively tinged S-O-A units onto
external figures and may develop pathologically intense fears of castration directed at abnormal

images of dangerous parents (Kernberg, 1975a). Or, children with organically based perceptual or

other learning problems may have introjections and identifications distorted by their faulty

apparatuses of primary autonomy. However, when considering adolescent and adult patients,

Kernberg’s position is that character pathology is best understood and interpreted as the result of
dynamic conflicts. Even if a learning disability is present in a borderline patient, only after considerable
treatment can its effect be differentiated from the results of pathological splitting and associated

primitive defenses (Kernberg, personal communication).

LEVELS OF CHARACTER PATHOLOGY

The developmental model previously outlined is the basis for a highly specific classification of

higher-level, intermediate, and lower-level (borderline) character pathology. This classification is

based on determining the level of instinctual development, superego development, defensive

operations, and internalized object relations (Kernberg, 1976). Kernberg (1980,1981c) has been a

vocal critic of the DSM-III (American Psychiatric Association, 1980) categorization of personality

www.freepsychotherapybooks.org 335
disorders, because it fails to consider these psychoanalytic perspectives and thereby omits certain
important diagnostic entities. In Kernberg’s system, higher-level character pathology is marked by the
achievement of genital primacy in the instinctual sphere; a well-integrated but excessively severe

superego; defense mechanisms organized around repression (including intellectualization,

rationalization, undoing, and higher forms of projection); and a stable, well-integrated concept of self

and others. Most hysterical, obsessive-compulsive, and depressive-masochistic personalities are in this

group—the classical neurotic patients.

At the intermediate level, pregenital fixation points are present, the superego is less well

integrated than in higher-level pathology, and sadistic superego precursors play an important role.

Defenses are organized around repression, but some more primitive defenses are present, with more

infiltration of instinctual impulses than is present in the more sublimatory or reactive traits
characteristic of higher-level pathology. Ego identity is established, and there is a stable concept of self

and others, but object relations are quite conflicted. Many oral, passive-aggressive, sadomasochistic,

and better-functioning infantile personalities and some narcissistic personalities are at this level.

Lower-level character pathology is characterized by borderline personality organization with,

in the instinctual realm, “pathological condensation of genital and pregenital instinctual strivings…with

a predominance of pregenital aggression” (Kernberg, 1976, p. 141). Lack of superego integration and
the continuing influence of sadistic superego forerunners are more marked than in the intermediate

group. Defenses are organized around splitting, with primitive forms of projection, denial, and other

mechanisms, which allow partial expression of the rejected impulse to a greater degree than in the
other levels of pathology. Object constancy is not firmly established, identity diffusion is present, and

object relationships are conceptualized in terms of part objects.

THE BORDERLINE CONCEPT

Kernberg is one of a very small number of investigators who have actually given a detailed

www.freepsychotherapybooks.org 336
definition of the term “borderline personality.” He provides a description of the intrapsychic structures
and other concepts he considers relevant to this diagnosis along with a sophisticated

phenomenological description of the patients. Kernberg prefers the term “borderline personality

organization” to “borderline state” or “borderline personality disorder,” underlining his belief that such
patients have a specific and stable personality organization characterized by ego pathology, which

differs from neuroses and less severe character disorders on the one hand, and the psychoses on the

other. These patients suffer from a particular type of psychic organization, which has a certain type of
history and resistance to rapid change. They are not in a transitory “state,” fluctuating between

neurosis and psychosis, nor are they defined solely by their obvious symptoms, as in psychiatric use of
the term “disorder.” Kernberg (1967, 1975a) stresses that similar symptomatology may occur as a
result of different intrapsychic configurations and conflicts, so that very detailed diagnostic study is

necessary.

Kernberg’s delineations of borderline and narcissistic patient groups rely on description of

symptoms and complaints presented by these patients, but also, just as important, on inferences about
types of psychic structure, defenses, and predominant conflicts. In his concern with “internalized

object relations,” he has devoted considerable work to explicating the method by which one makes
inferences about this and other hypothetical constructs, such as defenses or structures, on the basis of

a patient’s interview behavior, for example.

On a descriptive level, patients suffering from borderline personality organization present

symptoms that, if occurring in combination, suggest pathological ego structure: chronic, diffuse
anxiety; poly-symptomatic neuroses (severe phobias, rationalized obsessive-compulsive symptoms,

multiple, elaborate, or bizarre conversion symptoms, dissociative reactions, hypochondriasis with

chronic rituals and withdrawal, and paranoid trends with other symptoms); polymorphous perverse
sexual trends; impulse neurosis; and addictions (Kernberg, 1975a). Certain lower-level character

disorders (infantile, narcissistic, antisocial, and “as-if” personalities) and paranoid, schizoid,
hypomanic, or cyclothymic personalities also usually have borderline structure.

www.freepsychotherapybooks.org 337
Inferences about the patient’s psychological organization are based on other observations.

“Nonspecific manifestations of ego weakness” are noted by assessing lack of anxiety tolerance, as when

additional anxiety results in further symptom formation or regressive behavior; lack of impulse
control, where any increase in anxiety or drive pressure results in unpredictable impulsivity; the lack

of developed sublimatory channels (here the patient’s talents and opportunities must be considered).
A second sign is the appearance of primary process thinking, particularly in unstructured situations

such as projective psychological testing (Carr, Goldstein, Hunt, & Kernberg, 1979; Kernberg, 1975a).

The presence of the primitive defensive operations of splitting, projective identification, denial,
primitive idealization, and devaluation are important signs of borderline pathology. These may require

subtle inferences from interview behavior or interactions with the interviewer over a period of time to
establish their presence.

THE STRUCTURAL INTERVIEW

Aside from the presumptive diagnostic elements that may be indicated by a patient’s history or

presenting complaints, evidence for structural organization is found in the patient’s reactions to being

interviewed in a way that focuses on the ego functions and features characterizing neurotic,
borderline, and psychotic structures (Kernberg, 1981f).

In particular, the interviewer wishes to understand (1) the degree of identity integration, (2)

types of defenses, and (3) the capacity to test reality, including the subtle ability to “evaluate the self

and others realistically and in depth” (Kernberg, 1981f, p. 171). Borderline disorders may be

differentiated from psychoses by the borderline patient’s ability to test reality in the sense that
distinctions between internal and external and self and object representations remain. In contrast to

neurotic patients, however, persons with borderline structure will show identity diffusion, lower-level

defenses similar to those used by psychotic patients, and subtle alterations in the relationship to reality

and feelings of reality. Because their capacity to appreciate ordinary social reality is intact, however,

and because their defenses protect against the anxiety of intrapsychic conflict (rather than the anxiety

www.freepsychotherapybooks.org 338
of dedifferentiation, as in the psychoses), they respond to interpretations in the interview with better
functioning or, at least, without regression.

Thus, the interviewer seeks to assess the patient’s view of his or her problems, understanding of

self and others, and ability to make use of the interviewer’s questions and tentative interpretations.
The interviewer focuses on areas that seem odd, contradictory, or unclear to see if the patient can also

observe such contradictions and appreciate the possible explanations for these offered by the

interviewer.

The “pathology of internalized object relationships,” which contributes to the borderline

diagnosis, also relies on complex inferences from character traits and the patient’s behavior with the
interviewer. Kernberg (1976) states: “These patients have little capacity for a realistic evaluation of
others and for realistic empathy with others; they experience other people as distant objects, to whom

they adapt ‘realistically’ only as long as there is no emotional involvement with them” (pp. 36-37).
They do not empathize well with others, and are “ignorant of the higher, more mature and

differentiated aspects of other people’s personalities” (p. 37). Hence their relationships are shallow,

they are unable to experience guilt and concern, and they give evidence of exploitiveness and
unreasonable demands without signs of tact or consideration. In trying to control his or her

environment, the patient manipulates others. When they begin psychotherapy, these patients
immediately present chaotic and primitive object relations in the transference, as opposed to the

gradual unfolding of more mature and then less mature transferences found in neurotic patients

(Kernberg, 1976).

SPECIAL DIAGNOSTIC ISSUES

Several examples exemplify Kernberg’s contention that similar symptomatology may stem from
different types of underlying pathology and structure.

Hysterical versus Infantile Personality. Hysterical patients, while showing superficial similarities

www.freepsychotherapybooks.org 339
to infantile patients, have some conflict-free areas where their functioning is stable and appropriate.

They are impulsive or clinging only in certain relationships or areas of conflict. Their need to be loved

and admired, although it has oral, dependent components, is closer to an expression of genital needs.

Oedipal dynamics contribute to differential relationships with men and women, and the
provocativeness of these patients is usually not accompanied by promiscuity. Stable, if neurotic,

heterosexual relationships are present.

In contrast, infantile patients are more socially inappropriate and impulsive across all areas of

life. Oral, demanding elements are more prominent, so that the need to be loved is “more helpless” in

quality, and exhibitionistic trends have a primitive, narcissistic, exploitive quality. Promiscuity may be

present in conjunction with unstable, changing relationships (Kernberg, 1975a). Such patients
frequently are organized at a borderline level (Kernberg, 1981c).

Depression. Kernberg stresses the importance of differentiating depression as a symptom from

depressive-masochistic character traits. The higher-level depressive personality, for example, may

experience depression in connection with guilt over oedipal strivings or with true concern for the self
and others, because of the presence of superego integration. Depression that represents helpless rage
or disappointment in an ideal suggests less superego integration. Severe depression that causes

breakdown in ego functioning also suggests the presence of a sadistic superego, probably associated
with borderline organization. However, both the quantity and quality of depression must be
considered when making a structural diagnosis, as the absence of any depressive concern or guilt for

others may also be a sign of borderline organization in narcissistic and antisocial personalities

(Kernberg, 1975a, 1977a).

Adolescence. The stresses of identity consolidation in adolescence may, in conjunction with

environmental pressures (such as gang membership or cultural norms), suggest the presence of severe

personality disturbance. Kernberg recommends assessment of the presence or absence of whole-object

relationships, ideals, and the capacity for sublimation and work. Adolescents with borderline

personality structure will be far less able to describe themselves or their friends in depth and do not

www.freepsychotherapybooks.org 340
show evidence that they can invest themselves in ideals or goals that have meaning to them (Kernberg,
1978, 1979b, 1982e).

Borderline versus Schizophrenic Conditions. In the absence of clear signs of formal thought

disorder, hallucinations, or delusions, the primitive defenses present in both borderline and
schizophrenic conditions serve different functions, which can be used in interviewing to make this

distinction. In patients with borderline structure, these defenses protect the patients from the

experience of ambivalence, and “a feared contamination and deterioration of all love relationships by

hatred” (Kernberg, 1975a, p. 179). Schizophrenic patients use splitting and allied mechanisms to
prevent “total loss of ego boundaries and dreaded fusion experiences with others” (p. 179),

particularly under the stress of strong affects. This is because persons with psychotic structure do not

have clearly differentiated self- and object images. Since primitive defense mechanisms cause ego
weakness in patients with borderline structure, interpretations should strengthen the ego and lead to

better functioning in the interview—more reflectiveness and attempts at integration and better reality

testing. Interpretation of the same primitive defenses in schizophrenic patients reveals difficulty with
self-object differentiation and hence leads to regression—more overt primary process or delusional

thought, loosening of associations, or paranoid distortions of the interviewer—in response to the

interpretations given during the interview. Hence, the interview should be conducted with inquiry into

responses which are unusual or subtly inappropriate, to test the patient’s defensive functioning.

Transference psychosis, which may be present in both borderline and schizophrenic conditions,
is different in each group because of the different mechanisms involved. With borderline patients, the
transference psychosis is limited to the treatment hours and responds to Kernberg’s recommendations

for structuring the treatment. With psychotic patients, their psychotic behavior and lack of reality
testing in treatment is for a long time no different from that outside the treatment. Later on, they may
feel convinced that they and the therapist are one. This is in contrast to the transference psychosis of

borderline patients, who always maintain some sort of boundary, even if they feel themselves to be
interchanging aspects of identity with the therapist (Kernberg, 1975a, 1980c).

www.freepsychotherapybooks.org 341
TREATMENT IMPLICATIONS

Kernberg goes into considerable detail in his diagnostic system because he believes that
borderline structure, as well as certain other characterological features of diagnosis, have specific

implications for treatment and prognosis. For example, he views dishonesty by the patient as a

particularly unfavorable prognostic sign, which might lead to a recommendation for the use of major
environmental supports or other modifications in psychotherapy (Kernberg, 1975a). On the other

hand, he warns against supportive psychotherapy for schizoid patients. Narcissistic patients with

different types of functioning warrant different types of treatment.

RECOMMENDATIONS FOR PATIENT SUBGROUPS

Kernberg recommends expressive psychoanalytic psychotherapy, incorporating his

modifications, for patients with borderline personality organization, including patients with

narcissistic personality disorder who function on an overt borderline level. That is, the pathological
selfstructure in some narcissistic patients is sufficiently stable to allow the patient to function without

the impulsiveness, chaotic relationships and general manifestations of ego weakness that characterize

borderline functioning. Others, especially those who present with narcissistic rage, function in a

manner similar to borderline patients (Kernberg, 1975a, 1980a,c).

For patients with narcissistic personality disorders, Kernberg recommends unmodified

psychoanalysis if at all possible. Without the analytic setting, such patients tend to remain shallow,
empty, and uninvested in the treatment and do not develop very meaningful transference reactions.

Even if they do undertake analysis, however, they may wish to stop the treatment after amelioration of
some of their more painful experiences of envy or disruptive impulsiveness, feeling content to remain
somewhat shallow and unempathic. At such times, the analyst may need to shift to a partially

supportive technique to help the patient maintain a better adaptation by protecting some of the

narcissistic defenses when these cannot all be worked through (Kernberg, 1975a, 1979a).

www.freepsychotherapybooks.org 342
Some cases present the following contraindications for expressive psychotherapy: (1) inability

to work verbally with symbolic material; (2) a combination of low motivation and high secondary gain;

(3) severe negative therapeutic reaction; (4) severe cases of antisocial personality, so that the therapist
cannot assume the patient will be honest even most of the time; and (5) life circumstances that prevent

the patient from the frequency of sessions required for expressive treatment (usually two or three
times a week). These contraindications can include patients from across the diagnostic spectrum,

although the more disturbed borderline, narcissistic, and psychotic patients will fall into the first four

categories more often. Such patients should be treated with a frankly supportive treatment, with
rational, concrete treatment goals. The therapist should represent a commonsense point of view,

making suggestions, consulting with family members if necessary, and should interpret primitive
defenses and conscious negative transference only in the context of showing how these create
difficulties in the patient’s life. Idealization of the therapist should be discussed only if it interferes with

the work, for example, by inhibiting the patient’s questions or disagreements. The major focus is on
clear life goals. There are some patients who simply need a lifelong supportive relationship, but this
alternative should be chosen only after other treatments have been ruled out (Kernberg, 1980e, 1982i,

g). Kernberg (1977b) has also discussed indications and technique for brief psychotherapies.

There are two groups of patients who do not do well with supportive treatment, according to

Kernberg. These are well-functioning schizoid individuals, who would enter and leave a supportive
therapy untouched by the human interaction, and certain narcissistic patients who are lonely, isolated,

and empty. These characteristics are unlikely to change without exploration in detail of the primitive

defenses and representations of self and others that contribute to the shadowy quality of personality

conveyed by these patients. Patients who cannot experience much empathy for others cannot learn to
do so without the development of higher-order, more complex representations of self and others

interacting. For these patients—narcissistic patients functioning on a borderline level and most

patients with borderline personality organization—Kernberg (1980e, 1982d, i, g) recommends his

modified, expressive psychotherapy.

www.freepsychotherapybooks.org 343
RATIONALE FOR TECHNICAL RECOMMENDATIONS

Kernberg proposes that the model of development and psychopathology summarized earlier
explains the behavior of severely disturbed patients in various types of treatment as well as processes

in the traditional psychoanalysis of healthier, neurotic patients. The structural differences between
borderline and neurotic patients cause them to respond differently to classical psychoanalytic
technique. Neurotic patients, who have a well-formed tripartite structure, suffer from intrapsychic

conflict usually conceptualized as conflict among id, ego, and superego or between conflicting, higher-

order, relatively well-integrated identifications that represent various compromise solutions to the
basic conflicts. Kernberg (1980b) lays particular stress on this last point, insisting that there is no

impulse-defense configuration without an implied object relationship within which these defenses and

impulses are expressed. Borderline patients have primitive intrapsychic structures, which have not
been consolidated into the tripartite structure but instead have various split-off self-object-affect units,

so that these patients have little awareness that the loved and hated object is one and the same. Their

defenses are primitive and tend to weaken, rather than protect, the ego; and the superego is close to

being an internal persecutor, rather than a depersonified source of values and self-esteem. Id material
may be conscious.

In the psychoanalysis of neurotic patients, defenses are interpreted as they are manifested as

resistances, with a gradual unfolding of a regressive transference neurosis, which reveals the conflicts
that create the patients’ problems. Such patients’ defenses may be less than optimally adaptive, but

they do protect the ego; hence, their interpretation and undoing represents a stress that only patients

with intact ego functions can withstand. Id material becomes available only after considerable work,
and impulsive action is brief. As infantile conflicts are resolved, more flexible and efficient defenses

come into being.

When borderline patients are treated with standard psychoanalytic technique, the absence of

external structure to support reality-testing functions tends to lead to rapid emergence of primary
process material, transference psychosis, or, at least, intense early transference reactions prior to the

www.freepsychotherapybooks.org 344
development of any kind of working alliance. Thus, Kernberg (1982d, 1983) feels a need for a clear
distinction between psychoanalysis proper and modifications of technique that might be termed
“psychoanalytic psychotherapy.”

When patients with good ego strength are treated with one of the psychoanalytic
psychotherapies, the results are good in terms of behavioral change and alteration in character traits

(although not character structure). Kernberg believes this is a direct result of these patients’ greater

ego strength and capacity to develop a relationship in which they can accept help. However, when
borderline patients are treated with a type of psychotherapy that seeks to interpret only certain

defenses or to avoid interpretation of the negative transference, the patients’ severe psychological

problems persist, and a chronically shallow treatment relationship often develops, with acting out

elsewhere in the patient’s life.

In severe psychopathologies, in Kernberg’s (1980b) view, “what appear to be inappropriate,

primitive, chaotic character traits and interpersonal interactions, impulsive behavior, and affect storms

are actually reflections of the fantastic early object-relations-derived structures that are the building

blocks of the later tripartite system” (p. 187). These are not reflections of actual early relationships, in
most cases, but of their distorted internalization and continuation in the intrapsychic world without

integration into more accurate, complex representations and more mature intrapsychic structures. Ego
weakness results from the persistence of the defenses of splitting and of primitive forms of projection,

denial, idealization, and devaluation. Thus, pathology is seen as resulting from conflicts and defenses

rather than from a deficit.

Kernberg believes that the poor results when borderline patients are treated with
psychotherapy are due to the interaction of their pathological structures with the therapeutic

techniques. He makes the following argument:

1. Since patients with borderline pathology suffer from ego weakness as a result of their

primitive defenses, systematic interpretation of defenses is indicated to strengthen the

www.freepsychotherapybooks.org 345
ego. Interpretation of defenses will not lead to regression, but will aid the patient’s

capacity to observe and begin to integrate the defensively split S-O-A units.

2. Emphasis on developing a positive transference or providing the patient with a benign model
for identification does not accomplish its goal with seriously disturbed patients.

Borderline patients typically present strong negative, often paranoid, transferences at

some point in the treatment as the negatively experienced S-O-A units are activated in
the therapeutic relationship. More often than not, they will be unable to identify with the

therapist’s healthy ego without some interpretation and resolution of their negative

transferences. Without this, or with avoidance of negative transference material, the

therapist-patient dyad may simply come to be a reenactment of one of the positively

experienced S-O-A units, while other parts of the patient’s personality are expressed

outside the treatment. The treatment is rendered shallow and meaningless and has little

effect on the patient’s life.

3. Interventions that would gratify some of the patient’s transference demands, made with the
idea of lessening pressure on the weak ego, fail to help the patient, but rather tend to
support the enactment of one side of the patient’s conflicts as a defense against a

perception of the therapist as evil or devalued. Hence, such interventions contribute to

the patient’s distortions of the treatment situation. With healthier patients, gratification
of transference wishes is likely to have a more benign effect, as the patient’s capacity to

use what is good and to identify with a good parental figure is not so distorted

(Kernberg, 1980b, p. 194).

4. Since borderline patients present conscious conflicts that may involve primitive drive
content, efforts to avoid “deep” material are misguided. The therapist’s avoidance of

impulses that are conscious and troubling to the patient would tend to reinforce the

patient’s fear of these impulses and tendency to express them outside the treatment.

www.freepsychotherapybooks.org 346
Kernberg therefore recommends a modified form of expressive psychotherapy, not

psychoanalysis proper, for most borderline patients. Kernberg’s suggestions may be summarized as

follows:

Interpretation. Interpretation and clarification, rather than suggestion and manipulation, are the
major technical tools to be used. Very often, however, the patient’s interpretation of the interpretations
or other remarks must be explored, and this may often require the therapist to clarify what he or she

meant, as opposed to the patient’s distorted perception of what was said. Kernberg (1980b, p. 196;

1982g) believes that with these patients, such clarifications will be more frequent than interpretations,
thereby giving a different emphasis to the treatment.

Maintenance of Technical Neutrality. To be able to use interpretation, suggestion and

manipulation are contraindicated, and technical neutrality should be maintained as far as possible.

However, severely disturbed patients are often unable to observe the inappropriateness of their

behavior (for example, repeated verbal attacks on the therapist) or may act in such a way as to
endanger their lives or the treatment. It may be necessary to structure the treatment or the patient’s

life. For example, one might forbid shouting at the therapist, beyond a certain point, in a patient who

does not seem to be able to reflect in any way on the meaning of this behavior and who, on the

contrary, experiences some drive discharge and then seems unconcerned about this aggression.
Limiting this behavior would tend to make the patient anxious and might advance the treatment. Or

patients may be asked to live in a halfway house or to meet with another professional who would

monitor the patient’s activities and give advice, freeing the therapist from the need to “take over” in
this way so that an interpretive approach could still be maintained. These interventions would be

introduced, ideally, as parameters (Eissler, 1953), gradually eliminated, and their effect interpreted as

the therapist seeks to return to a position of technical neutrality. Less dramatic deviations will occur in

every session, when the therapist has to clarify the patient’s distortions of reality and, in so doing,
momentarily takes over an ego function and moves away from a neutral position.

Transference Analysis. Transference analysis will be partial because of the need for

www.freepsychotherapybooks.org 347
simultaneous consideration of the patient’s life situation and treatment goals. In addition, genetic
reconstructions are possible only very late in the treatment, if at all; earlier transference

interpretations should have a hypothetical quality (“You are acting as if you feel I am a cruel father

figure whom you anxiously need to placate”), to avoid premature assumptions about the reality of the
patient’s childhood experience. This is necessary to deal with the many shifting and fantastic S-O-A

units activated in the transference, not all of which will represent actual parent-child interactions. This

interferes with actual reconstruction; however, over the course of treatment, as these structures
become more integrated, part-object relations and part-object transferences should be transformed

into more mature relationships and transferences.

Kernberg suggests a face-to-face therapy that adheres as closely as possible to classical analytic

technique, within the constraints imposed by the differences that have been noted between
psychoanalysis and this type of psychoanalytic psychotherapy. In addition, the therapist should try to

clarify the use of splitting and the nature of the various S-O-A units that will be reenacted recurringly

in the treatment. When doing this, it is important to focus on both the current reactivation and the one
against which it functions as a defense. Thus, even in the course of discussing a patient’s hostile

transference attitude, the therapist should note other signs of positive feeling (for example, the patient
abuses the therapist, but comes faithfully on time to do so)—the more so because such positive

attitudes may form the basis for a working alliance. The cognitive aspect of such interpretations is
directed at the patient’s capacity to develop an observing ego and does not contribute to
intellectualization or rationalization, according to Kernberg. Rather, in primitively organized patients,

cognition is close to affect and psychic structures and helps to organize the patient’s chaotic
experience. In addition, such comments occur in the context of the therapist’s attempt to render a
confused, distant, or fragmentary patient-therapist interaction a meaningful human experience, even

though it may be based on bizarre fantasies in the patient’s mind (Kernberg, 1975a, 1979b, 1980b, c,
1982g).

THE THERAPEUTIC STANCE

www.freepsychotherapybooks.org 348
In psychotherapeutic treatment of seriously disturbed patients, Kernberg suggests, nonverbal

aspects of the patient’s communication play a larger role than they do in the treatment of healthier

patients. Patients with borderline or schizophrenic conditions may manifest nonverbal behavior that is
at odds with their remarks as a result of the use of splitting. Or they may express an S-O-A unit through

attempts to induce the therapist to play one of the roles in this unit, attempts that may be conveyed
through nonverbal means or through the use of words for their emotional effect. Kernberg (1975a,

1977c) recommends that the therapist follow Bion’s (1965, 1967, 1970) idea of the analyst as a

“container,” to try to integrate within himself or herself the disparate elements the patient presents, in
order to articulate the patient’s current experience and defenses in the transference. The analyst’s

willingness to tolerate great confusion, fragmentation, and aggression in the patient, while actively
seeking to explore it—thereby conveying an attitude of hope and acceptance—makes possible the
treatment of very seriously disturbed patients.

In a similar vein, Kernberg (1976a, 1981a) is a major proponent of what he terms the

“totalistic” view of countertransference, in which countertransference is defined as “the total

emotional reaction of the psychoanalyst to the patient in the treatment situation” (1975a, p. 49). While
advocating the resolution of countertransference reactions, Kernberg stresses the importance of

examining one’s reactions for information about the patient, a view characteristic of Kleinian and

interpersonalist theories. Kernberg claims that with more seriously disturbed patients, the therapist’s
reactions have more to do with his or her general capacity to tolerate stress and anxiety than with the

therapist’s neurotic needs. Since the patient often presents a very chaotic picture, the therapist’s

attempt to maintain empathic contact with the patient through partial identifications may lead to some

regression in the therapist’s ability to function (Kernberg, 1975a, 1977c). Kernberg (1977c, 1981a)
also describes very meaningfully the experience of a therapist in a stalemated treatment effort, and

offers suggestions for the resolution of chronic impasses.

GROUPS AND INSTITUTIONS

www.freepsychotherapybooks.org 349
Kernberg’s ideas about hospital treatment and psychotherapy and his creative application of

psychoanalytic thinking to psychiatric settings are based on his views about group and institutional

processes. Although less well known, his papers on these topics reflect a deep awareness of the
complexities of group life.

ANALYSIS OF GROUP AND INSTITUTIONAL PROCESSES

Following in the tradition of Miller and Rice (1967; Rice, 1965, 1967) and building on the

contributions of Freud and the British object relations group theorists such as Bion (1959), Kernberg

has sought to apply a psychoanalytically sophisticated open-systems theory to group and institutional

processes. He proposes that the tendency for normal individuals to behave and think regressively in
unstructured or large groups is due to the threat to personal identity posed by such groups. This threat

arises because such groups activate primitive internalized object relations in their members, with

associated primitive defenses and intense, pregenital, aggressive and sexual impulses (Kernberg,
1980b).

In order to understand institutional functioning, it is necessary to examine the institution’s task,

the resources available to it for this task, and the structure of authority and responsibility in the

institution. Kernberg (1973, 1975b, 1980c) discusses, for example, three types of problems that
prevent the accomplishment of an institution’s task: (1) the nature of the task may be unclear or

contradictory, or the task may be seen to be impossible when it is clearly defined; (2) the

administrative structure, that is, the structure that controls and maintains the institution’s internal and

external boundaries, may be unsuitable for the institution’s task, or the organization may be structured
to meet the emotional needs of administrators or staff, not to perform the task; and (3)

psychopathology in the leader or leaders within the institution may hinder the accomplishment of the

organization’s task.

Kernberg’s contributions in this area have focused particularly on the dilemmas of leadership

www.freepsychotherapybooks.org 350
and the interaction between leaders and groups or institutions. The leader is the individual who
manages the boundaries of the group—its time, membership, agenda, and utilization of resources—so

it can carry out its task. Because groups exist within organizations and consist of individuals who

themselves contain intrapsychic structures at different levels of organization, leaders must be aware of
boundary issues throughout these levels. In contrast to Miller’s (1969) view of systems as

hierarchically arranged in, as it were, concentric circles (society, institution, division of the institution,

individual, intrapsychic structures, and internalized object relations), Kernberg takes the position that
hierarchies in most group situations, cannot be reduced to this one-dimensional model. Usually the

leader must control the group’s contact with nonconcentric sets of systems that impinge on the group
in different ways. In therapeutic settings, in addition to administrative and political pressures on task
definition and resources, professional, personal and technical value systems are influential (Kernberg,

1975b). Kernberg suggests that the best way for a leader—particularly the leader of a therapy group or
hospital community—to understand the effect of these pressures and responsibilities on the group, is
to observe his or her own emotional and cognitive experience in the group. This view is similar to

Kernberg’s espousal of the usefulness of countertransference (defined broadly) in individual

psychotherapy.

Regressive pressures on staff members in organizations lead to a tendency to attribute the

causes of institutional problems to the leader’s incompetence or personality, so that the individual may
defend against awareness of problems with the institution’s task or structure. Organizational pressures
can affect the leader’s personality functioning, however, and some institutional problems are created

by individuals with particular types of psychopathology who actively seek positions of authority.
Hence, organizational consultants must combine the ability to define tasks and assess institutional
structures with the capacity to assess the personal qualities of leaders from a psychoanalytic

perspective (Kernberg, 1980b).

HOSPITAL TREATMENT

www.freepsychotherapybooks.org 351
Kernberg recommends hospitalization to protect the patient who might otherwise irreparably

damage his or her life, career, or relationships and to protect psychoanalytic psychotherapy by

allowing the therapist to maintain a position of technical neutrality, aside from the recommendation
for hospitalization. This might be necessary with a patient who had the capacity to benefit from an

expressive psychotherapy but who also needed external guidance and support. Hospitalization or a
period of residence in a halfway house would then serve to prepare the patient for outpatient

treatment in which the patient will take responsibility for his or her own life and would in other

respects maintain a therapeutic alliance. Some patients immediately threaten the continuation of their
psychotherapy with impulsive behavior, attempts to control the therapist, or attempts to force the

therapist to take responsibility for the patient’s life. In some such cases, an initial period of
hospitalization may help to clarify the patient’s psychological strengths and weaknesses (Kernberg,
1975a, 1976, 1981d, 1982b).

The group activities, rules, and regulations and the multiple, new interactions in which the

patient must engage in the hospital provide a way to diagnose the patient’s pathological internal object

relations. The combination of psychopathology in the patient and the many group situations in the
hospital allows the patient to replicate his or her internal conflicts in the social field (Kernberg, 1973,

1976). Kernberg (1973, 1975b, 1981d, e, 1982a) gives an outline for hospital administration which

provides a structure that maximizes the staff’s ability to gather and utilize such data therapeutically.
The hospital psychotherapist might then use such data to help patients explore the internal conflicts

that are causing them to act a certain way in the hospital. For example, borderline or schizophrenic

patients may quickly develop opposite relationships with different subgroups of staff, based on their

defensive use of splitting, with the tendency to create in the external world the “good” and “bad”
internalized object relationships that comprise their psychic worlds. Kernberg (1973, 1976) has also

provided detailed, sophisticated suggestions on the role of the various modalities of treatment (the

milieu, groups, nursing and medical management, activities, and hospital psychotherapy of various

kinds, with or without a separation between the therapist and administrator) in the psychoanalytic

hospital. Underlying his recommendations is the assumption that unmistakable evidence of the staff’s

www.freepsychotherapybooks.org 352
respect and concern for the patient is a crucial element in hospitalization, since patients who are

hospitalized are those who do not have sufficient respect or concern for themselves to manage their
lives. Ideally, through the hospitalization, the patient will develop a therapeutic alliance that will

sustain outpatient psychotherapy. This change occurs, in part, because so many aspects of

hospitalization are clearly and realistically helpful, in contrast to the patient’s fantasied, transference
distortions (Kernberg, 1973).

THE THEORY OF AFFECTS AND DRIVES

We will conclude this summary of Kernberg’s contributions with his theory of drives and
affects, which in many ways is his most carefully considered theoretical statement. We have already

summarized Kernberg’s model of the developmental stages of internalized object relations, the final
phase of which is the integration of contradictory S-O-A units into complex perceptions of self and

other, and the maturation of ego and superego into adaptive structures. We consider Kernberg’s theory

of drives and affects separately, even though it is intended to fit into the developmental model, because
it represents an additional focus in his work in which he interprets neurophysiological data and

reexamines the dual instinct theory (Kernberg, 1976, 1980d, 1982d, h). Kernberg proposes that

the units of internalized object relations (the S-O-A units) constitute

subsystems on the basis of which both drives and the overall psychic
structures of ego, superego and id are organized as integrating systems.
Instincts (represented by psychologically organized drive systems) and
the overall psychic structures (id, ego, superego) then become component
systems of the personality at large, which constitutes the suprasystem. In
turn, the units of internalized object relations themselves constitute an
integrating system for subsystems represented by inborn perceptive and
behavior patterns, affect dispositions, neurovegetative discharge patterns,
and nonspecific arousal mechanisms [p. 85].

Kernberg (1976) states that by conceptualizing the elements of this theory as subsystems and

suprasystems, he avoids proposing “a neurophysiological model of the mind or a mechanical model of

www.freepsychotherapybooks.org 353
body-mind equivalence” (p. 86). Thus, he speaks of hierarchies of organized systems. At some point,
however, there is a shift from “neuro-physiologically based functions” and “physiological units,” (which

would refer to changes in electrical patterns or neurotransmitters) to the integration of these units

into a “higher system represented by purely intrapsychic structures, namely, the primitive units of
internalized object relations (self-object-affect units)” (p. 86). These units are themselves eventually

integrated into id, ego, and superego.

“Affect dispositions,” which are inborn and determined by brain functioning, constitute primary

motivational systems, in that they represent dispositions to the subjective experience of pleasure and
unpleasure. These affect dispositions “integrate the perception of (1) central (pleasurable or

unpleasurable) states [that is, perception in the central nervous system], (2) physiological discharge

phenomena, (3) inborn perceptive and behavior patterns, and (4) environmental responses”
(Kernberg, 1976, p. 87). The Freudian concept of instinct may be included here. Affective patterns

communicate the infant’s needs to the mother and thereby initiate interactions, which are stored as

memory traces with affective and cognitive components. “Affects are the primary motivational system,
in the sense that they are at the center of each of the infinite number of gratifying and frustrating

events the infant experiences with his environment” (Kernberg, 1982h, p. 907), each of which leads to
an internalized object relation, fixed by memory.

Affect and cognition evolve together at first because their respective memory traces are

integrated in affective memory (Kernberg, 1976), but eventually differentiation of pleasurable and
unpleasurable experiences and of components of self and other takes place. At this point, Kernberg
(1982h) asserts, the “good” and “bad” experiences generate the overall organization of motivational

systems, which we term love and hate.

Kernberg (1982h) then suggests that love and hate become stable intrapsychic structures, “in
genetic continuity through various developmental stages” (p. 908), which can be equated with the

psychoanalytic concepts of the two drive systems, libido and aggression. At this stage of organization,

affects serve a signal function for the two drives, and increasingly complex subjective, behavioral, and

www.freepsychotherapybooks.org 354
cognitive elaborations of affects and drives develop. Drives will always be manifested by specific
wishes in the context of particular object relations, a phenomenon that is more precisely articulated

than an affect state.

Kernberg’s (1976) theory deals with economic issues as follows: Variations in the intensity of
drives or affects can be attributed to either constitutional variations in the innate components of the

system (the hypothalamus, genetically determined behavioral patterns, etc.), or to variations in the

environment (the responses of the mother and so forth). Neutralization (Hartmann, 1955) takes place

when positively and negatively valenced self-object-affect units are combined to form more complex
and realistic self- and object representations with the achievement of the depressive position.

Kernberg (1976) writes: “The synthesis of identification systems neutralizes aggression and possibly

provides the most important single energy source for the higher level of repressive mechanisms to come,
and implicitly, for the development of secondary autonomy in general” (pp. 45-46, italics in original).

What Hartmann termed fusion of drives is also included, according to Kernberg, in the combination or

integration of opposing affects as part of the integration of contradictory S-O-A units. Similarly,
sublimation is not simply a change in the use of drive derivatives in an economic sense; it, too, has an

object relations component: Sublimatory activity requires the capacity for some whole, integrated
object relationships, some genuine concern for oneself and others (Kernberg, 1975a, p. 134).

Nonetheless, despite the importance Kernberg assigns object relations in his theory of affects and
drives, he also argues for the importance of aggressive drive manifestations and the biologically based
changes in drives, which influence object relations (as in the genital strivings of the oedipal period).

Thus, he claims to support the proposition that drives, rather than object relations, constitute the
primary motivational system of the organism.

DISCUSSION

We will now offer commentary on the contributions of Kernberg that we have attempted to
summarize. Since we are not here comparing Kernberg’s positions with those of other analysts, such as

www.freepsychotherapybooks.org 355
Kohut or Brenner, we will restrict ourselves to a critical discussion of Kernberg’s clinical and
theoretical work.

At the very least, Kernberg has synthesized a good deal of the clinical observations of the object

relations school and helped to develop a nosology that orders these observations. Thus, for example,
Guntrip’s (1968, 1971) or Fairbairn’s (1952) observations of the schizoid person fit nicely into

Kernberg’s conceptualization of one type of patient with borderline personality structure. Kernberg is

able to show how some of the writings of Winnicott (1965, 1975), Melanie Klein (1946), Balint (1968),
and even Greenson (1954) can be understood within his concept of the borderline personality. His way

of thinking about the levels of severity of character pathology, based in part on object relations

concepts, may prove to be extremely useful. In addition, he has integrated the British object relations

school’s stress on aggression into his clinical and technical writings in a way that helpfully underscores
the importance of dealing with aggression, both in clinical situations and in theory development.

We consider it a strength of Kernberg’s writings that he frequently relates his theoretical points

to observable clinical phenomena. For example, he has not only shown in his attempts at theoretical

integration how a variety of authors (Balint, 1968; Fairbairn, 1952; Frosch, 1960; Greenson, 1954;
Guntrip, 1968,1971; M. Klein, 1946; Schmideberg, 1947; Winnicott, 1965, 1975) refer to the use of

primitive defenses such as splitting and projective identification by borderline patients (using
Kernberg’s definition of borderline, not necessarily those authors’ own), but he has also sought to

describe how one might infer the use of splitting or projective identification by a patient in a clinical

interview. Similarly, he is willing to claim that practical consequences follow from his theoretical
assumptions about diagnosis and particularly from assessment of level of defensive functioning. This

willingness to make predictions makes it easier for other investigators to test his inferences and
conclusions. As an example, Kernberg is remarkably specific and detailed in relating prognosis and
choice of psychological treatment method to diagnosis based on his nosology. A patient suffering from
a narcissistic personality disorder, without overt borderline-level functioning, should be treated with

unmodified psychoanalysis; a patient with narcissistic personality disorder who functions overtly on a

www.freepsychotherapybooks.org 356
borderline level should be treated with Kernberg’s modified psychoanalytic psychotherapy. The same
types of patients might require a shift to a supportive type of psychotherapy at some point in the
analysis or psychotherapy, but this would not result in the type of change to be expected from

psychoanalysis or from Kernberg’s modified form of psychoanalytic psychotherapy. Some narcissistic

patients present negative prognostic features (severe antisocial features, conscious enjoyment of

others’ suffering, chronic absence of human involvement, etc.), which indicate a need for supportive

psychotherapy from the onset (Kernberg, 1975a, 1979a, 1980c, 1982g, i).

