C

op
yr
ig
ht
©
20
11
by
th
e
Am
er
ic
an
Ps
y ch
ol
og
ic
and Violence
al
As
Human Aggression

so
ci
at
io
n
(A
P A)
C
op
yr
ig
ht
©
12051-00_FM-rev2.qxd

20
11
5/25/10

by
th
e
11:27 AM

Am
er
Page ii

ic
an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
A)
P
(A
n
Human Aggression

io
at
ci
so
and Violence

As
al
ic
Causes, Manifestations, and Consequences
og
ol
y ch
Ps
an
ic

Edited by Phillip R. Shaver and Mario Mikulincer

er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

American Psychological Association • Washington, DC

12051-00_FM-rev2.qxd 5/25/10 11:27 AM Page iv

Copyright © 2011 by the American Psychological Association. All rights reserved. Except as
permitted under the United States Copyright Act of 1976, no part of this publication may
be reproduced or distributed in any form or by any means, including, but not limited to, the
process of scanning and digitization, or stored in a database or retrieval system, without the
prior written permission of the publisher.

A)
Published by To order

P
American Psychological Association APA Order Department

(A
750 First Street, NE P.O. Box 92984
Washington, DC 20002 Washington, DC 20090-2984

n
www.apa.org Tel: (800) 374-2721; Direct: (202) 336-5510

io
Fax: (202) 336-5502; TDD/TTY: (202) 336-6123

at
Online: www.apa.org/pubs/books
E-mail: [email protected]

ci
so
In the U.K., Europe, Africa, and the Middle East, copies may be ordered from
American Psychological Association

As
3 Henrietta Street
Covent Garden, London

al
WC2E 8LU England

ic
og
Typeset in Goudy by Circle Graphics, Inc., Columbia, MD

Printer: Maple-Vail Manufacturing Group, York, PA

Cover Designer: Mercury Publishing Services, Rockville, MD
ol
y ch
The opinions and statements published are the responsibility of the authors, and such
Ps

opinions and statements do not necessarily represent the policies of the American
Psychological Association.
an

Library of Congress Cataloging-in-Publication Data

ic
er

Human aggression and violence : causes, manifestations, and consequences / edited by

Phillip R. Shaver and Mario Mikulincer.
Am

p. cm. — (The Herzliya series on personality and social psychology)

Includes bibliographical references and index.
ISBN-13: 978-1-4338-0859-3 (print)
e

ISBN-10: 1-4338-0859-5 (print)

th

ISBN-13: 978-1-4338-0860-9 (electronic)

by

ISBN-10: 1-4338-0860-9 (electronic)

1. Aggressiveness. 2. Violence. I. Shaver, Phillip R. II. Mikulincer, Mario.
11

BF575.A3H867 2011
20

155.2'32—dc22
2010005100
©

British Library Cataloguing-in-Publication Data

ht

A CIP record is available from the British Library.

ig
yr

Printed in the United States of America

op

First Edition
C
12051-00_FM-rev2.qxd 5/25/10 11:27 AM Page v

THE HERZLIYA SERIES ON PERSONALITY

AND SOCIAL PSYCHOLOGY

A)
Mario Mikulincer and Phillip R. Shaver, Series Editors

P
(A
Series Titles

n
io
Prosocial Motives, Emotions, and Behavior: The Better Angels of Our Nature

at
Edited by Mario Mikulincer and Phillip R. Shaver

ci
so
Human Aggression and Violence: Causes, Manifestations, and Consequences

As
Edited by Phillip R. Shaver and Mario Mikulincer

al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C
C
op
yr
ig
ht
©
12051-00_FM-rev2.qxd

20
11
5/25/10

by
th
e
11:27 AM

Am
er
Page vi

ic
an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-00_FM-rev2.qxd 5/25/10 11:27 AM Page vii

CONTENTS

PA)
(A
n
io
at
ci
so
As
al
ic
og
ol
y ch
Ps
an

Contributors ............................................................................................ xi
ic
er

Preface .................................................................................................... xv
Am

Introduction ............................................................................................. 3
Phillip R. Shaver and Mario Mikulincer
e
th

I. Major Theoretical Perspectives ...................................................... 13

by

Chapter 1. The General Aggression Model ................................. 15

11

C. Nathan DeWall and Craig A. Anderson

20

Chapter 2. I3 Theory: Instigating, Impelling, and Inhibiting

©

Factors in Aggression .................................................. 35

ht

Erica B. Slotter and Eli J. Finkel

ig

Chapter 3. Applying Adaptationism to Human Anger:

yr

The Recalibrational Theory ....................................... 53

op

Aaron Sell
C

vii
12051-00_FM-rev2.qxd 5/25/10 11:27 AM Page viii

Chapter 4. A Behavioral Systems Perspective on

Power and Aggression ................................................ 71
Phillip R. Shaver, Michal Segev, and Mario Mikulincer

A)
Chapter 5. Dispositional Influences on Human Aggression ........ 89
Jennifer L. Tackett and Robert F. Krueger

P
(A
Chapter 6. A Social Neuroscience Perspective on the

n
Neurobiological Bases of Aggression ........................ 105

io
Thomas F. Denson

at
ci
so
II. Genetic and Environmental Determinants ................................. 121

As
Chapter 7. The Transmission of Aggressiveness Across
Generations: Biological, Contextual,

al
and Social Learning Processes .................................. 123

ic
L. Rowell Huesmann, Eric F. Dubow, and Paul Boxer

og
Chapter 8. Genetic and Environmental Influences ol
ch
on Aggression ........................................................... 143
y

Soo Hyun Rhee and Irwin D. Waldman

Ps

Chapter 9. Social Information Processing Patterns as

an

Mediators of the Interaction Between

ic

Genetic Factors and Life Experiences in

er

the Development of Aggressive Behavior ................ 165

Am

Kenneth A. Dodge

Chapter 10. Violence and Character: A CuPS

e
th

(Culture × Person × Situation) Perspective ............. 187

by

Dov Cohen and Angela K.-y. Leung

11

III. Psychological and Relational Processes .................................... 201

20

Chapter 11. Mirror, Mirror, on the Wall, Who’s the

©

Most Aggressive of Them All? Narcissism,

ht

Self-Esteem, and Aggression .................................... 203

ig

Sander Thomaes and Brad J. Bushman

yr

Chapter 12. Evolved Mechanisms for Revenge and Forgiveness .... 221
op

Michael McCullough, Robert Kurzban,

C

and Benjamin A. Tabak

viii CONTENTS
12051-00_FM-rev2.qxd 5/25/10 11:27 AM Page ix

Chapter 13. Attachment, Anger, and Aggression ....................... 241

Mario Mikulincer and Phillip R. Shaver
Chapter 14. Attachment and Violence: An Anger Born of Fear ... 259
Donald G. Dutton

PA)
Chapter 15. Respecting Others and Being Respected

(A
Can Reduce Aggression in Parent–Child
Relations and in Schools .......................................... 277

n
io
Ofra Mayseless and Miri Scharf

at
ci
IV. Aggression at the Societal Level ................................................ 295

so
Chapter 16. An Existential Perspective on Violent Solutions to

As
Ethno–Political Conflict .......................................... 297

al
Gilad Hirschberger and Tom Pyszczynski

ic
og
Chapter 17. The Emotional Roots of Intergroup Aggression:
The Distinct Roles of Anger and Hatred ................. 315
ol
Eran Halperin
y ch

Chapter 18. Tension and Harmony in Intergroup Relations ....... 333

Ps

Tamar Saguy, Nicole Tausch, John F. Dovidio,

Felicia Pratto, and Purnima Singh
an
ic

V. Consequences of Aggression: The Victim Perspective ............. 349

er
Am

Chapter 19. Influence of Violence and Aggression on

Children’s Psychological Development:
e

Trauma, Attachment, and Memory ......................... 351

th

Sheree L. Toth, LaTonya S. Harris,

by

Gail S. Goodman, and Dante Cicchetti

11

Chapter 20. The Paradox of Partner Aggression:

20

Being Committed to an Aggressive Partner ............. 367

Ximena B. Arriaga and Nicole M. Capezza
©

Chapter 21. The Psychological Toll of Exposure to

ht

Political Violence: The Israeli Experience ............... 385

ig

Zahava Solomon and Karni Ginzburg

yr
op

Index .................................................................................................... 403

C

About the Editors ................................................................................ 421