To summarize at this point, Kernberg’s achievements in the areas of clinical writing and

observations seem particularly impressive:

1. He has synthesized the writings of a number of authors, particularly those of the British
object relations school but also including Jacobson and Mahler, and shown how their

clinical observations can be conceptualized in the context of his definition of the

borderline personality organization.

2. He has added a number of his own clinical observations and worked out a detailed

classificatory system, particularly for character pathology and the borderline

personality, within a five-level structure for describing the full range of

psychopathology.

3. He has specified a method of interviewing with stated criteria derived from the interview,

through which one can reach complex diagnostic determinations.

4. He has related his diagnostic categories to choice of treatment and to prognostic statements
about therapy outcomes.

Kernberg has covered a vast territory in his clinical writings, and he covers it in a systematic

fashion. We must join other writers (Calef & Weinshel, 1979), however, in wondering how he is able to

www.freepsychotherapybooks.org 357
make so many prognostic statements with such assurance.3 His level of specificity is rare in our field
and it would be virtually impossible for Kernberg to have personally diagnosed and treated (and

treated to the point of termination, in order to substantiate prognostic claims) all the different types
and subtypes of patients that are the subjects of his classification system, treatment recommendations,

and prognostic statements. Thus, his prognostic statements, for example, must come from a

combination of research findings, consultations, supervision, and his experience of being involved in
and directing a variety of clinical facilities.

Does Kernberg base his prognostic statements on research findings (Kernberg et al, 1972) from

the Menninger outcome studies or on his impressive clinical experience? It is often difficult to tell, but

most often he writes with the assurance and precision of someone who has a great deal of empirical
research to buttress his points. He understandably does not give extensive clinical examples, that is,

complete case studies, for if he did, given the range of categories and subcategories he discusses, he

would literally fill our journals with clinical examples. Though it is beyond the scope of our chapter to
evaluate the major outcome research with which he has been involved, we believe that Kernberg

would acknowledge that his assurance about all his prognostic statements could not reasonably be
based on this research. Moreover, although this research is of great interest, it is by no means free from

serious methodological criticisms, which affect the types of prognostic statements Kernberg has made.
It is our assumption, then, that a number of Kernberg’s statements and recommendations are based on

his clinical experience.

Given that this is the case, it is understandable that Kernberg has been criticized (Calef &
Weinshel, 1979) for his tone in his clinical writings. He writes as if he has sound evidence for his
assertions, but, at least up to this point, he has not fully indicated the nature and extent of his evidence.

We join in the criticism that has been leveled at Kernberg in this area, but we wish to note what we
believe are two mitigating considerations. First, one can criticize any number of psychoanalytic

authors for writing as if something had been “demonstrated,” when they were really stating their views

based on, perhaps very interesting, but nevertheless limited, clinical observations. Second, unlike the

www.freepsychotherapybooks.org 358
types of statements made by many other psychoanalytic authors, Kernberg’s statements are in a form
that makes them potentially testable (although to test his assertions would require a very elaborate
and difficult research undertaking).

A number of analysts have criticized Kernberg’s clinical concepts on other grounds than those
we have noted. Although it is beyond the scope of this exposition to enter into the type of detailed

criticism leveled by, for example, Calef and Weinshel (1979) or implied by the type of reconciliation

between Kohut and Kernberg attempted by Stolorow and Lachman (1980), we will comment briefly on
Calef and Weinshel’s critique.

We believe that Calef and Weinshel have brought up interesting and potentially devastating
criticisms. They include the ones we have previously discussed, and, most seriously, they cast doubt on
the validity of Kernberg’s contention that there are people with a stable personality organization which

he has labeled borderline. (A related criticism, that Kernberg claims premature diagnostic closure in a
very complex area, which still needs further exploration, is offered by Sugarman and Lemer (1980).

Calef and Weinshel also feel that Kernberg’s concepts tend to dilute basic psychoanalytic concepts such

as regression, and the very idea of intrapsychic conflict. However, a central point in their critique is
their attempt to question the borderline concept itself. They criticize Kernberg for discarding the idea

of a continuum that would include borderline and psychotic conditions and for maintaining that
conventional reality testing is either present or absent. Instead, Calef and Weinshel (1979) conclude

that “the relativity of reality testing…makes it a difficult area to establish hard and fast, categorical,

isolated criteria for the diagnosis of a psychosis” (p. 485) and, by extension, makes it difficult to
delineate people with borderline personality organization from people who are psychotic.

With respect to Calef and Weinshel’s criticisms, we would comment that many of their points

could be framed and tested or could at least be subject to empirical observation. We would hope that if
they are serious critics, they would endeavor to spell out the empirical justification for some of their

criticism. It hardly seems enough to doubt Kernberg’s observations. We are not asserting that they are

necessarily mistaken about some of their points, but that, they should attempt, as Stone (1980), for

www.freepsychotherapybooks.org 359
instance, has done, a more clinically and empirically oriented approach to some of their criticism. To
criticize Kernberg’s categorical formulation of the concept of reality testing, they might offer data that

support a continuum approach. Stone (1980) has provided examples of interviews in which

assessment of structure according to Kernberg’s criteria was extremely problematic, particularly in

patients with unusual types of affective illness or in recovering schizophrenic patients, leading him to

suggest that reality testing is not dichotomous in all situations. Our criticism of Calef and Weinshel is

that at times they seem to come close to simply saying Kernberg is wrong because he is not
“psychoanalytic.”

This brings us to consideration of criticisms of Kernberg’s theoretical endeavors. Calef and

Weinshel state that Kernberg’s theoretical position is close to, if not actually, a paradigm shift from

classical Freudian and ego psychoanalytic theories.4 Within the limits of their article, however, they do
not present convincing logical arguments for their assertion.

The question of Kernberg’s theoretical position is taken up more centrally in a paper by Klein

and Tribich (1981). In this article, Klein and Tribich are not specifically concerned with the idea of a

paradigm shift, but they state that from their point of view, “Kernberg’s rapprochement between

Freudian instinct theory and object-relations theory obscures the differences between these two
competing theories without taking any recognition of their differences” (p. 41). As is the case with Calef

and Weinshel (who criticize Kernberg’s more clinical positions), Klein and Tribich raise fundamental

questions concerning Kernberg’s theoretical positions. For example, they maintain that Kernberg’s
dismissal of “Bowlby, Fairbairn, Guntrip, and Winnicott is not based on any scientific discussion of

their theories but on the fact that these theories reject Freudian motivational theory” (p. 41). We will

not fully explore Klein and Tribich’s criticisms here, but we can comment that we find it strange to
maintain that Kernberg rejects all these theories. This in fact is not the case; Kernberg does attempt to

integrate aspects of Fairbairn, Guntrip, and Winnicott into his theoretical and clinical writings.

Before discussing more substantive criticisms of Kernberg’s theoretical work, however, we

would like to expand our introductory comments on the state of psychoanalytic theory and theoretical

www.freepsychotherapybooks.org 360
criticism. As we implied, we believe that much of the work in both areas leaves something to be
desired, when considered from the point of view of philosophy of science. Because the standards for

criticism typically seem to be so subjective (Ellman & Moskowitz, 1980; Moskowitz & Ellman, 1983),

any new psychoanalytic theoretical proposal or integration is vulnerable. We believe this statement
applies as we have noted, to some of Calef and Weinshel’s comments, and we would suggest that it

applies also to some, although not all, of Klein and Tribich’s remarks. It can be useful to discuss how

one theorist’s use of a concept differs from another theorist’s, but this does not constitute a criticism,
unless one discovers logical fallacies within the system or data that contradict the theory. To criticize

Kernberg for differing with Freud, for example, is not a theoretical criticism, but a value judgment.

However, we must also tender this and some other general criticisms in consideration of some

of Kernberg’s writings. We think that his points would be clearer if he would place greater emphasis on
stating his definitions, assumptions, and positions and less on cataloging theorists with whom he

agrees or disagrees. The clarity of the presentation of Kernberg’s theoretical propositions sometimes

suffers from his tendency to give such qualified and complex statements that it becomes difficult to use
his theoretical assertions to make definite predictions. In addition, the “catalogs” of theorists give

Kernberg’s theoretical work somewhat of an arbitrary feeling, akin to what we believe is an

arbitrariness in the writings of some of his critics, which seems to imply, “If you disagree with so and

so, then you are not psychoanalytic and, therefore, you are wrong.” This type of comment, although all
too prevalent in psychoanalytic writings, is not up to Kernberg’s standards. We would thus have to
agree with Klein and Tribich (1981, p. 39) when they criticize Kernberg’s rejection of Guntrip for his

“emotionally charged” attacks on instinct theory. Furthermore, we feel that Kernberg does not have a
strong position from which to censure another theorist for deviating from classical psychoanalytic
instinct theory.

A philosophical approach to the critical review of Kernberg’s theoretical contributions would

deal with different types of issues. We would wish to examine questions such as the following: How
well does Kernberg integrate object relations theory with Freudian theory? Aside from consideration

www.freepsychotherapybooks.org 361
of various psychoanalytic traditions, does Kernberg have a well-integrated theoretical position? And, in
a more general sense, is Kernberg’s theory a good theory, according to the requirements of theory
making such as logical structure, rules of inference, and so forth (Nagel, 1961; Popper, 1962)? We

would submit that no psychoanalytic theorist’s work could withstand this type of scrutiny. Hence,
again, the harsh tone of some of the criticism directed at Kernberg seems unwarranted.

It would be useful, however, to discuss briefly some of Kernberg’s contributions from the

standpoint of these questions to suggest directions for further work. We will comment on Kernberg’s
instinct theory, since he claims that in this work he integrates classical drive theory and object

relations concepts as well as newer data from neurophysiological studies. It is thus appropriate to ask

how well he succeeds in this theoretical integration. This question is separate from comments about

the validity, elegance, or heuristic value of Kernberg’s theory and from questions about whether or not
it is “psychoanalytic.”

In his discussion of instinct theory, Kernberg (1982h) goes over a familiar but nevertheless

important point: Freud’s term trieb, which is usually rendered as “instinct,” may more reasonably be

translated as “drive.” Kernberg is pointing out, as have others (Bibring, 1969; Hartmann, 1964; Holder,
1970; Schur, 1966), that by instinct Freud did not mean a fixed, prewired, behavioral pattern (which is

more of an ethological idea). Rather, in his concept of instinct or drive, a variety of behaviors or mental
events might emerge as a result of internal stimuli or excitation. Kernberg’s substantive attempt is to

link or translate Freud’s ideas into modern neurophysiological and neurobehavioral concepts. It is,

again, beyond the scope of this chapter to discuss fully this aspect of Kernberg’s writings, but in
summarizing Kernberg’s ideas we hope to give a sense of his position and our evaluation of this

position.

Kernberg (1976) places affect dispositions at the center of his statements on motivation. He
concludes:

Affect dispositions constitute the primary motivational systems which

www.freepsychotherapybooks.org 362
integrate the perception of (1) central (pleasurable or unpleasurable)
states, (2) physiological discharge phenomena, (3) inborn perceptive and
behavior patterns, and (4) environmental responses as they impinge on
specialized and general extroceptive and introceptive perceptions. The
earliest “self-object-affect” units are, I suggest, constellations of affectively
integrated and cognitively stored perceptions of affective, physiological,
behavioral, and environmental changes—perceptions within which the
“self” and “nonself” components are as yet undifferentiated [p. 87].

In this passage, Kernberg is attempting to link what he considers to be Freudian psychoanalytic

theoretical statements with neurophysiological statements through the use of an object relations

perspective. He goes on to specifically include MacLean’s (1967,1972) model of three concentric brains
as being relevant to the way he conceives of instinct as developing in the human being

gradually out of the assembly of these “building blocks,” so that the series
of pleasurable affect-determined units and the series of unpleasurable
affect-determined units gradually evolve into the libidinally invested and
aggressively invested constellations of psychic drive systems—that is, into
libido and aggression, respectively, as the two major psychological drives.
In other words, affects are at first primary organizers of instinctive
components such as specialized extroceptive perception and innate
behavior patterns and, later on, constitute the “signal” activator of the
organized hierarchy of “instinctually” determined behavior [pp. 87-88].

These two quotes give a reasonable flavor of the complexity and direction of Kernberg’s ideas
on instinct. We believe that, in fact, his theoretical compilation places him substantively closer to

Bowlby (1969, 1973,1980) and perhaps even Fairbairn and Guntrip than to Freud. Central to Freud’s

(1915) ideas about instinct is the formulation that it is generated internally and that the instincts
appear “as a constant force” (p. 119). Nowhere in Kernberg’s writings do we see this essential aspect of

Freud’s concept that instincts provide a form of constant internal stimulation that makes substantial

demands on the nervous system. To quote Freud (1915):

Instinctual stimuli, which originate from within the organism, cannot be

www.freepsychotherapybooks.org 363
dealt with by this mechanism. Thus they make far higher demands on the
nervous system and cause it to undertake involved and interconnected
activities by which the external world is so changed as to afford
satisfaction to the internal source of stimulation. Above all, they oblige the
nervous system to renounce its ideal intention of keeping off stimuli, for
they maintain an incessant and unavoidable afflux of stimulation. We may
therefore well conclude that instincts and not external stimuli are the true
motive forces behind the advances that have led the nervous system, with
its unlimited capacities, to its present high level of development. There is
naturally nothing to prevent our supposing that the instincts themselves
are, at least in part, precipitates of the effects of external stimulation,
which in the course of phylogenesis have brought about modifications in
the living substance [p. 120].

We have included this long quote from Freud in an attempt to capture what we believe is a
subtle but nevertheless important difference between Freud and Kernberg’s concept of instincts.

Certainly from at least 1915 on, Freud stressed the internal or endogenous nature of the instincts, not

only as a motivational concept but also, in higher-level organisms (particularly primates), as a system
that stimulated the development of the central nervous itself. Thus, the infant’s and child’s task of

“mastering” internal or endogenous stimulation is in fact a central task. Clearly, environmental factors
can make this task easier or harder, and clearly the environment is important in development, but the

“constant pressure” of endogenous stimuli will be there regardless of the type of “instinctual building
blocks” that are present in the infant’s environment. If we take Kernberg seriously in his attempted
neurophysiological integration, then he is moving toward more of an environmentalist position than

Freud held. By and large, Kernberg does not see endogenous stimulation as a central concern. Hence, in

this area of his theorizing, he has not really integrated Freud’s position into his own.

We would say that, in general, Kernberg has not fully integrated the various positions he uses;
that critics (Calef and Weinshel, Klein and Tribich) appear from both sides of the controversy between

Freudian and object relations theories is consistent with this view. At times, Kernberg merely places
together different theoretical positions rather than integrating these positions, for example, by
showing how a particular definition of a concept adds to the power of the theory. Similarly, he often

www.freepsychotherapybooks.org 364
presents his selections among possible points of view without giving the clinical or logical justification
as to why he has chosen certain positions and not others. It is never really clear that additional
explanatory power is gained by combining object relations and Freudian (or ego psychological)

concepts.

This brings us to a related logical criticism. Given that he has selected and defined certain

concepts in the formation of his theory, Kernberg provides little in the way of theoretical or logical

structures (rules of inference) to show how his theoretical positions link together in an overall
theoretical system. For example, he might begin to provide rules that would predict under what

circumstances active splitting replaces passive splitting and develop criteria independent of the

theoretical concepts to test the predictions implied by such rules. At this stage, he does not clarify the

explanatory power of his theory. To put this in another way, he does not show what the developmental,
affect, or instinctual aspects of his theory really add to our understanding of his clinical and nosological

observations and conceptions. In a sense, to use Rubinstein’s (1967) term, his theory often seems to be

“merely descriptive.” Although this is not necessarily a criticism, Kernberg obviously aspires to
something more. Yet often he does not show how this theory is more than a plausible restatement of

his clinical points.

We have been critical of Kernberg in the latter part of this review, but we reiterate that these
criticisms follow from the application of standards that, in our opinion, no psychoanalytic theorists

could meet. We have expected Kernberg to present a full-blown theory of the kind that not even Freud

managed to produce. Moreover, if Kernberg has not carried out the type of theory building or logical
analysis that would enable him to present more convincing arguments, neither have his critics. One

must sympathize with Kernberg to some extent, since his task is the harder one and since he has, at
times, attempted to alter or clarify his positions in response to points raised by critics.

In conclusion, we would say that Kernberg has raised fundamental issues and, more than any

other contemporary writer, he has pursued these questions with vigor and insight. The answers he

proposes are among the most interesting presented by today’s psychoanalytic theorists. He is also one

www.freepsychotherapybooks.org 365
of a relatively small number of psychoanalytic thinkers who devote considerable attention to research
issues and findings (Carr, Goldstein, Hunt, & Kernberg, 1979; Kernberg, 1981b; Kernberg et al., 1972).

Despite our critique, we are impressed with Kernberg’s attempts to develop a comprehensive and

systematic theory of development, psychopathology, and treatment, and he must be considered a

major psychoanalytic theorist. In many areas, one cannot begin to formulate appropriately a problem

without referring to Kernberg’s work. That is, by itself, no small achievement.

REFERENCES

American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd
ed.). Washington, D.C.: Author.

Balint, M. (1968). The basic fault. New York: Basic Books.

Bibring, E. (1969). The development and problems of the theory of the instincts. International Journal
of Psycho-Analysis, 50, 293-308.

Bion, W. R. (1959). Experiences in groups. New York: Basic Books.

Bion, W. R. (1965). Transformations. London: Heinemann.

Bion, W. R. (1967). Second thoughts: Selected papers on psychoanalysis. London: Heinemann.

Bion, W. R. (1970). Attention and interpretation. London: Heinemann.

Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York: Basic Books.

Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation. New York: Basic Books.

Bowlby, J. (1980). Loss: Sadness and depression. London: Hogarth.

Calef, V., & Weinshel, E. M. (1979). The new psychoanalysis and psychoanalytic revisionism.

www.freepsychotherapybooks.org 366
Psychoanalytic Quarterly, 48, 470-491.

Carr, A. C., Goldstein, E. G., Hunt, H. F., & Kernberg, O. F. (1979). Psychological tests and borderline
patients. Journal of Personality Assessment, 43, 582-590.

Eissler, K. R. (1953). The effect of the structure of the ego on psychoanalytic technique. Journal of the
American Psychoanalytic Association, 1, 104-143.

Ellman, S. J., & Moskowitz, M. (1980). An examination of some recent criticisms of psychoanalytic
“metapsychology.” Psychoanalytic Quarterly, 19, 631-662.

Fairbairn, W. R. D. (1952). An object-relations theory of the personality. New York: Basic Books.

Federn, P. (1947). Principles of psychotherapy in latent schizophrenia. American Journal of

Psychotherapy, 1, 129-139.

Frosch, J. (1960). Psychotic character. Journal of the American Psychoanalytic Association, 8, 544-551.

Freud, S. (1914). On narcissism: An introduction. Standard Edition, 14, 73-102.

Freud, S. (1915). Instincts and their vicissitudes. Standard Edition, 14, 117-140.

Freud, S. (1917). Mourning and melancholia. Standard Edition, 14, 243-258.

Freud, S. (1920). Beyond the pleasure principle. Standard Edition, 18, 7-64.

Greenson, R. (1954). The struggle against identification. Journal of the American Psychoanalytic
Association, 2, 200-217.

Guntrip, H. (1968). Schizoid phenomena, object relations, and the self. New York: International
Universities Press.

Guntrip, H. (1971). Psychoanalytic theory, therapy, and the self. New York: Basic Books.

Hartmann, H. (1955). Notes on the theory of sublimation. In H. Hartmann, Essays on ego psychology:
Selected problems in psychoanalytic theory (pp. 215-240). New York: International

www.freepsychotherapybooks.org 367
Universities Press, 1964.

Holder, A. (1970). Instinct and drive. In H. Nagera (Ed.), Basic psychoanalytic concepts of the theory of
instincts (Vol. 3, pp. 19-22). New York: Basic Books.

Jacobson, E. (1954). Contribution to the metapsychology of psychotic identifications. Journal of the

American Psychoanalytic Association, 2, 239-262.

Jacobson, E. (1971). Differences between schizophrenic and melancholic states of depression. In E.

Jacobson, Depression (pp. 264-283). New York: International Universities Press, (original
work published 1966)

Kernberg, O. F. (1966). Structural derivatives of object relationships. International Journal of Psycho-

Analysis, 1,7, 236-253.

Kernberg, O. F. (1967). Borderline personality organization. Journal of the American Psychoanalytic

Association, 15, 641-685.

Kernberg, O. F. (1973). Psychoanalytic object-relations theory, group processes, and administration:

Toward an integrative theory of hospital treatment. Annual of Psychoanalysis, 1, 363-
388.

Kernberg, O. F. (1975a). Borderline conditions and pathological narcissism. New York: Jason Aronson.

Kernberg, O. F. (1975b). A systems approach to priority setting of interventions in groups.

International Journal of Group Psychotherapy, 25, 251-275.

Kernberg, O. F. (1976). Object relations theory and clinical psychoanalysis. New York: Jason Aronson.

Kernberg, O. F. (1977a, May). Characterological determinants of depression. Paper presented at the

annual meeting of the American Psychiatric Association, Toronto, Ontario, Canada.

Kernberg, O. F. (1977b, December). Some objectives and technique in brief psychotherapy. Paper
presented at the symposium on Brief Psychotherapy, sponsored by the American
Psychoanalytic Association, New York.

www.freepsychotherapybooks.org 368
Kernberg, O. F. (1977c). Structural change and its impediments. In P. Hartocollis (Ed.), Borderline
personality disorders (pp. 275-306). New York: International Universities Press.

Kernberg, O. F. (1978). The diagnosis of borderline conditions in adolescence. Adolescent Psychiatry, 6,

298-319.

Kernberg, O. F. (1979a, November 6). Character structure and analyzability. In Character structure and
analyzability, panel at a scientific meeting of the Association for Psychoanalytic Medicine,
New York.

Kernberg, O. F. (1979b). Psychoanalytic psychotherapy with borderline adolescents. In S. C. Feinstein &

P. L. Giovacchini (Eds.), Adolescent psychiatry (Vol. 7, pp. 294-321). Chicago: University of
Chicago Press.

Kernberg, O. F. (1980a). Contemporary psychoanalytic theories of narcissism. Unpublished manuscript.

Kernberg, O. F. (1980b). Internal world and external reality. New York: Jason Aronson.

Kernberg, O. F. (1980c). Neurosis, psychosis, and the borderline states. In H. I. Kaplan, A. M. Freedman,
& B. J. Sadock (Eds.), Comprehensive textbook of psychiatry (Vol. 3, pp. 1079-1092).
Baltimore: Williams & Wilkins.

Kernberg, O. F. (1980d, May 4). The place of affects in psychoanalytic theory. In New directions in affect
theory, panel at the Annual Meeting of the American Psychoanalytic Association, San
Francisco.

Kernberg, O. F. (1980, September 5). Supportive psychotherapy. Presentation at clinical grand rounds,
The New York Hospital-Cornell Medical Center, Westchester Division.

Kernberg, O. F. (1981a, March). Countertransference, transference regression, and the incapacity to

depend. In Current concepts of transference and counter-transference, symposium
organized by the Association for Psychoanalytic Medicine, New York.

Kernberg, O. F. (1981b, June). Dilemmas in research on long-term psychotherapy. Paper presented at the
Annual Meeting of the Society for Psychotherapy Research, Aspen, Colorado.

www.freepsychotherapybooks.org 369
Kernberg, O. F. (1981c, September 11). Problems in the classification of personality disorders.
Presentation at clinical grand rounds, The New York Hospital-Cornell Medical College,
Westchester Division.

Kernberg, O. F. (1981d). Psychiatric hospital treatment in the United States. Nordisk Psykiatrisk
Tidsskrift, 5, 283-298.

Kernberg, O. F. (1981e). Some issues in the theory of hospital treatment. Tidsskrift for Den Norske
Loegeforening, 14, 837-843.

Kernberg, O. F. (1981f). Structural interviewing. Psychiatric Clinics of North America, 4, 169-196.

Kernberg, O. F. (1982a). Advantages and liabilities of the therapeutic community. In M. Pines & L.
Rafaelsen (Eds.), The individual and the group, Vol. 1. New York: Plenum.

Kernberg, O. F. (1982b, April 18). Clinical management of suicidal potential in borderline patients. In
Borderline personality disorders and suicidal behavior, panel at the 15th Annual Meeting
of the American Association of Suicidology, New York.

Kernberg, O. F. (1982c, May 8). The dynamic unconscious and the self. Presentation at a meeting of the
Association for Philosophy of Science, Psychotherapy and Ethics, New York.

Kernberg, O. F. (1982d). An ego psychology-object relations model of psychoanalytic psychotherapy.

Unpublished manuscript.

Kernberg, O. F. (1982e, May 16). Identity, alienation, and ideology in adolescence. Presentation at the
Annual Meeting of the American Society for Adolescent Psychiatry, Toronto, Ontario,
Canada.

Kernberg, O. F. (1982f). Narcissism. In S. Gilman (Ed.), Introducing psychoanalytic theory (pp. 126-136).
New York: Brunner/Mazel.

Kernberg, O. F. (1982g, May). The psychotherapeutic treatment of borderline personalities. Presentation

at the review-update program on narcissistic and borderline personalities at the
American Psychiatric Association, Toronto.

www.freepsychotherapybooks.org 370
Kernberg, O. F. (1982h). Self, ego, affects and drives. Journal of the American Psychoanalytic Association,
30, 893-918.

Kernberg, O. F. (1982i). Supportive psychotherapy with borderline conditions. In J. O. Cavenar & H. K.

Brodie (Eds.), Critical problems in psychiatry (pp. 180-202). Philadelphia: Lippincott.

Kernberg, O. F. (1983). To teach or not to teach. In E. Joseph & R. S. Wallerstein (Eds.), The influence of
the practice and theory of psychotherapy on education in psychoanalysis (Vol. 1).
International Psycho-Analytic Association Monograph Series, pp. 1-37.

Kernberg, O. F., Burstein, E. D., Coyne, L. Appelbaum, A., Horwitz, L., & Voth, H. (1972). Psychotherapy
and psychoanalysis: Final report of the Menninger Foundation's psychotherapy research
project. Bulletin of the Menninger Clinic, 36, (1 & 2).

Klein, M. (1946). Notes on some schizoid mechanisms. International Journal of Psycho-Analysis, 27, 99-
110.

Klein, M. (1948a). A contribution to the psychogenesis of manic-depressive states. In M. Klein,

Contributions to psycho-analysis, 1921-1945 (pp. 282-310). London: Hogarth. (Original
work published 1935)

Klein, M. (1948b). Mourning and its relation to manic-depressive states. In M. Klein, Contributions to
psycho-analysis, 1921-1945 (pp. 311-338). London: Hogarth. (Original work published
1940)

Klein, M., & Tribich, D. (1981). Kernberg’s object-relations theory: A critical evaluation. International
Journal of Psycho-Analysis, 62, 27-43.

Knight, R. P. (1954). Borderline states. In R. P. Knight & C. R. Friedman (Eds.), Psychoanalytic psychiatry
and psychology (pp. 97-109). New York: International Universities Press.

Kohut, H. (1966). Forms and transformations of narcissism. Journal of the American Psychoanalytic
Association, 14, 243-272.

Kohut, H. (1971). The analysis of the self. New York: International Universities Press.

www.freepsychotherapybooks.org 371
MacLean, P. D. (1967). The brain in relation to empathy and medical education. Journal of Nervous and
Mental Disease, 144, 374-382.

MacLean, P. D. (1972). Cerebral evolution and emotional processes: New findings on the striatal
complex. Annals of the New York Academy of Science, 193, 137-149.

Mahler, M. S. (1968). On human symbiosis and the vicissitudes of individuation: Infantile psychosis. New
York: International Universities Press.

Mahler, M. S., & Pine, F., & Bergman, A. (1975). The psychological birth of the human infant. New York:
Basic Books.

Miller, G. A., & Rice, A. K. (1967). Systems of organization. London: Tavistock.

Miller, J. G. (1969). Living systems: Basic concepts. In W. Gray, F. J. Duhl, & N. D. Rizzo (Eds.), General
systems theory and psychiatry (pp. 51-133). Boston: Little, Brown.

Moskowitz, M., & Ellman, S. J. (1983). A critical analysis of the concept of metapsychology in
psychoanalysis. Unpublished manuscript.

Nagel, E. (1961). The structure of science. New York: Harcourt, Brace & World.

Popper, K. (1962). Conjectures and refutations: The growth of scientific knowledge. New York: Basic
Books.

Rice, A. K. (1965). Learning for leadership. London: Tavistock.

Rice, A. K. (1969). Individual, group and intergroup processes. Human Relations, 22, 565-584.

Rubinstein, B. B. (1967). Explanation and mere description: A metascientific examination of certain

aspects of the psychoanalytic theory of motivation. New York: International Universities
Press.

Schmideberg, M. (1947). The treatment of psychopaths and borderline patients. American Journal of
Psychotherapy, 1, 45-55.

www.freepsychotherapybooks.org 372
Schur, M. (1966). The id and the principles of regulatory functioning. New York: International
Universities Press.

Stolorow, R. D., & Lachman, F. M. (1980). Psychoanalysis of developmental arrests. New York:
International Universities Press.

Stone, M. H. (1980). The borderline syndromes. New York: McGraw Hill.

Sugarman, A., & Lemer, H. (1980). Reflections on the current state of the borderline concept. In J.
Kwawer, H. Lemer, P. Lerner, & A. Sugarman (Eds.), Borderline phenomena and the
Rorschach (pp. 11-37). New York: International Universities Press.

Winnicott, D. W. (1965). The maturational processes and the facilitating environment. New York:
International Universities Press.

Winnicott, D. W. (1975). Through paediatrics to psychoanalysis. New York: Basic Books.

Zetzel, E. (1971). A developmental approach to the borderline patient. American Journal of Psychiatry,
127, 867-871.

Notes

1) In this chapter, we will use the terms “instinct” and “drive” interchangeably. Either term refers to
Freud’s more flexible use of the term trieb, as opposed to the ethologist’s use of “instinct”
as equivalent to a physiological process resulting in a fixed action pattern or a
stereotyped behavior pattern.

2) Kernberg adopts Jacobson's (1954) definition of refusion, as attempts to maintain absolute

gratification through fantasies that the self and object are merged, fantasies that ignore
realistic differences.

3) Although in one recent paper Kernberg (1982i) notes the necessity for caution in such statements
and urges further research, the preponderance of his writings imply greater surety about
these matters.

www.freepsychotherapybooks.org 373
4) We are considering the psychoanalytic version of ego psychology or the structural view as part of
the classical theory.

www.freepsychotherapybooks.org 374
11
WILFRED R. BION: AN ODYSSEY NINTO THE DEEP
AND FORMLESS INFINITE

JAMES S. GROTSTEIN, M.D.

Bion wrote less extensively than many of the other subjects of this volume or than many other

significant contributors to psychoanalysis. Yet, what he did write seems to have stirred profound

respect, much antipathy, considerable confusion, and even astonishment in his audiences. Most who
read him do not understand him. Many idealize him because of the experience his writing gave them,

to say nothing of his presence when he was alive—“an odyssey through the deep and formless infinite
void,” as I once heard it described. Having had an analysis with Bion, I can well empathize with this

extensive spectrum of feelings he inspired. Bion’s influence and reputation owe much to his capacity
for indirectly evoking experiences in his audience. As a particular instance of this “telescoped effect,”
some time after I finished my analysis with Bion I heard him deliver a lecture at the Los Angeles

Psychoanalytic Institute. I recall having been unimpressed, somewhat bored, and a little restless. When
I went home that evening, I found that my mind was in a whirl, and I could not sleep. I then felt

constrained to complete the outlines of three papers before I could lie down to rest.

Who was Wilfred Bion? Why and how did he evoke such disparate emotions in people? He was

www.freepsychotherapybooks.org 375
first of all an Englishman (actually Anglo-Indian) who seems to have been able to harmonize his Indian
childhood with a magnificent education at an English public school and at Oxford, to mix them further

with his capacity for wonderment and surprise, and to bring them all to psychoanalysis in a unique

way. Specifically, he was able to bring to psychoanalytic theory and practice the perspectives of Plato’s
theory of forms and Immanuel Kant’s Critique of Pure Reason, and he expanded the Freudian theory of

the unconscious from the narrower limitations of its biological foundations into a broader scope

consonant with the long tradition of Western—and even Eastern—epistemology.

Among the concepts that formed his work, Bion appreciated the concept of inherent
preconception from Plato’s theory of forms and formulated that they, under the impact of sensory-

emotional experience, link up with its external counterpart, its realization, to become a conception the

continuing affirmation and abstraction of this conception allow it to become a concept—like the
concept of a breast, for instance. Bion also called these inherent preconceptions “thoughts without a

thinker,” in the sense that they are thoughts that are older than the human race that now thinks them.

The mind had to be created, according to Bion, in order to think these primordial “thoughts.” He
invokes Kant in this way: Intuition without concept is blind; concept without intuition is empty. The

infant has the intuition of his or her experience but is as yet empty of the power to conceive of these
experiences by a notational system—to make sense of them. The task of analysis, Bion believed, is to

allow preconceptual experiences to be conceived so as to realize one’s intuition. Bion (1980) states: “A
‘marriage’ is taking place between you and you; a marriage between your thoughts and your feelings.
The intuition which is blind and the concept which is empty can get together in a way which makes a

complete mature thought” (p. 27).

Bion felt that, although language is one of the supreme accomplishments of human beings and is
essential for communication, it is also a vehicle for deception and inaccuracies. Language is static and
belongs to the sensual aspects of our development. It is therefore personal and ultimately misleading

or even suspect. Bion’s conception of the “suspension of memory and desire” reflects this belief on his
part that language in general and understanding in particular are vehicles of desire and therefore

www.freepsychotherapybooks.org 376
obtrude the ultimate experience of pure Truth. Language, and even knowledge (which he
mathematically symbolized as K), are only transient approximations to Truth (0), and we should not
confuse one with the other. This was one of the reasons, I believe, that Bion used mathematical

analogies and even spent a considerable amount of time trying to develop a mathematical grid on
polar-coordinated space with the aim of giving exact definition to psychoanalytic elements.

Mathematics are free of the sensuousness of memory and desire, he believed, and therefore are more

suitable in their unsaturation to describe phenomena from the internal world that are not describable
by a language, such as verbal language, that belongs to the sensory matrix.

Although this point may seem obscure to many, its merit lies in Bion’s attempt to help us get

beyond (behind, below) the language barrier so as to approximate pure experience before language. In

this regard, is closely in tune with the current work of Lacan and Derrida, the French
deconstructionists. Lacan in particular has called attention to the alteration of subjectivity, as “I”

descends into the “symbolic order of language.” Bion and Lacan seem to agree that words, like idols,

become static reifications of experience and progressively alienate oneself from it. Words seem to
grasp and enclose the experience so as to squeeze the life out of it and rob it of meaning. In addition to

mathematics, Bion also cited music, poetry, and art as generally superior ways of presenting the

domain of intuition.

Bion is perhaps best known to the mental health public for his Experiences in Groups (1961), in

which he revealed some of the most far-reaching innovations in the psychoanalysis of group process

since Freud. Like Freud before him, Bion viewed the group as a single entity, with a psychology that is
superordinate to the individuals who comprise it. By applying the principles of individual

psychoanalysis, however, he localized unique transferences to the group leader and special forms of
resistance unique to the group situation.

Bion’s vast experience with the psychoanalysis of psychotic patients allowed him to make

fascinating forays into psychotic thinking, and his metapsychological concepts owe much to these

experiences. One key concept he obtained from analyzing psychotics is the notion of the container and

www.freepsychotherapybooks.org 377
the contained, which designated a mother who was able to contain her infant’s projective
identifications. Her ability to do so (her reverie) allows the infant to internalize her as a mother who

can, through her reverie, contain the infant’s anguish and can thus form the basis of a “thinking

couple.” Bion’s theory of thinking distinguishes between thoughts and the mind that had to be created
to think them. Thinking in the normal individual comprises projective identifications of “thoughts

without a thinker” and sensory-emotional impressions onto an internalized object surface, which

endures the impact of these “thoughts” and then “thinks” them.

The concept of the container and the contained is often associated with the more passive aspect
of its action—that is, the passive absorption of the infant’s or patient’s mental pain. Bion meant far

more than that. The mother (or the analyst) must not only absorb the infant’s (or patient’s) pain

without being transformed by it—that is, yielding to the infant’s projections, identifying with them, and
responding reactively in tum—but must also delay them, sort them out as a prism does with a beam of

intense light, refracting them into a color spectrum of hierarchic meanings, and then, finally, act upon

them by relating to the infant’s specific needs. By doing this, mother turns the infant’s screams into
meaning, and, rather than thoughtlessly resorting to reflex action because of her hurt, uses the

containment experience for purposes of thoughtful translation. Not only does this become internalized
by the infant as a model for thinking, but it also becomes a model for permitting the experiencing of the

experience so as to “learn from experience,” which Bion believed to be the sine qua non of normal
development.

Before elaborating some of the concepts that are of importance in a survey of Bion’s work, I will
digress for a moment to give a few facts of Bion’s life. I must caution the reader, however, that these

facts are an “exercise in K,” and Bion would not for a moment want them to be confused with the truth
(O). He was born in Muttra, in the remote United Provinces of British India, on September 8,1897. His
father belonged to the British Civil Service and was an engineer. His mother was from a lower social

caste than his father. Bion was raised by two native women who read him stories from the
Mahabharata1, which made an everlasting impression on him. As was the custom with children whose

www.freepsychotherapybooks.org 378
parents were employed by the British Civil Service, he was sent back to England at age 8 to study at a
public school, Bishops Stortford, which he attended from 1906 to 1915. Shortly after World War I
broke out, he gained a commission in the Royal Tank Regiment, took part in the Battle of Cambrai (the

first major engagement in which tanks were deployed against the Germans), and was awarded the
Distinguished Service Order by King George V. He fought in every subsequent major battle on the

Western Front, poignant pictures of which can be read in his posthumously published autobiography,

The Long Week-End: 1897-1919 (1982).

Bion was demobilized late in 1918 and then went to Queens College, Oxford, in January of 1919,

where he studied modern history. It was there that he met H. J. Paton, a tutor of philosophy who

introduced him to the works of Kant and other philosophers. Following graduation, he returned to

Bishops Stortford College to teach history and French. He also had an active amateur career as a rugby
football player and coach for the swimming team.

In 1923 or 1924 he left to study medicine at University College Hospital in London and qualified

in 1929. While there, he came under the influence of Wilfred Trotter, the distinguished surgeon, who

was interested in the psychology of groups. This association was to be of great importance in Bion’s
future years when he made his important contributions to the theory of groups. After a short stint as a

medical officer in the Royal Air Force, he went to London in 1932 and began to practice psychiatry. He
entered analysis with John Rickman and began his training as a candidate in the British Institute for

Psycho-Analysis, but was interrupted by the outbreak of World War II in 1939. While Officer-in-Charge

of the Military Training Wing at Northfield Military Hospital, Bion seems to have arrived at the first
inklings of his conception of group psychology.

After the war Bion returned to the Tavistock Clinic and was appointed chairman of the

Executive Committee. After finishing his analysis with John Rickman, he was introduced by the latter to
Melanie Klein, with whom he began a second analysis.

He married in 1940, but the marriage ended with his wife’s tragic death after the birth of their

www.freepsychotherapybooks.org 379
only child, Parthenope, in 1945. Some years later he married Francesca, whom he met at the Tavistock

Institute of Human Relations and with whom he had two additional children, Julian, now a practicing

physician, and Nichola, a linguist who is currently working in publishing. Bion developed an excellent

reputation as an analyst in London and became president of the British Society for Psycho-Analysis. He
disliked this prominence, however, and often quipped “I was so loaded down with honors that I almost

sank without a trace!”

In 1966 he made a lecture tour of Los Angeles and returned the following year to remain for 12

years. During that time he exercised a profound and extensive impact on the psychoanalytic

community of that city. He also frequently traveled to Rio de Janeiro and Sao Paulo. It is interesting that

Bion seems to be more popular and his works are better known in the psychoanalytic community of
South America than in any other area.

Bion retired in 1979 and returned to England to be with his children. He died suddenly of
leukemia on November 8 of that year.

BION’S WORK

EXPERIENCES IN GROUPS

It is interesting to note that the mental health public knows more about Bion’s (1961) concepts
of group psychology than they do about him as a psychoanalytic theorist. What is now known as the
“Tavistock method” began during his experiences at Northfield Barracks during World War II and was

completed after the war at the Tavistock Clinic. As already noted, Bion, like Freud before him, observed
that groups behave with a psychology that is characteristic of the group as a unit, above and beyond
the psychologies of each of its members. A group convenes in order to focus on a common project; in

the original case at Northfield Barracks, this was a return to the battlefront. In the course of the group’s
progression, resistance, not unlike resistances in individual analysis, develops toward the progress of

www.freepsychotherapybooks.org 380
the group’s functioning. Members of the group seemed to cluster into resistance subgroups, which Bion
designated as (1) fight or flight, (2) pairing, and (3) dependence. Bion analyzed the expectations of the

individual and resistance subgroup members toward the leader as analogous to the transference

expectations in individual analyses. The Tavistock method is largely known in the United States
through the work of A. K. Rice (1965) and is used to study authority relationships in institutions. It has

never found its way into formal group psychotherapy to any significant extent.

EXPERIENCES WITH PSYCHOTICS AND THE ORIGINS OF A THEORY OF THINKING

Overlapping and succeeding Bion’s interest in the psychology of groups was his analytic work
with psychotics and borderline patients. His first contribution stemming from this interest was “The

Imaginary Twin” (1950). The patient discussed in this clinical paper was suffering from persecution by
an imaginary twin, which seemed to be derived from an earlier conception of the breast as the first

imaginary twin. In “Differentiation of the Psychotic from the Non-Psychotic Personalities” (1957a),

Bion offered the thesis that every psychotic demonstrates a normal or neurotic as well as a psychotic
personality, which long antedated Kernberg’s similar formulations. His papers, “Notes on the Theory of

Schizophrenia” (1954), “Development of Schizophrenic Thought” (1956), “On Arrogance” (1958a), “On

Hallucinations” (1958b), and “Attacks on Linking” (1959) all examined the shizophrenic experience of
attacking thoughts by attacking the links between objects and between object and self, the precursors
of thoughts.

Ultimately, Bion saw the psychotic experience as the result of a failure by the mother to contain
her infant’s fear of dying. This is, again, the concept of the container and the contained, which was to
have major significance not only for Kleinian metapsychology but also for psychoanalytic

metapsychology generally. For Kleinian metapsychology it added the adaptive principle, the formal
enfranchisement of the importance of external reality, a concept that previously was sadly lacking. It
added the concepts of “thoughts without a thinker” and inchoate emotional sense impressions that
need a thinker to think them—originally a mother—container whose ultimate internalization by the

www.freepsychotherapybooks.org 381
infant provides for this developing function. The concept of the container and the contained long
anticipated Kohut’s concept of the functions of selfobjects.

Another important contribution from Bion’s work on psychoses is his conception of alpha

function, alpha elements, and beta elements (Bion, 1962a, 1963). He used letters from the Greek
alphabet in order to avoid terms that were in ordinary use and therefore already saturated with

meaning. The second part of his terms, “functions” and “elements,” he borrowed from mathematics for

the same reason. Experience, according to Bion, begins as a beta element, which is a raw stimulus
confronting the sense organs in order to be experienced. If the sense organ allows itself to experience

the beta element stimulus, it does so through alpha function and therefore transforms the beta element

into an alpha element. The latter is analogous to metabolized food—it is suitable for mental digestion,

whereas the raw beta element is not. The alpha element comprises not just the impression of the
senses, however, but also the inherent and/or acquired preconceptions of that experience, which the

organism is prepared for beforehand; thus, the mating between preconception and beta element forms

the alpha element, the necessary ingredient for mental digestion. The alpha element may then be
transformed into dream or mythic elements for storage and/or may be processed by the mind for

immediate experience, to be thought about and acted upon.

Bion stated that alpha elements are able to produce an alpha screen, something akin to a
repressive barrier, which differentiates sleep from wakefulness. The psychotic, on the other hand, has

so much fear of experiencing his or her feelings because of being overwhelmed that he or she projects

out not only feelings and thoughts about these feelings, but also the mental apparatus that can accept,
absorb, and process these feelings—the psychotic’s very ego. As a consequence, the psychotic cannot

“alphabetize” sensory-emotional experiences—cannot allow them to be registered (cannot transform

them through alpha function into alpha elements). These experiences do not become properly
transformed for mental action; instead, they become pathologically transformed into altered beta
elements or bizarre objects, which comprise delusions and hallucinations. The psychotic seems to

develop a beta screen (rather than an alpha screen) of bizarre and persecuting objects which cluster

www.freepsychotherapybooks.org 382
around him or her and alienate him or her from the presence of and communication with others. The
absence of an alpha screen forecloses on the psychotic’s capacity to differentiate between waking and
sleep and therefore between dreaming and reality or between the delusional and real worlds.

Bion began to realize that the sense organs of the psychotic “do not talk to one another to make
common sense,” and that the psychotic uses these sense organs to project sensations onto objects,

which then become hallucinations because of their propensity for abnormal projective identification.

The “arrogance” of psychotic thinking is the defensive smugness of the psychotic in believing that he or
she can “think” by evacuating thoughts and the organ that thinks thoughts (the mind) into an object

that is then subjected to the patient’s curiosity—not for knowledge, but for control.

Bion formulated the notion of an infantile mental catastrophe as the basis for the development
of a psychotic personality. This catastrophe is a result of the infant’s hatred of experience in

conjunction with a defective maternal container, which can not soothe or contain the infant’s pain.
Bion invoked the concept of catastrophe for normal thinking as well, and his ideas in this respect seem

to have been misunderstood by many (see Hamilton, 1982). Catastrophe is a prerequisite for normal

thinking, Bion believed, because thinking requires overcoming the steady state of homeostasis and
therefore always involves pain. Thinking is in response to rents in the smooth surface of serenity; the

adjustments and adaptations we have to make inaugurate the need for thinking. Thus, to Bion,
catastrophe, in its theoretical sense, is a normal property of change, and thinking is our capacity to

anticipate, adjust to, and regulate it. This conception of “normal” catastrophe does not obviate its

counterpart, the child’s normal epistemophilic tendency, with the attendant enthusiasm and joy in
acquiring knowledge about the world.

During the years in which he worked on these concepts, Bion steadfastly endeavored to

mathematize psychoanalytic concepts so as to give them “scientific precision.” Bion’s mathematical

adventures occupy his first three major metapsychological works, Learning from Experience (1962a),

Elements of Psychoanalysis (1963), and Transformations (1965). Bion’s theory of transformations

borrows from Melanie Klein’s theory of the existence of the paranoid-schizoid and depressive

www.freepsychotherapybooks.org 383
positions, two basic sequential stages of development in early infantile mental life. The first stage is
characterized by states of persecutory anxiety in the infant, the latter by states of depressive concern

for the object, along with withdrawal of projections from the object back to the self. The depressive

position is considered a state of integration, and the paranoid-schizoid position is considered a state of
potential disintegration.

Bion proposed that all emotional states and thoughts begin in the paranoid-schizoid position as

definitory hypotheses. In other words, the infant experiences as a definitory hypothesis a sense of

certainty about its state of pain and feels persecuted by this pain. The infant thus experiences itself as
the innocent victim of pain superimposed upon it by some outside source, the nurturing object. The

infant who is able to tolerate this pain eventually may realize that the pain is like a hole or an absence

where a breast belongs and that it developed in the first place because the breast was not there when it
was needed. If the infant can tolerate the pain long enough, then the concept of an empty space

develops (like Kant’s “empty thoughts”). The infant (and, by corollary, the patient) can use this space as

an area of transformation in which a thought may alter from its original definitory hypothesis of
persecutory pain to a depressive awareness of the need for the breast and of the pain of mother’s

absence. Thus, the “thought” or “feeling” undergoes a transformation from the paranoid-schizoid to the
depressive position (P-S→D) on its way to integration and acceptance.

There are three major types of transformations: (1) rigid motion transformations, (2) projective

transformations, and (3) transformations in hallucinosis (–K). A rigid motion transformation, a term
borrowed from solid geometry, is one in which an experience from the past is experienced in the
present virtually intact. It corresponds to the general classical notion of transference, that is, of a

displacement of a past object experience to the analyst in the present. Projective transformations is
Bion’s term for Klein’s conception of projective identification; it designates the projective translocation
of aspects of the infant and/or patient in the present onto the image of the parent-analyst, who then is

believed to be identified with the projection (transformed and controlled by it).

Transformations in hallucinosis designate a much more extensive and abnormal change. Bion

www.freepsychotherapybooks.org 384
believed that the psychotic cannot bear the experience of pain and therefore does not develop the
space in which normal transformations can occur. Instead, he or she annihilates this space or never

develops it in the first place. The psychotic also projects out not only the intolerable feelings and

thoughts, but also the very mental apparatus, the ego, which feels the feelings and thinks the thoughts,
along with them. Insofar as this is experienced by the “psychotic” infant as not being able to contain,

feel, or think his or her thoughts, the infant correspondingly, thanks to projective identification,

believes the object he or she is projecting the thoughts onto also cannot contain these thoughts and
feelings or the infant’s mind, which is also being projected into the object. The consequences, according

to Bion, are as follows: The infant now is in a state of disorientation and is denuded of his or her mental
capacity to experience his or her feelings and think his or her thoughts. The infant no longer has the
alpha function to delay, sort out, and “alphabetize” feelings. He or she now lurks in the twilight of

confusion between sleep and dreams, where neither is distinguishable from the other.

The object to which the infant’s mind and feelings have meanwhile been projected has been

transformed into a bizarre object (transformation of rejected beta elements), which is unstable and
controlled by the psychotic projections within it. The latter seem to “swell up” and bizarrely distort the

configuration of the object. They then fragment, disintegrate, and reorganize as a beta screen of
psychotic impermeability, a pathological autistic shell “protectively” surrounding the denuded infant.

This beta screen is experienced as delusions and hallucinations, which circumscribe the infant or
patient as a “protectively menacing” envelope. It is protective in the sense that former painful relations
with external objects no longer occur, thanks to the beta screen, but menacing insofar as the infant

and/or patient is now in a veritible “concentration camp.”

Eventually, the patient may seemingly recover from this psychotic catastrophe but, as in the
example of the famous case of Schreber (see Freud, 1911), will reconstitute a private mental world of
his or her own which is a bizarre mock-up of the abandoned external world, totally cut off from and

impervious to it. This is the domain of –K, the final stage of a transformation in hallucinosis, otherwise
known as a “fixed” delusional system.

www.freepsychotherapybooks.org 385
A BRIEF OVERVIEW OF BION’S METAPSYCHOLOGY

Bion’s metapsychology is not only an elaboration of Freud’s basic tenets amalgamated with the
cartography of infantile mental life that Melanie Klein pioneered, but is also enriched with many

unique contributions from general epistemology and some innovative speculations about mental life
even beyond the caesura of birth. Bion believed that mental life may well begin when the sense organs
first become operant, early in fetal life, and he postulated that there may be some dim “awareness” of

the transition from the watery medium of the womb to the gaseous medium of postnatal life. These

early “experiences” are “empty” in the Kantian sense and must await language in order to become filled
as concepts. Before becoming “filled,” they may be registered as ideograms. Psychoanalysis is an

attempt to help the individual link up with his or her earliest preconceptions, giving language to those

primordial experiences that are still beyond words and may date to a time before there were words.

Bion gradually became dissatisfied with the language of Freudian and Kleinian theory, because
it dealt with objects that dwelled in the third dimension of external reality (the domain of senses).

Bion’s psychotic patients did not dwell in that third dimension but rather in dimensions alien to it.

Because the dimensionality of the internal world normally and of the psychotic world abnormally is far
different from that of the third dimension of external reality, it requires a language suitable to it. Today
we might call this the domain of the nondominant hemisphere, which can be thought of as the zero

dimension (see Grotstein, 1978).

Bion reminds us that Freud had postulated that the psyche’s capacity for consciousness
depends on mental processes that are “sensible” to the data of experience from external stimuli.

Pleasure and unpleasure are the original “codes” of differentiation of these processes. Even though
Freud hinted that the sense organ of consciousness was “sensible” to internal stimuli as well, there
seems to be very little in the psychoanalytic literature to designate the exact nature of that internal

sense organ system. Bion came to the rescue by reminding us that the sensual domain relates to the

external world and that intuition is the sense organ that is responsive to the internal world. In order to
be intuitively “sensitive” to this inner domain, one must blind oneself to the sensory capacities that are

www.freepsychotherapybooks.org 386
responsive to the external world of sense-dominated reality. Following a notion of Freud, Bion
advocated abandoning memory and desire in order to allow oneself to be intuitively sensitive to this
internal, dimensionless world. Ironically, we associate the pleasure principle with the id and therefore

with the unconscious. Pleasure seems to be the designation by the sense organs that are responsive to
external reality to code that form of information; therefore, paradoxically, we must eschew our

tendency toward pleasure and desire (suspend them) to allow the “thoughts without a thinker” to

emerge in the inner domain. These “thoughts without a thinker” do not have a language of their own
and must borrow the language of external reality via free associations (day residue experiences) to

become “visible” (sensible through intuition). Memory is the past tense of desire (of the senses), and

desire designates the future (of the senses). Thus, the analytic procedure requires suspension of
memory and desire so that there can be intuitive receptivity to the inner world. The ability to do this

requires the “man of achievement,” a designation Bion borrowed from Keats to connote the capacity
for patience in a field of doubts, mysteries, and half-truths while awaiting the selected fact. The
appearance of this selected fact rewards the “man of achievement” with intuitive security and

clarification about inner meaning.

From another point of view, we can see this formulation as Bion’s concern about the difference

between Truth (O) and knowledge (K). Our sense organs are “sensible” to knowledge about Truth but
are limited to the acquisition of knowledge about it. Words correspond to K; O is wordless and is the

thing-in-itself, unknowable. Psychoanalysis attempts to be a transformation in O, not by our

understanding of K, but rather via the experience of K. Thus, knowledge itself does not permit
transformation in O, only experience does. K is important only in being able, once accepted, to be used

to facilitate experience itself, the only route to O.

GENIUS AND THE “MESSIAH THOUGHT”

In Bion’s last metapsychological book, Attention and Interpretation (1970), he returned to his
earlier work on group formations and integrated that work with his earlier conceptions of elements,

www.freepsychotherapybooks.org 387
experiences, and transformations. Once again, Bion drew the analogy between the individual mind and
the group establishment and located within this group establishment the function of preserving and
conserving the group’s stability. The establishment resists change to defend against anticipated chaos.

The protectors of the establishment need to anticipate rebellion or challenge to the stability of the
group that may potentially endanger it. They therefore must locate the “enemies within,” stigmatize

them, and ultimately exile them from the group.

At the same time, the establishment must paradoxically anticipate the need for change so that
the group unit does not decay or disintegrate of its own accord. It therefore must prepare the way for a

messiah or genius (corresponding to the “messiah thought”), the new leader who is able to have a

“memoir of the future” and to be able, as a “genius”, to experience O directly without having to detour

through K. The genius and the messiah thought correspond to the “thought without a thinker,” the
inherent preconception that has not yet been thought but that is needed to be known and thought so as

to come to the rescue of the stalemated group establishment. The genius (and/or messiah thought) is

then conceived of as the definitory hypothesis, the apodictic message to the group, which then attacks
and challenges the veracity of this thought in an attempt to negate it.

When negation fails, the thought or feeling is accepted, notated, paid attention to, subjected to

inquiry, and, finally, acted upon. These functions (definitory hypothesis, negation, notation, attention,
inquiry, and action) occupy the horizontal axis of Bion’s mathematical grid. The vertical axis develops

in terms of the transformation from beta element → alpha element → dream thoughts or myths to

preconception → conception → concept to scientific deductive system → algebraic calculus. Thus,

whether in the group or the individual, the “thoughts without a thinker,” when allowed a

transformative space and time to be contained, thought about, and challenged, can then be accepted
and allowed to undergo their matriculation into ever-ascending conceptual schemes. This is how
individuals and cultures grow.

Bion meant this conceptualization of the messiah thought and its conflict with the very

establishment that summoned not only to be a statement in general about the evolution and

www.freepsychotherapybooks.org 388
maturation of the individual and culture, but also, undoubtedly, to be a generalization about the
difficulties he observed in the psychoanalytic establishment. The classical Freudian school, locked as it

was in the oedipal paradigm, seemed unconsciously to evoke the need for the messianic thoughts that

Melanie Klein brought to psychoanalysis about early infantile (preoedipal) mental life. The same battle
fought by Freud, the erstwhile messiah of another age, occurred again, with Klein in opposition to

Freud’s descendants. As it has turned out, the “messianic” ideas of Klein have not safely traversed the

challenge imposed by the psychoanalytic establishment—they have not cleared negation. Yet there are
ingredients in her discoveries that are essential to the normal progression of psychoanalytic theory. It

therefore seems as if a compromise formation has been instituted in classical analysis in which
Kleinian ideas have been extracted from her matrix, alienated from her, and now regrafted to classical
theory under a new name.

Transformations in which the genius and his or her messiah thoughts are accepted by the group

establishment are termed by Bion symbiotic, as both the group and the genius benefit from the

interchange. The fate of Kleinian ideas might correspond to what Bion calls a parasitic transformation,
insofar as the establishment did not recognize that it was in fact dependent on her ideas for its future

welfare and therefore “extracted” the ideas parasitically without full gratitude to their author. A third
form of transformation, which Bion calls commensal, designates the simultaneous presence of two

separate kinds of ideas or subgroups within a larger group that live in peace and harmony. They either
have not yet come into conflict or are able to live in harmony without the necessity of interaction or
conflict. In sociopolitical terms we might call this “pure democracy.”

The other fate of the messiah idea, especially when it comes before its time, is to ignite the

messianic idea in others via linear progression, much in the way that free associations transpire in a
seemingly endless chain. Suddenly, the selected fact once again emerges as the messiah thought which
is necessary for the survival of the person or group. The new messianic ideas may seem not to

“remember” their ancestry. One can see this phenomenon today in the work of Kohut and self
psychology with its emphasis on the empathic principle in psychoanalysis, which does not yet seem to

www.freepsychotherapybooks.org 389
know its ancestry in Sullivan, Fairbairn, Winnicott, Balint, Bowbly, and so many others, to say nothing
of Klein.

The coming of the genius and the messianic thought he or she expresses is apparently a

historical pattern with fortunate as well as revolutionary consequences. If the establishment cannot
bear the strain imposed on it by the messiah thought, then there is catastrophic revolution with violent

change (transformation in – K). If the thought is accepted by the establishment, it changes

correspondingly, in which case there is a transformation in K on its way to experiencing O (the thing-
in-itself, pure experience).

BION’S CONCEPTION OF THE UNCONSCIOUS

When Freud first discovered the system unconscious, it comprised the domain of traumatically

buried memories. Later, when he discovered the importance of fantasy, the unconscious became

composed of the instinctual drives, those elements of experience that had become secondarily
instinctualized and pulled into repression, the unconscious portion of the ego (especially the ego

defense mechanisms), and the superego. Bion’s invocation of inherent preconceptions modified this

picture. First, he saw the psychic apparatus as being composed, as did Freud, by the ego, superego, and
id. Unlike Freud, however, he saw them as three different vertices of experience of objects outside

them. In other words, the phenomenon of pain can be understood from a moral or religious point of
view (superego), from a rational or scientific point of view (ego), or from an esthetic or need-desire
(id) point of view, and correlations between these vertices of experience are required for integration.

Second, Bion’s notion of inherent preconceptions modified the concept of the unconscious in yet

another way by suggesting that what impelled its way into consciousness normally and abnormally

was not so much the instinctual drives per se, but inherent preconceptions of danger (“thoughts
without a thinker”) that traumatic experience has evoked. Thus, danger is not from the drives, but from

the driving force of the most atavistic reminder of imminent danger. This is a vastly different notion of

www.freepsychotherapybooks.org 390
psychic interaction, which can have enormous consequences in the treatment of patients, especially
psychotic and borderline patients. It makes a great deal of difference if the therapist conceives of a

patient’s psychotic break as being due to id irruptions rather than to urgent warning signals of ancient

preconceptions alarmed into readiness with an ego unable to listen or able to respond.

BION THE PHILOSOPHER

Bion was well versed in philosophy and was himself a philosopher as well. In terms of his

formal philosophical background, I have already mentioned Plato and Kant. To this list must also be

added Hume and many others, particularly the Intuitionistic mathematicians such as including
Poincare. Bion the philosopher, however, was another matter. One always felt when talking with him

that one was in the presence of a person who had thought profoundly and intimately about the nature
of relationships.

Although he revered the highly special and unique relationship between the patient and analyst,
Bion correspondingly deprecated the supervisory relationship. He generally refused to take on

supervisees for more than a few sessions. His rejection of the idea of supervision was based on his
belief that the therapist seeing the patient, no matter how inexperienced and ill trained, has more

authority about the experience that transpired than the “supervisor,” who is removed from the
experience. All the “supervisor” can do is share his or her own feelings as a “second opinion,” a favorite
expression of Bion’s.

A similar attitude was expressed in Bion’s memorable reply to a member of a group that had

been meeting with Bion who had presented some case material for his “second opinion.” The case
material referred to a patient’s relationship to his sister. Bion (personal communication) stated:

I don’t know why your patient feels guilty about his feelings toward his
sister. After all, she is a member of his father’s family, not his. The father’s
family is a temporary family, a rehearsal family, if you will, which has been
ordained, one must presume, to complete the rehearsal of childhood until

www.freepsychotherapybooks.org 391
such time as the human being is able to find his own family, the permanent
one, the thing-in-itself. His sister is no concern of his. It is a concern of her
father and mother, if they care to be concerned. On the other hand, if your
patient does desire to be concerned, then that is his business, and he may
be experiencing feelings from the time when he actually was a member of
his father’s family, along with his sister. Now that would be a different
matter. Natural affection has no rules.

I hope the reader can follow the twists, turns, ellipses, and zig zags of this thinking. Bion was a
magnificent tactician and strategist not only in combat, but also behind the couch.

Another example of Bion’s “philosophy of relationships” came out during one of my analytic

hours with him. I was complaining to him about how disappointed I was in myself. His reply was
instant and surprising:

You are the most important person you are ever likely to meet, therefore it
is very important that you be on good terms with this important person,
you. You appear more than willing to bear testimony against yourself, yet
are not supplying me with the evidence. Besides, whom am I to believe, the
accusor or the defendant? You haven’t yet presented evidence which
either I or the defendant can respond to.

On yet another occasion, when Bion gave me a particularly powerful and cogent interpretation I

(foolishly, in retrospect), said, “You know, you’re right; that’s a correct interpretation!” Bion
sarcastically replied:

“Oh yes, you would have me be right. How right I am! you state. I’m right
only because I uttered a second opinion about your associations to me. I
could just have easily have stated, ‘you’re right! By God, how right your
free associations are!’ ”

What I came to realize from this encounter was that Bion was enjoining me to be myself, respect
myself, reclaim my “power of attorney,” and use the mind God gave me—that is, to accept the

www.freepsychotherapybooks.org 392
responsibility of my own importance and the importance of consulting my feelings and listening to my
own responses to my experiences rather than trying to “understand” those who speak to me and whom

I am in danger of making mentors rather than “partners” with second opinions.

Although also an astute logician, Bion was superbly “right brain” as well. He was not only an

accomplished pianist and gifted artist in his on right, but he also had a high regard for the aesthetic

vertex of human experience and this was his genius, he revered imagination, which he often designated
as “image-ination.” “All that can be imagined is!” he was fond of saying.

BION IN PERSPECTIVE

Bion’s public language, both in his speeches and writings, closely epitomize his

metapsychological beliefs. He eschewed understanding because of his belief that understanding closed

off the experience and therefore foreclosed the transformation in O. He often cautioned that one should
not try to understand what he said or wrote but rather should be receptive to one’s individual

impressions and responses to what he said. “Do not listen to me, but listen to yourself listening to me,”
would be a succinct restatement of his view. He thereby clarified a theory of thinking whose

rationalistic roots go back to Plato and have coursed through Kant. It embraces a philosophical
conception of the human being as the innovator of imaginative conjecture, that intersects with the data

of external experience (K) to emerge as thought. He arrived at these ideas about thinking from many
years of psychoanalyzing psychotics who could not think. Psychoanalysis had previously concentrated
on the treatment of neurotics who could think but would not in selected areas of inhibition. By

clarifying that realm of psychotic transformation that is beyond repression and comprises the
mutilation of thoughts and thinking, Bion added a whole new domain to our clinical knowledge as
well.

The interested reader who wishes to become familiar with Bion for the first time, but who

might be afraid of becoming lost in the progression of his works, might well begin by reading one of his

www.freepsychotherapybooks.org 393
last publications, Bion in New York and Sao Paulo (1980). No background is required, and the reader
will be put quickly and effectively into Bion’s way of thinking. For the more intrepid reader, I

recommend all his works, especially his novel, A Memoir of the Future (1975,1977,1979), a trilogy that

reflects his incredible virtuosity in fictional form and that constitutes a summary of his psychoanalytic
thinking. For the reader who wishes to get to know Bion the man, I heartily recommend his

autobiography, The Long Week-End: 1897-1919. This work is graphic, direct, uncharacteristically lucid,

deeply personal, and moving.  

REFERENCES

Bion, W. R. (1950). The Imaginary twin. In W. R. Bion, Second Thoughts, London: Heinemann, 1967, pp.
3-22.

Bion, W. R. (1954). Notes on the theory of schizophrenia. In W. R. Bion, Second Thoughts, London:
Heinemann, pp. 23-35.

Bion, W. R. (1955). Language and the schizophrenic. New Directions in Psychoanalysis. (Eds.) Melanie
Klein, Paula Heimann, and R. E. Miney-Kyrle. N.Y.: Basic Books, 1957, pp. 220-239.

Bion, W. R. (1956). Development of schizophrenic thought. In W. R. Bion, Second Thoughts, London:

Heinemann, pp. 36-42.

Bion, W. R. (1957). Differentiation of the psychotic from the non-psychotic personality. In W. R. Bion,
Second Thoughts, London: Heinemann, pp. 43-64.

Bion, W. R. (1958a). On arrogance. In W. R. Bion, Second Thoughts (pp. 86-92), London: Heinemann,
1967.

Bion, W. R. (1958b). On hallucination. In W. R. Bion, Second Thoughts, London: Heinemann, pp. 65-85.

Bion, W. R. (1959). Attacks on linking. In W. R. Bion, Second Thoughts, London: Heinemann, 93-109.

Bion, W. R. (1961). Experiences in Groups. London: Tavistock Publications.

www.freepsychotherapybooks.org 394
Bion, W. R. (1962a). A theory of thinking. In W. R. Bion, Second Thoughts, London: Heinemann, pp. 110-
119.

Bion, W. R. (1962b). Learning from Experience. London: Heinemann.

Bion, W. R. (1963). Elements of Psycho-Analysis. London: Heinemann.

Bion, W. R. (1965). Transformations. London: Heinemann.

Bion, W. R. (1970). Attention and Interpretation. London: Tavistock Publications.

Bion, W. R. (1975). A Memoir of the Future: Book I: The Dream. Brazil: Imago Editora.

Bion, W. R. (1977). A Memoir of the Future: Book 2, The Past Presented.

Bion, W. R. (1979). A Memoir of the Future: Book 3, The Dawn of the Oblivion. Perthsire: Clunie Press.

Bion, W. R. (1980). Bion in New York and Sao Paulo. Perthsire: Clunie Press.

Bion, W. R. (1982). The Long Week-End: 1879-1919-Part of a Life. Abington: Fleetwood Press.

Freud, S. (1911). Psychoanalytic notes in an autobiographical account of a care of paranoia (dementia

paranoides). Standard Edition, 12, 3-84.

Grostein, J. S. (1978). Inner space: Its dimensions and its coordinates. International Journal of
Psychoanalysis, 58.

Hamilton, V. (1982). Narcissus and Oedipus. London: Routledge & Kegan Paul.

Rice, A. K. (1965). Learning for Leadership: Interpersonal and Intergroup Relations. London: Tavistock.

Notes

1) Sacred Indian epic poem dealing with the ideas of goodness and evil.

www.freepsychotherapybooks.org 395
12
PAUL RICOEUR: REPORTING, READING, AND
INTERPRETING

ROBERT S. STEELE, PH.D.

Hermeneutics is the reflective practice of interpretation. Although its primary concern is with

textual exegesis, its domain extends throughout the humanities, from the social sciences to the arts. In

his work, the eminent continental philosopher Paul Ricoeur has covered this territory. He has done
close textual analyses and enlightening readings of what he called The Symbolism of Evil (Ricoeur,

1967), of religious faith and atheism, of the phenomenologies of Husserl and Jaspers, of psychoanalysis
(1966, 1970, 1974), and of metaphor.

As a theory about the practice of interpretation, hermeneutic prescriptions can be made rather

explicit (see, for example, Radnitzky, 1973 and Steele, 1979). However, these many guidelines about

the text and reader relationship can be reduced to two conflicting demands that arise from the fact that

books are both closed and open. A text, or any being or thing that is interpreted, is enclosed. It has its

own boundaries, be they covers, the imaginary space inhabited by the “I,” or the symbolic and real
limits of our bodies. That closure or completeness must be respected, and a reading must be in part a

reporting that presents the text on its own terms; one must be faithful to the letter. But, one must also

www.freepsychotherapybooks.org 396
help the spirit speak. When one opens a book, one enters a new place and, if the reading is engaging,
the reader is changed by her or his immersion in the pages of another’s thought. One owes it to the text

and to the telos of modem consciousness to give back to the work the freedom it has given one.

Interpretive readings open enclosures by bringing out what is latent, hidden, shy, or self-effacing in
them.

Truth is opening. By reading Freud closely and sympathetically, Ricouer brings out of

psychoanalysis new ways of seeing it that have been buried by the sediment of too many debates about

the epistemological or, more specifically, scientific status of Freud’s research. In my report on Ricoeur’s
work I will review these findings, discoveries that, when I first read Ricouer, revolutionized my

thought about psychoanalysis.

In returning to Ricouer’s Freud and Philosophy (1970) after nearly a decade, I have found not

only that I have changed, but that the meaning of this book is also different. Of course, the words on the
page are the same and my underlinings and marginal comments are still there to remind me of the

joyous insights shared by author and reader, but what was once a manifestly brilliant work seems now
to have a latent content, which casts a darker light on the surface text.

In my reading of Ricouer, the second part of this essay, I will bring to light what Freud and
Philosophy does not say, and yet means. This reading will do violence to the text, because it breaks
open its enclosed discourse by identifying the text’s way of speaking as a symptomatic expression of

androcentrism. I hope by naming this rather common textual constriction I will help others see it and
thereby aid them in creating what Ricoeur many years ago helped me find: freedom.

The movement of the spirit is like a spiraling uroboros. In growing, consciousness continually
consumes its previous insights. A book, a way of thinking, or a certain style of performance creates a

new way of being in the world. Initially, perhaps, a book’s message is resisted as unpalatable, but one

comes to live its insights more and more, until one tires of the same fare day after day. Something new

comes along, which is initially quite foreign, but in opening to it, in tasting it, one comes to like it and

www.freepsychotherapybooks.org 397
live it. The new way dates the old, in fact makes it old. One now has perspective on one’s previous taste

and can reflect upon it, criticize it, and perhaps preserve what is left of it by combining it in a new

recipe for being.

Both the letter and the spirit of a work help us grow. The letter, the overt treatment of the

issues, does this when we accommodate ourselves to it by letting it help us see in a different way; the

spirit, what is manifestly unseen, does this by always promising new ways to be, even though within
our present enclosure we feel complete.

THE REPORT

The son of Jules Ricoeur and Florentine Favre, Paul Ricoeur was born in Valence, France, in
1913. He married Simone Lejas in 1935, and they have five children.

Ricoeur’s early work shows the influence of his mentor, Gabriel Marcel, but his intellectual

scope has greatly expanded in the nearly half a century he has been writing philosophy. His many
books and countless articles have made him a modem master. He holds appointments at the University

of Paris and the University of Chicago.

RICOEUR’S PLACE IN THE HISTORY OF PSYCHOANALYSIS

Ricoeur’s essays on Freud in The Conflict of Interpretations (1974) and Freud and Philosophy

(1970) are part of the “return to Freud” movement which began in France in the late 1950s, flourished
throughout the sixties, and was imported to America in the late seventies. Whereas Lacan was the

charismatic, enfant terrible, psychoanalytic spokesriddler for “Freud’s French Revolution” (Turkle,

1978), Ricoeur was the academic philosopher and scion of the rich phenomenological tradition.

“French Freud,” as this genre has been called, stresses textual meditations on Freud’s writings
(Mehlman, 1976). These close, rich, and complex readings explore the ambiguities of psychoanalysis.

www.freepsychotherapybooks.org 398
The scope of these enquiries is broad ranging from Laplanche and Pontalis’ (1973) marvelous essays
on Freudian terms, to Derrida’s (1976) at times baffling treatment of the meaning of inscription in
Freud’s writings, to Lacan’s (1976) poetic and playful oedipal interpretation of Poe’s “Purloined

Letter,” to the beautifully evocative prose of Irigaray (1980) on female sexuality.

For the French, American ego psychology is, if not anathema, at least in the dialectical position

of antithesis to their synthesis. Where the Americans have stressed assimilation to the rigors of
science, testing Freud’s thought empirically, clarifying it by simplifying its ambiguities, and

establishing sounder relations with biology, the French have abhorred the medical, scientific, and

normative use of Freud by ego psychologists. If Hartmann’s aspiration was to be a scientist, Lacan’s
was to be a poet.

As part of the French engagement with Freud, Ricoeur’s work shares these prejudices. His

reading reflects his tradition, which is phenomelogical and structural, but unlike Lacan’s ricocheting

potshots across the Atlantic, Ricoeur’s treatment of Anglo-Saxon Freudian research is careful,

concerned, and masterful. Like the Lacanian excavations of psychoanalysis, Ricoeur takes us into
Freud. Freud and Philosophy is a textual, experiential exploration of the depths of psychoanalysis.
Where the Americans point beyond Freud to a general psychology, the French, and Ricoeur in

particular, return to Freud and teach us how to read him.

THE PHENOMENOLOGICAL HERMENEUTIC TRADITION

Although, as noted, hermeneutics has traditionally been associated with textual exigesis, over
the last century we have become conscious of ourselves as the ones who are interpreting. We have

realized that humans are “hermeneutic animals” in that we read signs—be they tracks, traces, cries,

auguries, data or texts. In our natural science we follow Newton in “reading the book of nature,” and in
our cultural sciences we find meaning in the artifacts of our being. Hermeneutics is interpretation; it is

the practice of interpretation, the study of this practice, and reflection on such study. It is an

www.freepsychotherapybooks.org 399
articulation of the movement of consciousness from the inarticulate through to the well said.

Wilhelm Dilthey’s vision of the province of hermeneutics was modest when, late in the

nineteenth century, he declared that it was the methodological foundation of the Geisteswissenschaften.

Trying to save the humanities from the progressive encroachment of the natural sciences, Dilthey
asserted that the cultural interpreter had special access to his or her subject because, unlike the

natural scientist who must observe nature objectively from the outside, the hermeneut is a participant

within the historical field. Participant observation is at the heart of social analysis, because an

understanding of the culture that constitutes us is only gained in and through our participation in
civilization.

Unlike the scientist, who, through various cultural ritual, tries to separate him- or herself from

phenomena classified as natural and is therefore necessarily removed from participation with the
object of study, the interpreter is enmeshed in her or his humanity and thereby participates within the

phenomenon being analyzed. Whereas the orthodox natural scientist must be freed from co-

participation in nature and must block empathetic responses, the interpreter must begin with empathy
and use it as the source from which to articulate her or his work.

Unlike natural science, which, in trying to universalize its findings, resists attempts to relativize

its objective results by submitting them to sociohistorical critiques, hermeneutics is firmly rooted in its

history and constantly submits its seminal ideas and texts to reinterpretation. Whereas science orients
itself in the replication and extension of observations, hermeneutics locates itself within language and

our textual heritage.

Ricoeur’s lineage goes back to Descartes, and his work traces the evolution and descent of the
cogito and consciousness through the master works of phenomenology—the writings of Hegel,
Nietzsche, Husserl, Heidegger, and Merleau-Ponty. His reading of Freud is a confrontation of his

tradition—the hermeneutic phenomenology of consciousness—with the science of the unconscious—

psychoanalysis. Ricoeur (1950) writes: “I should say at the start that reading works on psychoanalysis

www.freepsychotherapybooks.org 400
has convinced me of the existence of facts and processes which remain incomprehensible as long as I

remain prisoner of a narrow conception of consciousness” (pp. 375-376). In exemplary hermeneutic

fashion, Ricoeur is intent on submitting the prejudices of his training, which he recognizes as

restricting, to the challenge of Freud’s attacks on the narcissism of consciousness. All good readings
are, however, dialectical: Not only was Ricoeur changed by reading Freud, but Freud, too, was altered.

This is because the product of Ricoeur’s years of Freudian study, Freud and Philosophy: An Essay on
Interpretation (1970), is a book that has created a new understanding of psychoanalysis, not as some

misfit science, but as a hermeneutic endeavor.

A SUMMARY OF RICOEUR’S READING OF FREUD

Freud and Philosophy is a master text. Its tone is one of reconciliation, restoration, and
exploration. Its arguments are complex and demand an understanding of both the phenomenological

and logico-empirical traditions. In reading it, it helps to have read Freud closely, because one then

understands more deeply how Ricoeur’s revisions are based both on the letter of Freud and the spirit
of his project, which is to make the latent manifest.

This review will summarize four major interrelated themes that organize Freud and Philosophy:
saving Freud from science, the place of consciousness after Freud, the semantics of desire, and

Ricoeur’s study of symbolism.

Saving Freud from Science. Although Freud located psychoanalysis within the domain of the

natural sciences and insisted on its scientific status, his work has long been exiled from that land to

which he was never granted a passport. His one prize, the Goethe, was in letters; he is studied in the
humanities, not in biology or scientific psychology.

Ricoeur provides Freud a haven from his scientific critics by granting them their criticism.

Ricoeur (1970) agrees that “psychoanalysis is not an observational science” (p. 358), but he uses this
admission to counter behaviorist, experimental, and logico-empiricist attacks on psychoanalysis.

www.freepsychotherapybooks.org 401
Drawing the line clearly between psychology as a behavioral science and psychoanalysis, Ricoeur

declares that the difference between them “comes at the beginning or never.” He continues:

“Psychology is an observational science dealing with the facts of behavior; psychoanalysis is an

exegetical science dealing with the relationships of meaning between substitute objects and the
primordial (and lost) instinctual objects” (p. 359). Whereas a fact in behaviorism is a datum that is

verifiable by multiple independent observers, there are no facts, as science understands them, in
psychoanalysis, “for the analyst does not observe, he interprets” (p. 365). Behaviors are significant in

psychoanalysis because they are “signifiers for the history of desire” and not because they are

“observables” (p. 364). For Freud, the focus of study is the meaning of symptoms, dreams, delusions,
and faulty actions in a life story that is being unfolded. The analysand’s speech and behaviors present

these, and the analyst and analysand articulate their significance through interpretation.