CONTENTS ix
C
op
yr
ig
ht
©
12051-00_FM-rev2.qxd

20
11
5/25/10

by
th
e
11:27 AM

Am
Page x

er
ic
an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-00_FM-rev2.qxd 5/25/10 11:27 AM Page xi

CONTRIBUTORS

PA)
(A
n
io
at
ci
so
As
al
ic
og
ol
y ch
Ps
an

Craig A. Anderson, PhD, Department of Psychology, Iowa State University,

ic

Ames
er

Ximena B. Arriaga, PhD, Department of Psychological Sciences, Purdue

Am

University, West Lafayette, IN

Paul Boxer, PhD, Department of Psychology, Rutgers University, Newark, NJ
e

Brad J. Bushman, PhD, Department of Psychology and Institute for Social

th

Research, University of Michigan, Ann Arbor, and VU University,

by

Amsterdam, The Netherlands

Nicole M. Capezza, PhD, Department of Psychological Sciences, Purdue
11

University, West Lafayette, IN

20

Dante Cicchetti, PhD, Institute of Child Development and Department of

Psychiatry, University of Minnesota, Minneapolis
©

Dov Cohen, PhD, Department of Psychology, University of Illinois at Urbana–

ht

Champaign
ig

Thomas F. Denson, PhD, School of Psychology, University of New South

yr

Wales, Sydney, Australia

op

C. Nathan DeWall, PhD, Department of Psychology, University of Kentucky,

C

Lexington

xi
12051-00_FM-rev2.qxd 5/25/10 11:27 AM Page xii

Kenneth A. Dodge, PhD, Center for Child and Family Policy, Duke Univer-
sity, Durham, NC
John F. Dovidio, PhD, Department of Psychology, Yale University, New
Haven, CT

A)
Eric F. Dubow, PhD, Department of Psychology, Bowling Green State Uni-
versity, Bowling Green, OH

P
(A
Donald G. Dutton, PhD, Department of Psychology, University of British
Columbia, Vancouver, Canada

n
Eli J. Finkel, PhD, Department of Psychology, Northwestern University,

io
Evanston, IL

at
Karni Ginzburg, PhD, The Bob Shapell School of Social Work, Tel Aviv

ci
University, Tel Aviv, Israel

so
Gail S. Goodman, PhD, Department of Psychology and Center for Public

As
Policy Research, University of California, Davis
Eran Halperin, PhD, School of Government, Interdisciplinary Center

al
Herzliya, Herzliya, Israel

ic
LaTonya S. Harris, EdM, Department of Psychology and Center for Public

og
Policy Research, University of California, Davis
ol
Gilad Hirschberger, PhD, School of Psychology, Interdisciplinary Center
ch
Herzliya, Herzliya, Israel
y

L. Rowell Huesmann, PhD, Institute for Social Research, University of

Ps

Michigan, Ann Arbor

Robert F. Krueger, PhD, Department of Psychology, Washington Univer-
an

sity, St. Louis, MO

ic

Robert Kurzban, PhD, Department of Psychology, University of Pennsylva-

er

nia, Philadelphia
Am

Angela K.-y. Leung, PhD, School of Social Sciences, Singapore Manage-

ment University, Singapore
e

Ofra Mayseless, PhD, School of Education, University of Haifa, Haifa, Israel

th

Michael McCullough, PhD, Department of Psychology, University of

by

Miami, Coral Gables, FL

Mario Mikulincer, PhD, School of Psychology, Interdisciplinary Center
11

Herzliya, Herzliya, Israel

20

Felicia Pratto, PhD, Department of Psychology, University of Connecticut,

Storrs
©

Tom Pyszczynski, PhD, Department of Psychology, University of Colorado,

ht

Colorado Springs
ig

Soo Hyun Rhee, PhD, Institute for Behavioral Genetics, University of Col-
yr

orado, Boulder
op

Tamar Saguy, PhD, School of Psychology, Interdisciplinary Center Herzliya,

Herzliya, Israel
C

Miri Scharf, PhD, School of Education, University of Haifa, Haifa, Israel

xii CONTRIBUTORS
12051-00_FM-rev2.qxd 5/25/10 11:27 AM Page xiii

Michal Segev, MA, Department of Psychology, Bar-Ilan University, Ramat

Gan, Israel
Aaron Sell, PhD, Department of Psychology, University of California, Santa
Barbara

A)
Phillip R. Shaver, PhD, Department of Psychology, University of California,
Davis

P
(A
Purnima Singh, PhD, Department of Psychology, Women’s College, Aligarh
Muslim University, Aligarh, India

n
Erica B. Slotter, MA, Department of Psychology, Northwestern University,

io
Evanston, IL

at
Zahava Solomon, PhD, The Bob Shapell School of Social Work, Tel Aviv

ci
University, Tel Aviv, Israel

so
Benjamin A. Tabak, MSc, Department of Psychology, University of Miami,

As
Coral Gables, FL
Jennifer L. Tackett, PhD, Department of Psychology, University of

al
Toronto, Toronto, Canada

ic
Nicole Tausch, PhD, Department of Psychology, Cardiff University, Cardiff,

og
Wales, United Kingdom
ol
Sander Thomaes, PhD, Department of Developmental Psychology, Utrecht
ch
University, Utrecht, The Netherlands
y

Sheree L. Toth, PhD, Mt. Hope Family Center, University of Rochester,

Ps

Rochester, NY
Irwin D. Waldman, PhD, Department of Psychology, Emory University,
an

Atlanta, GA
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

CONTRIBUTORS xiii
C
op
yr
ig
ht
©
12051-00_FM-rev2.qxd

20
11
5/25/10

by
th
e
11:27 AM

Am
er
Page xiv

ic
an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-00_FM-rev2.qxd 5/25/10 11:27 AM Page xv

PREFACE

PA)
(A
n
io
at
ci
so
As
al
ic
og
ol
We live in a violent world. Aggression, intimidation, and cruelty are
ch
part of daily life almost everywhere. Homicides, violent robberies, gang
y

warfare, political violence, terrorism, and international conflicts fill the daily
Ps

newspapers, TV news programs, and websites. Violence ruins many marriages,

damages countless children, and causes the deaths of police officers and
an

military recruits.
ic

Researchers and members of the general public often wonder what

er

makes a person violent. Is it a matter of genes? Parental abuse and neglect?

Am

The media? Alcohol and drugs that impair self-control and disrupt rational
problem solving? Access to guns?
e

Within the social and behavioral sciences, and now within the brain sci-
th

ences as well, investigators agree that human aggression is complex and multi-
by

faceted and that many kinds of research are necessary to understand its causes,
manifestations, and consequences. Some writers, for example, have considered
11

aggression and violence from evolutionary and genetic perspectives and have
20

produced solid evidence concerning the intergenerational transmission of

aggression and its stability across a person’s life course. Other writers have con-
©

sidered how parenting styles and parental modeling contribute to children’s

ht

aggression; still others have focused on intergroup conflicts and warfare.

ig

Unfortunately, findings from these different perspectives are rarely integrated

yr

or effectively applied in real life.

op

The time is therefore right for a collaborative, creative conversation

among experts on the variety of approaches to human aggression and vio-
C

lence. The timeliness and social relevance of the topic led us to dedicate the

xv
12051-00_FM-rev2.qxd 5/25/10 11:27 AM Page xvi

second volume of The Herzliya Series on Personality and Social Psychology to

the psychology of aggression, violence, and their effects as viewed at different
levels of analysis, from neural to societal. This series of volumes is designed to
explore socially significant topics with a group of world-class researchers who

A)
are invited to visit the Interdisciplinary Center Herzliya to engage in con-
structively critical but collaborative conversations and then to contribute to

P
(A
a focused, gracefully written volume that can be shared with other members
of the discipline and understood by a wide range of educated readers outside

n
the discipline. The first volume in the series, Prosocial Motives, Emotions, and

io
Behavior (published by the American Psychological Association [APA] in

at
2009), was dedicated to prosocial behavior. The present volume is the second

ci
in the series.