If significant behaviors are operationally defined and recorded in settings that do not allow the

ambivalence of human action to be shown or the ambiguities of speech to be expressed, then there is
no need for psychoanalysis. This is because what is manifest in the observational situation is defined as

the datum; it need not be read, but only recorded. Any such situation is neither entirely human nor

psychoanalytic. Any capitulation on this point is to Ricoeur an abandonment of what he sees as Freud’s
central project: the explication of meaning through discourse. It is in the illusions and disillusionments
of exchange between analyst and analysand, reader and text, ourselves and others as well as between

us and our artifacts—paintings, music, machines, and dreams—that hermeneutics locates itself and in

which Ricoeur places Freud’s work.

The Place of Consciousness after Freud. Freeing epistemology from the dictates of scientific
rationalism, liberating language from the demands of rational discourse, and saving the person from

the rationalizations of false consciousness are three variations on one historical theme: “the

dispossession of the ego” (Ricoeur, 1970, p. 55).

For Descartes the cogito, “I think, therefore I am,” is a transparent certainty in a world of things

and beings that are opaque and resistant to immediate understanding. However, if consciousness is not

www.freepsychotherapybooks.org 402
pellucid, if it “is not what it thinks it is, a new relation must be instituted between the patent and the
latent” (Ricoeur, 1970, p. 33). Ricoeur continues: “After the [Cartesian] doubt about things, we have

started to doubt consciousness” and those “masters of suspicion”—Marx, Nietzsche, and Freud—have

fostered our distrust of the purity of consciousness, which is a given for empiricism and
phenomenology. “All three [men] clear the horizon for a more authentic word, for a new reign of Truth,

not only by means of a ‘destructive’ critique, but by the invention of an art of interpreting” (p. 33). To

be suspicious means to doubt the given—be that the evidence of our senses, our instruments, our
consciousness, or the text before us—and to create via interpretation from the latent, the unseen, the

unconscious, and the unsaid a context that illuminates the ambiguities of the obvious.

Ricoeur began his reading of Freud in order to challenge the epistemologies of consciousness in

which he was schooled, and so he is very careful in locating the position of consciousness in Freud’s
work. Freud displaces consciousness in two ways: He makes its position relative to other psychic

processes in the mind, and he discounts the veracity of its testimony.

Freud not only removes consciousness from the center of mental being, he keeps changing its
location and redefining its relations to the ego as he creates new representations of the psyche. Ricoeur

painstakingly records and comments on these moves in Freud’s texts, because with these models
Freud is not only trying to locate consciousness, he is also redefining its relationship to knowing.

The first representation that Ricoeur considers (he does not examine Freud’s psychic model in

Studies on Hysteria), is the neuronal ego of The Project for a Scientific Psychology and the ω system with
which Freud (1895) unsuccessfully tries to represent consciousness. Ricoeur calls Freud’s
neuropsychological model “a nonhermeneutic state of the system.” This biophysics machine did not

run, because within it Freud could not represent meaning; it did not explain consciousness. It was

replaced by the mental apparatus of chapter 7 of The Interpretation of Dreams (1900), which is a
topographical spatialization of the psyche with three regions—unconscious, preconscious, and

conscious—and boundaries of censorship between them.

www.freepsychotherapybooks.org 403
This topography improves on that of the Project because it not only pictures intrapsychic

relations, it also helps to explain how we come to know. In it Freud combines the energetics of the

Project with the hermeneutics of dream interpretation to describe the process of making the
unconscious conscious. The dream work at the behest of the censorship between the unconscious and

the preconscious distorts, by the mechanisms of condensation and displacement, unacceptable wishes
in order to preserve the repose of the sleeping ego, which would be shocked by such desires. This

dream work is undone by a countereffort of interpretation, as consciousness comes to know of these

impulses retrospectively through interpreting their disguised expression in dreams. Freud has linked
his energetics to his hermeneutics because the mechanisms of condensation and displacement not only

signify transformations of energy, but they also provide interpretive concepts for understanding the
distortion in dreams. Any picture or text will be garbled if its scenes are compacted and confused and
its emphasis is misplaced. One brings out what is latent in it by unpacking and sorting out its images

and relocating its emphasis. Knowing, within this model, means making evident to ego consciousness,
through interpretation or undoing of the dream work, what it has been denied by its own censorship.

The next significant “dispossession of the ego” from its reign as all-knowing consciousness
comes in Freud’s papers on metapsychology. Working with the topography of the dream book, Freud

(1914) further displaces the omnipotent ego by showing first that its esteem comes from its self-

cathexis. Therefore the ego is, in part, narcissistic, self-absorbed, and infantile. He next explored the
complexities of the relations between the conscious and the unconscious and linked these to concepts

of instinctual representation and verbal inscription (Freud, 1915). He thereby cast more doubt on

whether ego consciousness has unmediated access to its desires, its past, or the world, since what it

knows directly is censored transcriptions of experience. Finally, in Mourning and Melancholia (1917),
Freud forever violates the integrity of the ego by showing how it is structured by its identifications

with significant others and altered by its incorporations of lost love objects.

In his papers on technique (circa 1911-15), Freud explored the implications for analysis of

dealing with a consciousness that is an agent of the ego’s defenses. No longer is the analysand’s

www.freepsychotherapybooks.org 404
knowing simply dependent on an insightful interpretation that enlightens an anxious ego; insight now
depends on working through in the analytic session all those traumas that have distorted one’s
relations with reality. Thus, the increasing complexity of Freud’s representations of the energetics and

topography of the psyche is mirrored by a corresponding complexity in what it means to make an

interpretation that creates insight. Interpretation now comes to be imbedded in the transference

relationship as the analysand relives via projection onto the analyst the scenes of a life which have

worked to make consciousness resistant not only to the unconscious but to knowledge about itself
(Freud, 1912).

These insights, along with Freud’s work on the ontogeny and phylogeny of the Oedipus

complex, are incorporated into Freud’s last and most radical revisioning of the psyche in the structural

model of The Ego and the Id (1923). In this work the solipsistic energy system of the Project is gone.
The psyche is now a scene inhabited by near mythological personifications of nature (the “it”), culture

(the superego), and identity (the “I”), which take their roles and masks from those that the “it” and the

“I” have loved and lost. Consciousness in this model has been moved very far from center stage; it is
now just a facet of the ego, which in itself incorporates nearly the entire psyche of the first topography.

The “I” has its own unconscious, preconscious, and conscious regions.

The implications for epistemology of this last model are profound, because there is no
grounding for positive knowledge in an ego consciousness free of the conflicts of life. The “I” has as its

heritage, and built into its structure, an individual and cultural history of defense, censorship, and

distortion. To undo this dream work of a lifetime, these oneiric deposits of civilization, becomes an
interminable task of interpretation guided by the principles of psychoanalytic exegesis. The analytic

setting still has its locus in the consulting room, but the analysand is now not only the individual
patient with his or her fantasies and symptoms, but civilization with its religious delusions, sexual
repression, and artifacts, which, like the ego, are the sediment of unfulfilled desire.

The Semantics of Desire. The French psychoanalytic project is to articulate a semantics of desire,

about which Ricoeur (1970) writes: “The semantics of desire…is bound up with [the] postponement of

www.freepsychotherapybooks.org 405
satisfaction, with the endless mediating of pleasure” (p. 322). In addition, it is tied to the never-ending
postponement of meaning and the mediation of being through language. Freud’s coupled discourses,

the energies of pleasure and the hermeneutics of meaning, are dialectically interwoven in his attempts

to represent the vicissitudes of longing. The energy metaphors—and they became metaphors when the
psychic apparatus replaced the neuronal machine—are used to give an accounting of the disjunction

between one meaning and another. The hermeneutics of desire involves the replacement of one

meaning (the manifest) with another, more fundamental and authentic articulation of the wish (the
latent). The energetics or economics of desire uses a system of interrelated concepts like cathexis,

displacement, and condensation to account for the movement of forces from one place to another,
movements that displace and disguise meaning. Force, place, and meaning, then, are the terms of
Freud’s thought, and every concept is determined by its coordinates in his topographical energic

system of reading signs.

Desire arises from a lack, a void. In its generation it is already a substitute, which covers over

with longing what cannot be said: the place of nothing, of mute death. Displacement and replacement
are the two terms that are joined in the semantics of desire and in the homeopathic treatment of

psychoanalysis. If neurosis arises from the displacement of psychic conflict into symptoms, and the
symptoms replace the conflict with a symbiotic representation in the speech of the body or behavior,

then analysis replaces the original conflict, restores the latent, by displacing the manifest symptoms via
interpretation. This restoration, however, does not occur in the original context, the traumatic scene,
but in its reproduction in the transference relationship. The analyst takes the place of significant

others, as scenes of frustration are restaged in a situation where insight—intellectual pleasure—takes

the place of desire. The desire to know, to have a life history without lacunae in its narration, is the
substitute satisfaction offered by psychoanalytic interpretation for those carnal pleasures that can

never be realized. The articulation of desire through interpretation is a sublimation of an unnameable

longing.

The Study of Symbolism. This longing is, perhaps, to become an “I,” to be an identity that is not

www.freepsychotherapybooks.org 406
haunted by imagoes from long ago and is not a fabric of fantasies that serve as a gloss for the past.
However, this very task remains the unnamed project in Freud’s work, and “the empty concept of
sublimation is the final symbol of this unspoken factor” (Ricoeur, 1970, p. 492). Freud could never give

a satisfactory economic or energic account of sublimation, for it arises not from defense but from
reflection. It is a transmutation of the natural into the cultural, of the carnal into the spiritual.

Sublimation is a hermeneutic term which stands for a transcendental movement of the spirit toward

the realization of consciousness. Freud, who did not speak of the self or things transcendental, could
not, of course, say this. Ricoeur in his writing on symbolism, tries to name what is missing in Freud, to

say what Freud cannot.

Symbols are products of desire; in fact, Ricoeur (1970) asserts, “If man could be satisfied…he

would be deprived of symbolization” (p. 322). The symbol stands for desire, but unlike the symptom,
which is but a disguised signifier for the repetitive insistence of desire to be signified, the symbol

captures, contains, and transforms desire into a living sign in which signifier and signified are held

together in a sublime icon.

Religion is a collective neurosis because its expression is a symptomatic repetition of the

longing for the father. Its iconography requires belief, thereby blocking the process of individual

participation and reflection essential to sublimation. Art does not repetitively recapitulate a man’s or
mens’ past, because the work of art is not simply a projection of the artist’s or the culture’s conflicts; it

is “the sketch of their solution.” Dreams, symptoms, and religion “look backward toward infancy, the

past; the work of art goes ahead of the artist; it is a prospective symbol of his personal synthesis and of
man’s future, rather than a regressive symbol of his unresolved conflicts” (Ricoeur, 1970, p. 175).

That “progression and regression are carried by the same symbols” is the Ricoeurian insight

which mediates his hermeneutic phenomenology with Freud’s psychoanalysis. The symbol arises from
unfilled desire, and, therefore, points to the past; but it also takes one forward into the future,

providing a guide for the movement of reflection. That Freud only reads symbols backward to their

ontogenetic and phylogentic origins was pointed out long ago by Jung (1913) and has been elaborately

www.freepsychotherapybooks.org 407
critiqued by him (1916), by Ricoeur (1970), and by Steele (1982). The past Freud thereby creates,
however, is itself symbolic, because Freud uses all his primal events and primary processes to delimit

the boundaries of the imaginary and to provide narrative guidelines by which to organize his analyses

of the present (Laplanche and Pontalis, 1968).

Ricoeur (1970) insists, as did Jung, that symbols must also be read progressively, because “the

emergence of the self is inseparable from its production through a progressive synthesis.” This is

because “the truth of a given moment lies in the subsequent moment” (p. 464) and the significance of a

symbol always lies in the future developments of its meanings, in the trajectory of interpretations and
in the realization of the spirit. However, reflection on all of this always proceeds retrogressively. The

past—the archaeology of the subject—and the future—the teleology of the spirit—meet in symbols

whose interpretation engenders the development of self-consciousness, which arises by making the
past present through retrospective analysis and the future imminent in the present through imagining

the meanings of the symbols.

It is in this temporal duality of the symbol that Ricoeur finds hope not only for the synthesis of
the self through the development of self-consciousness, but for a joining of the two styles of

hermeneutics that have been at odds for years: the hermeneutics of suspicion and of restoration. The
two come together in what serves as Ricoeur’s epigram for Freud and Philosophy (1970): “Thus the

idols must die—so that symbols may live” (p. 531).

Why psychoanalysis is necessary to phenomenology is that analytic suspicion is needed to

break the thrall that makes us the slave of the idols of the past, be they parental imagoes, castration
anxieties, or simple narcissistic egocentrism. Why psychoanalysis needs a hermeneutics of the spirit is

to free it from its bondage to the past and to aid it in reconstructing a past, which serves as a ground

for the present on which to build a future—a future that is not an illusion, because it comes from a less-
distorted past.

www.freepsychotherapybooks.org 408
THE READING

Consciousness grows by cannibalistic criticism of its grounds, its prejudices, and its embedded,
unseen ways of being. For anyone with a history who grows, that which was once liberating becomes

constricting, and that which was once a criticism of orthodoxies becomes an orthodoxy to be criticized.

Freud and Philosophy helped free me from the prohibitions against thought that are fostered by an
American scientific education and that make it hard to appreciate what is truly revolutionary in

Freud’s science: its pursuit of knowledge through dialogue, its concern with meanings over

observables, and its devotion to the development of critical self-consciousness. However, in returning
to Ricoeur after writing my own “conflicts of interpretation” (Steele, 1982), after teaching many

brilliant and radical students, and after becoming a feminist, I feel that Freud and Philosophy

promulgates many of the same oppressive values that are dear to both science and psychoanalysis. The
common perspective of these becomes visible when one steps outside of it and sees that

psychoanalysis, science, and hermeneutics share a masculinist world view, a Weltanschauung, in which

most of us were reared. One of the joys of feminist hermeneutics is standing apart from this very old
tradition and showing how this embedded way of doing things is restrictive, antilibertarian, and often

just plain wrong.

The masculine voice, which is shared by Ricoeur, Freud, and most scientists and academics, is

rigorous, objective, concerned with authority, determined to debate the issues, mute or opinionated on

the subject of women, and utterly positive about what is natural. In what follows I will show how
Ricoeur’s immersion in this way of speaking creates several interrelated problems in Freud and

Philosophy and is responsible for various errors in it. None of the four topics I will consider are

manifestly central to Ricoeur’s text. Their position is latent, so their importance will only be

established through an interpretive reading in which I shall show that Ricouer’s dismissal of Jung, his
overweening concern with authority, his disregard for women, and his reductive, demeaning

conceptualization of nature are all aspects of the androcentric bias that dominates his book.

THE REJECTION OF JUNG

www.freepsychotherapybooks.org 409
To anyone who has read Jung, Ricoeur’s modifications of the psychoanalytic theory of

symbolism will be repetitive, not innovative. Jung’s Symbols of Transformation (1911-12), which

helped speed the break with Freud, was about the regressive and progressive function of symbols and
about their function of transforming carnal into spiritual fantasies. In his “On the Psychology of

Unconscious” (1943), Jung does both analytic-reductive (Freudian) and synthetic-prospective

(Jungian) interpretations of the same case; and his studies of the interrelations of transference,

sublimation and symbolism (Jung, 1946) would have aided Ricoeur in his discussion of these.

There are no citations to Jung’s works in Freud and Philosophy. Ricoeur dismisses Jung because
he is confusing and not a rigorous thinker like Freud. Ricoeur (1970) says: “With Freud I know where I

am going; with Jung everything risks being confused: the psychism, the soul, the arthetypes, the

sacred” (p. 176). For Ricoeur, Freud is a strong, sure leader, whereas with Jung he fears being lost. This

craving to always know where one is is typical of thinkers who are concerned with mastery and with
following a master. It doesn’t hurt to be lost or confused; one might just find something new. Jung

explored the female symbolism of the unconscious and the importance of goddesses. This is something

neither Ricoeur nor Freud, both of whom were unwilling to risk being lost in fantasy, in the “realm of
the mothers,” could find.

Although Jung’s writing is associative, symbolic, suggestive, and often mythopoeic, it is usually
only confusing to those who expect causal, argumentative prose and are uncomfortable with loose

thinking. For the confused, however, Jung even provides a dictionary (1921).

I doubt if Ricouer ever really tried to read Jung. If he had, he would have soon discovered that

his ideas on symbols were thoroughly Jungian. There has been, ever since Freud exiled Jung, a

compulsive quality to psychoanalysis’ dismissal of Jung’s work. Ricoeur merely repeats Freud’s actions,
but he seems to have read even less Jung than did Freud.

This dividing into camps, schools, teams, and disciplines who worship a totemic founder is so
obviously a primitive male bonding ritual that one would think that men of reason, like Ricoeur and

www.freepsychotherapybooks.org 410
Freud, would have renounced its practice. In modern times, however, the rite has merely been

transformed through the cunning of reason into a proper and reasonable respect for authority.

ISSUES OF AUTHORITY

Freud makes rules and leads. This makes him an authority to Ricoeur. Citing another great

authority in order to justify his own feelings on what legitimates psychoanalysis, Ricoeur adopts Kant’s

view that a system is limited by what justifies it. Freud’s determination to explain the most complex
phenomena from the topographic-economic point of view is, according to Ricoeur (1970), a restriction

“which gives psychoanalysis its rights” (p. 153). Such limits serve to facilitate Ricoeur’s project, which

is to conduct “a rigorous debate with the true founder of psychoanalysis” (p. xi).

If the boundary lines of knowledge claims are not clearly drawn, then debate about ideas, which

are an intellectual’s property, cannot be judiciously conducted, and the lineage of a thought cannot be
unambiguously traced back to the father. In a short space it is difficult to critique the notion that ideas

are discovered, owned, claimed, and adjudicated, except to suggest that if the outlines of such a critique

are not obvious, then the reader is not aware of how much his or her thought is dominated by the
tropes and practices of capitalism. That ideas come from a founder and are passed on to his followers
is so obviously totemic, and so germane to both Freud’s and Ricoeur’s work, that I will take time to

develop its connections with the biases of the masculinist perspective.

Much of Freud and Philosophy is about the significance of the Oedipus complex and the

symbolism of the father. The father, for Ricoeur (1970), is “the name-giver and the lawgiver,” with the

institution of the father—patriarchy—serving the son by directing his education in the culture. Freud
serves these purposes for Ricoeur: He named psychoanalysis and established the rules of its practice.
More than this, however, psychoanalysis is an education in culture. Ricoeur (1970) says Freud’s work

is “a monument of our culture” and a place “in which our culture is expressed and understood” (p. xi).

To acquire culture then, one must be educated by Freud, and Ricoeur (1970) opens his work

www.freepsychotherapybooks.org 411
with the simple declaration that, “This work is a discussion or debate with Freud” (p. 3). This single

line is a clear expression of the symbolism of the father in modern academic totemism. For “Freud,”

here, is a trope, the name “Freud” being a metonymy for the master’s work. Debates with dead men are

only possible in societies that revere their elders and have a set of cultural practices that preserve their
deeds after their deaths. Our reverence for the immortal works of genius is just such a mechanism, and

the idea that we can debate with these men is an obvious illusion. While Sigmund Freud was alive, few
people held successful debates with him. Jung, with ideas very similar to Ricoeur’s, tried and failed. But

Ricoeur’s metonymic Freud is much more the ideal or totemic father. He is not the primal tyrant that

Wittels (1924) describes, but the embodiment of rigor, suspicion, and closely reasoned debate.

It is not some anomaly in Ricoeur’s character that makes Freud and Philosophy an extended
intellectual oedipal drama. It is that authority in our culture is paternal, and a man, to be a scholar,

must make his place among the fathers. Ricoeur’s scholarship is impeccable in this regard; he cites,

critiques, modifies, and expands on the ideas of one master after another (Hegel, Kant, Marx, Nietzsche,
Heidegger, etc.) in exemplary academic fashion, thereby displaying his expertise and ensuring that his

work will be commented on by future generations. Indeed, Freud and Philosophy has become a master

text upon which an ambitious son, displaying all his scholarly expertise, is commenting. This must stop,
because identification with the masters means an acceptance of their discourse and their rules of
debate.

THE OMISSION OF WOMEN

There are almost no references to women in Ricoeur’s work; in fact, the few places they are

present in the text they are identified with the absent or the lost. Whereas the father is a strong
presence throughout, and Ricoeur has much to say about him, he accepts Freud’s portrait of the mother

as an “archaic object…who bore us, nursed us and cared for us” (Ricoeur, 1970, p. 445). That is all he

says, and he is speaking Freud’s words.

www.freepsychotherapybooks.org 412
The only other significant reference to women is in a discussion of Freud’s (1910) analysis of

Leonardo da Vinci’s relationship with his mother. In this instance, Ricoeur’s contact with females is

mediated not by one man, but by two, and Ricoeur’s theme is that symbols signify absence born of
desire. The “unreal smile” of Mona Lisa is a symbol for “the smile of the lost mother” of Leonardo

(Ricoeur, 1970, p. 177). If women are archaic, lost, and symbols of absence, this is not so much a
description of them as it is a comment on their place in Ricoeur’s discourse. They are simply absent

from his text, and he seems to know about them only through their representation in the works of

other men.

This omission of women, however, is no simple oversight; it is a near blindness born of

masculine myopia. Ricoeur returns to Freud’s analysis of Leonardo when he takes up the topics of

religion and the Oedipus complex. Following Freud, Ricoeur (1970) writes,

If religious illusion stems from the father complex, the “dissolution” of the
Oedipus complex is attained only with the notion of an order stripped of
any paternal coefficient, an order that is anonymous and impersonal.
Ananke is therefore the symbol of disillusion…Ananke is the name of a
nameless reality, for those who have “renounced their father.” It is chance,
the absence of relationship between the laws of nature and our desires or
illusions.…Ananke, it seems to me, is a symbol of a world view...in it is
summed up a wisdom that dares to face the harshness of life [pp. 327-
328].

In this passage the omission of women is oppressive. First, “an order of things stripped of any

paternal coefficient” is not anonymous and impersonal. Women exist. As mothers they provide one of
the most complex and personal relationships we will ever have. They lay down the law, and they are

usually the person who gives us our first name and who we name first. As lesbians, feminist

separatists, and as people whose identities come neither from their fathers nor husbands, women also

exist.

Women are made invisible, anonymous, and impersonal by denying their existence or

www.freepsychotherapybooks.org 413
transforming them into things, into “its.” Ananke is a female goddess, a she, not an it. But Ricoeur, twice

calls her “it,” thus ensuring her anonymity.

DENATURED NATURE

Ricoeur, like most other hermeneuts, has accepted science’s grossly unnatural representation of

nature. He speaks of “the laws of nature” and refers to “the conditions of objectivity of nature” (1970, p.

48). He characterizes Freud’s energic tropes as nonhermeneutic, because such language describes the
transformations of the natural order. Ricoeur nowhere sees that science interprets nature in some odd

ways. Science has turned nature into a thing in order to investigate and exploit her (Griffin, 1978).

There are no laws of nature, only laws of men, which are used to tame nature’s unruly ways. Science

uses nature as a stage on which to strut its prowess, but feminist critics have shown what is being
done. A subset of existence has been set apart by us as an other, a mother, a “she” and an “it” and

designated as natural (Dinnerstein, 1977). The “objectivity of nature” is pure projection onto this

other, and science sees reflected back from this mirror its own projections onto her. Finally, nature is
no more a system of energic transformations than it is God’s creation or a giant turtle. The

representation of the natural as energic is of recent origin and is the animism of a materialist,
mechanistic culture (Merchant, 1980).

I think Ricoeur accepts the natural scientists’ representation of nature because of the general

acquiescence of authorities in one field to experts in another. The deal struck by Dilthey with science,
“You take nature, we’ll take culture,” is still honored by his descendant, Ricoeur. Ricoeur follows other

twentieth century covenants of rationality: He is respectful of genius and wary of mystics, and he

believes what other men say about women.  

REFERENCES

Derrida, J. (1976). Freud and the scene of writing. In J. Mehlman (Ed.), French Freud: Structural studies
in psychoanalyists (pp. 73-117). Millwood, NY: Kraus Reprint Co.

www.freepsychotherapybooks.org 414
Dinnerstein, D. (1977). The mermaid and the minotaur: Sexual arrangements and human malaise. New
York: Harper.

Freud, S. (1895). Project for a scientific psychology. Standard Edition, 1, 295-397.

Freud, S. (1900). The interpretation of dreams. Standard Edition, 4.

Freud, S. (1910). Leonardo da Vinci and a memory of his childhood. Standard Edition, 11, 63-137.

Freud, S. (1912). The dynamics of transference. Standard Edition, 12, 99-108.

Freud, S. (1914). On narcissism: An introduction. Standard Edition, 14, 73-102.

Freud, S. (1915). The unconscious. Standard Edition, 14, 166-215.

Freud, S. (1917). Mourning and melancholia. Standard Edition, 14, 243-258.

Freud, S. (1923). The ego and the id. Standard Edition, 19, 12-66.

Griffin, S. (1978). Woman and nature: The roaring inside her. New York: Harper.

Irigaray, L. (1980). When our lips speak together (Carolyn Burke, Trans.). Signs, 6,69-79.

Jung, C. G. (1911-12). Symbols of transformation. Collected Works, 5, 1956.

Jung, C. G. (1913). General aspects of psychoanalysis. Collected Works, 4, 229-242.

Jung, C. G. (1916). The psychology of the unconscious (B. Hinkle, Trans.). New York: Moffat, Yard & Co.,
1916.

Jung, C. G. (1921). Psychological types. Collected Works, 6, 3-495.

Jung, C. G. (1943). On the psychology of the unconscious. Collected Works, 7.

Jung, C. G. (1946). The psychology of the transference. Collected Works, 16, 163-323.

www.freepsychotherapybooks.org 415
Lacan, J. (1976). Seminar on “The Purloined Letter.” In J. Mehlman (Ed.), French Freud: Structural
studies in psychoanalysis (pp. 38-72). Millwood, N.Y.: Kraus Reprint Co.

Laplanche, J., & Pontalis, J.-B. (1968). Fantasy and the origins of sexuality. International Journal of
Psycho-Analysis, 49, 1-18.

Laplanche, J., & Pontalis, J.-B. (1973). The language of psychoanalysis (D. Nicholson-Smith, Trans.). New
York: Norton.

Mehlman, J. (Ed.). (1976). French Freud: Structural studies in psychoanalysis. Millwood, N.Y.: Kraus
Reprint Co.

Merchant, C. (1980). The death of nature: Women, ecology, and the scientific revolution. New York:
Harper.

Radnitzky, G. (1973). Contemporary schools of metascience. Chicago: Henry Regnery.

Ricoeur, P. (1950). Freedom and nature: The voluntary and the involuntary (E. Kohak, Trans.). Evanston,
IL: Northwestern University Press, 1966.

Ricoeur, P. (1967). The symbolism of evil (E. Buchanan, Trans.). New York: Harper.

Ricoeur, P. (1970). Freud and philosophy: An essay on interpretation (D. Savage, Trans.). New Haven:
Yale University Press.

Ricoeur, P. (1974). The conflict of interpretations: Essays in hermeneutics (D. Ihde, Ed.). Evanston, IL:
Northwestern University Press.

Steele, P. (1979). Psychoanalysis and hermeneutics. International Review of Psycho-Analysis, 6, 389-

411.

Steele, R. with S. Swinney (Consult. Ed.). (1982). Freud and Jung: Conflicts of interpretation. London:
Routledge & Kegan Paul.

Turkle, S. (1978). Psychoanalytic politics: Freud's French Revolution. Cambridge, MA: MIT Press.

www.freepsychotherapybooks.org 416
Wittels, F. (1924). Sigmund Freud: His personality, his teaching, and his school. New York: Dodd, Mead.

www.freepsychotherapybooks.org 417
13 JACQUES LACAN: PSYCHOANALYST,
SURREALIST, AND MYSTIC

JEANINE PARISIER PLOTTEL, PH.D.

Jacques Lacan’s contribution to psychoanalytic theory and practice is and has been the subject

of intense controversy. The quarrels between various factions of both enemies and disciples, the

counterculture quality of his teaching, and the political implications of some of his positions have cast
shadows on a correct appraisal of his work. The notoriety that came to Lacan in old age, his links with

linguistics and structuralism, and his role as trend setter of the Paris intelligentsia have obscured his
significant legacy to French psychoanalysis, psychiatry, and neurology. Although many facets of Lacan’s
approach to psychoanalysis may seem heretical, in fact, its archaeology, in the sense of Michel Foucault,

leads to the nineteenth century French tradition of psychiatry and neurology—to Jean-Martin Charcot
and other French masters of Freud. Indeed, when considering Lacan’s evolution, it is important to
remember that this very same tradition was one of the catalysts in Freud’s development that led to the

creation of psychoanalysis. A brief review of how the Viennese disciple viewed Charcot, his French
teacher, will provide the first key to Lacan’s texts.

It is common knowledge that Freud’s studies with Charcot at the Salpétrière in Paris from
October 1885 to the end of February 1886 marked a turning point in the direction of his interests.

What may not be so well remembered is how much Freud admired Charcot’s clinical presentations of

www.freepsychotherapybooks.org 418
patients. We cannot assert that Freud went so far as to give up the traditional German way in favor of
French clinical technique, but this technique was surely integrated in his method. His description of

how Charcot presented his patients (Freud, 1887-88) emphasizes the “concepts of the ‘entité morbide’,

of the series, of the ‘type’ and of the ‘formes frustes’ ” (p. 135). Such concepts are important in French
clinical method and were quite foreign to the German perspective.

What especially struck Freud, however—and I am certain that the psychoanalytic

infrastructures bear traces of this to this day—was Charcot’s friendliness and openness, his

responsiveness to students, whom he considered his peers. Freud (1893) ascribes “the intellectual
significance” of this man

to the magic that emanated from his looks and from his voice, to the kindly
openness which characterized his manner as soon as his relations with
someone had overcome the stage of initial strangeness, to the willingness
with which he put everything at the disposal of his pupils, and to his
lifelong loyalty to them. The hours he spent in his wards were hours of
companionship and of an exchange of ideas with the whole of his medical
staff [p. 16]...

Freud went on to elaborate:

As a teacher, Charcot was positively fascinating. Each of his lectures was a

little work of art in construction and composition; it was perfect in form
and made such an impression that for the rest of the day one could not get
the sound of what he had said out of one’s ears or the thought of what he
had demonstrated out of one’s mind [p. 17].

I am not going to delve into the substance of Charcot’s science and art—a recent history of

psychoanalysis in France, La bataille de cent ans, by Elisabeth Roudinesco (1982) has already done this

—but I want to stress the oral aspect of his legacy. There is an analogy between knowledge transmitted
in such a way and the transference that takes place in the course of an analysis. Spectacle and

encounter captivated Freud, just as they had many other scientists and laymen. It can also be argued

www.freepsychotherapybooks.org 419
that the significance of Lacan’s manner should be sought in the traditional mediums of Charcot and
other French alienists that had struck Freud: oral presentation of clinical cases, lectures, and teaching
in an asylum setting.

That Lacan wrote very little and published even less—in fact, only his thesis and a few articles

—has been pointed out by several critics. In her recent book, Vies et legendes de Jacques Lacan,

Catherine Clément (1981), a philosopher turned journalist, observes that most of the essays included
in Lacan’s Ecrits (1966) are papers and communications that were first read at meetings and

congresses. The six volumes published to date in the Séminaire series (Lacan, 1953-54, 1954-55, 1955-

56, 1964, 1972-73), transcripts of Lacan’s so-called seminar (“lecture” is the American term), were
edited not by Lacan himself but by Jacques-Alain Miller, his son-in-law. This Séminaire that is Lacan’s

major achievement, and we must always bear in mind that its essence is essentially oral. Although

these lectures were very carefully prepared, ideas came to Lacan as he spoke before an audience, and
some of the best parts were improvised. These improvisations were charismatic, even inspired, in the

literal sense of the word. Their effect on the audience was comparable to the frenzy of an extraordinary

bullfight, to the ecstasy of the mystics, and to the passion of absolute love. Then, little by little, as the

year went by, the language miracle failed and the spell loosened. Inspiration ceased; the magician on
the podium lost his power and turned into an old, hollow man.

In old age, Lacan became a Parisian celebrity, a household word in households where nobody
had read a single one of his paragraphs. With his friend Claude Lévi-Strauss, he was the representative

of the new structuralism, the “ism” that had followed Jean-Paul Sartre’s existentialism. For more than
20 years, attendance at Lacan’s Séminaire was de rigueur for anyone who wanted to be in the
mainstream of French thought—Barthes, Derrida, Leiris, Jakobson, Kristeva, and Sollers

(Schneiderman, 1983), for example, and not merely out-of-town intellectuals. If Lacan happened to

dine at Maxim’s or some such place, his presence was noted. For instance, Stuart Schneiderman (1983)
tells a story in which Lacan managed to upstage Roman Polanski, who was sharing his table. But I

believe that Lacan’s serious achievements belonged to the fortieth, fiftieth, and sixtieth decades of his

www.freepsychotherapybooks.org 420
life, before he actually attained notoriety and an international reputation.

It is obvious that the texts of Lacan’s old age are as elusive as those of many certified psychotics.

Are they poetry? Creations of a psychoanalytic Zen master? Do they signal a revolution in
psychoanalytic form? Or have these texts been edited in such a way that they take on the stamp of the
meanderings of the unconscious? The cliché, “Only time will tell,” is in order here. However, although

Lacan is indeed a difficult and precious writer, most of us find that, read in chronological order, he is

quite accessible. Most of his writings are no more arcane than those of Melanie Klein or Heinz
Hartmann. And most of Lacan’s significant ideas were present at a time when he still wrote in an easily

intelligible way. To my mind, the complicated mathematical knots, the abstruse formulas, the complex

formal symbolism added little if anything to the substance of the most important psychoanalytical

theorist since Freud.

What explanation can be offered? Clèment (1981) puts it well when she states that for a long
time, the author was Jacques-Marie Lacan, and when he was Jacques-Marie Lacan, he was

comprehensible. We can apply to him his offhand remark about Napoleon (Lacan, 1950, p. 39; 1966, p.

171). “What is the difference between a madman who takes himself for Napoleon and Napoleon
himself?” he asked. The obvious answer is that unlike the madman, Napoleon never believed he was

Napoleon, but knew he was Bonaparte, and remembered very well what he had done in order to turn
Bonaparte into Napoleon. So perhaps Jacques-Marie Lacan knew how he had become Lacan, the guru

of French psychoanalysis. Perhaps only his disciples, those who call themselves Lacanians, take the

legend seriously. It is likely that had the International Psycho-Analytical Association not cast him out,
he would have remained an orthodox professional, but that is another story. I suspect that his

exclusion from traditional psychoanalytic societies caused him enormous pain and anguish. His
attempts to be reinstated by the IPA, his pleas with his former friends and colleagues—for example his
letters to Loewenstein and Hartmann1—make this abundantly clear.

Jacques-Marie Lacan was born in Paris on April 13, 1901, and his career ran the usual obstacle
course of a French doctor of medicine, psychiatrist, and psychoanalyst. His psychiatric curriculum

www.freepsychotherapybooks.org 421
vitae, printed in his thesis (1932), indicates that he had impeccable clinical credentials and the highest
possible pedigree in the field. He worked with Henri Claude, an expert on schizophrenia and one of the

foremost French psychiatrists of the early century, at the Clinique des Maladies Mentales et de

l’Encéphale (Clinic for Mental Illnesses and Illnesses of the Encephalus) in 1927-28. In 1928-29, he was
attached to the Infirmerie Spéciale Près de la Prefecture de Police (Special Police Headquarters

Infirmary) and trained with Georges de Clérambault, whose theory of mental automatism was a

decisive influence. “Our only master in psychiatry” is Lacan’s appraisal of his role. From 1929 to 1931
he continued his training at the Henri Rousselle Hospital and spent the summers in Zurich at the

Burgholzi, Eugen Bleuler’s and Carl Jung’s clinic. He obtained a diploma in forensic medicine, and in
1931-32, he returned to the Clinique des Maladies Mentales et de l’Encephale.

Lacan co-authored his first articles with leading senior psychiatrists and neurologists, and he
published in psychiatric journals, for example, L’evolution psychiatrique, whose contributors became

early recruits of psychoanalysis. His doctoral thesis (Lacan, 1932), which we shall examine in more

detail presently, was a traditional work, with meticulous references, careful research, and detailed
clinical observations, written in a clear and straightforward style. The young doctor was well on his

way to a successful psychiatric career. At this time there appear to be at least two developments in
Lacan’s professional vitae that must be taken into account to explain his deviations from the

psychiatric and medical mainstream.2 I am referring to his connections with surrealism and his
contacts with psychoanalysis.

Further research is needed about actual relations between Lacan and surrealism. We do know

that he published several fascinating articles (Lacan, 1933a,b) in Le minotaure, a surrealist journal, and
that it was Lacan’s ideas that prompted Salvador Dali’s famous critical paranoia theory. He had
contacts with René Crevel, the poet who shot himself playing Russian roulette with a loaded pistol

(Lacan, 1966, p. 65) and he was a good friend of André Breton. His second wife, Sylvia Maklès, the star

of Jean Renoir’s film Une partie de campagne, attended the same school as the sisters Simone and
Jeanine Kahn, who respectively married André Breton and Raymond Queneau. Sylvia’s own first

www.freepsychotherapybooks.org 422
husband was Georges Bataille, a writer whose style Lacan imitated (Roudinesco, 1982).

The stamp of this movement is discernible in Lacan’s own texts in several ways. First, many

characteristics of automatic writing—for example, the use of puns, and arbitrary and striking
comparisons and making verbal associations the organizing structure of an expository piece—are also
characteristics of Lacan’s own manner. A sentence such as “A casser I’oeuf sefait l'Homme, mais aussi

I’Hommelette”(roughly translated, “In breaking an egg homme (man) is made, but also an [h]omelet”)

and the allusion to “a large crepe moving about like an amoeba” in the sentence that follows (Lacan,
1966, p. 845) are pure surrealism.

Second, Lacan’s contacts with poets led him to interpret the utterances of his psychotic patients
just as he might interpret a surrealist poem, or for that matter any poem at all. For example, he
analyzed (Lèvy-Valensi, Migault, & Lacan, 1931, p. 376) the following apparently senseless sentence

from the writing of Marcelle C., a paranoiac patient: “A londoyer sans meurs on fait de la becasse”
(“Londoning without morals one makes woodcocks”). Meurs is a kind of portemanteau word composed

of moeurs (customs, morals) and meure (from the verb mourir, to die). Lacan showed that underlying

this ponderous formula is the rhythm of a famous line of poetry by the seventeenth century dramatist
Pierre Corneille that is known by every French schoolchild: “A vaincre sans pèril on triomphe sans

gloire” (“In conquering without peril one triumphs without glory”). What appears to be an original
verse is in fact generated by a stereotypical automatic auditory mechanism. Familiarity with poets such

as Robert Desnos, Philippe Soupault, and André Breton led Lacan to notice that patients gave different

graphic renditions of the same phonic material in different places and poems: “la mais l’as, Vame est
lasse, et la mélasse”(“the but, the ace, the soul is tired, and molasses”). Or, “le merle à fouine, la mère, la

fouine”(“The weaseled blackbird, the pitchforked mother”). We can give an English approximation of
this mechanism by playing on the word molasses to produce “Moe’s lassies, more losses, my asses.”
The result of Lacan’s juxtapositions of such phrases is an awareness that psychotic productions may or
may not have poetic value and that the substratum of a poem is often material that may be given the

label “psychotic” in a clinical context and perspective. (For a fuller discussion, see Lacan, 1933a.)

www.freepsychotherapybooks.org 423
Likewise, Lacan might have learned from surrealism and not necessarily from Freud how to

interpret a literary work as though it were a living being. The seminar comes to mind that deals with

Edgar Allen Poe’s “The Purloined Letter” (Lacan, 1966), in which the letter stolen from the Queen by
the minister is restored to her by Dupin, but many other instances can be given. For example, in his

lecture of March 2, 1960, Lacan (1959-60) quoted a stanza by Arnaut Daniel, a great troubadour that
Dante ranked with Virgil. His point was that this poem about courtly love embodied “the central void

around which is organized and articulated whatever it is that sublimates desire” (p. 29). The same void

and sense of nothingness is revealed in his appraisal of André Gide. When Gide’s wife Madeleine took
revenge on her husband by burning all the letters he had ever written to her, she knew what she was

doing. The letters had been Gide’s way of filling up his own sense of emptiness, the literal hole that he
stuffed with all kinds of games, which allowed him to watch himself pretending to be himself. In Et
NUNC Manet in Te (Lacan, 1966), written after the death of Madeleine, his wife, Gide confessed that

after the letters’ destruction, his relationship with her, “n’offre plus, à la place ardente du coeur, qu’un
trou”(“left but a hole in the ardent part of his heart”) (p. 762). The loss of this correspondence, of which
Gide had no copy meant that whereas previously his mirror had been the substance of words, phrases,

sentences, and paragraphs, it had turned into the vertigo of a ditch, a gap, nothing, and nothingness.

Lacan’s sense of play and games would of course have delighted the surrealists. He liked using

everyday imagery, slang, and ordinary words of our childhood and adolescence, anything from
mustard pots to Picasso’s ostrich cabbages, to illustrate philosophical and psychoanalytic concepts. He

himself referred to “this seriousness that I always develop further and further to its punchline,” (“ce

sérieux queje développe toujours plus en pointe”). Elsewhere he says that he is the Gòngora of

psychoanalysis. When he spells the French word raison, (reason) r-e-s-o-n, following the example of
Francis Ponge, to show how the sound itself suggests something that resonates; when he puns on the

French word poubelle (garbage can), referring to psychoanalytic publications—his own included—as

Poubellications; when he dismisses the “Lacanians” by reminding them that he himself is a Freudian,

Lacan is playing. But he is also playing when he ridicules his opponents and his disciples, when he

applies linguistic and mathematical concepts to psychoanalysis. A surrealist is never more serious than

www.freepsychotherapybooks.org 424
when he is playing, of course, so in that sense Lacan remained a surrealist to the end.

Finally, Lacan is a surrealist because his own formulas are themselves short poems, or so they
would have been defined by his friends Paul Eluard and André Breton. I am thinking of aphorisms such

as “Ton désir c'est le désir de l’Autre” (“Your desire is the desire of the Other”); “L’Inconscient est

structuré comme un langage”(‘‘The Unconscious is structured like a language”); and “Moi, la vérité je

parle”(“Me, I speak the truth”).

To stress Lacan’s surrealism is to remain true to French intellectual history. The so-called

surrealist revolution coincided with the introduction of psychoanalysis. André Breton was one of the
first French writers to read and write about The Interpretation of Dreams. Public opinion often attacked

both surrealism and psychoanalysis for being foreign and hostile to “la clarté française,”—French

clarity. Indeed, just as surrealists were drawn to the study of dreams and the exploration of the
unconscious, so psychoanalysts were drawn to the surrealists. Lacan was not alone in being close to

them. For example, Adrien Borel, one of the founders of the Société Psychanalytique de Paris (SPP) in

1926, analyzed Georges Bataille and Michel Leiris (Roudinesco, 1982, pp. 358-360). René Allendy,
author of 200 articles on various occult subjects was one of Antonin Artaud’s psychiatrists and was

also Anaîs Nin’s analyst. In a general way, many of the first- and second-generation French analysts
were writers and had contacts with the world of arts and letters. Marie Bonaparte was a prolific
author, and her book on Edgar Allen Poe was widely read. Eugénie Sokolnicka was André Gide’s model

for the character of Madame Sophroniska, the analyst who unsuccessfully treated Boris in Les faux-

monnayeurs (The Counterfeiters). Edouard Pichon, the president of SPP, was co-author with his uncle,

Jacques Damourette, of a monumental seven-volume study of French grammar, De la langue à la

pensée, a book that Lacan often cites.

During Lacan’s formative years, in the Paris of the 1920s and early 1930s, many young

psychiatrists were drawn to the study of Freud and became psychoanalysts. These same

psychoanalysts were interested in language, literature, and the arts; and artists and writers, in turn,
took up psychoanalysis. The fact that Lacan had contacts with Breton, Crevel, Eluard, and Dali did not

www.freepsychotherapybooks.org 425
make him an isolated figure, but rather one who was very much in the mainstream of his avant-garde
milieu. Psychoanalysis was itself a marginal discipline, but within it, Lacan was a member of the

reigning establishment and a very classical, orthodox Freudian analyst. He was analyzed by Rudolph

Loewenstein, and the analysis seems to have lasted a long time, from about 1932 to 1939. The two men
remained on very cordial terms. As noted earlier, when Lacan left the Société Psychanalytique de Paris

and began to have difficulties with the IPA, he wrote “Loew” a long letter justifying his position and

asking him to intervene on his behalf with Hartmann, who was then president of the IPA.

An examination of Lacan’s first book, De la psychose paranoîaque dans ses rapports avec la
personnalité (1932), his doctoral thesis, completed before his own analysis, will show the synthesis of

these various themes in a clinical case history, the case of Aimée.

AIMÉE

At eight o’clock one evening, a well-known Parisian actress arrived at the theater where she was
scheduled to perform and was greeted by a nicely dressed woman whom she mistook for one of her

many fans. This woman asked the actress whether she was Madame Z., and when the answer was yes,
the woman pulled out a knife out of her handbag and turned the blade toward the star. Madame Z.

managed to grab it, cutting two tendons in her fingers in the course of the scuffle. The woman,
henceforth called Aimée A., was duly restrained and carted off to jail. Madame Z. did not press charges,
and her assailant was moved to Ste. Anne Asylum, where Lacan observed her for a year and a half. At

first, Aimée continued to have hallucinations, obsessions, and to heap abuse on her intended victim.
But suddenly, 20 days after the incident, at seven o’clock in the evening, she began to weep as she
realized that the actress was totally innocent of any wrongdoing. Her delirium dissipated completely

and the vanity of her megalomanic intentions and the inaness of her fears struck her all at once. She
had recovered.

This 38-year-old woman was originally from Dordogne, born into a large peasant family, with

www.freepsychotherapybooks.org 426
three brothers and two sisters. She had a tenured job with a rail transport company; her record was
outstanding, and her superiors were pleased with her performance and tolerated some of her

idyosincrasies. She was married to another employee of the same company, but the couple lived in

different towns. Her husband took care of their 8-year-old son, and she visited them more or less
regularly. The patient herself had organized this life-style at the end of a previous voluntary

commitment to a mental institution a year and a half earlier. At that time she had believed that a

number of highly placed celebrities, including several writers, were going to have her son killed, and
she had written a letter of resignation on behalf of her husband to their mutual employer. Then, forging

his signature, she had applied for a passport to the United States.

The fixation on Madame Z. was not an isolated episode. Aimée had set her sights on celebrities

before. For example, she had tried to establish contact with a well-known novelist, Mr. P. B., the initials
of Paul Bourget, and with the Prince of Wales. She sent them letters and miscellaneous writings,

including a weekly sonnet and a novel called Le Dètracteur; in turn, she collected newspaper and

magazine clippings reporting their activities. Her initial infatuation for P. B. had turned to hatred, and
she was now convinced that he was plotting to kill her son.

The changing of love into hate was another pattern of her relationships. Her first love, for
example, was characteristic in this respect. She had become infatuated with the local Don Juan a month

before she was transferred to another town. For three years she wrote him regularly and spent most of

her leisure daydreaming about him, hiding her passion from everyone. She never saw him again, and
one day her love changed to hatred and scorn: “I went from love to hate abruptly,” she admitted
spontaneously to Lacan (1932, p. 225). The same mechanism played in her friendship for

Mademoiselle C. de la N., a fellow worker from an impoverished aristocratic family who influenced her
deeply. It was this woman, in fact, who introduced Madame Z., a neighbor of one of her relatives, into
Aimée’s life. “You are not like the other girls,” Mlle. C. de la N. is reported to have said. “I feel that I am

masculine,” was Aimée’s response. “You are masculine,” agreed her friend. Lacan characterized the
manner in this book as midire (literally, to “midsay”—to speak in half tones). The suggestion that

www.freepsychotherapybooks.org 427
Aimée’s attraction for her own sex may be a factor here would be readily accepted today, but in the
early 1930s an observer might have neglected to note that at the time of her attempted crime Aimée
had broken all contacts with her old friend. The circumstances of her change of heart went back ten

years, when Aimée had given birth to a stillborn baby girl, strangled by the umbilical cord. Her friend
had telephoned for news. The patient immediately felt that Mlle. C. de la N. was responsible for this

calamity and that she had conspired to kill the little girl.

Throughout his account, Lacan took care to include long excerpts from Aimée’s writing and to
present her aspirations for the improvement of the social and human condition in such a way that his

readers come to esteem rather than belittle this patient. He avoided the patronizing tone of the

superior judge, the medical boss, or even the average Frenchman or Frenchwoman. The diagnosis was

that she suffered from self-punitive paranoia (paranoia auto-punitive). Madame Z. represented an
idealized version of herself, a mirror of her ideal ego. Like Aimée, Madame Z. had a career, and being a

wife, mother, and homemaker was not the focus of her daily life. Her activites were covered by

reporters, so that there was a connection between her life and print. Aimée herself aspired to literary
renown, to a place in the newspaper. In many circles, the morality of actresses is questionable; and it

seems likely that Aimée’s own code of ethics would classify her in the category of fallen woman and

sinner. That is just what Aimée felt herself to be; in her family’s mythology she was the brightest child,
the intellectual star, but also the one who was always late and kept everyone waiting, the one who

could not pull her act together, the one who was disorganized and undependable. The feeling was that
she should never have gotten married at all. Aimée incorporated Madame Z. into this image, and the
stab wound that punished her was but a punishment inflicted upon herself. When Aimée came to

realize the senselessness of her attempted aggression, she was in a sense cured. She had been
punished, and now she had no more use for her delusions.

The root of this illness was found in her relationship with her older sister. Aimée recognized the
virtues of this sister but nevertheless hated her and felt herself the victim of this woman, who had

achieved her equilibrium at Aimée’s expense. A childless widow, this sister now had an ersatz husband

www.freepsychotherapybooks.org 428
and child, that is to say, she lived with Aimée’s husband and child. When Lacan interviewed the sister,
she made it clear that her younger sibling’s illness and incarceration suited her well, and she feared
that a pardon would jeopardize her life. Aimée understood this, yet although her feelings could hardly

have been more ambivalent, she rejected all criticism leveled against her rival. Lacan was especially
struck by the sharp contrast between her words expressing hyperbolic praise and the icy tone in which

she uttered them. Lacan (1932, pp. 232-233) characterized her attitude as a Verneinung (denial)

reaction of the purest kind.

The interpretation here follows Freud’s in The case of Schreber, quoted by Lacan. We can shape

the famous paradigm of denial in paranoia so that it applies to females rather than males, and we can

see how apt it is for Aimée: “I love her” may be denied to produce “I do not love her.” This is equivalent

to “I hate her” and leads to the projection, “She hates me,” which is a leitmotif of the persecution theme
here. A second type of denial, “I do not love her, but I love him,” can be turned into “He loves me.” We

can thus interpret Aimée’s infatuation with the male figments of her imagination—the Prince of Wales,

the writer P. B., and her first love. In other words, she was able to mask her attachment to her own sex
by denying it and substituting a “him” for a “her.” The third denial structure, “It is not I who love the

women—he loves them” (Freud (1911), p. 64 leads to the theme of jealousy, whether there is

projection or not. “Delusions of jealousy, added Freud, contradict the subject, delusions of persecution,
contradict the verb, and erotomania contradicts the object” (Freud (1911), p. 64-5. Recall that Aimée

believed that the objects of her attention want to kill her son. Her unfounded fears were meant to hide
the fact that it is not her child she loved, but the woman she connected with him. Finally, the fourth
type of denial is an absolute denial: “I do not love her. I do not love anyone at all. I love only myself.”

This leads to megalomania and to a regressed narcissistic stage (Lacan, 1932, pp. 261-262).

The symptoms of Aimée’s illness were but denials, displacements, and substitutes of a
prototype, the sister persona. However, her actual choices of love-hate objects were determined by the
conjunction of random coincidences and deep analogies of affect (Lacan, 1932, p. 234). The sister was

the mirror that reflected an image that erased and displaced any other image of herself. Killing the

www.freepsychotherapybooks.org 429
sister meant wiping out the image that was but a reflection of her own self. The actress embodied
Aimée’s ideal ego insofar as she was a projection of her artistic endeavors, of her desire to better
herself, to be in the public limelight, and to gain fame and glory. Madame Z. was only a shell, an image,

an object. Aimée denied her otherness and perceived her only as an extension of Aimee’s own
imagination.

Lacan’s (1937) looking-glass theory provides the tool for further elaboration of these

mechanisms. At the heart of this theory is the observation that the human child goes through a mirror
phase from 6 to 18 months. Unlike the chimpanzee, a human baby who sees himself or herself in a

mirror is able to perceive that the baby in the mirror is indeed himself or herself, and the sight of his or

her image fills the baby with joy. The baby will begin to laugh, to move with glee, and to express elation

in every possible way. To describe this as jubilation is hardly an overstatement. What has happened is
that the child has put himself or herself on: The child has fit himself or herself into the image in the

mirror, and that structure becomes the identification—in the psychoanalytic sense—of the child’s self.

The “I” shapes itself before objectifying itself as an ego in the dialectic of identification with the imago
of the double and before language assigns it the function of subject in the realm of the universal (Lacan,

1966, p. 94). In French, this fact becomes obvious when we consider the distinction grammar makes

between je and moi, a distinction that roughly approximates the difference between “I” and “me” in
English. When the baby recognizes his or her image in the mirror, the baby has a notion that he or she

is an “I.” The awareness of being an ‘I” means that the image of a whole body, a body that is a totality
replaces the image of a body in pieces in the Kleinian sense, in which the baby is part an organ of his or
her own body and part an organ of another body. Indeed, when a patient’s sense of self has utterly

disintegrated, he or she will often dream that his or her body is cut up and its organs separated and
disjointed with the wings and limbs like those represented in paintings by Hieronymous Bosch. When
the “I” attempts to build itself up, however, dreams represent the id as fortified buildings, castles with

elaborate walls, moats, towers, and other metaphors of inversion, isolation, duplication, annulment,
and displacement characteristic of obsessional neurosis.

www.freepsychotherapybooks.org 430
At the end of this mirror phase, another dialectical mechanism inaugurates the insertion of the

“I” into the “me,” and this takes place in situations that are elaborated by social relationships. Human

knowledge is mediated through identification with the imago of the desire of the other. Perhaps the
mirror also reflected another image; someone else may have been holding the infant—a mother

and/or a father. The constructs that follow will be socially determined, and language will be the
mediator.

The looking-glass phase provides an inkling of why Lacan rejected the positions of American

ego psychology promoted by Rudolph Loewenstein, Ernst Kris, and Heinz Hartmann. It is doubtful that
one of the reasons for his criticism of the “New York troika,” as he often called it, was his sense of

abandonment when Loewenstein set up residence in the United States during the war. In fact, Lacan’s

rejection of ego psychology lies at the very root of his thinking.

In America, Lacan claimed, psychoanalysis was a therapy whose goal was to make the citizen

adjust to the environment. Put in a political perspective, members of society should behave and lead
their lives according to the values of that society. But if we substitute the term “dominant ideology” for

the term “values,” then whether abiding by this ideology is a sign of equilibrium is highly debatable.

Lacan held that this was not the goal of psychoanalysis. His position toward the use of psychoanalysis

in the United States was similar to the position many Americans take about the use of psychiatry in the
Soviet Union. It is possible that from a Soviet perspective, the mere fact of being a dissident is a sign

that one is not “right in the head,” that one is unhinged, and that treatment is needed. But it also seems

quite clear that the purpose of psychiatry or of psychoanalysis is not to adjust these dissidents to the
society in which they live. Today, it is difficult to argue with Lacan’s position that the purpose of

psychoanalysis is psychoanalysis—or, in other words, a quest for truth—rather than making patients

adjust to the cultural mainstream.

Many of Lacan’s most moving pages make this point over and over. In a sense, his most

debatable technical innovation, the variable analytic hour, is a consequence of this quest for truth. He
himself explained that closing off a session meant that an obsessional patient would not go on for

www.freepsychotherapybooks.org 431
months on end making small talk about Dostoevski’s novels while his or her life wasted away. Forcing
such a patient to pay more for less can be an effective truth serum! Be that as it may, the ultimate goals

of analysis for Lacan is the moment of truth, an ineffable sense of unity and plenitude of one’s being.

Lacan took great care to separate the various planes and relations that he expressed with the
words “imaginary,” “symbolic,” and “real.” These terms become intelligible when we examine the

perception we have of ourselves. On a very literal level, since I have never seen myself, and since the

only “me” I can actually “see” is an image of “me” in a mirror, this “me,” this “ego” is an imaginary

function. It is the discovery of an experience, and not an a priori category (Lacan, 1954-55, p. 50).
Furthermore, this imaginary function will intervene in psychic life as if it were a symbol. “One uses the

ME the way the Bororo uses a parrot. The Bororo says I AM A PARROT; we say, I AM ME” p. 52). (The

Bororo are South American Indians found along the upper Paraguay River.)

The imaginary differs from the symbolic. Lacan’s symbolic function is a transcendental function,
beyond any image, and it is inscribed in memory. That is, one of its characteristics is that it is a

presence in absence and an absence in presence. For example, when the baby takes a ball, hides, it, and
takes it back again, all the while saying “here,” “gone,” “here,” the baby is learning that the ball is

present even though he or she cannot see it. When the baby does see it, when it is present, he or she
knows that it may disappear and that its absence is a possibility. In Freud, of course, the disappearance

of the object is linked to the disappearance of the mother. The paradox as Lacan sees it, is that the baby

misses his or her mother when he or she notices she is not present. The mother’s presence is
acknowledged when she has gone. And when the mother is absent, the child learns that he or she can
keep her image present in his or her mind symbolically. Making the ball appear and disappear is a

symbolic expression of learning to cope with the mother as other.

In life, we cannot see the symbolic, of course, but it is present nevertheless. We build it and we
learn how to build it just as, in order to play ball, we have to learn how to do so. For example, the baby

boy sees himself in the mirror, and he also sees his father and mother. When he perceives his parents

as images of his own projections, he functions in the realm of the imaginary. But his parents also exist

www.freepsychotherapybooks.org 432
as the other (l’autre) beyond their images in the mirror. They are parents, but they are also children
and grandchildren of their parents and ancestors. In a sense, siring a child does not make a man a

“father.” A father becomes a “father” only when he takes on for himself the symbolic function of the

“father” and is able to pass this Other on to his child. The child integrates the Other, (l’Autre), with an
initial capital letter. His past, that is to say his history, is inserted into the present as well as the future

—not only his own history, the history he knows, such as the childhood he remembers, but also the

history he has forgotten and the history that his ancestors repressed but that he himself continues to
perpetuate. When I claim that my cat Jeffrey is a devoted and caring father, I am guilty of

anthropomorphism that attributes to the cat the feelings of a human father. My statement is articulated
on Lacan’s imaginary level in which my words reflect what I see in my mirror. When I write that the
horse Prince William V may win the famous X derby because Prince William IV, an X derby winner—

himself sired by Prince William III, also a winner—was his father, I am speaking on a level Lacan would
call symbolic. My example may be imaginary—after all, I have made up the names of the racehorses—
but because the racing world itself is a symbolic realm and because its customs and conventions make

sense in a historical and human perspective, the racehorse as father is a symbolic entity.

Lacan’s “real,” the third element in the tryptich, is not reality. It is likely that Lacan uses this
term in the same sense as Jeremy Bentham did when he meant that the “real” was the opposite of the

“fictitious” (see Lacan, 1959-60, p. 60). The concept includes what is neither symbolic nor imaginary. It
refers to very stuff that is structured by the symbolic. Applied to the concept of fatherhood, for
example, the real would be the physiological act of procreating without any interpretation whatsoever.

Anything at all that we say about the act, the very words I use to convey the information, immediately
draw the reader and me onto the symbolic plane. The word “father” itself is a sublimation and a
spiritual act. As Lacan (1959-60, p. 14), explained many times, the king is naked. The unconscious itself

is structured around the symbolic function.

We are now in a position to understand why and how the unconscious is structured like a
language. The real cannot be apprehended at all except through a symbolic operation. If there is no

www.freepsychotherapybooks.org 433
symbolic level, the real remains organic and dead, as it were. The initial perception is in a sign, and this
sign is itself both a signifier and a signified, that is, an element of language.

A lot of ink has been spilled about Lacan’s debt to Ferdinand de Saussure’s (1915) Cours de

linguistique générale. In fact, the ideas of the Cours are and were quite familiar to all French-speaking
linguists and psychoanalysts of the early twentieth century. Ferdinand’s own son, Raymond de

Saussure, was a psychoanalyst, a member of the Sociacete Psychanalytique de Paris, and he knew

Lacan well. It has been claimed (Roudinesco, 1982) that Raymond was totally ignorant of his father’s
contribution to linguistics, but I cannot believe this at all. When Lacan takes up Saussure’s distinction

between the “signifier” (signifiant)—the acoustic image, the sound of an utterance—and the “signified

(signifié)—the concept or concepts expressed by the utterance, he is using a linguistic shorthand that

was widely used. Likewise, he is using appropriate modem terminology when he refers to the
paradigmatic chain of thought—the principle of “clang” associations whereby “big” leads to “dig,” and

“dig” leads to “rig” or another such sound—and to syntagmatic associations, in which “big” may lead to

“great,” “Alexander,” “Philip,” and “Macedonia.” In this perspective, the conclusion that the unconscious
is structured like a language means simply that there are no innate ideas, and that the unconscious is a

cultural rather than an organic entity.

The same point can be made about other applications of linguistics to psychoanalysis. Freud’s
analysis of dreams, his mechanisms of displacement, denial, and similarity, are themselves tropes. An

attempt to determine whether a given symptom is expressed linguistically by, for example, a

synecdoche (the trope that suggests a part for the whole, less for more, or more for less) or by a
metalepsis (the phrase whereby an indirect expression is substituted for a direct expression3 may lead
to an accurate descriptions of given speech pattern corresponding to given clinical configurations.
Perhaps that is the significance of Lacan’s coinage of the word Lalangue, a linking in one word of the

article la and the noun langue, meaning tongue, in the sense of speech or language. The word also

suggests André Lalande, the author of a famous French dictionary of philosophy, a book philosophy
students refer to with the metonymy or synecdoche, Lalande. The reasoning goes something like this: I

www.freepsychotherapybooks.org 434
speak English just as you speak English, but my speech is different from yours, although it is also the
same, so that my Lalangue is like your Lalangue, yet the two are not the same. Just like Humpty
Dumpty, I make my words mean something different than you make your words mean, but we have to

use the same words. Even when the unconscious coins new words—Lacan’s Lalangue—it adapts
signifiers of the linguistic and puts them to its own use. This new sign may remain a private term, or it

may enter the linguistic mainstream. When it does, it modifies the Lalangue of everyone and in some

way it changes the cultural unconscious, that is, the linguistic substratum of our culture.

Perhaps herein lies the explanation for Lacan’s deliberate use of a language that bares his own

linguistic associations. As a student of Saussure and a reader of Hegel and Heidegger, he knew that in a

sense, each one of our utterances changes the total language of our linguistic community and that some

utterances change it more than others. For example, his theory of the “Nom du Père” certainly modified
the theoretical assumptions French psychiatrists and psychoanalysts have about psychosis. Here, in a

sense, Lacan’s Lalangue has begun to change not only clinical theory, but also its practice. Very simply

put, the “Nom du Père” means not only the father’s name, but also the father’s “no,” that is to say, the
act whereby the father severs the symbiotic bond between mother and child. This “no” must take place

if the child is to develop into an autonomous being. The name of the father cannot be transferred to the

child unless the child receives it and accepts it on the symbolic level. In Lacan’s terminology, the image
in the mirror, my other, must have achieved a link with the Other, who is not myself, but who is

constituted by my recognition of how my history can be integrated in the world in which I live—that is,
the Name-of-the-Father. Why are the N in “Nom” and the P in Père capitalized? These capital letters
suggest a symbolic level, and they are allusion to the Father in the Scriptures. The signifier goes beyond

the actual daddy, and suggests that the Name-of-the-Father is sacred insofar as it gives a meaning to
our lives and sustains the ideas and ideals of society, culture, and civilization.

“What makes a psychosis come about?” was the question Lacan asked himself. Years of clinical
experience (it must always be kept in mind that Lacan’s theory and his reading of Freud took place in

the context of his extensive clinical experience with psychotics) led him to perceive that in every case

www.freepsychotherapybooks.org 435
there was a Verwerfung. Lacan translates Verwerfung into French as foreclusion, a term he borrowed
from Damourette & Pichon’s monumental grammar book, Des mots à la pensèe.4 I would translate this
into English as “shut out, forclosed, and excluded,” suggesting something that might have opened, but
remained closed. Pichon used foreclusion to describe characteristics of the second term in the French

negative, for example, the words pas (not), plus (not), rien (nothing), jamais (never), aucun (none), and
personne (nobody) in such phrases as Je ne sais pas (I don’t know), Je ne sais plus (I no longer know),

Cela ne me dit rien (That doesn’t mean anything to me), Elle ne sait rien (She knows nothing), Il ne va

jamais au cinèma (He never goes to the movies), Il n’a aucun devoir (He has no homework), and
Personne n’est venu (Nobody came). In each of these sentences, and in this type of French sentence

generally, the second negative casts out definitively something that might have been. Likewise, in
psychosis, the ‘Nom du Père’ signifier is itself excluded.

In order for psychosis to manifest itself, the Name-of-the-Father, must be verworfen, excluded,
foreclosed; it must have failed to reach the Other’s place, and must now be called there in a symbolic

opposition with the subject. The failure of the Name-of-the-Father at that place, by the hole that it

opens in the signified, begins the cascade of signifiers whereby proceeds the growing disaster of the
imaginary, until the level is reached where signifier and signified stabilize in a metaphor of delirium.

Lacan’s theory, then, is that in psychosis the central signifier, that is, the Name-of-the-Father,

has failed to inscribe itself in the subject’s language register. At the place where it should have been
incorporated, there is a gap, a hole, a void. When the occasion presents itself—for example, when an

ersatz signifier happens to make its way into the appropriate chain—this vacuum will suck up any
signifier at all that happens to come along, and an elaborate delusional system will come to occupy the
place of the missing Name-of-the-Father. For example, in the case of Schreber, Geheimrat Professor

Flechsig, remained for him the chief instigator during the entire course of his illness. Freud (1911)
quotes Schreber: “Even now the voices that talk with me call out your name to me hundreds of times
each day. They name you in certain constantly recurring connections, and especially as being the first
author of the injuries I have suffered” (p. 38). God Almighty comes to play a part as Flechsig’s

www.freepsychotherapybooks.org 436
accessory, as does the soul of the chief attendant of Pierson’s asylum, the clinic to which Schreber
moves. They are but substitutes for the Name-of-the-Father; and the divine rays, the special birds, the
nerves of God, and Schreber’s own transformation into a woman are generated to fill the emptiness

created by the absence of the transcendental signifier.

Lacan’s theory of the unconscious is a materialistic theory: The unconscious is structured like a

language—that is, a concept, a signified, is linked to a signifier, an acoustic image, and in turn this

signifier suggests another signified, so that an idea is immediately turned into matter. It is paradoxical,
therefore, that Lacan speaks like a theologian. The psychotic—and Schreber is an excellent example—

makes God Almighty into the image of the father, but in fact, the correct stance and the condition of

sanity is that the father be created in the image of God Almighty. The unconscious may be structured

like a language, but if this language is to sustain interhuman relations, culture, and civilization, then it
must itself rest on a transcendental signifier in the image of the Great Other (le Grand Autre), Lacan

often said.

The dedication of Lacan’s doctoral thesis to his brother, Reverend Father Marc-François Lacan,

Benedictine monk of the Congregation of France, makes us wonder whether both brothers did not
follow a similar path. Lacan was not a man of the church, but nevertheless he preached a gospel. In his

gospel the tropes of psychoanalysis incorporated tropes of other disciplines—philosophy, theology,

literature, art, linguistics, and anthropology—characteristic of the culture of a given time and place: the

middle of the twentieth century in France, an anticlerical country with a strong Catholic tradition.

Lacan’s Christian Parisian cosmopolitanism may be the counterpart of Freud’s Jewish middle European
universalism.

ACKNOWLEDGMENT

The author gives grateful thanks to the Research Foundation of the City University of New York
for a Faculty Research Fellowship.

www.freepsychotherapybooks.org 437
REFERENCES

Bertin, C. (1982). La dernière Bonaparte. Paris: Librairie Académique Perrin.

Clement, C. (1981). Vie et légendes de Jacques Lacan. Paris: Grasset.

Damourette, J., & Pichon, E. (1911-1950). Des mots à la pensie. Essai de Grammaire de la langue
francçaise. 7 vol. Paris: d’Artrey.

Freud, S. (1887-88). Preface to Charcot. Standard Edition, 1.

Freud, S. (1893). Charcot. Standard Edition, 3.

Freud, S. (1911). Psycho-Analytic Notes on An Autobiographical Account of a Case of Paranoia.

Standard Edition, 12.

Gide, A. (1925). Les faux-monnayeurs. Paris: Gallimard.

Lacan, J. (1932). De la psychose paranoîaque dans ses rapports avec la personnalité. Thèse pour le
doctorat en médecine, diplóme d’état. In J. Lacan, De la psychose paranoiaque dans ses
rapports avec la personnalité, suivi de Premiera écrits sur la paranoîa. Paris: Seuil, 1975.

Lacan, J. (1933a). Le problème du style et la conception psychiatrique des formes paranoîaques de

l’expérience. In J. Lacan, De la psychose paranoiaque dans ses rapports avec la
personnalité, suivi de Premiera écrits sur la paranoîa. Paris: Seuil, 1975.

Lacan, J. (1933b). Motifs du crime paranoîque. In J. Lacan, De la psychose paranoiaque dans ses rapports
avec la personnalité, suivi de Premiera écrits sur la paranoîa. Paris: Seuil, 1975.

Lacan, J. (1936, July). The looking-glass phase. Paper presented at the 14th International
Psychoanalytic Congress, July 31 Marienbad. (1947) 16th International Congress, July
17; Zurich . (1966) Ecrits, Paris: Seuil.

Lacan, J. (1950). Propos sur la causalité psychique. In L. Bonnafe, H. Ey, S. Follin, J. Lacan, J. Rouart Le
Problème de la psychogenèse des névroses et des psychoses. Paris: Desclée de Brouver,
(1966). Ecrits.

www.freepsychotherapybooks.org 438
Lacan. J. (1953-54). Le séminaire: Livre I. Les écrits techniques de Freud (J.-A. Miller, Ed.). Paris: Editions
du Seuil, 1975.

Lacan, J. (1954-55). Le séminaire: Livre 2. Le moi dans la théorie de Freud et dans la technique de la
psychanalyse (J.-A. Miller, Ed.). Paris: Editions du Seuil, 1978.

Lacan, J. (1955-56). Le séminaire: Livre 3. Les psychoses (J.-A. Miller, Ed.). Paris: Seuil, 1981.

Lacan, J. (1957). Le séminaire sur “La lettre volée,” April 26,1955. In Ecrits. Paris: Seuil, 1966.

Lacan, J. (1958). Jeunesse de Gida ou la lettre et le désir. Critique (April, 1958), 131, pp. 291-315. (1966,
Ecrits), Paris: Seuil.

Lacan, J. (1959-60). L’éthique. Paris: Editions du Piranha. Unauthorized transcript.

Lacan, J. (1964). Le séminaire: Livre 11. Les quatre concepts fondamentaux de la psychanalysee (J.-A.
Miller, Ed.). Paris: Seuil, 1973.

Lacan, J. (1966). Ecrits. Paris: Edition, du Seuil.

Lacan, J. (1966). Position de l’inconsicent au congrès de Bonneval repris de 1960 en 1964. In Ecrits.
Paris: Seuil.

Lacan, J. (1972-73). Le séminaire; Livre 20. Encore (J.-A. Miller, Ed.). Paris: Seuil, 1975.

Lacan, J. (1980). Le Séminaire de 1980: Dissolution. Ornicar? No. 20-21.

LaScisson de (1953, July 14). Ornicar?. 1956, p. 136.

Levy-Valensi, J., Migault, P., & Lacan, J. (1931). Ecrits “inspirés: Schizographie. In J. Lacan, De la psychose
paranoiaque dans ses rapports avec la personnalité, suivi de Premiera écrits sur la
paranoîa. Paris: Editions du Seuil, 1975.

Miller, J.-A. (Ed.). (1976). La Scission de 1953 [Special supplement], Ornicar? 7.

Roudinesco, E. (1982). La Bataille de cent ans: Histoire de la psychanalyse en France. Paris: Ramsay.

www.freepsychotherapybooks.org 439
Saussure, F. de (1915). Cours de linguistique générale. Ed. Bally, C., Sechehaye, A. Geneva.

Schneiderman, S. (1983). Jacques Lacan. The Death of an Intellectual Hero. Cambridge, Mass. and
London, England: Harvard University Press.

Notes

1) The letter to Lowenstein, dated July 14, 1953, contains the following:

I want you to feel how bitter this experience has been for us, and also how decisive. I give
you authority to communicate this [letter]—in spite of the tone of the confessional that is
found and in spite of our special relationship—to Heinz Hartmann whose person I have
always held in the highest esteem [p. 135].

The end of his letter to Hartmann on July 21, 1953, reads as follows:

Dear Heinz Hartmann, I regret that the chaotic events of past years, as well as the
extreme isolation that is conditioned by our professional life prevented me from making
myself better known to you. But I count on your authority to make it possible for the
authentic and deeply caring effort that is the foundation of my work in bringing Freud’s
teaching alive to be respected; to bring back the tone of reason to a fight that is as sterile
in its forms as it is base in its motives, and to take the equitable measures necessary to
preserve the audience that psychoanalysis is presently conquering in France and that
this fight can only hinder, (p. 136)

Evidence that Marie Bonaparte might have been behind Lacan’s exclusion from the IPA is
apparent from excerpts of her own letters to Lowenstein published in the biography, La
dernière Bonaparte (Bertin, 1982).

2) It is likely that more information will become available in the near future, particularly with the
publication of the second volume of Roudinesco’s history of psychoanalysis in France.

3) For example, in the French play Phèdre, by Jean Racine, the heroine in love with her stepson,
Hippolytus, expresses her desire by pretending that she loves her husband Theseus,
Hippolytus’ father, not the way he is now, but the way he was when he was his son’s age
(Roudinesco, 1982, p. 158).

www.freepsychotherapybooks.org 440
4) Roudinesco (1982, pp. 392-395) points out that although Lacan is usually given credit for this term
in psychoanalysis, in fact, he borrowed it from his colleague.

www.freepsychotherapybooks.org 441
14
ADOLF GRÜNBAUM: PSYCHOANALYTIC
EPISTEMOLOGY

BARBARA VON ECKARDT, PH.D.

Adolf Grünbaum was born on May 15,1923 in Cologne, Germany. He received his B.A. from

Wesleyan University in 1943 with high distinction in mathematics and philosophy, his M.S. in physics

from Yale University in 1948, and his Ph.D. in philosophy from Yale University in 1951. He began his
teaching career at Lehigh University in 1950. Five years later he was appointed William Wilson

Selfridge Professor of Philosophy. In 1960 he accepted a position at the University of Pittsburgh as

Andrew Mellon Professor of Philosophy, where he has been ever since. In 1979 he was also appointed

Research Professor of Psychiatry at the same university on the basis of his work on psychoanalytic
epistemology.

Grünbaum is currently one of the leading figures in contemporary philosophy of science. He has

been president of the Philosophy of Science Association for two terms, 1965-67 and 1968-70, and was

elected president of the American Philosophical Association (Eastern Division) for 1982-83. In
addition, he has received numerous honors and awards for his work, the most recent of which is a

festschrift in his honor (Cohen & Laudan, 1983) containing essays by 14 of today’s principal

www.freepsychotherapybooks.org 442
researchers in philosophy of science as well as two leading psychoanalysts.

Grünbaum’s interest in psychoanalysis is relatively recent. His past work primarily concerned

philosophical problems of space and time and the theory of scientific rationality (see Cohen & Laudan,

1983 for a complete bibliography). Since 1976, however, when his first paper on psychoanalytic
epistemology appeared (Grünbaum, 1976), he has produced at least 10 papers as well as a book on the

subject, which have succeeded in completely changing the state of the art. The purpose of this essay,

then, is to provide a summary and critique of this work.

The two fundamental questions that Grünbaum’s work on psychoanalysis addresses are these:

1. What sort of standards of assessment ought we to invoke in evaluating psychoanalysis? That

is, ought we to regard it as making knowledge claims, and, if so, what kind?

2. Given that we have chosen certain standards of assessment, how does psychoanalysis

measure up to those standards?

With respect to the first question, Grünbaum has argued emphatically that (a) the most

appropriate standards of assessment for psychoanalysis are those derived from empirical science,

contrary to the claims of the hermeneuts, Jurgen Habermas, Paul Ricoeur, and George Klein
(Grünbaum, 1983c, 1984); and (b) psychoanalysis meets the minimal conditions necessary for

applying those standards, contrary to the claims of Karl Popper (1963) (who accepts Grünbaum’s first

statement but denies the second on the grounds that psychoanalysis is unfalsifiable) (Grünbaum, 1976,
1977, 1979). With respect to the second question, however, his stance has been severely critical. In his
view, there are serious difficulties in the way of regarding psychoanalysis as good science. These stem

not only from serious liabilities involved in the use of clinical data but also from the modes of
reasoning that Freud used to provide evidential support for his theory (Grünbaum, 1983b, 1984).

It should be clear that any attempt to argue convincingly either for or against the scientific

www.freepsychotherapybooks.org 443
status of psychoanalysis ought to be informed by both a thorough understanding of the psychoanalytic
literature and a sophisticated conception of the nature of science. The literature prior to Grünbaum’s

recent outpouring on the subject suffers, in my view, in both of these respects. That is, either it exhibits
a very superficial understanding of psychoanalysis or it is naive about the nature of science. The

importance of Grünbaum’s contribution in the area of psychoanalytic epistemology rests on the fact

that his work is unparalleled on both counts. Not only does he bring to bear a very great sophistication
in the philosophy of science but, in addition, he has done his psychoanalytic homework.

In 1959, the philosopher John Hospers summed up the results of one of the first major

conferences on philosophy and psychoanalysis as follows:

As I try to get a composite picture of the results of the conference, the

thing that stands out most in my mind is the lack of genuine
communication between the psychoanalysts and the philosophers.
Psychoanalysts are, quite understandably, too busy treating patients to
have acquainted themselves with the latest guns in the arsenal of
epistemology and philosophy of science, and are therefore at a loss to
reply to the charges leveled at them by the philosophers in the way the
philosophers want. The philosophers, for their part, are—equally
understandably—ignorant of the vast amount of empirical detail garnered
by psychoanalysts in the last half-century as well as the complexity of
many of the theoretical concepts employed in psychoanalysis. The
inevitable result is that each party to the dispute only feels confirmed in
his previous suspicion, namely that the other party’s remarks are either
incompetent or irrelevant, given to making either scandalously overblown
claims or excessively demanding systematic requirements [p. 336].

I believe that Grünbaum has gone more than halfway toward closing this communication gap

from the philosophical side. Not only is his work impressively learned with respect to the
psychoanalytic literature, as already mentioned, but he has also worked very hard at establishing lines

of communication with the psychoanalytic community. For all of this, however, his writing may not be

easily accessible to psychoanalysts and students of psychoanalysis, for it does presuppose a

www.freepsychotherapybooks.org 444
considerable sophistication in the philosophy of science and the techniques of philosophical
argumentation. It is chiefly this consideration that has dictated the style of the present essay.