so
The chapter authors generously agreed to deliver lectures, participate in

As
hours of discussion of the lectures, and then return home and prepare chap-
ters based on these lectures and discussions. The meeting was cohosted by the

al
two of us, and we worked together with the chapter authors to make the book

ic
as accessible, coherent, and readable as possible, so that it would be suitable

og
for researchers and application-oriented professionals as well as university
ol
classes and the educated public. The book provides a state-of-the-art review
ch
of the major theoretical perspectives on human aggression and violence; the
y

basic psychological processes underlying destructive behaviors; and recent

Ps

research on the consequences of aggression in couple and family relation-

ships, groups and organizations, and intergroup relations.
an

We are grateful to everyone who made the preparation of this volume

ic

so enjoyable and successful. We thank all of the chapter authors, an amazing

er

group of scholars and admirable human beings who care about both their sci-
Am

ences and the betterment of human lives. We especially wish to thank Pro-
fessor Uriel Reichmann, president of the Interdisciplinary Center Herzliya,
e

who provided financial and staff support for an annual series of conferences
th

on personality and social psychology. We thank the staff of the Interdiscipli-

by

nary Center Herzliya—Shulli Sardes, Lior Lev-Ari, and Tsachi Ein-Dor—

who handled all of the arrangements for the conference, dealt effectively with
11

the many on-site details, and coped masterfully with the inevitable glitches
20

and emergencies. We would also like to thank Maureen Adams, senior acqui-
sitions editor at APA Books, for seeing the value of this book and the series
©

in which it appears and for being a generous, thoughtful, and supportive

ht

friend during this book’s preparation. Finally, we thank Tyler Aune and Katie
ig

Funk, development editor and production editor at APA Books, respectively,

yr

and copyeditor Erin O’Brien for their careful efforts.

op
C

xvi PREFACE
C
op
yr
ig
ht
©
20
11
by
th
e
Am
er
ic
an
Ps
y ch
ol
og
ic
and Violence
al
As
Human Aggression

so
ci
at
io
n
(A
P A)
C
op
yr
ig
ht
©
20
11
12051-01_Introduction-rev2.qxd

by
th
e
5/25/10

Am
er
11:28 AM

ic
an
Page 2

Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-01_Introduction-rev2.qxd 5/25/10 11:28 AM Page 3

INTRODUCTION
PHILLIP R. SHAVER AND MARIO MIKULINCER

PA)
(A
n
io
at
ci
so
As
al
ic
og
ol
y ch
Ps
an

In a world riddled with greed, intimidation, and violence, there is an

ic

urgent need to understand and control these destructive forces. Within the
er

social and behavioral sciences, and recently within genetics and neuroscience
Am

as well, there are deep and empirically supported insights into aggression and
violence, but they are rarely included in a single conversation. We know more
e

than ever about the evolutionary history and functions of aggression; several
th

scientific disciplines have explored its genetic, developmental, interpersonal,

by

and cultural causes. But the resulting knowledge has not been integrated and
effectively applied in real life. It is time for an open, collaborative, and creative
11

discussion among experts about the most promising perspectives on human

20

aggression and violence, as well as ways to apply scientific insights in effective

interventions.
©

In social psychology, aggression is defined as behavior performed by one

ht

person (i.e., an aggressor) with the intent of physically or psychologically harm-

ig

ing another person who wants to avoid the harm (i.e., a victim), and violence is
yr

defined as an extreme, especially destructive, and cruel form of aggression (for

op

examples and discussions of violence at the dyadic, familial, and societal levels,
C

see Chapters 14, 16, 17, and 19–21). In psychological science, aggression and

3
12051-01_Introduction-rev2.qxd 5/25/10 11:28 AM Page 4

violence are viewed as different from assertiveness and power. Although aggres-
sion can be used to assert one’s authority and control over precious resources,
there are many cases in which power can be attained without inflicting any
harm to another person (for discussions of the relations among aggression, vio-

A)
lence, and power, see Chapters 3, 4, and 13). Researchers also make distinc-
tions between subtypes of aggression (e.g., proactive vs. reactive aggression,

P
(A
thoughtful vs. thoughtless aggression), but they realize that aggressive acts in
the real world are often complex combinations of these subtypes that can

n
change over time and as a function of the victim’s responses (for discussions of

io
dyadic-level effects on aggressive behavior, see Chapters 1, 2, and 20).

at
Historically, aggressive behavior was viewed either as the result of an

ci
inborn instinct aimed at the destruction of life (e.g., Freud’s, 1920/1961, the-

so
ory of the death drive) or as a learned response to the frustration of one’s needs

As
(e.g., Dollard, Doob, Miller, Mowrer, & Sears’s, 1939, well-known frustration–
aggression hypothesis). With the advance of psychological research, several

al
theories of aggression and its causes have emerged. For example, neoassociation-

ic
ist theory (Berkowitz, 1993) posits that aversive events (e.g., frustration, provo-

og
cation) can automatically trigger aggressive behavior, whereas a cognitive
ol
revision of this theory emphasizes that higher level cognitive processes (e.g.,
ch
attributions, appraisals) moderate these automatic aggressive responses. Social
y

learning theory (Bandura, 1973) proposes that people develop aggressive

Ps

behavior when they observe others behaving aggressively, particularly if the

others are likable, have high social status, or are rewarded for their aggressive
an

behavior. According to script theory (Huesmann, 1998), individuals learn

ic

aggression-related mental representations and action plans from the wider

er

world around them, including the mass media. Social interaction theory (Tedeschi
Am

& Felson, 1994) proposes that aggression is an instrumental means to achieve

important goals, such as obtaining something of value or establishing a desired
e

social identity.
th

Recently, Anderson and Bushman (2002) proposed a broad, integrative

by

theory of aggression, the general aggression model (GAM; for a detailed discus-
sion of this theory, see Chapter 1), encompassing the common features of
11

previous theories of aggression. In the GAM, the causes of aggression are

20

analyzed at two different levels of causation, distal and proximal, and within
each level there are personal and situational risk factors as well as affective,
©

motivational, and cognitive mechanisms through which the risk factors influ-
ht

ence aggression. (These factors are discussed in detail throughout this vol-
ig

ume.) For example, at the distal level, the GAM emphasizes genetic risk
yr

factors (see Chapters 7–9), exposure to parental aggressive behavior during

op

childhood (Chapters 7 and 9), cultural norms and values (Chapter 10),
socioeconomic risk factors, and child abuse and neglect (Chapters 13, 14, and
C

19). The GAM also emphasizes personality predispositions (see Chapter 5),

4 SHAVER AND MIKULINCER

12051-01_Introduction-rev2.qxd 5/25/10 11:28 AM Page 5

neural mechanisms (Chapter 6), narcissism and self-esteem (Chapter 11),

cognitive biases in attributions and appraisals (Chapter 9), and proximal (i.e.,
situational) risk factors, such as frustration, incentives, aversive stimulation,
stress, alcohol and drugs, and salient opportunities for aggression. In addition,

A)
the GAM emphasizes the role of self-regulatory processes in triggering and
resisting aggression (see Chapters 2, 6, and 12).

P
(A
Theory and research on aggression have also informed efforts to prevent
and treat aggression. The methods include reducing exposure to events that

n
teach or reward aggressive behavior or hostile appraisals and attributions (see

io
Chapters 7 and 9), diminishing personality and situational forces that can trig-

at
ger aggression such as narcissism and alcohol abuse (see Chapters 2 and 11),

ci
strengthening prosocial virtues and action tendencies (Chapters 12 and 15),

so
and increasing the quality of parental care and the resulting sense of emotional

As
security (Chapters 13 and 14).
The need for understanding and reducing aggression, violence, and their

al
destructive consequences inspired us to dedicate the second volume of The

ic
Herzliya Series on Personality and Social Psychology to the psychology of aggres-

og
sive behavior, viewed from various levels of analysis, from the neural to the
ol
societal. We invited experts on the science of aggression and violence to
ch
explain their ideas and research findings, and encouraged them to trace the
y

historical and theoretical background of their research and speculate about

Ps

where it might lead in the future. Readers of this volume will gain a deeper and
broader understanding of aggression, violence, and their social and psycholog-
an

ical causes and effects, whether the readers are anchored in the disciplines of
ic

personality, social, and clinical psychology; the close relationships field; devel-
er

opmental psychology; neuroscience; or political psychology. They will also

Am

learn about methods used to study aggression and violence, and about prom-
ising psychological and social interventions that can reduce violence and deal
e

with its destructive effects.

th

The volume is organized into five sections. The first, focusing on major
by

theoretical perspectives, includes six chapters on evolutionary, neuroscientific,

developmental, personality-oriented, and social psychological perspectives on
11

aggression and violence. The second section, focusing on research into the
20

genetic and environmental determinants of aggressive behavior, includes four

chapters on the intraindividual and intrafamilial causes and correlates of aggres-
©

siveness, and the genetic, social information processing, and cultural factors
ht

that mediate the development of aggression and its transmission across gener-
ig

ations. The third section includes five chapters dealing with psychological and
yr

social-relational processes underlying aggression and violence. The fourth sec-

op

tion includes three chapters on the existential, emotional, and structural roots
of intergroup aggression. The final section considers the consequences of aggres-
C

sion viewed from the perspective of its victims: the abused child, the battered

INTRODUCTION 5
12051-01_Introduction-rev2.qxd 5/25/10 11:28 AM Page 6

romantic partner, and citizens and soldiers exposed to political violence, acts
of terrorism, and wars.
In Chapter 1, Nathan DeWall and Craig Anderson present a current
account of the influential GAM. They conceptualize aggression in terms of

A)
cycles—the single-episode cycle, the developmental cycle, the violence esca-
lation cycle—and specify causes and processes, both biological and environ-

P
(A
mental, that contribute to aggressive behavior. They also discuss ways to
reduce violence at the individual, community, national, and international lev-

n
els. This overview of the classic research on aggression provides a foundation

io
for the following chapters.

at
In Chapter 2, Erica Slotter and Eli Finkel expand on the GAM by pro-

ci
viding a new theoretical framework, the I3 (I-Cubed) theory, which organizes

so
aggression risk factors into three categories: (a) instigating triggers—discrete

As
events that arouse action tendencies conducive to aggression; (b) impelling
forces—factors that increase the likelihood of an aggressive impulse follow-

al
ing an instigating trigger; and (c) inhibiting forces—factors that increase the

ic
likelihood that aggression will be contained or held in check rather than

og
being expressed.
ol
In Chapter 3, Aaron Sell explains how anger and aggression may have
ch
evolved biologically to solve certain adaptive problems. Applying the meth-
y

ods of evolutionary psychologists’ “adaptationist program” to human anger,

Ps

Sell argues that anger is the output of a cognitive mechanism “designed” by

natural selection to negotiate conflicts of interest, influence the weight placed
an

by a target of anger on an angry person’s interests, and cause the target of

ic

anger to treat the angry person better. Using this framework, Sell reviews
er

recent studies of anger and aggression and shows that their findings can be
Am

explained by the process of recalibration that occurs in the mind of an anger

target when his or her interaction partner becomes noticeably angry.
e

In Chapter 4, Phillip Shaver, Michal Segev, and Mario Mikulincer con-

th

sider an attachment-theoretical perspective on aggression. Attachment theory

by

parses human motivation into several life domains and considers the biological
function of the behavioral systems that evolved to solve adaptive problems
11

within each of these domains. Shaver et al. propose that a “power” behavioral
20

system evolved to encourage the attainment and maintenance of dominance

and influence in social relations. They describe a new Power Behavioral System
©

Scale that measures hyperactivation and deactivation of the power system, and
ht

present new research findings concerning how these two dimensions provide a
ig

framework for understanding individual differences in aggressive behavior.

yr

In Chapter 5, Jennifer Tackett and Robert Krueger present a personality

op

trait perspective on aggression, showing how aggression is situated within cur-

rent models of personality and psychopathology. They review research on sta-
C

ble individual differences in aggressive tendencies and the associations of these

6 SHAVER AND MIKULINCER

12051-01_Introduction-rev2.qxd 5/25/10 11:28 AM Page 7

tendencies with other personality constructs and with “externalizing” (e.g.,

aggressive or antisocial) forms of psychopathology. Tackett and Krueger explain
the current state of theory and research concerning how genetic and environ-
mental factors combine to influence aggression while emphasizing that genetic

A)
factors are best understood in relation to particular environmental contexts.
In Chapter 6, Thomas Denson moves to the neural level of analysis and

P
(A
considers how some of the processes discussed by other chapter authors are
manifested in the brain. He shows that the limbic system and the dorsal ante-

n
rior cingulate are involved in the arousal of anger, and that regions associated

io
with emotion regulation are involved in rumination and displaced aggression

at
(i.e., harming people who are not responsible for the aggressor’s problems).

ci
In addition, neural regions involved in reward processing are active during

so
acts of aggression, helping to explain the difficulty of down-regulating aggres-

As
sion. Many provocative exciting ideas for future research are proposed in this
chapter.

al
The second section of the volume begins with a chapter by Rowell Hues-

ic
mann, Eric Dubow, and Paul Boxer (Chapter 7), who consider the prevalence

og
of aggressiveness from childhood to adulthood and the extent to which this
ol
trait is transmitted across generations. They also provide a road map for under-
ch
standing the relative importance of genetic factors, continuity in environ-
y

ments, and social learning processes in determining within-individual and

Ps

cross-generational forms of continuity in aggression. Huesmann et al. propose

a model that includes all of these factors and then apply it to data from a three-
an

generation longitudinal study.

ic

In Chapter 8, Soo Hyun Rhee and Irwin Waldman explore genetic influ-
er

ences on aggression and compare the patterns and magnitudes of genetic and
Am

environmental influences on different forms of aggression. On the basis of a

meta-analysis of twin and adoption studies of antisocial behavior in general,
e

and of specific forms of aggressive behavior in particular, Rhee and Waldman

th

conclude that aggression is partially attributable to genes and that the heritabil-
by

ity of aggression and antisocial behavior depends on the type of aggression

assessed, the assessment method, and the age and sex of the people under study.
11

In Chapter 9, Kenneth Dodge presents his well-researched social infor-

20

mation processing model of aggression. Dodge argues that (a) aggression-related

social information processing patterns (e.g., hypervigilance to threat cues, hos-
©

tile attributional biases) are influenced by genetic and environmental factors,

ht

(b) these patterns account for the effects of genetic and environmental factors
ig

on aggression, and (c) interventions to alter processing patterns can prevent

yr

aggressive behavior. Dodge evaluates his model with data from the ongoing lon-
op

gitudinal Child Development Project. His work has important implications for
the design of interventions to impede the development of aggressive and vio-
C

lent behavior.

INTRODUCTION 7
12051-01_Introduction-rev2.qxd 5/25/10 11:28 AM Page 8

In Chapter 10, Dov Cohen and Angela Leung combine an individual-

differences approach with a cultural perspective on variations in aggressive
behavior. They describe three kinds of cultures, those of honor, dignity, and
face, and consider within-culture variations in the extent to which individu-

A)
als position themselves in alignment with the dominant value system of their
culture. They review recent studies of people from honor, dignity, and face

P
(A
cultures, showing that both within- and between-culture forms of variation
are important in explaining a person’s propensity to behave aggressively.

n
The third section of the volume begins with Sander Thomaes and Brad

io
Bushman’s chapter (Chapter 11) on the relations among narcissism, self-

at
esteem, and aggression. In the first section of the chapter, the authors present

ci
evidence against the familiar hypothesis that low self-esteem leads to aggres-

so
sion. In the second section, they argue that aggressive individuals are typi-

As
cally self-absorbed, believing they are better than others and overestimating
their own valuable qualities; in other words, they are narcissistic. Thomaes

al
and Bushman review recent findings showing that the combination of narcis-

ic
sism and insult leads to exceptionally high levels of aggression. Their work is

og
important for designers of school programs that foster self-views that deter
aggression. ol
ch
In Chapter 12, Michael McCullough, Robert Kurzban, and Benjamin
y

Tabak discuss the adaptive functions of the evolved mechanisms underlying

Ps

humans’ capacities for revenge and forgiveness. They argue against character-
izations of revenge as something “gone wrong” in human nature, and against
an

characterizations of forgiveness as an “antidote” for the “poison” of revenge.

ic

They review evidence suggesting that both revenge and forgiveness are behav-
er

ioral adaptations that helped our human ancestors solve prevalent social prob-
Am

lems. McCullough et al. enumerate the selection pressures that probably gave
rise to both revenge and forgiveness and describe the psychological processes
e

that activate these processes in contemporary humans.

th

Chapters 13 (by Mario Mikulincer & Phillip Shaver) and Chapter 14 (by
by

Donald Dutton) use attachment theory to conceptualize the psychodynamic

mechanisms and interpersonal processes that underlie functional and dysfunc-
11

tional forms of aggression. We have included two chapters about this perspec-
20

tive because there are two different lines of work combining attachment
theory with (a) personality and social psychological research on social motives,
©

emotions, cognitions, and behavior (covered by Mikulincer & Shaver) and

ht

(b) issues in clinical and forensic psychology (covered by Dutton). In the first
ig

of these chapters (Chapter 13), Mikulincer and Shaver focus on attachment

yr

insecurities in adulthood and their effects on human aggression. They review

op

research on the links between these insecurities and anger, domestic violence,
antisocial behavior, and intergroup hostility. They also consider the main
C

goal of human aggression—to maintain power and dominance—and present

8 SHAVER AND MIKULINCER

12051-01_Introduction-rev2.qxd 5/25/10 11:28 AM Page 9

exploratory research on the influence of attachment-related processes on cog-

nition and action when a person’s sense of power is experimentally enhanced.
Dutton (in Chapter 14) focuses on the clinical implications of attach-
ment theory and argues that violence toward relationship partners (i.e.,