Grünbaum’s work merits serious attention from anyone interested in the cognitive status of

psychoanalysis. My principal concern, therefore, has been to make the most important of his ideas and
arguments accessible to the reader. This approach has had several consequences. First, I have devoted

a certain amount of space to providing background that seemed to me essential to understanding

either the content of Grünbaum’s writing or its importance. Second, I have had to strike a compromise
between the demands of depth and breadth in the discussion of Grünbaum’s work itself. Grünbaum’s

writing is exceedingly rich. In attempting to present clearly the central lines of argumentation, much of
this richness has necessarily been lost. I thus urge the reader interested in his work to consult the
original. In addition, certain topics have simply not been touched on at all. Where this is the case, I have

tried to indicate what has been omitted in the appropriate place in my discussion.

SHOULD FREUDIAN PSYCHOANALYSIS BE ASSESSED AS SCIENCE?

Grünbaum’s approach to this question has been twofold. First, he has repeatedly emphasized

that Freud himself regarded psychoanalysis as scientific. In support of this claim, he cites passages
such as the one in which Freud states that the explanatory gains from positing unconscious mental
processes “enabled psychology to take its place as a natural science like any other” (Freud, 1940a, p.

158, see also 1925, p. 58; 1933, p. 159; 1940b, p. 282). Second, Grünbaum has devoted considerable
effort to providing counterarguments to those who have suggested that, for one reason or another,
Freudian theory ought not be regarded as scientific on the grounds that it fails to satisfy certain

minimal requirements for scientific candidacy. These arguments have been directed, in particular,
against Karl Popper, Jurgen Habermas, Paul Ricoeur, and George Klein.

Although it might appear that Grünbaum has simply adopted the strategy of shifting the burden

of proof to those who wish to deny scientific status to Freudian theory, it is possible to view his
discussion as part of an overall implicit positive argument as follows:

www.freepsychotherapybooks.org 445
1. A body of work should be judged by the standards of adequacy subscribed to by the author or

creator unless there is compelling reason not to.

2. Freud took himself to be doing science.

3. The reasons that have been offered in the literature against assessing psychoanalytic theory

in terms of the standards of science are uniformly uncompelling.

4. Therefore, Freudian psychoanalytic theory ought to be assessed as science.

Since the second premise is not difficult to establish, the bulk of Grünbaum’s discussion on this

matter has been devoted to justifying the third premise. In the discussion that follows, we shall focus

on his consideration of the arguments of Popper and Habermas. Readers interested in his discussion of
Ricoeur and Klein should consult Grünbaum, 1984, pp. 43-93.

PSYCHOANALYSIS AS PSEUDO-SCIENCE

POPPER’S CHALLENGE

In 1953, in a paper reviewing his philosophical work of the past 30 or more years, Karl Popper

challenged the scientific status of psychoanalysis, claiming that it was nothing more than a
pseudoscience. His reasoning was this: To be scientific, a theory must be falsifiable; however,

psychoanalytic theory is not falsifiable. Therefore, psychoanalytic theory is not scientific. Interestingly
enough, it was in part the case of psychoanalysis that led Popper to see the importance of falsifiability
in the scientific process in the first place.

When the problematic nature of psychoanalysis first occurred to him, Popper’s principal

concern was the so-called “problem of demarcation.” This is the problem of distinguishing theories that
can legitimately be considered candidates for scientific evaluation from those that cannot, in particular,

www.freepsychotherapybooks.org 446
from those “pseudoscientific” theories such as astrology that share certain superficial characteristics
with genuine scientific theories but that lack some essential feature. The accepted demarcation
principle at the time was an inductivist one: A theory is scientific just in case it is inductively well

confirmed on the basis of empirical evidence. It was in part the contrast between Freud’s

psychoanalytic theory and Einstein’s theory of gravitation that led Popper to believe that this was an
incorrect view. On intuitive grounds, something seemed to be wrong with psychoanalysis, but the
problem could not be its lack of “verifications” because these seemed to be rampant. Popper began to

suspect that the difficulty was precisely that psychoanalytic theory could always be verified no matter
what. In contrast, a genuine scientific theory like Einstein’s was distinguished by the fact that, if it were
false, it could be falsified so easily, because potentially falsifying test outcomes were readily imagined.

Popper (1963) wrote:

I found that those of my friends who were admirers of Marx, Freud, and
Adler, were impressed by a number of points common to these theories,
and especially by their apparent explanatory power. These theories
appeared to be able to explain practically everything that happened within
the fields to which they referred. The study of any of them seemed to have
the effect of an intellectual conversion or revelation, opening your eyes to
a new truth hidden from those not yet initiated. Once your eyes were thus
opened you saw confirming instances everywhere: the world was full of
verifications of the theory. Whatever happened always confirmed it. Thus
its truth appeared manifest; and unbelievers were clearly people who did
not want to see the manifest truth; who refused to see it, either because it
was against their class interest, or because of their repressions which were
still ‘un-analyzed’ and crying aloud for treatment. The most characteristic
element in this situation seemed to me the incessant stream of
confirmations, of observations which “verified” the theories in question.
...It began to dawn on me that this apparent strength was in fact their
weakness [p. 34].

In contrast, the situation with Einstein’s theory was “strikingly different.” On the basis of his
theory of gravitation, Einstein had predicted that light from a distant star would be bent near the sun.

What was impressive about this case, according to Popper (1963), was the risk involved in a prediction

www.freepsychotherapybooks.org 447
of this kind. For

if observation shows that the predicted effect is definitely absent, then the
theory is simply refuted. The theory is incompatible with certain possible
results of observation—in fact with results which everybody before
Einstein would have expected. This is quite different from the situation I
have previously described, when it turned out that the theories in question
were compatible with the most divergent human behaviour, so that it was
practically impossible to describe any human behaviour that might not be
claimed to be a verification of these theories [p. 36].

It was this purported insight that led Popper to his well-known principle of falsifiability. In

addition, he proposed that the method of science is essentially one of bold conjectures and attempted
refutations whose rationality lies in the facts that first, scientists are always seeking to falsify their
theories and, second, they accept their theories only (and always only tentatively) when they have

successfully resisted numerous attempts at falsification.

THE RESPONSE TO POPPER’S CHALLENGE

The philosophical response to Popper’s challenge over the past 20 years has taken a variety of
forms. In order to understand Grünbaum’s contribution to this discussion, it will be useful to indicate

briefly the major positions that have been taken.

It was noted quite early on that there is an important ambiguity in the claim that psychoanalysis

is not falsifiable. Kennedy (1959), for example, pointed out that psychoanalysis can be considered

unfalsifiable for two very different reasons: first, because of the attitude of its proponents in the face of
allegedly unfavorable evidence; and, second, because of the logical structure of the theory. Martin

(1964b) refined this distinction further by introducing four possible senses of the notion of

refutability, two of which concerned the attitudes of proponents of the theory, and two of which

concerned its logical structure. He wrote:

www.freepsychotherapybooks.org 448
When we ask whether a theory T is a refutable theory, we may be asking
any of the following questions:

1. Are people who are advocates of theory T willing to specify what

evidence could count against theory T?

2. Are people who believe in theory T willing to accept some of the

evidence brought forth to refute theory T instead of explaining it all
away?

3. Is the relation between the theoretical language and the observational

language of theory T clear and unambiguous?

4. Is it possible to give theory T, in which the relation between the

theoretical and observational language is extremely vague and
ambiguous, clear and unambiguous formulation [p. 81]?

Martin claimed, however, that the fourth question is not an interesting sense of ‘refutable,’ since

any theory can be considered refutable in that sense, including those that we consider paradigm cases
of unrefutable theories (such as that the absolute is perfect and developing in history).

If we subdivide Popper’s challenge into two parts, one directed at the attitudes of its proponents
and one at the logical structure of psychoanalytic theory, we find endorsements of both positions in the
literature. For example, a number of people have argued that the proponents of psychoanalytic theory

typically exhibit a very unscientific attitude with respect to putative disconfirming data. After
proposing the four senses of ‘refutability,’ Martin (1964b) claimed that the answer to the first two
questions is no. Typically, psychoanalysts are unwilling to specify what evidence will count against

their theory. Furthermore, they tend to discount any allegedly disconfirming evidence. A similar view
had been voiced earlier by Hook (1959a).

Cioffi (1970) took the charge much further. Psychoanalysis is a pseudoscience, he wrote,

principally because it uses methodologically defective procedures:

www.freepsychotherapybooks.org 449
For an activity to be scientific it is not enough that there should be states of
affairs which would constitute disconfirmation of the theses it purports to
investigate; it must also be the case that its procedure should be such that
it is calculated to discover whether such states of affairs exist. I use the
word “calculated” advisedly. For to establish that an enterprise is pseudo-
scientific it is not sufficient to show that the procedures it employs would
in fact prevent or obstruct the discovery of disconfirmatory states of
affairs but that it is their function to obstruct such discovery. To claim that
an enterprise is pseudo-scientific is to claim that it involves the habitual
and willful employment of methodologically defective procedures (in a
sense of willful which encompasses refined self-deception) [p. 472].

Cioffi goes on to argue that Freudian psychoanalysis is pseudoscientific in precisely this sense.

For it is characterized by a “host of peculiarities…which are apparently gratuitous and unrelated, but

which can be understood when once they are seen as manifestations of the same impulse: the need to
avoid refutation” (p. 473). The principal devices that Freud uses to accomplish this end, according to

Cioffi, are these: First, hypotheses presented prior to the discovery of apparently disconfirming

evidence are, typically, formulated in a narrow and determinate sense; afterwards, however, they are
construed in a “broader and hazier” way so as to avoid the disconfirmation. Second, prior to the

discovery of apparently disconfirming evidence, Freud allows for the relevance of evidence from a

number of intersubjective sources, including observation of the behavior of children, inquiry into the
distinctive features of the current sexual lives or actual infantile sexual history of neurotics, or

determination of the outcome of therapy based on his theory. In the face of apparently disconfirming

evidence, however, he typically retreats to the claim that the only reliable source of evidence is

material obtained during the psychoanalytic session and subjected to interpretation by a trained
analyst. Third, his theory contains such a variety of mechanisms and interpretative principles that it is

possible for him to interpret any phenomenon in a way consistent with his theory. Thus, “he typically

proceeds by beginning with whatever content his theoretical preconceptions compel him to maintain
underlies the symptoms, and then, by working back and forth between it and the explanandum,

constructing persuasive but spurious links between them” (Cioffi, 1970, p. 497). Finally, his

interpretations are not even constrained by considerations of logic; for it is not even necessary for the

www.freepsychotherapybooks.org 450
various meanings of a symptom to be compatible with one another.

The principal early supporter of Popper’s position with respect to the nonfalsifiability of
Freudian theory in the logical sense was Nagel. In a classic paper (Nagel, 1959) he offers us an analysis

of precisely why psychoanalytic theory is problematic:

The theory does not seem to me to satisfy two requirements which any
theory must satisfy if it is to be capable of empirical validation.…In the first
place, it must be possible to deduce determinate consequences from the
assumptions of the theory, so that one can decide on the basis of logical
considerations, and prior to the examination of any empirical data,
whether or not an alleged consequence of the theory is indeed implied by
the latter. For unless this requirement is fulfilled, the theory has no
definite content, and questions as to what the theory asserts cannot be
settled except by recourse to some privileged authority or arbitrary
caprice. In the second place, even though the theoretical notions are not
explicitly defined by way of overt empirical procedures and observable
traits of things, nevertheless at least some theoretical notions must be tied
down to fairly definite and unambiguously specified observable materials,
by way of rules of procedure variously called “correspondence rules,”
“coordinating definitions,” and “operational definitions.” For if this
condition is not satisfied, the theory can have no determinate
consequences about empirical subject matter [p. 40].

Nagel argued that Freudian theory failed both of these conditions primarily because of its

vagueness and metaphorical character.

Freudian formulations seem to me to have so much “open texture,” to be

so loose in statement, that while they are unquestionably suggestive, it is
well-nigh impossible to decide whether what is thus suggested is
genuinely implied by the theory or whether it is related to the latter only
by the circumstance that someone happens to associate one with the other
[p. 41].

Martin (1964b) provided further support for Nagel’s view. As the quote from Nagel makes clear,

www.freepsychotherapybooks.org 451
the accepted view at the time was that the empirical import of a genuinely scientific theory (i.e., the
link to its observation base) is mediated by so-called correspondence rules, consisting either of explicit
or partial definitions of the theoretical vocabulary in terms of an observational vocabulary. Since the

existence of such correspondence rules is a necessary condition of a theory being falsifiable, one way to
ascertain the scientific status of psychoanalytic theory, according to Martin (1964b), is to try one’s best

“to separate the observational basis of the theory from the theoretical structure, and to extract rules of
correspondence from the context of the uses of the two languages” (p. 85). When Madison (1961) used

this strategy, he concluded that for some aspects of psychoanalysis, there was no associated

observational language and rules of correspondence, whereas for others, there was. Martin, however,
argues that Madison’s allegedly positive results are incorrect. What Madison actually found, according

to Martin (1964b), are “the rudiments of an observational language and rules of correspondence” (p.

86). Madison takes these rudiments and reformulates them into a clearer and more precise form, but
he fails to distinguish his formulations from Freud’s. Thus, he only shows that Freudian theory is

falsifiable in Martin’s fourth and, presumably, uninteresting sense.

There have been, however, a few dissenting voices. Salmon (1959) argued that psychoanalytic

theory appears to be unfalsifiable only if one assumes that “a few restricted items of behavior can
constitute evidence for or against the hypothesis” (p. 262). It is true, according to Salmon, that any
single item of behavior may be compatible with a hypothesis, for example, that the patient suffers from
unconscious hostility toward his father, for according to psychoanalytic theory, unconscious hostility

can be expressed in a variety of ways and is served by a variety of mechanisms. This does not mean,
however, “that every total behavior pattern is compatible with the hypothesis of unconscious hostility”

(p. 262). A similar point was made by Hospers (1959). There are no “crucial experiments” for

psychoanalysis, but neither do they exist for physics. What validates or invalidates psychoanalytic
hypotheses are patterns of behavior. Correspondence rules do not take the form of “If p, then q”; rather

they look like “If p, then q or r or s or…” followed by a finite disjunction of propositions. And since the

disjunction is finite, Hospers (1959) argues, “it is emphatically not true that the Oedipus complex
would be believed in no matter what the empirical facts are: if none of the items q, r, s... occurred, it

www.freepsychotherapybooks.org 452
would have to be concluded (and would be) that the individual in question had no Oedipus complex”

(p. 343). More recently, Glymour (1974, 1980) has argued that there is a rational strategy for testing
important parts of psychoanalysis and that this strategy was immanent in at least one of Freud’s

(1909) case studies, that of the Rat Man. In particular, the best available evidence concerning the actual

life history of the Rat Man, Paul Lorenz, had refuted the hypothesis Freud held at the time concerning
the sexual etiology of adult obsessional neurosis.

A number of philosophers came to Freud’s defense in a quite different way. They agreed that if

the falsifiability of a theory requires that the theory alone (mediated only be correspondence rules)

entails a falsifiable observation statement, then Freudian psychoanalytic theory is, strictly speaking,

unfalsifiable. However, this does not necessarily make it a pseudoscience. Why not? Farrell (1963,
1964) suggested that there was another option available. Psychoanalysis is, on his view, a protoscience.

That is, it is an “empirical and speculative synthesis, which is premature in that it runs far ahead of the
evidence that can upset or support it with reasonable certainty” (Farrell, 1963, p. 24). Nevertheless,

there is reason, he claims, to take it seriously as a tentative basis for future research. The

psychoanalytic method has produced an enormous amount of factual material, which the theory has to
some degree succeeded in ordering, describing, and explaining. In addition, a lot of experimental work
by psychologists attempting to test psychoanalytic theory seems to show that in places, at least,

Freudian theory is “on to something.”

Another, far more damaging reason for rejecting Popper’s claim that psychoanalytic theory is

pseudoscientific because it is unfalsifiable was offered by Lakatos (1970, 1971). He suggested that

Popper’s demarcation criterion can be assessed in terms of the following metacriterion: “If a
demarcation criterion is inconsistent with the “basic’ appraisals of the scientific elite, it should be

rejected” (Lakatos, 1971, p. 125). Given this metacriterion, Popper’s demarcation principle is clearly

problematic. For “exactly the most admired scientific theories simply fail to forbid any observable state
of affairs” (Lakatos, 1970, p. 100). The principal reason for this is that most scientific theories are

normally interpreted as containing a so-called ceteris paribus clause. That is, they “forbid an event

www.freepsychotherapybooks.org 453
occurring in some specified finite spatio-temporal region…only on the condition that no other factor…has
any influence on it” (p. 101). But then if a prediction of the theory is not borne out, the theory is not
automatically falsified because “by replacing the ceteris paribus clause by a different one the specific

theory can always be retained whatever the tests say” (p. 101-102). In the philosophy of science
literature of recent years, a more general version of this point has become commonplace. As we noted,

in the early responses to Popper’s challenge, it was generally assumed that theory and observation are

mediated by correspondence rules. In recent years, however, it has been argued that the so-called
“received view” of correspondence rules vastly oversimplifies the relationship between a theoretical

hypothesis undergoing test and the observable evidence adduced in its behalf. Careful examination of

scientific case studies has revealed that theory and data are often mediated by a complex array of
auxiliary propositions: hypotheses from related theories, theories of measurement and theories of the

data, assumptions about the experimental situation, and assumptions about the ways in which the
putative theoretical states causally influence the observable states of affairs (see Schaffner, 1969;
Suppes, 1962, 1967; and more recently, Hempel, 1970, 1973).

That theories are connected with observable results only via a mediating link of auxiliary

hypotheses has important implications for the testing of theories. For if theories confront data only in

conjunction with other theories or hypotheses, then if a theory’s prediction is not borne out, the most
one can conclude is that either the theory or one of the auxiliary hypotheses is wrong. As the

nineteenth century philosopher and physicist, Pierre Duhem (1906) wrote:

The physicist can never subject an isolated hypothesis to experimental test

but only a whole group of hypotheses; when the experiment is in
disagreement with is predictions, what he learns is that at least one of the
hypotheses constituting this group is unacceptable and ought to be
modified; but the experiment does not designate which one should be
changed [p. 187].

The point again is that if one takes Popperian falsifiability to require that the theory whose

status is being determined can in itself make falsifiable predictions, then very few legitimate scientific

www.freepsychotherapybooks.org 454
theories, if any, will be falsifiable by themselves. Hence—so the argument goes—falsifiability ought to
be rejected as a demarcation criterion.

Whether a revised version of falsifiability can be formulated that will be serviceable as a

demarcation criterion is still a matter of controversy. Popper (1963, p. 112) himself briefly considers
the matter (see Grünbaum, 1976 for a discussion of this passage). I have made some positive

suggestions in this area (Von Eckardt, 1982) as has Lakatos (1970). In contrast, Laudan (1983) has

recently argued that no satisfactory demarcation principle will be forthcoming, especially not one

formulated along Popperian lines. However this issue is resolved, it is important to keep in mind that
even if falsifiability cannot function as a demarcation principle (which requires it to be both necessary

and sufficient for a theory’s being scientifically entertainable), it may well constitute simply a

necessary condition. In any case, it certainly behooves us to appraise Popper’s rejection of

psychoanalysis as pseudoscience on the alleged ground of unfalsifiability. Thus, if we are interested in

the scientific status of Freudian psychoanalytic theory, it remains a worthwhile project to inquire into

its falsifiability.

GRÜNBAUM’S CONTRIBUTION

Grünbaum has had something to say on virtually every aspect of the issue of the falsifiability of
Freudian theory. What makes his discussion so noteworthy is that it takes place against a background

of serious consideration of the importance and relevance of the requirement of falsifiability in the
scientific enterprise in general. Thus, before we turn to a discussion of his response to the

pseudoscience challenge, I shall briefly summarize his work in philosophy of science that pertains to

falsifiability.

As the previous discussion should have made clear, there are two extreme positions that
someone can take on the importance of falsifiability in science. On the one hand, it can be argued, as

Popper has done, that falsifiability is the “touchstone of scientific rationality.” On the other hand, there
is the view, inspired by Duhem, that falsifiability is completely unimportant in science because no

www.freepsychotherapybooks.org 455
scientific theory is ever, strictly speaking, falsifiable. In a series of important papers in the 1960s and

1970s, Grünbaum took on both of these extreme positions, advocating instead a more reasonable,

middle-of-the-road view.

Grünbaum (1969) considered what he calls the “D-thesis,” a view, that if not historically

attributed to Duhem, represents the Duhemian philosophical legacy in contemporary philosophy of

science. The D-thesis consists of the following two claims.

Dl. No constitutent hypothesis H of a wider theory can ever be sufficiently

isolated from some set or other of auxiliary assumptions so as to be
separately falsifiable observationally. H is here understood to be a
constituent of a wider theory in the sense that no observational
consequence can be deduced from H alone.

It is a corollary of this subthesis that no such hypothesis H ever lends itself

to a crucially falsifying experiment any more than it does to a crucially
verifying one.

D2. In order to state the second subthesis D2, we let T be a theory of any
domain of empirical knowledge, and we let H be any of its component
subhypotheses, while A is the collection of the remainder of its
subhypotheses. Also, we assume that the observationally testable
consequence O entailed by the conjunction H & A is taken to be empirically
false, because the observed findings are taken to have yielded a result O'
incompatible with O. Then D2 asserts the following: For all potential
empirical findings 0' of this kind, there exists at least one suitably revised
set of auxiliary assumptions A' such that the conjunction of H with A' can
be held to be true and explains O'. Thus D2 claims that H can be held to be
true and can be used to explain O' no matter what O' turns ought to be, i.e.,
come what may [p. 1070-1071].

Grünbaum (1966) argued the following three points with respect to the D-thesis: (1) There are

quite trivial senses in which D1 and D2 are uninterestingly true and in which no one would wish to
contest them (see pp. 276-280); (2) In its nontrivial form, D2 has not been demonstrated (see pp. 280-

www.freepsychotherapybooks.org 456
281); and (3) D1 is false, as shown by counterexamples from physical geometry (see pp. 283-295;
Grünbaum, 1968, 1969). Grünbaum (1969) discusses this further and, in response to criticism,
introduces a qualification with respect to the third point (3). Griinbam concedes to Hesse (1968) that if

the falsification of H denied by D1 is construed as irrevocable, then his geometrical example does not
succeed as a counterexample. However, he insists that it does succeed if falsification is construed in a

scientifically realistic sense, that is, if one requires “only falsification to all intents and purposes of the
scientific enterprise” (p. 1092). In sum, then, in Grünbaum’s view, falsifiability is a meaningful notion in

science.

Falsifiability is not, however, the only possible basis for a demarcation principle or the only
possible ground for a theory of scientific rationality, as Popper has claimed. In Grünbaum’s view, the

alternative—inductivism—which Popper summarily dismisses, merits serious consideration as well.

Grünbaum (1976, 1977, 1979) argues that Popper’s rejection of inductivism rests on a serious
misportrayal.

Inductivism offers the following demarcation principle: A theory is scientific (“I-scientific”) if

and only if it qualifies as empirically well supported by neo-Baconian standards of controlled inquiry.
Note that this is quite different from a demarcation principle based on falsifiability, in that the focus is

on the credibility of the theory rather than simply its entertainability. Thus, a speculative theory in
physics, for example, for whom evidence has not yet been gathered, would not count as actually I-
scientific, but only as potentially so, although the latter would qualify it as scientifically entertainable.

Grünbaum’s dispute with Popper concerns what sorts of theories inductivism would count as

being empirically well supported. According to Grünbaum (1977), Popper attributes the following to

inductivism: “If a theory T can explain a sufficiently large number of observational results or has a
suitably large number of so-called positive instance, then T automatically qualifies as well-supported by

the evidence” (p. 224). The distinction between being a positive instance and being a supportive one is

crucial here. According to Grünbaum (1976), “an instance is a “positive’ one with respect to a non-

statistical theory T, if its occurrence or being the case can be deduced from T in conjunction with

www.freepsychotherapybooks.org 457
suitable initial conditions. But an instance is supportive of T, if it is positive and has the probative
significance of conferring a stronger truth presumption on T than T has without that instance” (p. 217).

Thus, Popper would claim that inductivism requires a positive instance to be sufficient for being a

supportive one. This claim overlooks two important features of the inductivist position, in Grünbaum’s
view:

1. The “declared consequence restriction.” Grünbaum (1979) states: “If, at a particular time, S is

declared to be a logical consequence of T under the assumption of stated initial conditions, or is

declared not to be such a consequence, then neither declaration is allowed to depend on knowing at the
time whether S is true” (p. 133). The point is that what counts as a consequence of a theory T (and,

hence, as a positive instance) is a function solely of the logical relations between T and this

consequence; whether it is true or false is completely irrelevant. According to Grünbaum (1977) this
requirement is “at least implicitly imposed by inductivists to preclude ‘retroactive’ tampering with the

construal of T as follows: S is only ex post facto held to have followed from T after having been found to

be true” (p. 227).

2. The need for controls with respect to causal hypotheses. According to Grünbaum, this need has

been emphasized by inductivists ever since Francis Bacon wrote three centuries ago. Consider a causal
hypothesis of the form, “Events of kind X are causally relevant to (either causally necessary for,

causally sufficient for, or stochastically relevant to) events of kind Y.” A merely positive instance for

such a hypothesis will be an event of kind X coupled with an event of kind Y. For example, if the
hypothesis (H) in question is “Ceteris paribus, daily consumption of at least one-fifth pound of coffee for
two weeks [X] is causally sufficient as well as causally necessary for the remission of colds [Y],” then a

positive instance of H would be one case of a person with a cold drinking at least one-fifth pound of
coffee for two weeks and getting rid of his or her cold at the end of that period. Such a positive instance
would not, however, count as supportive, in the inductivist view, unless it is conjoined with findings

from an appropriate control group. For, as Grünbaum (1977) states “Even a large number of cases of X
which are also cases of Y does not preclude that an equally large number of cases of non-X are also

www.freepsychotherapybooks.org 458
cases of Y. But being an X should make a difference with respect to being a Y,” (p. 232) given the claim
of causally sufficiency. In addition, “if there is to be inductive warrant for deeming coffee to be
remedially necessary, every known case of non-X would have to be a case of non-Y” (p. 232). In sum,

Grünbaum concludes, “only the combination of positive instances with instances of non-X and non-Y
could constitute inductively supportive instances of our strong causal hypothesis H” (p. 232).

Note that given this more accurate portrayal of inductivism, any theory containing causal

hypotheses that is I-scientific will necessarily be falsifiable as well, although, of course, the converse
will not be true. Thus, in Grünbaum’s view, Popper was completely wrong in claiming that, in contrast

to falsifiability, inductivism is powerless to impugn the scientific credentials of a theory like

psychoanalysis. In fact, as we shall discuss later, one of Grünbaum’s principal theses is that the

weakness of Freudian theory lies not in its unfalsifiability but in the fact that it fails to satisfy neo-
Baconian standards of inductive credibility.

Let us turn now to Grünbaum’s response to the challenge of unfalsifiability. His principal points

are the following: First, the arguments that have been offered by Popper and others to show that

psychoanalytic theory is unfalsifiable are inadequate. Second, although there is some merit to the
charge that the majority of Freud’s defenders, and even sometimes Freud himself, have exhibited a

“tenacious unwillingness…to accept adverse evidence” (Grünbaum, 1979, p. 138), Cioffi’s global
indictment of Freud’s methodology as pseudoscientific cannot be sustained. And, third, given any

reasonable scientific sense of falsifiable (that is, modulo revocable auxiliary assumptions and initial

conditions), there are clear counterexamples to the thesis of unfalsifiability.

The thesis of unfalsifiability says that there does not exist even one way in which Freudian
theory could, in principle, be falsified. As Grünbaum (1983b) points out, since a negative claim is here

being made about an infinite class of consequences of the theory, it is not even clear what a good
argument for this claim would look like. Certainly, what Popper offers us is not satisfactory. For instead

of providing a general argument to support his general claim, he simply gives us a single alleged

example of how Freudian theory could explain the facts no matter how they turn out. Popper (1963)

www.freepsychotherapybooks.org 459
describes two cases: that of a man who pushes a child into the water with the intention of drowning it;
and that of a man who sacrifices his life in an attempt to save the child. He writes: “Each of these two

cases can be explained with equal ease in Freudian…terms. According to Freud the first man suffered

from repression (say, of some component of his Oedipus complex), while the second man had achieved
sublimation” (p. 35).

As an argument for the unfalsifiability thesis, this example fails miserably, according to

Grünbaum (1979). First, “why would it necessarily be a liability of psychoanalysis, if it actually could

explain the two cases of behavior with equal ease? Presumably there actually are such instances of self-
sacrificing child-rescuing behavior no less than such cases of infanticidal conduct. And a fruitful

psychological theory might well succeed in actually explaining each of them” (pp. 134-135). Second,

even if this case were cogent, it is certainly not clear how it is supposed to generalize to cover the
infinite class of cases which fall under the thesis. Popper seems to be relying on the method of

“induction by enumeration,” which he himself has rejected as inadequate. Third, the example is totally

contrived. Popper should, at least, have chosen an example based on the Freudian text. Finally, Popper
claims that Freudian theory could explain both of these cases. However, such explanations are

forthcoming only if the psychoanalytic theorist is at liberty to posit initial conditions at will. But, asks
Grünbaum (1979): “Is it clear that the postulation of initial conditions ad libitum without any

independent evidence of their fulfillment is quite generally countenanced by that theory to a far greater
extent than in, say, physics, which Popper deems to be a bona fide science?” (p. 135). Certainly, Popper
gives us no argument to that effect. Eysenck (Eysenck & Wilson, 1973) puts forth another similar

argument, which Grünbaum (1979, pp. 138-139) discusses and dismisses as inadequate.

Grünbaum considers Cioffi’s (1970) claim that Freud’s methodology was prompted chiefly by
the need to avoid refutation. After carefully reexamining the textual passages on which Cioffi builds his
case, Grünbaum (1980b) concludes that Cioffi “mishandled” his examination of Freud’s reasoning and

“was thereby driven to the gratuitous or mistaken conclusion that concern with pertinent evidence had
played no essential role in Freud’s rationale for espousing psychoanalysis” (p. 84). Freud was willing to

www.freepsychotherapybooks.org 460
acknowledge both the possibility and, on several occasions, the fact of falsification, according to
Grünbaum. In support of this contention, he cites the following cases:

1. In his “Reply to Criticisms of My Paper on Anxiety Neurosis” Freud (1895) stated explicitly

what sort of finding he would acknowledge to be a refuting instance for his hypothesis
concerning the etiology of anxiety neurosis.

2. In 1897 Freud abandoned his hypothesis that actual episodes of traumatic seduction in

childhood were responsible for the occurrence of hysteria in adulthood. Among the
reasons that he explicitly cites (see Freud, 1954, pp. 215-216) is the fact that the

hypothesis had extremely implausible consequences; in particular, the required

incidence of perverted acts against children would have had to have been preposterously
high (Grünbaum, 1979, p. 135).

3. In 1909 Freud recognized that the best available evidence concerning the actual life history of

his “Rat Man,” Paul Lorenz, refuted his prior hypothesis concerning the etiology of adult
obsessional neurosis (Grünbaum, 1979, p. 137).

4. In “A Case of Paranoia Running Counter to the Psychoanalytic Theory of the Disease” Freud

(1915) considered the case of a young woman who appeared to be paranoid but who
initially failed to give any indication of the underlying homosexual attachment that Freud

had hypothesized to be causally necessary for paranoia. At this point, he reasoned:

“Either the theory must be given up or else, in view of this departure from our

[theoretical] expectations, we must side with the lawyer and assume that this was no
paranoic combination but an actual experience which had been correctly interpreted” (p.

266; Grünbaum, 1983b, p. 155).

5. Freud’s (1933) “Revision of the Theory of Dreams” presents an acknowledged falsification on

the basis of the recurrent dreams of war neurotics.

www.freepsychotherapybooks.org 461
These cases not only suffice to undermine Cioffi’s pseudoscience charge, they also function as

counterexamples to the claim that, from a logical point of view, Freudian theory is unfalsifiable. To

further emphasize the incorrectness of the logical unfalsifiability thesis, Grünbaum mentions a number
of additional cases of either possible or actual (revocable) falsification:

1. In Freud’s theory of personality types, both personality traits and a specific childhood
etiology are associated with each character type. Thus, for example, Freud claims that

the “oral” character is associated with dependency, submissiveness, need for approval,

and pessimism and originates in such unfavorable childhood experiences as premature

weaning. Grünbaum (1979, p. 137) suggests that this coupling of certain personality

traits with certain childhood experiences is at least prima facie falsifiable.

2. Grünbaum (1979, p. 137) notes that experimental work has provided evidence counter to

both Freud’s doctrine of repression (see Holmes, 1974) and his theory of dreams (see

Fisher & Greenberg, 1977).

3. Certain of Freud’s hypotheses entail “statistical” predictions that might be tested. For

example, Grünbaum (1983b) writes, Freud’s hypothesis that repressed homosexuality is

the specific etiologic factor for paranoia entails that

the decline of the taboo on homosexuality in our society should be

accompanied by a decreased incidence of male paranoia. And by the same
token, there ought to have been relatively less paranoia in those ancient
societies in which male homosexuality was condoned or even sanctioned,
for the reduction of massive anxiety and repression with respect to
homosexual feelings would contribute to the removal of Freud’s conditio
sine qua non for this syndrome [p. 157].

PSYCHOANALYSIS AS CRITICAL THEORY

HABERMAS'S READING OF FREUD

www.freepsychotherapybooks.org 462
Like Popper, Habermas wants to hold that psychoanalysis cannot appropriately be regarded as

natural science. However, his attitude toward psychoanalysis is quite different. Contrary to Popper

who, as we have seen, wants to relegate Freudian theory to the epistemological dustbin of
pseudoscience, Habermas seeks to make it an object of profound study. He believes that Freudian

theory and practice represent a prototype (along with Marxian theory and practice) of a completely
new form of knowledge—one he has chosen to designate “critical theory.” Habermas’ interest in Freud

is part of a much larger concern with the nature of knowledge in general. Although I cannot here do

justice to his views, it will be useful for our purposes to attempt a rough characterization of some of his
basic doctrines. First, Habermas assumes there to be three fundamentally different kinds of knowledge

(Wissenschaft): (1) empirical-analytic sciences, of which the natural sciences are the paradigm; (2)
historical-hermeneutic sciences, including the humanities (Geisteswissenschaften) and the historical and
social sciences insofar as they aim at interpretive understanding of their subject matter; and (3)

critically oriented sciences, in which he includes psychoanalysis as well as the critique of ideology
(critical social theory)1. Each kind of knowledge is distinguished, in his view, by both the cognitive
structure of its theories and the mode of “testing” appropriate to it. Note that in taking this position,

Habermas is consciously going counter to one of the principal theses of the logical positivist unity of
science movement, namely, that the logic of inquiry of any science (Wissenschaft) is the same.

Second, Habermas has emphasized the importance of locating knowledge in the course of
human life. According to McCarthy (1978), Habermas’ central thesis is that “the specific view points

from which we apprehend reality,” the “general cognitive strategies” that guide systematic inquiry,

have their “basis in the natural history of the human species” (p. 55). In particular, Habermas believes
that any search for knowledge is guided by certain cognitive interests and that distinct forms of

knowledge are associated with distinct cognitive interests. Thus, Habermas (1971) assumes that each

of the three kinds of knowledge he distinguishes is associated with its own kind of cognitive interest:
“The approach of the empirical-analytic sciences incorporates a technical cognitive interest; that of the

historical-hermeneutic sciences incorporates a practical one; and the approach of critically oriented

sciences incorporates the emancipatory cognitive interest” (p. 308). Roughly speaking, the technical

www.freepsychotherapybooks.org 463
interest is an interest in making use of causal knowledge of nature for the purposes of prediction and

control; the practical interest is an interest in establishing reliable intersubjective understanding in

ordinary language communication; and the emancipatory interest is an interest in freeing oneself from

ideological delusion and establishing social or intrapsychic relations “organized on the basis of

communication free from domination” (McCarthy, 1978, p. 93). Furthermore, Habermas says that the
specific kind of cognitive interest associated with a specific kind of theory shapes the cognitive

structure of that theory to a large extent.

Much of Habermas’ intellectual effort over the past 15 years has been devoted to elucidating

and arguing for the existence of the third category of knowledge, critical theory. His first attempt to

articulate the logic, methodology, and structure of a critical theory were published in 1967 and 1971.
His more recent views on the topic are to be found in Communication and the Evolution of Society

(1979). But it is the earlier Knowledge and Human Interests (1971) which is of most concern to us, for it
is here that Habermas’ most extended treatment of Freud is to be found.

Habermas’ (1971) two principal claims about Freudian psychoanalysis are stated in the opening
passage of his discussion of Freud:

The end of the 19th century saw a discipline emerge, primarily as the work
of a single man, that from the beginning moved in the element of self-
reflection and at the same time could credibly claim legitimation as a
scientific procedure in a rigorous sense.…Psychoanalysis is relevant to us
as the only tangible example of a science incorporating methodological
self-reflection. The birth of psychoanalysis opens up the possibility of
arriving at the dimension that positivism closed off.…This possibility has
remained unrealized. For the scientific self-misunderstanding of
psychoanalysis inaugurated by Freud himself, as the physiologist that he
originally was, sealed off this possibility [p. 214].

Habermas’ claim that psychoanalysis involves self-reflection is, as we shall see shortly, essential

to his construing it as a critical theory. “The dimension that positivism closed off” I take to be a

www.freepsychotherapybooks.org 464
reference to the possibility of a science existing (Wissenschaft) that differs in important ways from the

natural sciences. Thus, Habermas is making two claims: (1) psychoanalysis is a “tangible example” of a

critical theory; and (2) this fact has not been recognized because Freud himself was guilty of

perpetuating a misunderstanding of his own enterprise, namely, the mistaken view that what he was
doing was empirical-analytic science rather than critical theory and practice.

Habermas attempts to argue for his first claim by providing us with a description of Freudian
doctrine that makes salient its “critical” features. To understand his reading of Freud, we need to say a

bit more about the aims of a critical theory. We have already noted that, for Habermas, a critical theory

is essentially tied to the emancipatory interest. More specifically, it has as its aim the emancipation of

the agents that make use of it by means of their self-enlightenment. We can gain a clearer picture of
what this emancipation and enlightenment is supposed to come to by viewing it as a transition from an

initial to a final state. Geuss (1981) characterizes these states as follows:

(a) The initial state is one both of false consciousness and error, and of
‘unfree existence.’...

(b) In the initial state false consciousness and unfree existence are
inherently connected so that agents can be liberated from one only if
they are also at the same time freed from the other.…

(c) The “unfree existence’ from which the agents in the initial state suffer is
a form of self-imposed coercion; their false consciousness is a kind of
self-delusion.…

(d) The coercion from which the agents suffer in the initial state is one
whose ‘power’...or ‘objectivity’…derives only from the fact that the
agents do not realize it is self-imposed.

(e) The final state is one in which the agents are free of false consciousness
—they have been enlightened—and free of self-imposed coercion—
they have been emancipated [p. 58].

www.freepsychotherapybooks.org 465
A critical theory is supposed to achieve such enlightenment and emancipation by inducing what

Habermas calls “self-reflection.” It is by reflecting, Geuss (1981) says, that the agents in question “come

to realize that their form of consciousness is ideologically false and that the coercion from which they
suffer is self-imposed. But, by (d) above, once they have realized this, the coercion loses its ‘power’ or

‘objectivity’ and the agents are emancipated” (p. 61).