A)
domestic violence) or oneself (e.g., self-cutting or suicide) can be caused by
attachment-related threats (e.g., threats of separation, rejection, or abandon-

P
(A
ment). Dutton also discusses symbolic attachment, the promise of reunion with
and high praise from loved ones after death, and considers its role in religion-

n
based aggression, military violence, and other forms of violence against mem-

io
bers of outgroups.

at
In Chapter 15, Ofra Mayseless and Miri Scharf discuss the importance

ci
of respect in social relationships as a buffer against aggression. They consider

so
various definitions of respect and draw distinctions between respect and other

As
emotions or attitudes thought to be buffers against aggression, such as trust,
empathy, and acceptance. They review studies that illuminate the role of

al
respect in reducing aggression in unequal relationships, such as the respect

ic
shown by parents for their children and by children for their parents, and the

og
respect shown by teachers for their students and by students for their teach-
ol
ers. The chapter provides important insights into the role of respect in inter-
ch
ventions aimed at reducing aggression in homes and schools.
y

In the first chapter in the fourth section of the volume (see Chapter 16),
Ps

Gilad Hirschberger and Tom Pyszczynski provide an existential perspective on

intergroup aggression. Expanding terror management theory, the authors argue
an

that basic existential fears and awareness of one’s own mortality affect the
ic

extent to which people rely on aggression and violence to solve intergroup con-
er

flicts. They review studies from their own and other laboratories showing that
Am

experimentally increasing death awareness increases support for violent actions

against enemies. Hirschberger and Pyszczynski also review evidence regarding
e

contextual factors that inhibit the effects of existential concerns on the escala-
th

tion of violence. The chapter ends with an integrative model showing how
by

patriotism, vengeance, and pacifism are all rooted in basic existential concerns.
In Chapter 17, Eran Halperin probes the emotional roots of intergroup
11

aggression—anger and hatred—and reveals their different and interactive

20

roles. Equipped with an integration of theories from social psychology and the
field of conflict resolution, Halperin reviews studies focused mainly on the
©

Israeli–Palestinian conflict, showing that neither anger nor hatred alone is suf-
ht

ficient to arouse intergroup violence. Instead, only the co-occurrence of these

ig

two emotions incites large-scale aggression. The optimistic message of this

yr

chapter is that the link between anger and escalating cycles of violence is
op

avoidable if hatred is reduced or contained.

Tamar Saguy, Nicole Tausch, John Dovidio, Felicia Pratto, and Purn-
C

ima Singh explore in Chapter 18 ways of reducing intergroup aggression while

INTRODUCTION 9
12051-01_Introduction-rev2.qxd 5/25/10 11:28 AM Page 10

discounting the basic premise of contact theory, which proposes that inter-
group aggression can be reduced by bringing members of opposing groups
together under conditions that stress commonalities. Instead, Saguy et al.
focus on power-related dynamics and review findings from both laboratory

A)
and real-world studies showing that stressing commonalities between strong
and weak groups can create tension rather than harmony. The authors show

P
(A
that well-intentioned interventions aimed at reducing intergroup hostility
may paradoxically reduce the likelihood of reconciliation.

n
The final section of the volume deals with the consequences of aggres-

io
sion viewed from the perspectives of its victims. In Chapter 19, Sheree Toth,

at
LaTonya Harris, Gail Goodman, and Dante Cicchetti discuss the long-term

ci
consequences of aggression against children (child abuse and maltreatment).

so
The authors focus on emotion regulation, attachment insecurity, and mental

As
health, as well as effects of maltreatment on memories of abusive experiences
and events. They suggest policies and practices that can benefit children, ado-

al
lescents, and adults who were abused, neglected, or maltreated during infancy

ic
and childhood. They also suggest practices that may increase the accuracy of

og
such people’s memories and reports in forensic contexts.
ol
In Chapter 20, Ximena Arriaga and Nicole Capezza focus on intimate
ch
partner violence and consider the paradox of being a victim of a partner’s
y

aggression while still being committed to the aggressive partner and the trou-
Ps

bled relationship. Using the frameworks of interdependence theory and cogni-

tive consistency theories, Arriaga and Capezza explain how committed victims
an

of partner aggression justify or downplay their partner’s destructive acts. These

ic

authors then present findings from several studies showing that a victim’s well-
er

being and the maintenance of the victim’s relationship are sometimes at odds.
Am

The chapter emphasizes the need for special interventions to protect the vic-
tims of relational aggression.
e

In Chapter 21, Zahava Solomon and Karni Ginzburg discuss the long-
th

term pathogenic consequences of exposure to political violence, acts of terror,

by

and warfare. Using life in Israel as an example of recurrent and prolonged

exposure to military and political violence, Solomon and Ginzburg review
11

findings from 3-decade longitudinal studies of (a) veterans of the First Lebanon
20

War in 1982 and (b) ex-POWs from the Yom Kippur War in 1973. The stud-
ies reveal that for many soldiers and civilians the war does not end when the
©

shooting stops but continues to plague them in diverse and complicated ways.
ht

In some cases, the violence is brought home from the battlefield and adversely
ig

affects families and civil society.

yr
op

As can be seen from these brief thumbnail sketches, the chapters in this
volume cover a broad array of ideas and research on aggression and violence,
C

their detrimental effects, and interventions to dampen or eliminate these

10 SHAVER AND MIKULINCER

12051-01_Introduction-rev2.qxd 5/25/10 11:28 AM Page 11

harmful effects. The book is realistic, recognizing that aggression arises partly
from human evolution, genes, and natural interpersonal and intergroup con-
flict and is therefore not likely to be completely eliminated. But the book is
also hopeful, showing that scientific insight and a variety of research-based

A)
interventions can reduce the development, expression, and detrimental con-
sequences of aggression and violence. The authors have provided a com-

P
(A
mendable service to readers by writing clearly and compellingly about their
areas of expertise, discussing their ideas in person with each other at a vibrant

n
and creative series of meetings in Israel, and cross-referencing their chapters

io
to help readers pursue useful connections. Together they offer a deep, mature

at
portrait of human aggression and violence and suggest concrete ways in which

ci
it can be reduced in future generations.

so
As
REFERENCES

al
ic
Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of Psy-

og
chology, 53, 27–51. doi:10.1146/annurev.psych.53.100901.135231
ol
Bandura, A. (1973). Aggression: A social learning analysis. Englewood Cliffs, NJ:
ch
Prentice-Hall.
y

Berkowitz, L. (1993). Aggression: Its causes, consequences, and control. New York, NY:
Ps

McGraw-Hill.
Dollard, J., Doob, L., Miller, N. E., Mowrer, O. H., & Sears, R. R. (1939). Frustration
an

and aggression. New Haven, CT: Yale University Press. doi:10.1037/10022-000

ic

Freud, S. (1961). Beyond the pleasure principle (J. Strachey, Ed. & Trans.). New York,
er

NY: Norton. (Original work published 1920) doi:10.1037/11189-000

Am

Huesmann, L. R. (1998). The role of social information processing and cognitive

schema in the acquisition and maintenance of habitual aggressive behavior. In
e

R. G. Geen & E. Donnerstein (Eds.), Human aggression: Theories, research, and

th

implications for social policy (pp. 73–109). San Diego, CA: Academic Press.
by

Tedeschi, J. T., & Felson, R. B. (1994). Violence, aggression, and coercive actions.
Washington, DC: American Psychological Association. doi:10.1037/10160-000
11
20
©
ht
ig
yr
op
C

INTRODUCTION 11
C
op
yr
ig
ht
©
20
11
12051-01_Introduction-rev2.qxd

by
th
e
5/25/10

Am
er
11:28 AM

ic
an
Page 12

Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 13

A)
P
(A
MAJOR THEORETICAL

n
io
PERSPECTIVES

at
ci
so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C
C
op
yr
ig
ht
©
20
12051-02_PT1-CH01-rev2.qxd

11
by
5/25/10

th
e
Am
11:29 AM

er
ic
an
Page 14

Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 15

PA)
THE GENERAL AGGRESSION MODEL

(A
n
io
C. NATHAN DEWALL AND CRAIG A. ANDERSON

at
ci
so
As
al
ic
og
ol
y ch
Ps
an

Marwan Abu Ubeida contradicts general stereotypes of would-be mass

ic

murderers. He had a privileged upbringing, is deeply religious, and shows none

er

of the signs of psychopathology typically used to identify violent people. Yet

Am

he has an insatiable drive to kill as many people as he can. Marwan is an Iraqi

suicide bomber. When asked what will happen in the last moments of his life,
e

Marwan outlined a two-step process (Ghosh, 2005). First, he will ask Allah to
th

bless his holy mission with a high rate of American casualties. Second, he will
by

ask for a pure soul that is suitable to see Allah and his mujahideen brothers
who are already in paradise. Marwan’s final wishes are both chilling and puz-
11

zling, suggesting, as they do, that Allah will approve and assist Marwan’s mur-
20

der of many Americans. How does such a privileged youth become a suicide
bomber?
©

Further puzzlement comes from the behavior of Lynndie England. She

ht

joined the American National Guard not to inflict pain on others but to pro-
ig

vide a means to pay for her education. She has a son toward whom she behaves
yr

with love and kindness. As a guard in the Abu Ghraib prison in Iraq, however,
op

she wreaked havoc on Iraqi prisoners through the use of cruel and humiliating
torture practices. Her acts gained worldwide attention through the publication
C

of pictures showing her apparent glee over inflicting pain and humiliation on

15
12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 16

naked, shackled prisoners. In interviews, Lynndie deflected responsibility for

her actions, noting that she was merely following orders from superiors and
seeking the approval of a fellow soldier with whom she was in love. What leads
a person such as Lynndie England to perform such aggressive acts?

A)
Scholars, politicians, and the general public are often perplexed not only
by aggressive acts committed by individuals but also by aggression between

P
groups large and small. The escalating conflict between Israelis and Palestini-

(A
ans, for example, has claimed the lives of thousands of people, including many

n
unarmed civilians. Both Israeli and Palestinian leaders have offered logical sug-

io
gestions for ways to end the conflict, and other governments have done the

at
same. Yet the end of the conflict is nowhere in sight. The dispute appears

ci
intractable (see Chapters 16 and 17). On December 29, 2008, Israel launched

so
a major military offensive designed to stop Hamas militants from firing mis-

As
siles into the Jewish state. Over 1,000 people died, with more than 4,500 addi-
tional people wounded. After the assault ended, the Hamas rockets continued

al
to land in southern Israel. Why does the Israeli–Palestinian conflict continue

ic
to escalate despite recurrent efforts to end it peacefully?

og
The purpose of this chapter is to demonstrate how the general aggression
ol
model (GAM; Anderson & Bushman, 2002; Anderson & Huesmann, 2003)
ch
helps to answer these perplexing questions (and many others) regarding the
y

causes and conditions of aggression and violence. The GAM is a dynamic,

Ps

social–cognitive, developmental model that provides an integrative frame-

work for domain-specific aggression theories. It includes situational, persono-
an

logical, and biological variables. The GAM draws heavily on social–cognitive

ic

and social learning theories that have been developed over the past 40 years
er

by social, personality, cognitive, and developmental psychologists (e.g., Ban-

Am

dura, 1977; Berkowitz, 1989; Crick & Dodge, 1994; Dodge, 1980; Huesmann,
1988; Mischel, 1973; Mischel & Shoda, 1995; Chapters 7 and 9, this volume).
e

These perspectives paved the way for understanding the learning and devel-
th

opmental processes involved in shaping aggressive behavior and understand-

by

ing how aggression operates under the control of intrapsychological processes

aimed at overriding impulses to remain in agreement with standards for appro-
11

priate behavior (see Chapter 2, this volume).

20

The chapter is organized into seven sections. First, we offer definitions of

antisocial, aggressive, and violent behavior. Second, we provide a brief descrip-
©

tion of the GAM. Third, we discuss the dynamic process by which persono-
ht

logical and situational factors establish and sustain aggression: the violence
ig

escalation cycle. Fourth, we use the GAM to understand how seemingly ordinary
yr

citizens become terrorists, suicide bombers, torturers, and other doers of aggres-
op

sion and violence. Fifth, we discuss the implications of the GAM for aggression
between groups of people. Sixth, we apply the GAM to show how certain gov-
C

ernment actions designed to promote peace can increase aggression and violent

16 DE WALL AND ANDERSON

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 17

behavior. Seventh, we discuss useful suggestions based on the GAM regarding

ways to reduce aggression and violence.

A)
ANTISOCIAL BEHAVIOR, AGGRESSION, AND VIOLENCE

P
(A
Much has changed since the Mesolithic period, during which human
evolution presumably formed many of our current innate tendencies. Agri-

n
culture now dominates people’s access to food, allowing people to settle in

io
communities instead of hunting and gathering in nomadic groups. Cultural

at
progress enables people to depend on others for food, clothing, and shelter

ci
instead of having to provide for themselves. Technological advances provide

so
people with the means to travel easily and to transmit knowledge to each

As
other quickly. Despite these revolutionary changes in human lives, aggression
and violence remain important topics in modern society, just as they must

al
have been to our evolutionary ancestors.

ic
Archeological and historical evidence indicates that aggression and vio-

og
lence were prevalent among our hunter/gatherer ancestors 25,000 years ago.
ol
Aggression and violence was widespread among Greek, Egyptian, and Roman
ch
societies 2,000 to 3,000 years ago. Just as modern citizens of the world ingest
y

violent media, ancient Romans had their own form of “media violence” in
Ps

which gladiators inflicted physical injury and death on each other in the pres-
ence of thousands of viewers. Aggression and violence continue to be wide-
an

spread. In short, they remain ubiquitous parts of human life. Before we can
ic

understand the causes and conditions of aggression and violence, it is necessary

er

to discuss what we mean when we say that a person or group acts antisocially,
Am

aggressively, or violently.
e

Antisocial Behavior
th
by

Antisocial behavior refers to any action that violates personal or cultural

standards for appropriate behavior. Antisocial behavior often involves aggres-
11

sion and violence, but not always. In societies with norms prohibiting physical
20

violence between romantic partners, punching, kicking, or biting one’s spouse

would be considered antisocial behavior. Littering, lying, and stealing also rep-
©

resent antisocial behaviors, although none of these behaviors involves physical

ht

aggression or violence. People with antisocial personality disorder (Hare, 1996)

ig

often engage in aggressive and violent actions, but they also violate standards
yr

for appropriate behavior in nonaggressive ways such as cheating, stealing, and

op

breaking other laws. Thus, antisocial behavior can involve aggression, violence,
or any other type of response that defies cultural standards for desirable behav-
C

ior. This chapter focuses primarily on aggressive and violent behavior.

THE GENERAL AGGRESSION MODEL 17

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 18

Aggression and Violence

Aggression refers to behavior carried out with the proximal (i.e., immedi-
ate) intention to inflict harm on another person who is motivated to avoid the

A)
harm (e.g., Anderson & Bushman, 2002; Baron & Richardson, 1994). Harm-
ful behavior that is accidental or an incidental by-product of helpful actions is

P
not aggressive. From a social psychological perspective, violence usually refers

(A
to the most severe types of physical aggression, those that are likely to cause

n
serious bodily injury. Occasionally, researchers in this domain use “emotional”

io
or “psychological” violence to indicate severe forms of nonphysical aggression.

at
All acts of violence fit our definition of aggression, but not all aggressive acts

ci
are violent. By our definition, violent actions need not involve illegal behav-

so
ior. Note, however, that other behavioral sciences (e.g., criminology) define

As
violence in somewhat different ways, such as by requiring the act to be illegal
(Neuilly, 2007).

al
ic
og
GENERAL AGGRESSION MODEL
ol
Several dichotomous distinctions among various forms of aggression have
ch
been proposed. Although these distinctions (i.e., proactive vs. reactive, instru-
y

mental vs. hostile, impulsive vs. premeditated) have yielded important insights,
Ps

we argue for a more flexible understanding of aggression based on a knowledge

an

structure approach (Bushman & Anderson, 2001). We also argue that differ-
ent forms of aggression can be distinguished in terms of proximate and ultimate
ic

goals (Anderson & Bushman, 2002). Furthermore, we can characterize any

er

aggressive behavior according to four dimensions, each of which fits well with
Am

research on the development, use, and automatization of knowledge structures.