It is not difficult to see how Freudian psychoanalysis can fit in with Geuss’ schema. The first four

statements constitute a quite straightforward (if abstract) description of certain of the central features

of psychoanalytic therapy. Thus, we find Habermas arguing his thesis “that psychoanalytic knowledge
belongs to the category of self-reflection” by reference to Freud’s papers on analytic technique (see

Habermas, 1971, pp. 228-236). The important point, however, is this: Because of his doctrine of

cognitive interests, Habermas’ view of psychoanalytic therapy as emancipatory self-reflection has

certain consequences for his reading of the psychoanalytic theory of personality. That is, because he, in
effect, subordinates the theory to the therapy, he ends up representing Freud’s theoretical claims in a

certain idiosyncratic way. It is not only this idiosyncratic reading of Freudian theory but also his

fundamentally mistaken views about the nature of (natural) science that become the target of
Grünbaum’s criticisms.

THE HABERMAS-GRÜNBAUM DISPUTE

Like most interpreters of Freud, Habermas divides Freud’s theoretical claims into two parts: the
metapsychology and the clinical theory. As I read him (which is not always a straightforward matter),

in arguing that Freud was guilty of misunderstanding his own enterprise, Habermas provides us with
two sets of arguments to the effect that Freudian psychoanalysis cannot correctly be regarded as an
empirical-analytic science. The first of these considers the relationship of the clinical theory to the

metapsychology; the second considers the scientific characters of the clinical theory itself.

Habermas begins by arguing that Freud took psychoanalysis to be scientific because

psychoanalytic assumptions could be “reformulate[d]…in the categorical framework of a strictly

www.freepsychotherapybooks.org 466
empirical science” (p. 252), namely, the energy model of the metapsychology. That is, he attributes to

Freud two beliefs: first, that the clinical theory could be “reduced” to the metapsychology, and, second,

that the metapsychology was a “strictly empirical science.” With respect to the second point, Habermas

(1971) writes: “Freud surely assumed tacitly that his metapsychology, which severs the structural
model from the basis of communication between doctor and patient and instead attaches it to the

energy-distribution model by means of definitions, represented an empirically rigorous scientific

formulation of this sort” (p. 253). However, in Habermas’ view, Freud “erred” in adopting this

reductionistic approach, because “psychology, insofar as it understands itself as a strict empirical

science, cannot content itself with a model that keeps to a physicalistic use of language without
seriously leading to operationalizable assumptions” (p. 253). That is, the metapsychology is not

genuinely scientific unless its underlying energy model is operationalizable. But, Habermas continues,

this is not the case:

The energy-distribution model only creates the semblance that

psychoanalytic statements are about measurable transformations of
energy. Not a single statement about quantitative relations derived from
the conception of instinctual economics has ever been tested
experimentally. The model of the psychic apparatus is so constructed that
metapsychological statements imply the observability of the events they
are about. But these events are never observed—nor can they be observed
[p. 253; italics added].

Grünbaum’s first point against Habermas effectively undercuts this whole line of

argumentation. For, according to Grünbaum (1984), careful examination of the Freudian text (Freud,

1914, p. 77; 1915a, p. 117, 1925, p. 32) reveals clearly

that when Freud unswervingly claimed natural science status for his theoretical constructions
throughout his life, he did so first and foremost for his evolving clinical theory of personality and

therapy, rather than for the metapsychology. For he had been chastened in his early reductionistic

exuberance by the speedy demise of his Project. And, once he had repudiated his ephemeral

neurobiological model of the psyche after 1896, he perenially saw himself entitled to proclaim the

www.freepsychotherapybooks.org 467
scientificity of his clinical theory entirely on the strength of a secure and direct epistemic warrant from
the observations he made on his patients and on himself. In brief, during all but the first few years of his

career Freud’s criterion of scientificity was methodological and not ontologically reductive, (p. 6)

The consequence of Grünbaum’s exegetical position here is that he simply passes over
Habermas’ first set of arguments, presumably on the grounds that they are simply irrelevant to the

issue at hand. Implicitly, Grünbaum’s reasoning seems to be something like this: When Freud claimed

that psychoanalysis was scientific, what he chiefly had in mind was that the clinical theory was

scientific. And since the status of the clinical theory does not depend in any essential way on the status
of the metapsychology, any argument that assumes that the scientificity of the clinical theory depends

on that of the metapsychology is irrelevant to the question of whether the theory in general is

scientific. Grünbaum therefore turns his attention to Habermas’ second set of arguments.

To be in a position to understand this second set of arguments and to appreciate Grünbaum’s

replies, we must briefly consider Habermas’ conception of the clinical theory. The standard reading of

Freud is that the clinical theory consists of a large number of universal generalizations about the
human psyche. Habermas’ view is somewhat different. Rather than viewing the theory of psychosexual

development, say as a set of universal claims about the ontogenesis of human personality, Habermas
(1971) takes it to consist of a set of narrative schemata. He writes:

A general interpretation…has the form of a narrative, because it is to aid

subjects in reconstructing their own life history in narrative form. But it
can serve as the background of many such narrations only because it does
not hold merely for an individual case. It is a systematically generalized
history, because it provides a scheme for many histories with foreseeable
alternative courses [p. 263],

Furthermore, in keeping with his (misplaced) emphasis on the centrality of the therapy to the
psychoanalytic enterprise as a whole, Habermas takes the primary function of Freud’s general
interpretations to be their role in self-reflection. For it is by the application of such general

www.freepsychotherapybooks.org 468
interpretations to the individual case that patient and physician together create the interpretative

constructions, by means of which the self-reflective process takes place. Habermas (1971) states:

Only the…systematically generalized history of infantile development with

its typical developmental variants puts the physician in the position of so
combining the fragmentary information obtained in analytic dialogue that
he can reconstruct the gaps of memory and hypothetically anticipate the
experience of reflection of which the patient is at first incapable [p. 260].

On the basis of this rather one-sided conception of Freud’s clinical theory, Habermas offers us a

number of arguments that the clinical theory ought not to be regarded as science of the empirical-
analytic sort. I label these “the argument from therapeutic application,” “the argument from

explanation,” and “the argument from validation.”

1. The Argument from Therapeutic Application. I pointed out earlier that, in Habermas’ view,
empirical-analytic theories are always associated with a technical interest in manipulating nature. The
argument from therapeutic application relies heavily on the further assumption that such

manipulation always occurs by means of the exploitation of causal laws. We can reconstruct the
argument as follows:

1. If psychoanalytic theory were scientific (empirical-analytic), its application would consist in

the manipulation of its domain by the exploitation of causal laws.

2. The application of psychoanalytic theory consists in the doing of psychoanalytic therapy.

3. However, psychoanalytic therapy does not work by the exploitation of causal laws; rather “it

owes its efficacy to overcoming causal connections themselves” (Habermas, 1971, p.

271).

4. Thus, psychoanalytic theory cannot be scientific.

www.freepsychotherapybooks.org 469
Habermas (1971) defends the key third premise as follows:

Psychoanalysis does not grant us a power of technical control over the sick
psyche comparable to that of biochemistry over a sick organism. And yet it
achieves more than a mere treatment of symptoms, because it certainly
does grasp causal connections, although not at the level of physical events
—at a point “which has been made accessible to us by some very
remarkable circumstances” [Freud, 1971, p. 436], This is precisely the
point where language and behavior are pathologically deformed by the
causality of split-off symbols and repressed motives. Following Hegel we
can call this the causality of fate, in contrast to the causality of nature. For
the causal connection between the original scene, defense, and symptom is
not anchored in the invariance of nature according to natural laws but only
in the spontaneously generated invariance of life history, represented by
the repetition compulsion, which can nevertheless be dissolved by the
power of reflection [p. 271].

Habermas’ point seems to be that the power of reflection can “overcome” the causal
connections responsible for the patient’s neurosis, because these causal connections are of a different

sort than those posited by the empirical-analytic sciences. They constitute the “causality of fate” rather

than the “causality of nature.” What Habermas has in mind by this term is far from clear, although I

suspect that it is, in some way, a consequence of his reading of the clinical theory as consisting of
narrative schemata. Whatever it is, however, it is irrelevant. For, as Grünbaum argues, the kind of

causality avowed by psychoanalytic etiologic and therapeutic theory does not permit this kind of

“dissolution.” In addition, careful examination of the causal assertions made by the theory exhibits the
complete folly of this sort of talk of dissolution. In other words, Habermas has a case only by blatantly
misconceptualizing psychoanalytic theory.

To be more precise, Habermas’ account, in Grünbaum’s view, “flatly repudiates the

psychoanalytic explanation for the patient’s therapeutic transition from unconsciously driven behavior

to more consciously governed conduct” (Grünbaum, 1984, p. 10). This psychoanalytic explanation, first

articulated in Breuer and Freud’s (1893, pp. 6-7) “Preliminary Communication,” rests on the etiological

www.freepsychotherapybooks.org 470
principle that repression is causally necessary not only for the initial development of a neurotic
disorder, but also for its maintenance. The explanation of why therapy is efficacious then is as follows:

1. Repression of type R is the causal sine qua non of a neurosis of kind N.

2. Therapy largely consists of ridding the patient of R.

3. Therefore, therapy has the effect of obliterating N.

Grünbaum points out that, in this explanation, therapy involves the instantiation or

exemplification of the etiologic causal relationship rather than its dissolution. For it is precisely because

after the fulfillment of the second condition the patient no longer satisfies the sine qua non state that
the symptoms are claimed (predicted) to disappear. Paradoxically, Habermas appears to accept both

the etiological principle and the explanation; thus, he is guilty not only of contradicting the
foundational postulate of Freudian theory but also of confusing the dissolution of the neurosis with the
dissolution of its causal link to its original pathogen.

To further bring home his objection, Grünbaum (1984, p. 14) offers us a reductio ad absurdum

argument to show that if Habermas’ reasoning were legitimate, then thermal elongation in physics
could also be shown to rest on the dissolution rather than the instantiation of a causal law:

For consider a metal bar that is isolated against all but thermal influences.
It is subject to the law ∆L = α∆T·L0, where L0 is its length at the fixed
standard temperature, ∆T the length increase or decrease due to this
temperature change, and a the coefficient of linear thermal expansion
characteristic of the particular material composing the metal bar. Now
suppose that the bar, initially at the standard temperature, is subjected to
a “pathogenic” temperature increase ∆T, which produces the elongation
∆T as its “pathological” effect. In addition to supplying this “aetiology,” the
law of linear thermal elongation also provides a basis for a corresponding
“therapy”: It tells us that if the bar’s temperature is reduced to its “healthy”
standard value, the “pathological” effect DL will be wiped out. Thus, we can

www.freepsychotherapybooks.org 471
correlate the “therapeutic intervention” of temperature reduction with the
patient’s remedial lifting of his own repressions. Similarly, we correlate
the bar’s “neurotic symptom” ∆L with the patient’s repetition compulsion.

By parity with Habermas’ reasoning, we could then draw the following

ludicrous conclusion: When the temperature reduction “therapeutically”
wiped out the endurance of the “pathological” effect ∆L generated by the
“pathogenic” temperature increase, this thermal termination also
“dissolved” the stated law of thermal elongation.

What is overcome here is clearly the “pathological” effect, not the causal connection itself. And
the same is true, according to Grünbaum, in the psychoanalytic case (that is, assuming the Freudian

story is correct, as Habermas does). In sum, Habermas’ claim that psychoanalytic therapy owes its
efficacy to “overcoming causal connections” rather than “making use” of them is totally
unsubstantiated.

2. The Argument from Explanation. Habermas’ (1971) second argument concerns the kind of

explanation that results from the application of Freud’s clinical theory to a specific case:

In its logical form…explanatory understanding differs in one decisive way

from explanation rigorously formulated in terms of the empirical sciences.
Both of them have recourse to causal statements that can be derived from
universal propositions by means of supplementary conditions: that is,
from derivative interpretations (conditional variants) or lawlike
hypotheses. Now the content of theoretical propositions remains
unaffected by operational application to reality. In this ease we can base
explanations on context-free laws. In the case of hermeneutic application,
however, theoretical propositions are translated into the narrative
presentation of an individual history in such a way that a causal statement
does not come into being without this context....Narrative explanations
differ from strictly deductive ones in that the events or states of which
they assert a causal relation is [sic] further defined by their application.
Therefore general interpretations do not make possible context-free
explanations [pp. 272-273].

www.freepsychotherapybooks.org 472
The passage is somewhat confusing because Habermas uses the term ‘theoretical propositions’

in both a narrow and broad sense. I assume that the first reference to such theoretical propositions is

meant to refer to the theoretical propositions of empirical-analytic science, whereas the second
reference includes also those that can have a “hermeneutic application.” Given this reading, the basic

structure of the argument seems to be the following:

1. The explanation of a particular phenomenon by means of the causal laws of an empirical-

analytic science always results in a “context-free” explanation.

2. However, this is not the case for the application of the general interpretations of
psychoanalytic theory; “general interpretations do not make possible context-free

explanations.”

3. Therefore, these general interpretations cannot be part of an empirical-analytic science.

Recall that, in responding to Habermas’ first argument, Grünbaum took issue with Habermas’
grasp of Freudian theory, in particular, his failure to see that the therapeutic conquest of a neurosis

instantiates rather than dissolves its etiologic linkage to its pathogen. In this case, he objects that
Habermas relies on a false view of natural science. In particular, Grünbaum offers an array of

counterexamples from physics to the first premise of our reconstruction of Habermas’ argument. In
Grünbaum’s view, Habermas is simply wrong that explanations in the natural science are never context
dependent; thus, this cannot be used as a reason for distinguishing the Freudian enterprise from that

of natural science. In arguing his point, Grünbaum (1984) again draws on his knowledge of physics,
specifically, the physical theory of classical electrodynamics. He writes: “For that major physical theory
features laws that embody a far more fundamental dependence on the history and/or context of the

object of knowledge than was ever contemplated in even the most exhaustive of psychoanalytic
explanatory narratives…” (p. 17; for a briefer version of this argument, see Grünbaum, 1983c).
Grünbaum’s (1984) specific counterexample is the following:

www.freepsychotherapybooks.org 473
Consider an electrically charged particle having an arbitrary velocity and
acceleration. We are concerned with the laws governing the electric and
magnetic fields produced by this point charge throughout space at any one
fixed time t. In this theory, the influence of the charge on any other test
charge in space is postulated to be propagated with the finite velocity of
light rather than instantaneously, as in Newton’s action-at-a-distance
theory of gravitation. But this non-instantaneous feature of the
propagation of the electrodynamic influence contributes to an important
consequence as follows: At any space point P, the electric and magnetic
fields at a given time t depend on the position, velocity and acceleration
that the charge had at an earlier time t0. That earlier time has the value t
−r/c, where r is the distance traversed by the influence arriving at P at
time t after having traveled from the charge to P with the velocity c of
light.

Clearly, the greater the distance r that was traversed by the influence by
the time t of its arrival at point P, the earlier its origination time t0. Thus,
for space points at ever larger such distances r in infinite space, the
origination time t0 = t−r/c will be ever more remotely past. In short, as the
distance r becomes infinitely large, the origination time goes to past
infinity.

It follows that at ANY ONE INSTANT t the electric and magnetic fields
produced throughout infinite space by a charge moving with arbitrary
acceleration depend on its own PARTICULAR ENTIRE INFINITE PAST
KINEMATIC HISTORY! (p. 17).

This is not at all a unique case, according to Grünbaum. There are other cases that exhibit
“hysteresis” in the sense that “a property of a physical system induced by a given present influence

upon it depends not only on that present influence, but also on the past history of variation of that
influence” (Grünbaum, 1984, p. 18; see also 1983c for a briefer discussion). These cases include the

hysteresis behavior of highly magnetizable metals (e.g., iron, cobalt, nickel, etc.), the elastic hysteresis

of certain solids, the electric hysteresis exhibited by dielectric substances in electric fields, and the
hysteresis of a radiation counter tube. Even rubber bands exhibit like behavior, and metal fatigue in

www.freepsychotherapybooks.org 474
airplanes is a similar phenomenon. These cases clearly show, in Grünbaum’s view, that some of the
important laws of nature, and, hence, any explanation that makes use of them, exhibit context

dependence. On the basis of these considerations, Grünbaum’s (1983c) summary judgment of

Habermas’ second argument is a harsh one: Habermas (as well as Gadamer (1975) who echoes
Habermas’ view) have simply succeeded in “parlay[ing] the severe limitations of their own personal

scientific horizons into a pseudo-contrast between the humanistic disciplines and the natural sciences”

(p. 11).

3. The Argument from Validation. What I call “the argument from validation” consists of two
subarguments—one concerning supposed differences between how psychoanalytic theory and

empirical-analytic theories are confirmed; the other concerning how they are disconfirmed. The first

subargument rests on the fact that, according to Habermas, there is the following “specific difference”
between empirical-analytic theories and the general interpretations of psychoanalysis (that is, Freud’s

clinical theory):

In the case of testing theories through observation…the application of

assumptions to reality is a matter for the inquiring subject. In the case of
testing general interpretations through self-reflection...this application
becomes self-application by the object of inquiry, who participates in the
process of inquiry. The process of inquiry can lead to valid information
only via a transformation in the patient’s self-inquiry. When valid, general
interpretations hold for the inquiring subject and all who can adopt its
position only to the degree that those who are made the object of
individual interpretations know and recognize themselves in these
interpretations. The subject cannot obtain knowledge of the object unless
it becomes knowledge for the object—and unless the latter thereby
emancipates itself by becoming a subject [pp. 261-262].

I take it that Habermas is here assuming that the general interpretative schemata of the clinical
theory are confirmed only to the extent that they are inductively supported by valid individual

constructions. The claim, then, is that the latter are confirmed, in turn, only if they become a part of the

self-reflection of the analysand. That is, the analysand has, as Grünbaum puts it, complete “epistemic

www.freepsychotherapybooks.org 475
privilege” with respect to these constructions, even as against the analyst him- or herself. In contrast,
according to Habermas, the objects of standard empirical-analytic inquiry do not have this kind of

epistemic privilege. Here confirmation occurs on the basis of observations of the object by the scientist

(the so-called subject of inquiry). We can reconstruct the argument thus:

1. Statements relevant to the confirmation of clinical psychoanalytic theory (for example,

individual constructions) can be accepted by the researcher only if they have first been

accepted as valid by the subject.

2. No such requirement holds for statements relevant to the confirmation of empirical-analytic

theories, which are typically accepted on the basis of observation by the researcher.

3. Therefore, clinical psychoanalytic theory is not an empirical-analytic theory.

In replying to this argument, Grünbaum again attacks Habermas’ conception of psychoanalysis,

this time on the grounds that the thesis of privileged epistemic access expressed in Habermas’ first

premise is ill-founded. His first point is that the only argument Habermas supplies for his first premise

is a question-begging one. For in the above quote, Habermas construes the “otherwise innocuous
phrase ‘testing through self-reflection’ so as to stipulate that only the patient’s own appraisal can carry

out the application of general interpretations to his particular life situation.” (Grünbaum, 1984, p. 23)

Second, the epistemic privilege that Habermas assigns to the analysand does not accord with Freud’s
own views concerning when an individual construction ought to be regarded as true. In particular,

Freud (1937) explicitly rejects recollection by the patient as essential.

Quite often we do not succeed in bringing the patient to recollect what has
been repressed. Instead of that if the analysis is carried out correctly, we
produce in [the patient] an assured conviction of the truth of the
construction which achieves the same therapeutic result as a recaptured
memory [pp. 265-266].

www.freepsychotherapybooks.org 476
Habermas might reply at this point that perhaps he was wrong about the need for recollection;

however, this quote from Freud shows that the patient’s conviction is necessary, which is enough to

maintain some form of an epistemic privilege doctrine. This reply is inadequate, however. For, as
Grünbaum also points out, in Freud’s paper on “Constructions in Analysis” (1937), he argues (from the

confluence of clinical induction) that the analyst could justify an individual construction on the basis of
the totality of the patient’s productions, even in the face of the patient’s denial. (See also Freud, 1920,

on the treatment of a young lesbian as a case in point.) Finally, Grünbaum points out that Habermas’

attribution of epistemic privilege to the analysand has also been impugned by the contemporary
psychoanalysts Thomä and Kächele (1973, pp. 315-316) and Eagle (1973) as being untrue to the

psychoanalytic situation (treatment setting).

One might expect the subargument from disconfirmation to run exactly parallel to that from

confirmation. That is, one might expect Habermas to argue that the difference between the
disconfirmation of psychoanalytic theory and empirical-analytic theory is that the former relies on the

failure of self-reflection whereas the latter relies on the failure of observable prediction. But this is not

the case for the following reason: Although Habermas regards the acceptance of a construction C by the
analysand during self-reflection to be sufficient for the correctness of C, the absence of self-reflection in

the face of C does not falsify it. The patient’s resistances might simply be too strong. Thus, Habermas

focuses instead on the logic of disconfirmation in the two cases and claims that there is a fundamental
contrast between them on the purported grounds that an unsuccessful prediction in the natural

sciences automatically refutes the hypothesis used to make it. In fact, Habermas (1971) takes the

existence of an alternative to disconfirmation in the face of apparently disconfirming evidence to be the

distinguishing feature of the psychoanalytic case. He argues as follows:

General interpretations do not obey the same criteria of refutation as

general theories. If a conditional prediction deduced from a lawlike
hypothesis and initial conditions is falsified, then the hypothesis may be
considered refuted. A general interpretation can be tested analogously if
we derive a construction from one of its implications and the
communications of the patient. We can give this construction the form of a

www.freepsychotherapybooks.org 477
conditional prediction. If it is correct, the patient will be moved to produce
certain memories, reflect on a specific portion of forgotten life history, and
overcome disturbances of both communication and behavior. But here the
method of falsification is not the same as for general theories. For if the
patient rejects a construction, the interpretation from which it has been
derived cannot yet be considered refuted at all....[T]here is still an
alternative: either the interpretation is false (that is, the theory or its
application to a given case) or, to the contrary, the resistances, which have
been correctly diagnosed, are too strong [p. 266].

But, as Grünbaum points out, it has become a commonplace of the philosophy of science, ever
since Pierre Duhem’s work before World War I, that precisely the same ambiguity of refutation holds

for science in general. By and large, it is not theories alone that are at issue in prediction but theories in
conjunction with a statement of initial conditions and various collateral hypotheses. This means that if

a prediction is not borne out, the blame cannot be pinned on the theory with certainty. Thus, again, the

alleged difference between psychoanalysis and empirical-analytic science rests on a false view of the
latter.

IS FREUDIAN PSYCHOANALYSIS GOOD SCIENCE?

In considering the merits of psychoanalytic theory as a scientific theory, Grünbaum has been

concerned with the extent to which Freud’s theoretical claims are supported by the available evidence.
He has focused, in particular, on the sort of evidence that Freud invoked, namely, evidence obtained

“from the couch.” In making his assessment, Grünbaum has relied both on logical considerations and

on various canons of inductive support that have become standard since the time of Bacon. He makes

three basic claims:

1. The therapeutic effectiveness of the characteristic constituent factors of Freudian

psychoanalytic therapy is in serious question.

2. Clinical data are subject to so many epistemological liabilities as to render them virtually

www.freepsychotherapybooks.org 478
useless in supporting the cardinal hypotheses as Freudian theory.

3. Even if clinical data were not epistemologically contaminated and could be taken at face
value, they would fail to sustain any of the central postulates of Freud’s clinical theory as
well as the investigative utility of the method of free association.

Let us consider each of these claims in turn.

THE QUESTION OF THERAPEUTIC EFFECTIVENESS

The effectiveness of Freudian therapy has been under attack at least since Eysenck (1952,

1966) published his classic challenge. Contending that available evidence does not adequately support

the claim that psychoanalysis is therapeutically effective, Eysenck claimed to have telling evidence that
psychoanalysis did no better than simply having people go on about their lives without therapy. Erwin

(1980) has reconstructed Eysenck’s argument as follows:

1. If there is no adequate study of psychoanalytic therapy showing an improvement rate of

better than two thirds or better than that of a suitable no-treatment control group, then

there is no firm evidence that the therapy is therapeutically effective.

2. There is no adequate study showing either rate of improvement.

3. Therefore, there is no firm evidence that the therapy is therapeutically effective.

Originally, Eysenck made use of an overall spontaneous remission rate across all varieties of
neurotic disorder. In response to criticism, however, Eysenck (1977) has recently emphasized that

different types of neurotic disorder have different incidences and/or time courses of spontaneous
remission. He now claims that any comparative evaluation must focus on a particular diagnostic

grouping and a diagnostically matched untreated control group.

www.freepsychotherapybooks.org 479
In the light of much subsequent literature, Grünbaum proceeds on the assumption that the

superiority of the outcome of analytic treatment over that of rival treatment modalities has not been

demonstrated. However, in his essay, “How Scientific is Psychoanalysis?” (Grünbaum, 1977), he

stresses the following additional fact which is frequently overlooked: If psychoanalytic treatment

outcomes do exceed the spontaneous remission rate, this alone does not suffice to establish that
psychoanalytic treatment gains are due to mediation of analytic insight. It would not rule out an

important rival hypothesis, namely, that such treatment gains are due to an inadvertent placebo effect.

In defining this term, Grünbaum (1981, 1983a) notes that of the various constituent factors that make
up a treatment process, we can distinguish those that are characteristic, that is, claimed by the theory

to be remedial, from others it regards as incidental. Grünbaum (1980) continues:

A treatment process t characterized by having constituents F, will be said

to be an inadvertant placebo with respect to target disorder D and
dispensing physician P just in case each of the following conditions is
satisfied: (a) none of the characteristic treatment factors F are remedial for
D, but (b) P credits these very factors F with being therapeutic for D and
indeed he deems at least some of them to be causally essential to the
remedial efficacy of t, and (c) the patient believes that t derives remedial
efficacy for D from constituents belonging to t’s characteristic factors [p.
330].

The point is that in assessing the effectiveness of psychoanalytic therapy or of any of its rivals,
“one must try to disentangle from one another (i) the effects, if any, indeed due to those factors that the

relevant therapeutic theory postulates as being genuinely remedial, and (ii) purportedly lesser changes
due to the expectations aroused in both patients and physicians by their belief in the therapeuticity of
the treatment” (Grünbaum, 1977, p. 238). As Grünbaum reads the relevant literature on treatment

effectiveness (in particular, Fisher & Greenberg, 1977; Luborsky, Singer, & Luborsky, 1975; Meltzoff &
Komreich, 1970; Sloan et al., 1975), there is good reason to suspect that insofar as Freudian therapy is

effective, it is, in fact, “placebogenic.” The studies seem to point to two conclusions; (1) psychotherapy

of a wide variety of types and for a broad range of disorders is better than nothing, but (2) there is
either no difference between different treatment modalities or the behavioral treatment is better.

www.freepsychotherapybooks.org 480
EPISTEMOLOGICAL LIABILITIES OF CLINICAL DATA

Eysenck (1963) not only impugned the effectiveness of Freudian therapy, he also raised serious
questions about the epistemic validity of clinical data as had Wilhelm Fliess (see Freud, 1954) before

him. In contrast, Freud himself, as well as most of his advocates (see Luborsky & Spence, 1978, for a
recent statement) have regarded clinical evidence as the basis for the claims of psychoanalytic theory
to truth.

In considering how clinical material is supposed to bear evidentially on Freudian theory, it is

important to distinguish three levels of clinical material. At the lowest level, we have what we can call

the patient’s productions. These include their dream reports, slips of the tongue, memory reports, and
free associations as well as assents or dissents to interpretations offered by the analyst. In addition, we
have facts concerning the presence or absence of behaviors or bodily states that are regarded as

symptoms. At the second level, we have the interpretations provided either by the analyst or by
patients themselves of these productions and symptoms as expressions of unconscious wishes,

resistance, and so forth. Finally, we have what Freud (1937) later called a construction, a whole

psychoanalytic story about the patient’s psyche from the patient’s early infantile history to the present
state, including, of course, an etiological account of the symptoms. Although it is possible to maintain
that the patient’s productions bear directly on Freud’s universal theoretical claims, a more plausible

epistemological reconstruction is roughly as follows: Most productions, such as dream reports, slips of
the tongue, free associations, and expressions of feeling toward the analyst during transference are

taken to be relevant insofar as they provide the raw material for interpretations, which, in turn,

provide the building blocks for the ultimate construction. Some productions may also be taken to attest
to therapeutic success. In contrast, others, such as the patient’s assent to or protest against a proposed

construction, are often taken as direct evidence for the truth of that construction. The constructions

themselves, clearly, are supposed to bear on the theory in the way that a particular instantiation of a

universal claim bears on the universal claim.

The principal epistemological liability to which clinical data are subject is that the analyst, who

www.freepsychotherapybooks.org 481
presumably is committed to the truth of Freudian theory, unwittingly influences both patients’
productions and the course of the analysis. This point has been recognized for some time (by Fliess, as
is clear from Freud, 1954; as well as Christiansen, 1964; Glover, 1952; Martin, 1964a; Nagel, 1959).

What appears not to have been recognized, as Grünbaum (1983b) points out, is that Freud himself was
aware of this problem and, in addition, had a very sophisticated, albeit unsuccessful, strategy for

dealing with it. Freud (1917) acknowledges the so-called problem of suggestion in his Introductory

Lectures:

It must dawn on us that in our technique we have abandoned hypnosis

only to rediscover suggestion in the shape of transference.

But here I will pause, and let you have a word; for I see an objection boiling
up in you so fiercely that it would make you incapable of listening if it were
not put into words: “Ah! so you’ve admitted it at last! You work with the
help of suggestion, just like the hypnotists! That is what we’ve thought for
a long time. But, if so, why the roundabout road by way of memories of the
past, discovering the unconscious, interpreting and translating back
distortions—this immense expenditure of labour, time and money—when
the one effective thing is after all only suggestion? Why do you not make
direct suggestions against the symptoms, as the others do—the honest
hypnotists? Moreover, if you try to excuse yourself for your long detour on
the ground that you have made a number of important psychological
discoveries which are hidden by direct suggestion—what about the
certainty of these discoveries now? Are not they a result of suggestion too,
of unintentional suggestion? Is it not possible that you are forcing on the
patient what you want and what seems to you correct in this field as well?”
[pp. 446-447].

By this time in his career, Freud had clearly recognized the importance of transference as a
motive force in therapy. Thus, the challenge was that, as Freud (1917) so nicely put it, “what is

advantageous to our therapy is damaging to our researches” for “the influencing of our patient may
make the objective certainty of our findings doubtful” (p. 452). His reply was as follows:

www.freepsychotherapybooks.org 482
Anyone who has himself carried out psycho-analyses will have been able
to convince himself on countless occasions that it is impossible to make
suggestions to a patient in that way. The doctor has no difficulty, of course,
in making him a supporter of some particular theory and in thus making
him share some possible error of his own. In this respect the patient is
behaving like anyone else—like a pupil—but this only affects his
intelligence, not his illness. After all, his conflicts will only be successfully
solved and his resistance overcome if the anticipatory ideas he is given tally
with what is real in him [italics added]. Whatever in the doctor’s
conjectures is inaccurate drops out in the course of the analysis; it has to
be withdrawn and replaced by something more correct (p. 452).

Grünbaum has dubbed the underlined statement the “necessary condition thesis,” NCT for

short. (Elsewhere, Grünbaum, 1983c, calls it—more honorifically—“Freud’s master proposition”.) This

assertion plays the key role in Freud’s attempted solution to the problem of suggestion. What he is
claiming, according to Grünbaum (1983c), is tantamount to the following: “(1) only the psychoanalytic

method of interpretation and treatment can yield or mediate to the patient correct insight into the

unconscious pathogens of his psychoneurosis, and (2) the analysand’s correct insight into the etiology
of his affliction and into the unconscious dynamics of his character is, in turn, causally necessary for the

therapeutic conquest of this neurosis” (p. 184). NCT can then be used to vindicate the validity of the

clinical data furnished by patients in analysis by means of what Grünbaum dubs the “tally argument”
(referring to Freud’s assumption that ideas given patients tally with what is real in them). The

argument runs as follows:

1. The analysis of patient P was therapeutically successful.

2. NCT.

3. Therefore, the psychoanalytic interpretations of the hidden causes of Ps behavior given to

him by his analyst are indeed correct.

Freud’s strategy was brilliant, according to Grünbaum. But was it successful? It should be clear

www.freepsychotherapybooks.org 483
from our discussion of the therapeutic efficacy question, that Grünbaum does not think so. For,
although the tally argument is logically valid, there is a serious question concerning the truth of its
premises, in particular, the crucial NCT. NCT claims that therapeutic success is mediated only by

psychoanalytic insight. Insofar as there is either spontaneous remission of symptoms or there exist
rival successful treatment modalities, NCT is false. As Grünbaum (1980a) argues—after extensive

review of the relevant literature—there appears to be strong evidence for both. (Interestingly enough,
Freud himself explicitly conceded the existence of spontaneous remission [Grünbaum, 1983c]).

Grünbaum (1983b) concludes: “Since the Tally Argument is thus gravely undercut, any therapeutic

successes scored by analysts, even if spectacular, have become probatively unavailing to the validation
of psychoanalytic theory via that argument” (p. 208).

Grünbaum (1980a) considers one possible alternative to the use of the tally argument. This is to

make use of a patient’s introspections once he or she has been successfully analyzed. It might be thought
that, if reliable, such introspections could provide the needed validation for two sorts of claims: (1)

claims concerning the etiology of the patient’s affliction, and (2) claims concerning the necessary role

of the analyst’s constructions in the therapeutic process. The validation of such claims could, in turn,

provide direct evidence for Freud’s psychogenetic theory as well as help to discredit the rival
therapeutic hypothesis of placebogenesis. Unfortunately, however, these “hopeful speculations” are
“fundamentally impugned” in Grünbaum’s view by the findings reported by Nisbett and Wilson (1977)
on the extent to which we have introspective access to the dynamics of our mental life. Nisbett and

Wilson do not apply the results of their findings to the case of psychoanalysis. Grünbaum (1980a)
believes, however, that they are directly relevant and that “they marshal telling empirical support” for

the following conclusions:

1. Far from justifying the prevalent belief in privileged access to the

dynamics of our psychic responses, the findings strongly indicate the
following: Purportedly introspective self-perception of causal
connections between one’s own mental states is just as liable to
theory-induced errors as is drawing causal inferences about
connections between purely external events from apparent

www.freepsychotherapybooks.org 484
covariations among their properties.…

2. When asked how, if at all, a particular stimulus influenced a given

response, the persons in the experimental studies, and ordinary
people in their daily lives did not and do not even attempt to
interrogate their memories of the mediating causal process.
Although it may feel like introspection, what they actually do is draw
on the causal theories provided by their culture or pertinent
intellectual subculture for a verdict as to the effect, if any, of that kind
of stimulus on that kind of response....

3. As N & W remark: “Subjective reports about higher mental processes are

sometimes correct, but even the instances of correct report are not
due to direct introspective awareness. Instead, they are due to the
incidentally correct employment of a priori causal theories” [Nisbett
& Wilson, 1977, p. 233] [p. 363-364].

(See Rothstein, 1980, for some criticisms of Grünbaum’s discussion of the epistemological

liabilities of patient introspection and Grünbaum, 1981, for a reply.)

Grünbaum’s point, then, is that neither the tally argument nor the use of patients’ introspective

judgments subsequent to successful analysis can be used to guard against the very real possibility that

both patients’ productions and therapeutic outcomes are due more to the suggestive influence of the

analyst than to the causal mechanisms and states of affairs posited by Freudian theory. Grünbaum
(1983b) considers the suggestion hypothesis to be more than a mere logical possibility. He discusses in

detail three of the major kinds of clinical findings that Freud deemed either initially exempt from

contamination or, at least, unmarred when gathered with proper precautions. These are the products
of “free” association, the patient’s assent to analytic interpretations that were initially resisted, and

memories recovered from early life. Grünbaum finds “solid” evidence in the psychological literature
that each of these instances is subject to “considerable epistemic contamination.” Grünbaum (1983b)
concludes:

Thus, generally speaking, clinical findings—in and of themselves—forfeit

www.freepsychotherapybooks.org 485
the probative value that Freud claimed for them, although their potential
heuristic merits may be quite substantial. To assert that the contamination
of intraclinical data is ineradicable without extensive and essential
recourse to extraclinical findings is not, of course, to declare the automatic
falsity of any and every analytic interpretation that gained the patient’s
assent by means of prodding from the analyst. But it is to maintain—to the
great detriment of intraclinical testability!—that, in general, the epistemic
devices confined to the analytic setting cannot reliably sift or
decontaminate the clinical data so as to identify those that qualify as
probative [p. 270].

THE LOGICAL FOUNDATIONS OF THE THEORY OF REPRESSION

The problem of the contamination of clinical evidence is not the only epistemic problem to

which Freudian theory is subject. In his most recent work, Grünbaum (1983b, 1984) has charted a
number of further, even more serious, difficulties, the upshot of which is that even if clinical data could

be taken at face value, they would not support the basic tenets of Freud's theoretical structure.

Grünbaum argues for this conclusion by considering the reasoning that Freud used at various
stages of his career to support “the cornerstone” of his theoretical edifice. This is the hypothesis that it

is repressed material that initially causes and continues to maintain psychoneurotic symptoms as well

as other psychic phenomena such as dreams and parapraxes. Grünbaum begins by considering the

reasoning used by Freud and Breuer to support the original version of this “repression hypothesis” for
psychoneurosis. Although the evidence they adduced to support their theory was not completely

unflawed, it did come up to a relatively high standard, according to Grünbaum. As it turned out, Freud

himself discovered that this evidence was spurious. Rather than abandoning the repression hypothesis
at this point, however, Freud substituted a new version. The difficulty with this—and the basis of

Grünbaum’s complaint—is that Freud never succeeded in providing new evidence that was anywhere

near as cogent as his original observations with Breuer. In addition, he proceeded to extrapolate from

his repression hypothesis of the psychoneuroses to a more general repression hypothesis covering

both parapraxes and dreams. But in neither case was there any new, compelling, evidence that would

www.freepsychotherapybooks.org 486
warrant the extrapolation.

The original Freud-Breuer hypothesis was that (1) the therapeutic conquest of hysterical

symptoms is effected by the abreactive lifting of the repression of a traumatic memory, and (2) this
posited therapeutic efficacy can be explained deductively by the etiologic hypothesis that the
repression of the traumatic event was causally necessary for the formation and maintenance of the

given hysterical symptom. Freud and Breuer’s (1893) evidence for these claims was that “each

individual hysterical symptom immediately and permanently disappeared when we had succeeded in
bringing clearly to light the memory of the event by which it was provoked and in arousing its

accompanying affect” (p. 6; emphasis in original). Grünbaum (1983b) reconstructs their reasoning as

follows:

First, they attributed their positive therapeutic results to the lifting of

repressions. Having assumed such a therapeutic connection, they wished to
explain it. Then they saw it would indeed be explained deductively by the
following etiological hypothesis: the particular repression whose undoing
removed a given symptom S is causally necessary for the initial formation
and maintenance of S. Thus, the nub of their inductive argument for
inferring a repression etiology can be formulated as follows: the removal
of a hysterical symptom S by means of lifting a repression R is cogent
evidence that the repression R was causally necessary for the formation of
the symptom S [p. 218].

The beauty of their appeal to separate symptom removal was this. To support their hypothesis,
Freud and Breuer had to show that removal of the repression was sufficient for the removal of the

symptom. This would count as cogent inductive grounds for the claim that the repression was a

causally necessary condition of the symptom. The difficulty was that given merely the conjunction of
removal of the repression with removal of the symptom, there was a rival explanation—namely, that

the therapeutic efficacy of the cathartic method was placebogenic. But, they argued, if the symptom
removal were a placebo effect wrought by suggestion, one would expect all the symptoms to be

removed at once. Thus, in their view, the fact that they were removed one by one was evidence against

www.freepsychotherapybooks.org 487
the rival placebo hypothesis and in support of their own view.

Although Freud and Breuer deserve considerable credit for realizing the importance of the

alternative rival placebo hypothesis and attempting to rule it out, their line of reasoning was not totally

successful, in Grünbaum’s (1983a) view. Precisely because of the analyst’s evident focus on a specific
memory for each symptom, the patient’s conquest of the given symptom might be affected by

suggestion. That is, on the basis of the analyst’s behavior, the patient might come to believe that

uncovering a memory associated with a given symptom would cause that symptom to disappear. Thus,

as a consequence of this belief, rather than the state of affairs posited by the Freud-Breuer hypothesis,
the symptom might then actually disappear.

As it turned out, the Freud-Breuer hypothesis had a far more serious problem to contend with

—namely, that the crucial evidence concerning therapeutic success was spurious. As Freud (1925) put
it:

Even the most brilliant [therapeutic] results were liable to be suddenly

wiped away if my personal relation with the patient became disturbed. It
was true that they would be reestablished if a reconciliation could be
effected; but such an occurrence proved that the personal emotional
relation between doctor and patient was after all stronger than the whole
cathartic process [p. 27].

Freud, however, continued to maintain a version of the repression hypothesis, substituting

repression of infantile sexual wishes for the Freud-Breuer repression of a traumatic event in adulthood

and calling on the full array of clinical material in support of his claims.

We have already seen that, according to Grünbaum, this clinical material is probatively hopeless
because of the failure of the tally argument to protect against the ever-present problem of suggestion

and because of the unavailability of any other vindication of the probity of clinical data. Let us suppose,
however, that this is not the case. In the face of the demise of the therapeutic vindication of the tally

www.freepsychotherapybooks.org 488
argument, can clinical data nevertheless provide the support that Freudian theory so badly needs
without relying on therapeutic success? According to Grünbaum, the answer is no; Freud’s clinical data
suffer from serious epistemic limitations as support for causal hypotheses, even if they are regarded as

uncontaminated. Consider, for example, products of the method of free association. According to
Grünbaum (1983b), the epistemic legitimation of free association as a reliable means of identifying and

certifying pathogenic causes as such collapsed with the demise of the Breuer-Freud cathartic method.
Thus, the most that the method of free association can come up with is the expression of a thought or

wish that was previously repressed. But this is a far cry from the etiologic claim that the pertinent

repression had been the pathogen P of the patient’s neurosis N on the strength of its emergence as an
association to the symptom. For, Grünbaum (1983b) argues:

to support Freud’s etiologic hypothesis that P is causally necessary for N,

evidence must be produced to show that being a P makes a difference to
being an N. But such causal relevance is not attested by mere instances of
Ns that were Ps, i.e., by patients who are both Ps and Ns. For a large
number of such cases does not preclude that just as many non-Ps would
also become Ns, if followed in a horizontal study from childhood onward!
Thus, instances of Ns that were Ps may just happen to have been Ps. Then
being a P has no etiologic role at all in becoming an N.…Thus, to provide
evidence for the causal relevance claimed by Freud, we need to combine
instances of Ns that were Ps with instances of non-Ps who are non-Ns.
Indeed, since he deemed P to be causally necessary for N—rather than just
causally relevant—his etiology requires that the class of non-Ps should not
contain any Ns whatever, and the class of Ps is to have a positive (though
numerically unspecified) incidence of Ns [p. 277].

Furthermore, for the purpose of supporting etiologic (causal) hypotheses, the absence of such
controls undermines the probative value of not only data collected by the method of free association,

but also evidence based on memories such as those discussed by Glymour (1974) in Freud’s Rat Man
case.

Freud’s causal explanations of dreams and parapraxes fare no better, as it turns out (see

www.freepsychotherapybooks.org 489
Grünbaum, 1983b, pp. 222-265). There are two basic difficulties. First, Freud’s claim that dreams and
parapraxes are like neurotic symptoms in the sense of being compromises between the demands of

our unconscious and conscious life is simply an extrapolation from his theory of psychoneurosis. In

fact, Grünbaum (1983b) argues that it is a misextrapolation because there is nothing akin to the
therapeutic base of the latter. With respect to parapraxes, for example, “Freud did not adduce any

evidence that the permanent lifting of a repression to which he had attributed a parapraxis will be

‘therapeutic’ in the sense of enabling the person himself to correct the parapraxis and to avoid its
repetition in the future” (Grünbaum, 1983b, p. 225). Second, the method Freud used to identify the

particular unconscious determinants of dreams and parapraxes is simply the method of free
association. Hence, even assuming that it is free from epistemic contamination, the method is
powerless to provide support for any causal hypothesis, including those pertinent to dreams and

parapraxes.

EXAMINATION OF A RADICAL CRITIQUE OF GRÜNBAUM’S VIEWS

Because most of Grünbaum’s work on psychoanalysis is so recent, there has, as yet, been little

time for critical reply. One exception is the strongly negative reaction of Flax (1981), who argues that

“neither Popper nor Grünbaum offers an adequate philosophy of science by which psychoanalysis may
be judged” (p. 561). Furthermore, she chastises Grünbaum for restricting his discussion to Freud,
contending that “this is like confining a discussion of physics to Newton because contemporary physics

is in such disarray and then throwing out physics because there are unresolved problems in Newton’s
theory” (p. 564). More specifically, she seems to believe that the more contemporary version of
psychoanalysis embodied in object relations theory is immune from the epistemic difficulties

Grünbaum attributes to Freud. In fact, she makes the astounding claim that “all the phenomena that
Grünbaum counts as the clinical liabilities of psychoanalysis on empiricist grounds—epistemic
contamination (i.e., intersubjectivity), suggestion, the placebo effect, etc.... are evidence that object-

relations theory is correct” (p. 567). Since these points would, if correct, strike at the heart of
Grünbaum’s work, I will conclude my discussion of Grünbaum with a consideration of Flax’s principal

www.freepsychotherapybooks.org 490
contentions.

Flax’s strategy of attack involves isolating a number of assumptions “suppressed within this
debate” that she takes to be problematic. To make her case, then, she must show both that these

assumptions are problematic and that they are, in fact, essential to Grünbaum’s arguments. I suggest

that she does neither, with the failure on the second count the more serious. It is to this point that I will

direct my remarks, for it suggests that she has seriously misunderstood the character of Grünbaum’s
arguments. Let me give a few examples.

“Empiricism,” Flax (1981) claims, “is simply untenable as a methodology of philosophy of

science. A datum is never observed as it is in itself.…Thus fact and theory cannot be totally distinct.

Empirical experience loses its special status as the most privileged and unproblematic evidence.... All

data are ‘epistemically contaminated’ ” (p. 563). I take it that the assumption of concern here is that
data gathered on the basis of observation are somehow epistemically privileged and independent of

theory. Suppose, however, we take this to be false. Suppose we agree that observation is always

“theory laden” in the sense that it always involves interpretation, and such interpretation is relative to
a person’s conceptual apparatus, beliefs, expectations, and so forth. Furthermore, we take it that

observational claims, like any other, are subject to controversy and revision and must be supported if
contested. Does it then follow that Grünbaum’s epistemological liabilities arguments fail? Flax’s
reasoning seems to be that if all observation is theory laden, then all data are “epistemically

contaminated,” including the data of our most esteemed scientific theories. Hence, any argument based

on the implicit assumption that a theory cannot be scientific if it is based on contaminated data will be

an argument based on a totally unreasonable demarcation principle.

The difficulty with this line of reasoning is that it is perfectly possible to agree that all
observation is theory laden and still maintain a distinction between data that are biased in a

damagingly relevant sense and those that are not so biased. The ideal of objective data is possible at

least to the extent that data relevant to a given theory T can be collected by someone whether he or she
believes in Tor even, in fact, whether he or she has knowledge of T. Scientists have become increasingly

www.freepsychotherapybooks.org 491
aware of the ways in which experimenters’ bias toward their pet hypotheses can affect the outcome of
experiments. With animal subjects, bias often operates in the recording of observations; with human

subjects, it can be unintentionally conveyed in the communication of experimental instructions. But is

it important to note that the result of this increasing knowledge about the potential pitfalls of
experimenter bias has not been despair over the inevitable irrationality and arbitrariness of scientific

theorizing. Rather, it has been the adoption of new and more stringent controls to minimize or

eliminate such bias. For example, the use of so-called “double blind” experimental procedure has
become standard for experimentation with human subjects in drug and other medical research.

Grünbaum’s quarrel with the use of clinical evidence as support for psychoanalytic theory is

precisely that it consists of data subject to the charge of investigator bias. Not only are the data being

gathered in the clinical setting obtained by someone firmly committed to the truth of the theory, but
they are gathered in such a way—during the course of a therapeutic process in which transference

plays a major role—that even Freud (1917, pp. 446-447) worried about the charge of suggestion.

Furthermore, as Grünbaum (1983d) points out, Flax fails to distinguish between data that is
merely theory-laden and data generated by the self-fulfilling use of the theory in their production. As

Merton’s (1949) studies of self-fulfilling and self-defeating predictions in the social sciences have
shown, “identifiable alterations of the presumed initial conditions, rather than mere theory-ladenness,

generate phenomena that furnish demonstrably spurious confirmations and disconfirmations” (p. 50).

Again, it’s not mere theory ladenness but the occurrence of precisely such alterations of the presumed
initial conditions that is the object of Grünbaum’s concern. In Grünbaum’s (1983d) view, this
occurrence “has been tellingly demonstrated experimentally in studies [reported in Marmor, 1970] of

the purportedly “free” associations produced by patients in analysis” (p. 50). Grünbaum can perfectly
well grant that all data are theory laden and still maintain that certain forms of theory-ladenness are
epistemically unacceptable.

Another one of Flax’s (1981) objections is that “a purely internal philosophical analysis of

theories and theory shifts is not adequate…[to] explain why a theory is accepted as “credible” or when

www.freepsychotherapybooks.org 492
this acceptance occurs” (p. 563). The problematic assumption she has in mind is obviously that such a
purely internal philosophical analysis of theories and theory shifts is adequate for such purposes. What

Flax means by such a “purely internal philosophical analysis” can be gleaned from the sorts of

considerations she thinks are left out of account. She writes: “At least equally important and under
dispute is what counts as a fact, how data are to be interpreted and which data must be explained” (p.

563). I take it then that such a purely internal analysis, then, is one that focuses solely on the

relationship of theory to evidence as the basis of theory choice. And the putatively problematic
assumption would be that it is possible to give a complete account of why scientists in fact accept

theories as credible at particular moments in the history of science solely in terms of the logical
relations between theory and evidence. Now, I have grave doubts as to whether most normative
philosophers of science, including Grünbaum, would accept this assumption. But the main point again

is: What of it? Suppose we agree that the assumption is wrong. It seems to me that Grünbaum’s
discussion of the problems inherent in the use of clinical data rests on no assumptions whatsoever
concerning the sorts of considerations that must be invoked to explain particular historical

occurrences of accepting particular theories. Someone interested in the assessment of the evidential
grounds claimed to provide support for a theory is simply interested in a different question than

someone interested in explaining why those who have accepted the theory did so. The former is a

question about epistemic merit, the second about human psychology. Comments by Flax (1981) such

as “Neither Grünbaum’s nor Popper’s philosophy, can provide an adequate account of the scientific
process” (p. 563) indicate that she has no real understanding of the normative project. Certainly one

can argue that the facts of scientific practice bear on one’s choice of normative principles. And given

certain views about what down-to-earth normative philosophy of science should be like, one can fault

a particular exercise in appraisal for using utopian standards. However, all this in no way affects the
point that normative philosophy of science is not concerned with giving a psychological or political or

sociological or historical explanation of why particular episodes in the history of science occurred as
they did, and, therefore, ought not be criticized if it does not do so.

Perhaps Flax in some sense realizes this, for her last criticism concerning the putatively

www.freepsychotherapybooks.org 493
problematic nature of Grünbaum’s grounds goes for the jugular. Flax (1981) writes:

Some of the greatest weaknesses of both Popper’s and Grünbaum’s

accounts of science stem from the attempt rationally and arbitrarily to
reconstruct the nature of scientific practice. Integrally connected with
rationalization is their claim to legitimately legislate what counts as
science and to evaluate how well it is done. Neither Popper nor Grünbaum
give a scientific or philosophic justification for this claim, and there are
good philosophical grounds for questioning its validity [p. 564].

Flax does not tell us here what those grounds are but simply refers us to Rorty (1979). Surely,
one might think, I cannot charge Flax with irrelevance here. For certainly, Grünbaum’s criticism of

Freudian theory does at least presuppose that normative philosophy of science is a legitimate
enterprise. My reply is, yes and no. Normative philosophy of science is not just one sort of thing, but
many. I believe that what Flax is attacking is a far more ambitious form of the enterprise than the one

Grünbaum undertakes in his recent writings. The ambitious form aims at a global rational
reconstruction of at least those parts of scientific practice that seem to be governed by reason. This

involves an attempt to find a set of principles that serve to rationalize the decisions, acts, and heuristic

rules that belong to actual scientific practice.

This ambitious form of normative philosophy may well not be possible. But the enterprise can
be made more modest in a number of ways. First, rational reconstruction can be done in a piecemeal

rather than a global way. Second, the philosopher of science can engage in the appraisal of specific
scientific contributions not as an external critic, invoking, as Scheffler (1967) puts it, norms based on
“an abstract epistemological ideal” but rather as a participant whose norms are “an ideal which,

regulating the characteristic activities of science, may enter into its very description” (p. 73). Flax fails
to understand that in Grünbaum’s various epistemological liabilities arguments, he is playing it very
close to the ground. The normative principles he invokes do not stem from any philosophical rational

reconstruction of scientific practice. They are part of that practice itself. This is particularly true of the
inductivist principles he marshals in his criticism of Freud’s attempts to establish his causal claims.

www.freepsychotherapybooks.org 494
Furthermore, Grünbaum mounts a persuasive case that these normative principles are ones Freud
himself explicitly avowed. Thus, it seems quite true, as Grünbaum (1983c) himself says, that the verdict

he reaches concerning the scientific merit of psychoanalysis “is hardly predicated on the imposition of

some extraneous methodological purism” (p. 13). The point, then, is this. Even if Flax were to convince
us that a global rational reconstruction of science were impossible, I do not see how this would

undercut Grünbaum’s critique in any way.

Grünbaum (1983d) himself has replied to Flax’s other charges. He is particularly insistent that

the critique he has offered of Freud’s claims is equally applicable to more contemporary
psychoanalytic theorists such as Heinz Kohut and the object relations school. These latter-day

theorists

all claim clinical sanction for the generic repression-aetiology of neuroses.

And they hold that free association has the epistemic capability of
identifying the unconscious causes of all kinds of thought and behavior,
such as dream content and parapraxes. Moreover, qua being
psychoanalytic, the post-Freudian versions also deem the successful lifting
of repressions to be the decisive agency in the postulated insight dynamics
of the therapy [p. 47].

Eagle, a psychoanalytically oriented clinical psychologist, has recently voiced full support for

these claims of Grünbaum’s. After examining recent formulations in psychoanalytic object relations
theory and self psychology, Eagle (1983) concludes:

Contrary claims notwithstanding, Grünbaum’s criticisms of Freudian

theory are neither vitiated nor undone by these recent developments. In
no way do current formulations somehow manage to weaken or even
constitute a response to these criticisms. The clinical data generated by an
object relations theory or self psychology approach are as
epistemologically contaminated as data generated by the more traditional
approach. There is a little, or perhaps even less, evidence available on
therapeutic process and therapeutic outcome. And finally, the etiological
claims made in more current formulations are perhaps even more logically

www.freepsychotherapybooks.org 495
and empirically flawed than Freud’s etiological formulations [pp. 49-50].

Thus, Flax’s analogy to the case of physics completely misfires, Grünbaum (1983d) asserts, “if

only because the much vaunted post-Freudian versions have not remedied a single one of the

methodological defects” (p. 48) that Grünbaum charges against the psychoanalytic method of clinical
investigation.

SUMMARY

In his recent work on psychoanalytic epistemology, Grünbaum has exhibited an extremely

impressive command of both the psychoanalytic literature and the philosophy of science. His views
thus ought to be taken very seriously by anyone interested in the epistemic status of psychoanalytic

theory. I have attempted here to extract the principal points and arguments contained in that body of

work.

Grünbaum addresses himself to two fundamental questions: (1) What sorts of standards of
assessment ought we to invoke in evaluating psychoanalysis? and (2) How does psychoanalysis

measure up relative to those standards? Because Freud himself insisted that psychoanalysis was a
natural science, and because, in Grünbaum’s view, there are no good arguments to the contrary,

Grünbaum has insisted that psychoanalysis ought to be assessed as an empirical science. To support

his position, he has engaged in debate with Popper, Habermas, Ricoeur, and George Klein, although we
have restricted our attention here to his consideration of the views of Popper and Habermas.

Popper’s famous contention that psychoanalysis is not scientific because it is unfalsifiable was

historically important not only because it raised interesting questions about the epistemic status of
psychoanalysis but also because it raised fundamental issues about what makes something scientific.
In the context of replying to Popper’s challenge, Grünbaum has argued that (1) falsifiability is a

meaningful notion in science, although it is not the touchstone of scientific rationality as Popper
maintains; (2) in particular, Popper is completely wrong in claiming that inductivism is powerless to

www.freepsychotherapybooks.org 496
impugn the scientific credentials of a theory like psychoanalysis; (3) Popper has no good arguments for

his claim that psychoanalysis is unfalsifiable; and (4) in fact, psychoanalysis is falsifiable, if one applies

a scientifically reasonable notion of falsifiability.

Habermas has argued that psychoanalysis ought to be regarded as a critical science rather than

as an empirical-analytic one. Habermas rests his case on two sets of arguments. The first concerns the

relationship of the clinical theory to the metapsychology. The second concerns the epistemic
properties of the clinical theory itself. Grünbaum has addressed these arguments as follows: First,

Freud rightly saw that the scientific status of the clinical theory is not dependent on that of the

metapsychology; hence, any argument which assumes that there is such a dependence is irrelevant to

the question of whether the clinical theory is scientific. Second, the specific arguments that Habermas
advances to show that the clinical theory is not appropriately regarded as an empirical science fail,

either because Habermas does not correctly understand psychoanalysis or because he is ignorant of

certain features of the natural sciences.

In considering the merits of psychoanalysis as a scientific theory—that is, in reply to the second
of the questions he sets himself—Grünbaum has argued for three points: (1) the therapeutic
effectiveness of the characteristic constituent factors of psychoanalytic treatment is seriously in

question; (2) all known attempts to save clinical data from the charge of contamination from
suggestion fail, so that such data are virtually useless in providing support for the cardinal hypotheses
of Freudian theory; and (3) even if clinical data were not epistemologically contaminated, they would

not support the basic tenets of Freud’s theoretical structure, because Freud's major clinical arguments

are basically flawed.

The most explicit critique of Grünbaum’s views to date is to be found in the work of Flax (1981).
Flax argues that Grünbaum’s discussion of psychoanalysis makes use of a number of implicit

assumptions regarding the nature of science which, in her view, are seriously questionable. To make

her case, she must show both that these assumptions are problematic and that they are essential to

Grünbaum’s arguments. I have argued that, in fact, she does neither. The second failing is the more

www.freepsychotherapybooks.org 497
serious because it indicates that Flax does not clearly understand the sort of normative philosophy of
science in which Grünbaum is engaged.  

REFERENCES

Breuer, J., & Freud, S. (1893). On the physical mechanism of hysterical phenomena: Preliminary
communication. Standard Edition, 2, 1-18.

Christiansen, B. (1964). The scientific status of psychoanalytic clinical evidence: III. Inquiry, 7, 47-79.

Cioffi, F. (1970). Freud and the idea of a pseudo-science. In R. Borger & F. Cioffi (Eds.), Explanation in
the behavioural sciences (pp. 471-499). Cambridge: The University Press.

Cohen, R. S., & Laudan, L. (Eds.). (1983). Physics, philosophy and psychoanalysis: Essays in honor of Adolf
Grünbaum. Dordrecht, Holland: D. Reidel.

Duhem, P. (1906). The aim and structure of physical theory [La théorie physique son objet et sa
structure]. Princeton, NJ: Princeton University Press, 1954.

Eagle, M. N. (1973). Validation of motivational explanation in psychoanalysis. Psychoanalysis &

Contemporary Thought, 2, 265-275.

Eagle, M. N. (1983). The epistemological status of recent developments in psychoanalytic theory. In R.

S. Cohen & L. Laudan (Eds.), Physics, philosophy and psychoanalysis: Essays in honor of
Adolf Grünbaum (pp. 31-56). Dordrecht, Holland: D. Reidel.

Erwin, E. (1980). Psychoanalytic therapy: The Eysenck argument. American Psychologist, 35, 435-443.

Eysenck, H. J. (1952). The effects of psychotherapy: An evaluation. Journal of Consulting Psychology, 16,
319-324.

Eysenck, H. J. (1963). Uses and abuses of psychology. Baltimore, MD: Penguin.

Eysenck, H. J. (1966). The effects of psychotherapy. New York: International Science Press.

www.freepsychotherapybooks.org 498
Eysenck, H. J. (1977). You and neurosis. London: Temple Smith.

Eysenck, H. J., & Wilson, G. D. (1973). The experimental study of Freudian theories. London: Meuthen &
Co.

Farrell, B. A. (1963). Psychoanalytic theory. In S. G. M. Lee & M. Herbert (Eds.), Freud and psychology
(pp. 19-28). Middlesex, England: Penguin, 1970.

Farrell, B. A. (1964). The status of psychoanalytic theory. Inquiry, 7, 104-123.

Fisher, S., & Greenberg, R. P. (1977). The scientific credibility of Freud's theories and therapy. New York:
Basic Books.

Flax, J. (1981). Psychoanalysis and the philosophy of science: Critique or resistance? Journal of
Philosophy, 78, 561-569.

Freud S. (1895). Reply to criticisms of my paper on anxiety neurosis. Standard Edition, 3, 123-129.

Freud, S. (1909). Notes upon a case of obsessional neurosis. Standard Edition, 10,155-318.

Freud, S. (1914). On narcissism: an introduction. Standard Edition 14, 73-102.

Freud, S. (1915a). Instincts and their vicissitudes. Standard Edition, 14, 117-140.

Freud, S. (1915b). A case of paranoia running counter to the psychoanalytic theory of the disease.
Standard Edition, 14, 263-272.

Freud, S. (1917). Introductory lectures on psychoanalysis. Part III: General theory of the neuroses.
Standard Edition, 16, 243-463.

Freud, S. (1920). The psychogenesis of a case of homosexuality in a woman. Standard Edition, 18, 147-
172.

Freud, S. (1925). An autobiographical study. Standard Edition, 20, 7-70.

Freud, S. (1933). New introductory lectures on psychoanalysis, xxix-xxxv. Standard Edition, 22, 5-182.

www.freepsychotherapybooks.org 499
Freud, S. (1937). Constructions in analysis. Standard Edition, 23, 254-269.

Freud, S. (1940a). An outline of psychoanalysis. Standard Edition, 23, 141-208.

Freud, S. (1940b). Some elementary lessons in psychoanalysis. Standard Edition, 23, 281-286.

Freud, S. (1954). The origins of psychoanalysis. New York: Basic Books.

Gadamer, H. G. (1975). Truth and method. New York: Seabury.

Geuss, R. (1981). The idea of a critical theory: Habermas and the Frankfurt School. Cambridge:
Cambridge University Press, 1981.

Glover, E. (1952). Research methods in psychoanalysis. International Journal of Psychoanalysis, 33, 403-
409.

Glymour, C. (1974). Freud, Kepler, and the clinical evidence. In R. Wollheim (Ed.), Freud (pp. 285-304).
Garden City, NY: Anchor.

Glymour, C. (1980). Theory and evidence. Princeton, NJ: Princeton University Press.

Grünbaum, A. (1966). The falsifiability of a component of a theoretical system. In P. K. Feyerabend & G.

Maxwell (Eds.), Mind, matter and method: Essays in philosophy and science in honor of
Herbert Feigl (pp. 273-305). Minneapolis: University of Minnesota Press.

Grünbaum, C. (1968). Geometry and chronometry in philosophical perspective. Minneapolis: University

of Minnesota Press.

Grünbaum, A. (1969). Can we ascertain the falsity of a scientific hypothesis? In M. Mandelbaum (Ed.),
Observation and theory in science (pp. 69-129). Baltimore: Johns Hopkins Press, 1971.

Grünbaum, A. (1976). Is falsifiability the touchstone of scientific rationality? Karl Popper versus
inductivism. In R. S. Cohen, P. K. Feyerabend, & M. W. Wartofsky (Eds.), Essays in memory
of Imre Lakatos (Boston Studies in the Philosophy of Science, Vol. 38 (pp. 213-252).
Dordrecht; Holland: D. Reidel, 1976.

www.freepsychotherapybooks.org 500
Grünbaum, A. (1977). How scientific is psychoanalysis? In R. Stern, L. Horowitz, & J. Lynes (Eds.),
Science and psychotherapy (pp. 219-254). New York: Haven Press.

Grünbaum, A. (1979). Is Freudian psychoanalytic theory pseudoscientific by Karl Popper’s criterion of

demarcation? American Philosophical Quarterly, 16, 131-141.

Grünbaum, A. (1980a). Epistemological liabilities of the clinical appraisal of psychoanalytic theory.

Noús, 14, 307-385.

Grünbaum, A. (1980b). The role of psychological explanations of the rejection or acceptance of

scientific theories. In Transactions of the New York Academy of Sciences: Vol. 39 A
Festschrift for Robert Merton, (pp. 75-90).

Grünbaum, A. (1981, March 5). How valid is psychoanalysis? An exchange. New York Review of Books,
28, 40-41.

Grünbaum, A. (1983a). Logical foundations of psychoanalytic theory. Festschrift fur Wolfgang

Stegmuller, Erkenntnis, 1983,19,109-152. In J. Reppen (Ed.), Future directions of
psychoanalysis. Hillsdale, NJ: Lawrence Erlbaum Associates. (Also reprinted in T. Millon,
Ed., Theories of psychopathology and personality, 3rd edition. New York: Holt, Rinehart &
Winston, 1983.)

Grünbaum, A. (1983b). The foundations of psychoanalysis. In L. Laudan (Ed.), Mind and medicine
(Pittsburgh Series in the Philosophy and History of Science, Vol. 8, pp. 143-309).
Berkeley: University of California Press.

Grünbaum, A. (1983c). Freud’s theory: the perspective of a philosopher of science. Presidential

Address to the American Philosophical Association, Eastern Division. In Proceedings and
Addresses of the American Philosophical Association, 57, 5-31.

Grünbaum, A. (1983d). Is object relations theory better founded than orthodox psychoanalysis? A reply
to Jane Flax. Journal of Philosophy, 80, 46-51.

Grünbaum, A. (1984). The foundations of psychoanalysis: A philosophical critique. Berkeley: University

of California Press.

www.freepsychotherapybooks.org 501
Grünbaum, A. (1985). Explication and implications of the placebo concept. In L. White, B. Tursky, & G.
F. Schwartz (Eds.), Placebo: clinical phenomena and new insights. New York: Guilford
Press.

Habermas, J. (1967). Zur Logik der Sozialwissenschaflen. (Philosophische Rundschau, Belheft 5).
Tubingen, West Germany: Siebeck & Mohr, 1967.

Habermas, J. (1971). Knowledge and Human Interests. (J. J. Shapiro, Trans.). London: Heinemann.

Habermas, J. (1979). Communication and the evolution of society. (Thomas McCarthy, Trans.). Boston:
Beacon Press.

Hempel, C. G. (1970). On the standard conception of scientific theories. In M. Radner & S. Winokur
(Eds.), Minnesota Studies in the Philosophy of Science, Vol. 4, (pp. 142-163). Minneapolis:
University of Minnesota.

Hempel, C. G. (1973). The meaning of theoretical terms: A critique of the standard empiricist construal.
In P. Suppes, L. Henkin, A. Joya, & G. C. Moisil, (Eds.), Logic, methodology and philosophy of
science, Vol. 4 (pp. 351-378) Amsterdam: North Holland Publishers.

Hesse, M. (1968). Induction, confirmation and philosophical method. [A review of P. K. Feyerabend & G.
Maxwell (Eds.), Mind, matter, and method: Essays in philosophy and science in honor of
Herbert Feigl], The British Journal for the Philosophy of Science, 18, 330-335.

Holmes, D. S. (1974). Investigations of repression: Differential recall of material experimentally or

naturally associated with ego threat. Psychological Bulletin, 81, 632-653.

Hook, S. (Ed.) (1959). Psychoanalysis, scientific method and philosophy. New York: New York University
Press.

Hospers, J. (1959). Philosophy and psychoanalysis. In S. Hook (Ed.), Psychoanalysis, scientific method
and philosophy (pp. 336-357). New York: New York University Press.

Kennedy, G. (1959). Psychoanalysis: protoscience and metapsychology. In S. Hook (Ed.),

Psychoanalysis, scientific method and philosophy (pp. 267-281). New York: New York
University Press.

www.freepsychotherapybooks.org 502
Lakatos, I. (1970). Falsification and the methodology of scientific research programmes. In I. Lakatos &
A. Musgrave (Eds.), Criticism and the Growth of Knowledge. Cambridge, England:
Cambridge University Press, 1970.

Lakatos, I. (1971). History of science and its rational reconstruction. In R. Buck & R. Cohen (Eds.), PSA
1970: In memory of Rudolf Carnap. (Boston Studies in the Philosophy of Science, Vol. 8).
Dordrecht, The Netherlands: D. Reidel, 1971.

Laudan, L. (1983). The demise of the demarcation problem. In R. S. Cohen & L. Laudan (Eds.), Physics,
philosophy and psychoanalysis: Essays in honor of Adolf Grünbaum. Dordrecht, Holland: D.
Reidel.

Luborsky, L., Singer, B., & Luborsky, L. (1975). Comparative studies of psychotherapies: Is it true that
“Everyone has won and all must have prizes”? Archives of General Psychiatry, 32, 995-
1008.

Luborsky, L., & Spence, D. P. (1978). Quantitative research on psychoanalytic therapy. In S. L. Garfield &
A. E. Bergin (Eds.), Handbook of psychotherapy and behavior change (pp. 408-438). New
York: Wiley.

Madison, P. (1961). Freud’s concept of repression and defense: Its theoretical and observational
language. Minneapolis: University of Minnesota Press.

Marmor, J. (1970). Limitations of free association. Archives of General Psychiatry, 22, 160-165.

Martin, M. (1964a). The scientific status of psychoanalytic clinical evidence. Inquiry, 7, 13-36.

Martin, M. (1964b). Mr. Farrell and the refutability of psychoanalysis. Inquiry, 7, 80-98.

McCarthy, T. A. (1978). The critical theory of Jurgen Habermas. Cambridge, MA: MIT Press.

Meltzoff, J., & Komreich, M. (1970). Research in psychotherapy. New York: Atherton Press.

Merton, R. K. (1949). Social theory and social structure. Glencoe, IL: Free Press.

Nagel, E. (1959). Methodological issues in psychoanalytic theory. In S. Hook (Ed.), Psychoanalysis,

www.freepsychotherapybooks.org 503
scientific method and psychoanalysis (pp. 38-56). New York: New York University Press.

Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on mental
processes. Psychological Review, 84, 231-59.

Popper, K. (1963). Conjectures and refutations: The growth of scientific knowledge. New York: Harper &
Row. (First presented as a paper, Science: Conjectures and reflections, 1953.)

Rorty, R. (1979). Philosophy and the mirror of nature. Princeton, NJ: Princeton University Press.

Rothstein, E. (1980, October 9). The scar of Sigmund Freud. New York Review of Books, 27, 14-20.

Salmon, W. C. (1959). Psychoanalytic theory and evidence. In S. Hook (Ed.), Psychoanalysis, scientific
method and philosophy (pp. 252-267). New York: New York University Press.

Schaffner, K. (1969). Correspondence rules. Philosophy of Science, 36, 280-290.

Scheffler, I. (1967). Science and subjectivity. Indianapolis: Bobbs-Merrill.

Sloane, R. B., Staples, F. R., Gristol, A. H., Yorkston, N. J., & Whipple, K. (1975). Psychotherapy vs.
behavior therapy. Cambridge, MA: Harvard University Press.

Suppes, P. (1962). Models of data. In E. Nagel, P. Suppes, & A. Tarski (Eds.), Logic, methodology, and
philosophy of science: Proceedings of the 1960 International Congress. Stanford, CA:
Stanford University Press, 1962.

Suppes, P. (1967). What is a scientific theory? In S. Morgenbesser (Ed.), Philosophy of science today (pp.
55-67). New York: Basic Books.

Thoma, H., & Kachele, H. (1973). Wissenschaftstheoretische and methodologische Probleme der
klinisch-psychoanalytischen Forschung. Psyche, 27, 205-236, 309-355. [Translated as
Problems of metascience and methodology in clinical psychoanalytic research. The
Annual of Psychoanalysis, 1975, 3, 49-119.]

Von Eckardt, B. (1982). The scientific status of psychoanalysis. In S. Gilman (Ed.), Introducing
Psychoanalysis (pp. 139-180). New York: Brunner/Mazel.

www.freepsychotherapybooks.org 504
Notes

1) We are confronted, unfortunately, with a terminological difficulty concerning the word “science.”
English renditions of Habermas use the word ‘science’ as the translation of the German
‘Wissenschaft.’ Hence, it is used in the more inclusive sense, which encompasses not only
the natural sciences but also the hermeneutic and critical sciences. In contrast, when we
ask Grünbaum whether psychoanalysis is a science, we are using the term to refer
paradigmatically to what physicists, chemists, and biologists do, and it becomes an open
question whether the so-called cultural and critical “sciences” in fact count as science. I
alert the reader to this fact so as to minimize possible confusion. I will try to make it
clear in context which sense is intended.

www.freepsychotherapybooks.org 505
NOTES ON CONTRIBUTORS

ANNI BERGMAN, Ph.D. is a research scientist for the Margaret S. Mahler Psychiatric Research
Foundation and a faculty member of the Clinical Psychology Department of the City University of New

York. She is coauthor of The Psychological Birth of the Human Infant as well as two films and has
written several papers on aspects of the separation-individuation process.

MONICA CARSKY, Ph.D., is Instructor of Psychology in Psychiatry, Cornell University Medical

College, and Staff Psychologist at the New York Hospital-Cornell Medical Center, Westchester Division.

A graduate of Swarthmore College, she received her Ph.D. in clinical psychology from the City
University of New York. She completed an internship at Bronx Psychiatric Center, Albert Einstein
College of Medicine, and a postdoctoral fellowship at the New York Hospital-Cornell Medical Center,

Westchester Division, and is currently a student at the New York Center for Psychoanalytic Training.

MORRIS N. EAGLE, Ph.D., is Professor and Chairman of the Department of Psychology at York

University, Toronto. He recently completed six months as Visiting Professor of Psychiatry at Western

Psychiatric Institute and Clinic and Fellow at the Center for the History and Philosophy of Science. In
private practice of psychotherapy and a consultant at the Clarice Institute of Psychiatry in Toronto, he

is author of Recent Developments in Psychoanalysis: A Critical Evaluation. Dr. Eagle received his Ph.D. in
1958 at New York University with George Klein and worked at the New York University Research
Center for Mental Health, followed by a directorship of the Clinical Program and chairmanship of the

Psychology Department at Yeshiva University.

www.freepsychotherapybooks.org 506
STEVEN ELLMAN, Ph.D., is Professor and Director, Clinical Psychology Doctoral Training

Program of the City University of New York.

JAMES S. GROTSTEIN, M.D., is Associate Clinical Professor of Psychiatry at the University of

California, Los Angeles; Attending Staff Physician at Cedars-Sinai Medical Center; and Director of the
Interdisciplinary Group for Advanced Studies in Psychotic, Borderline, and Narcissistic Disorders.

VICTORIA HAMILTON was born in Aberdeenshire, Scotland in 1941, studied at Glasgow School

of Art, and graduated in philosophy at University College, London. Her psychoanalytic training was

undertaken at the Tavistock Clinic, London, during which time she attended John Bowlby’s seminars on
ethology, child development, and attachment. She has worked as a child psychotherapist for the

National Health Service and the Child Guidance Service in London and in private practice in Los
Angeles. She currently lives in London.

IRWIN Z. HOFFMAN, Ph.D., is Assistant Professor, Department of Psychiatry, University of

Illinois College of Medicine, Chicago; Consultant, Illinois State Psychiatric Institute; faculty member at

the Chicago School for Professional Psychology; and candidate at the Chicago Institute for

Psychoanalysis. He is the author of several articles on adaptation to loss, psychotherapy, and the
psychoanalytic process. He was awarded the Chicago Institute’s Edwin Eisler Prize in 1980.

FREDERIC J. LEVINE, Ph.D., is Associate Professor and Coordinator of the Internship Training in
Psychology in the Department of Mental Health Sciences at Hahnemann University. He is a Diplomate

in Clinical Psychology (ABPP), an Associate Member of the American Psychoanalytic Association, and a
member of the Division of Psychoanalysis of the American Psychological Association.

ZVI LOTHANE, M.D., is in private practice of psychoanalysis in New York; an Assistant Professor
of Psychiatry at Mount Sinai School of Medicine, City University of New York; and Training Analyst at

the Institute for Psychoanalytic Training and Research and the National Psychological Association for
Psychoanalysis. He has written papers on psychoanalytic methodology, listening as an instrument in

www.freepsychotherapybooks.org 507
psychoanalysis, and hallucinations.

HYMAN L. MUSLIN, M.D., is a Professor of Psychiatry at the University of Illinois, College of

Medicine at Chicago and a practicing psychoanalyst. His interests have become focused on the
application of self psychology to the transformation of the self in patients with cancer; the application

of self psychology in literature and history; and the learning and teaching of self psychology.

STANLEY R. PALOMBO, M.D., practices psychoanalysis in Chevy Chase, Maryland. He is

Associate Clinical Professor of Psychiatry at George Washington University and a faculty member of

the Washington School of Psychiatry. His book, Dreaming and Memory: A New Information-Processing

Model, redefines the process of dream construction through the application of information theory.

JEANINE PARISIER PLOTTEL, Ph.D., is Professor at Hunter College and the Graduate Center, City

University of New York, where she is director of the Twentieth Century Conference. She writes about

literature and psychoanalysis, is on the editorial advisory boards of the Review of Psychoanalytic Books

and Dada/Surrealism, and is publisher of New York Literary Forum.

JOSEPH W. SLAP, M.D., is Training and Supervising Analyst at the Institute of the Philadelphia
Association for Psychoanalysis, and Clinical Professor in the Department of Mental Health Sciences at

Hahnemann University where he participates in the training of psychiatric residents and doctoral
candidates in clinical psychology. He is in private practice in Philadelphia.

DONALD P. SPENCE, Ph.D., is Professor of Psychiatry at Rutgers Medical School, University of

Medicine and Dentistry of New Jersey, and Visiting Lecturer at Princeton University. He received his
Ph.D. from Teachers College, Columbia University, was certified in psychoanalysis at the New York
Psychoanalytic Institute, and was formerly a professor of psychology at New York University.

ROBERT S. STEELE, Ph.D., teaches at Wesleyan University, where he is co-coordinator of the

Women’s Studies Program and an Associate Professor of Psychology. He gardens, raises cats, and has

www.freepsychotherapybooks.org 508
written several pieces on psychoanalysis, including Freud and Jung: Conflicts of Interpretation. His
recent work has been in feminist criticism.

BARBARA VON ECKARDT, Ph.D., is Assistant Professor in Philosophy at Yale University. She
received her Ph.D. from Case Western Reserve University in 1974, and subsequently spent three years

at the Massachusetts Institute of Technology as a postdoctoral fellow in the Department of Psychology.

Her major research interests are in philosophy of psychology and philosophy of science.

www.freepsychotherapybooks.org 509
Share this Book with your Friends!

www.freepsychotherapybooks.org 510