Any aggressive act (i.e., proximal intent to harm, target motivated to avoid the
e

act) can be characterized along each of the following dimensions: degree of hos-
th

tile or agitated affect present, automaticity, degree to which the primary (i.e.,
by

ultimate) goal is to harm the victim versus benefit the perpetrator, and degree
to which consequences are considered. Because many aggressive acts involve
11

mixed motivations or are sensitive to specific consequences, considering aggres-

20

sion along these four dimensions rather than relying on dichotomous category
systems provides researchers with a better means of understanding aggression
©

and of creating useful interventions (Anderson & Bushman, 2002; Anderson

ht

& Huesmann, 2003).

ig
yr

Basic Model
op

The GAM takes into account how aggression depends on cognitive fac-
C

tors within the individual. Aggression depends on how an individual perceives

18 DE WALL AND ANDERSON

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 19

and interprets his or her environment and the people therein, expectations
regarding the likelihood of various outcomes, knowledge and beliefs about
how people typically respond in certain situations, and how much people
believe they have the abilities to respond to a variety of events (see Chap-

A)
ters 7 and 9). By understanding these cognitions, researchers have a basis for
understanding both within-person and situation-specific stability in aggres-

P
(A
sion because people show similarity in how they respond to similar events
over time, and because situations frequently impose realistic demands that

n
limit the number of options regarding how people can construe the situation.

io
Furthermore, such social–cognitive models also account for variability in

at
aggression across time, people, and contexts, as different knowledge struc-

ci
tures develop and change, and as different situational contexts prime differ-

so
ent knowledge structures.

As
The GAM also focuses heavily on how the development and use of
knowledge structures influence both early (e.g., basic visual perception) and

al
downstream (e.g., judgments, decisions behaviors) psychological processes

ic
(e.g., Bargh, 1996; Fiske & Taylor, 1991; Wegner & Bargh, 1998). People

og
develop knowledge structures from their experience. Within the context of
ol
aggression, knowledge structures can influence toward whom a person directs
ch
visual attention as a function of possible threat, affective responses to provo-
y

cation or cues linked to aggression in memory, attributions regarding the

Ps

causes of a provocateur’s behavior, judgments regarding the costs and benefits

of various behavioral options, memory for people who do and do not represent
an

potential threat, and actual behavior directed toward a target.

ic

Of particular interest are findings showing that through repeated practice

er

and exposure, complex judgments and choices become automatized, requiring

Am

little or no mental energy or conscious awareness (Bargh & Pietromonaco,

1982). A “shoot first, ask questions later” mentality may result from learning
e

through repeated experience or cultural teachings that members of various

th

groups represent threats and therefore should be perceived as dangerous even

by

in neutral or ambiguous situations. For example, repeated experience and cul-

tural teachings that African Americans are likely to be hostile and pose a phys-
11

ical threat have been shown to affect decisions to shoot unarmed African
20

American crime suspects—decisions made by both college student research

participants (Correll, Park, Judd, & Wittenbrink, 2002) and police officers
©

(Plant & Peruche, 2005). The effect that knowledge structures can have on
ht

violence was demonstrated in the tragic incident of Amadou Diallo, an African

ig

American man who was shot 19 times by New York City police officers as he
yr

reached for his wallet (Cooper, 1999). Thus, knowledge structures set the stage
op

for understanding how people identify objects, people, and complex social
events as relevant or irrelevant to aggression; how beliefs about specific peo-
C

ple (e.g., Osama bin Laden) or groups (e.g., Nazis, Hutus) shape perceptions

THE GENERAL AGGRESSION MODEL 19

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 20

of relevance to aggression; and how people use behavioral scripts to guide their
behavior under various circumstances (e.g., responding with retaliation to an
insult when that insult increases hostile affect).

A)
Single Episode Cycle

P
(A
At the most basic level, the GAM focuses primarily on how aggression
unfolds within one cycle of an ongoing social interaction. At this level the

n
model emphasizes three main issues: person and situation inputs, present

io
internal state (i.e., cognition, arousal, affect), and outcomes of appraisal and

at
decision-making processes (see Anderson & Bushman, 2002, Figure 2).

ci
so
Person and Situation Inputs

As
The GAM considers both situation and person factors—relatively

al
enduring traits, motivations, attitudes, beliefs, and other chronic knowledge

ic
structures and less enduring cognitive, affective, and arousal states that arise

og
in particular contexts (see Chapters 2, 7, and 9). Person and situation inputs
ol
are proximate causes in that they provide the most direct guiding force
ch
behind aggression behavior, although the behavior may also serve an ulti-
mate goal (see Tooby & Cosmides, 1992). Social psychologists have identi-
y
Ps

fied a variety of situational factors that promote aggressive behavior, such as

provocation, exposure to weapons, a hot environment, unpleasant odors,
an

loud noises, violent media, and physical pain (for a review, see Anderson &
ic

Bushman, 2002). Examples of person factors known to increase aggression

er

are hostile attribution bias, narcissism, being male, and a host of beliefs, atti-
Am

tudes, values, and behavioral scripts (see Chapters 2, 5, 9, 10, 11, 13, and 14,
this volume).
Situation and person factors are not mutually exclusive. Some situa-
e
th

tional factors give rise to states that closely resemble person variables; for
example, social rejection or playing violent video games can strengthen hos-
by

tile cognitive biases (Bushman & Anderson, 2002; DeWall, Twenge, Gitter,
11

& Baumeister, 2009). In addition, situational variables often interact with

person variables to predict aggression. In response to provocation, for exam-
20

ple, narcissistic people tend to behave quite aggressively, whereas narcissists

©

do not show high levels of aggression in response to praise (Bushman &

ht

Baumeister, 1998; Chapter 11, this volume). Similarly, exposure to hunting

and assault weapons influences the mental accessibility of hostile cognitions
ig
yr

and aggressive behavior differently according to whether people have devel-

op

oped knowledge structures through experience to certain kinds of weapons

(e.g., hunters compared with people who have less differentiated knowledge
C

about types of weapons; Bartholow, Carnagey, & Anderson, 2005).

20 DE WALL AND ANDERSON

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 21

Internal States
Person and situation variables influence aggression through the internal
states they create. That is, internal states serve as mechanisms underlying the
relationship between person and situation variables and outcomes of appraisal

A)
and decision-making processes. Affect, arousal, and cognition represent the

P
three most significant internal states. A specific person variable (e.g., high

(A
trait hostility) or situational variable (e.g., viewing violent media) may influ-
ence one, two, or all three types of internal states. Violent media, for exam-

n
io
ple, affect all three states. Moreover, the three internal states can influence

at
each other.

ci
so
Outcomes

As
A large body of literature within social psychology suggests that com-
plex information processes can involve reliance on the automatic system or

al
the controlled system (Smith & Lazarus, 1993). In the GAM (see Anderson

ic
& Bushman, 2002, Figure 3), the third stage includes complex appraisal and

og
decision processes that range from automatic to heavily controlled (Strack &
ol
Deutsch, 2004; Chapter 6, this volume). Therefore, inputs (Stage 1) affect
ch
internal states (Stage 2), which in turn influence appraisal and decision
y

processes (Stage 3). The appraisal and decision processes include automatic
Ps

processes referred to as immediate appraisal and more controlled processes

referred to as reappraisal. Based on the outcomes of immediate appraisal or
an

reappraisal processes, people are impelled to act in either thoughtful or impul-

ic

sive ways. These actions enter a feedback loop that becomes part of the input
er

for the next episode.

Am

Immediate appraisals depend heavily on the automatic system and influ-

ence affective, goal, and intention information. Appraising environmental
e

threat, for example, occurs effortlessly and without conscious awareness (e.g.,
th

Öhman, Lundqvist, & Esteves, 2001). This immediate appraisal process may
by

include fear and anger-related affect, goals related to aggression, and the for-
mation of intentions to carry out aggression-related acts. Person and situation
11

inputs guide immediate appraisals in ways that are congruent with a person’s
20

social learning history (i.e., personality) and current psychological and

physiological state. Because immediate appraisal is effortless and requires few
©

resources, some aggressive acts occur so fast that it may seem that appraisal
ht

has not even occurred, and indeed some behavioral scripts may be so closely
ig

linked to the perception of a particular stimulus that the behavioral response

yr

is functionally a part of perceiving the stimulus.

op

Reappraisal processes, in contrast, depend on whether people have ade-

quate resources and whether the immediate appraisal is judged (automati-
C

cally) to be both important and unsatisfactory. A growing body of literature

THE GENERAL AGGRESSION MODEL 21

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 22

suggests, for example, that the ability to override unwanted impulses depends
on a limited energy resource that becomes depleted after prior exertion (Gail-
liot et al., 2007; Chapter 2, this volume). If a person has recently engaged in
an act involving the expenditure of self-regulatory energy, that person will be

A)
less likely to engage in reappraisal (DeWall, Baumeister, Stillman, & Gail-
liot, 2007; Finkel, DeWall, Slotter, Oaten, & Foshee, 2009; Chapter 6, this

P
(A
volume). Likewise, if a person’s immediate appraisal indicates that the prob-
able outcome is either satisfying or unimportant, then the person will be less

n
likely to engage in reappraisal. Other resource limitations, such as time and

io
cognitive capacity, may also preclude reappraisal.

at
Thus, aggression results from the proximate convergence of situations

ci
and personological inputs. Situations can impel or inhibit aggression, whereas

so
personological factors enhance or diminish a person’s propensity to behave

As
aggressively. These situational and personological inputs activate affective,
arousal, and cognitive internal states, which in turn influence aggression by

al
means of appraisal and decision processes. Once the individual has performed

ic
the impulsive or thoughtful action, the behavior feeds back to the situation

og
and personological inputs to guide the next episodic cycle.
ol
ch
Aggression Before and After the Single Episode Cycle
y

Is the GAM stuck in the present? At first glance, the GAM appears to
Ps

focus most of its attention on how current internal states determine aggres-
an

sion, neglecting the importance of the past and future. However, the persono-
logical input factors bring the past to the present in the form of knowledge
ic

structures and well-rehearsed cognitive and affective processes that have been
er

influenced by biological factors (e.g., genes, hormones) and past history (see
Am

Anderson & Bushman, 2002, Figure 4).

Similarly, the GAM details how the present influences and is influenced
e

by the future through the knowledge structures used to perceive, react, and
th

learn. The present influences the future in at least two different ways: by
by

changing the person’s relatively enduring beliefs, attitudes, expectations (i.e.,

personality) and by changing the person’s social environment (i.e., the atti-
11

tudes, beliefs, expectations other people have of the person). Chronic exposure
20

to violent media, for example, can increase aggressive attitudes, beliefs, expec-
tations of others as hostile, and desensitization to future violence (Anderson,
©

Carnagey, & Eubanks, 2003; Bartholow, Bushman, & Sestir, 2006; Carnagey,
ht

Anderson, & Bushman, 2007). The present is influenced by the future through
ig

the person’s beliefs and expectations about how others will act, their goals,
yr

and other plans. Thus, the GAM focuses on internal states as they relate to
op

what people bring with them to the present episode from the past and also
shows how the present episode can influence future personological and situ-
C

ational factors that will influence future internal states and subsequent appraisal

22 DE WALL AND ANDERSON

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 23

and decision processes that guide aggression. It even provides a simple process
by which personality influences situations.

A)
VIOLENCE ESCALATION CYCLE

P
(A
Most incidents involving aggression and violence occur after a series of
conflict-based interactions in which the two parties trade retaliatory behav-

n
iors back and forth in an escalating cycle. Such escalating cycles include what

io
some refer to as “ordinary” violent crimes between individuals (e.g., assault

at
and murder) and between larger groups and nations. Figure 1.1 illustrates the

ci
violence escalation cycle.

so
The violence escalation cycle begins with an initial triggering event

As
that may be serious or relatively benign. The triggering event can influence
any kind of dyad, including two people, two groups, two religions, or two

al
nations. Whereas person or group A considers retaliation to the event to be

ic
justified and relatively mild, person or group B considers the retaliation to be

og
unjustified and severe, leading to retaliation toward person or group A. The
ol
cycle persists through several iterations of violent actions in which one unit
ch
perceives its retaliation to be appropriate and justified, whereas the second
y

unit perceives it to be inappropriate and exaggerated (see similar analyses in

Ps

Chapters 10 and 12).

Consider an example of the violence escalation cycle within the context
an

of street gangs. Members of gang A venture to a part of town normally consid-

ic

ered to be gang B’s turf. Gang B perceives this lack of respect for gang bound-
er

aries as an affront to their power and influence. As a result, members of gang

Am

B retaliate in a manner that they perceive to be both justified and relatively

mild: They destroy several cars belonging to gang A’s leaders and assault sev-
e

eral members of gang A who try to stop the destruction of the cars. Gang B’s
th

retaliation therefore becomes gang A’s provocation, leading them to shoot and
by

kill several members of gang B. The escalation cycle continues over the course
of several weeks or months, with dozens of members of both gangs experienc-
11

ing serious physical injury or death. Real-world examples of the violence esca-
20

lation cycle abound in contemporary society. The Israeli–Palestinian conflict

and the “preemptive” war on Iraq by the United States are examples of vio-
©

lence escalation cycles.

ht

Why do violence escalations persist? We propose three reasons. First,

ig

violence often produces violence as a result of faulty attributions. Whereas

yr

neutral third parties can make accurate causal inferences regarding violence
op

between two parties, the parties themselves usually cannot. In a version of the
fundamental attribution error, people tend to explain the causes of others’
C

behaviors as due to dispositions and their own behavior as due to situational

THE GENERAL AGGRESSION MODEL 23

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 24

Inappropriate Appropriate
over-retaliation retaliation

B harms A

A)
P
Appropriate Inappropriate

(A
retaliation over-retaliation

n
io
A harms B

at
ci
Inappropriate Appropriate

so
over-retaliation retaliation

As
B harms A

al
ic
Unintentional Intentional

og
Justified Unjustified
Relatively mild
ol Relatively harmful
y ch
Ps

A harms B
an
ic
er

A’s perspectives Events B’s perspectives

Am

Figure 1.1. Violence escalation cycle. Adapted from “Creating Your Own Hostile
e

Environment: A Laboratory Examination of Trait Aggression and the Violence

th

Escalation Cycle,” by C. A. Anderson, K. E. Buckley, and N. L. Carnagey, 2008,

Personality and Social Psychology Bulletin, 34, pp. 462–473. Copyright 2008
by

by Sage. Adapted with permission.

11
20

forces (Anderson, Krull, & Weiner, 1996); people become caught in a web
in which members perceive the other party as acting out of malice or evil and
©

perceive their own behavior as appropriate responses to the situation at hand.

ht

Outgroup homogeneity effects may also prevent members of both parties from
ig

making accurate attributions. U.S. citizens, for example, may perceive all
yr

members of the Islamic faith as jihadists, when in reality this term represents
op

a minority of Muslims. Second, retaliation often exceeds prior levels of aggres-

sion. A push turns into a punch, a punch turns into shooting someone, shoot-
C

ing one person turns into shooting many people, and so on. Such escalation

24 DE WALL AND ANDERSON

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 25

is often an attempt to signal to the other side that it should back down (see
Chapters 3 and 12, this volume). Third, increased levels of retaliatory vio-
lence stem from perspective biases in which the most recent perpetrator views
the harmful behavior as appropriate and justified, whereas the most recent

A)
victim perceives the perpetrator’s act as an inappropriate overreaction.
The violence escalation cycle corresponds to the GAM’s assertion that

P
(A
personological and situational factors can influence each other in a dynamic
manner. Social psychology frequently demonstrates that powerful situational

n
factors can override personality traits. But personality traits (including beliefs,

io
attitudes, and insecurities) sometimes dramatically influence the situation

at
(see Chapters 4 and 5). People who characteristically perceive the world as

ci
hostile and who resolve conflict through the use of aggression can turn a sit-

so
uation that involves potentially mild conflict into a severely hostile one that

As
gives rises to escalating violence (Anderson, Buckley, & Carnagey, 2008;
Chapters 11, 13, and 14, this volume).

al
ic
og
USING THE GENERAL AGGRESSION MODEL TO UNDERSTAND
HOW PERPETRATORS OF VIOLENCE ARE CREATED ol
y ch

Scholars have often approached violence from two different perspec-

Ps

tives, one focusing on the development of aberrant individuals who become

violent criminals, the other focusing on how large segments of a population
an

become involved in “institutionalized” violence. The habitual violent offender

ic

is the prototype of the former, whereas various genocidal events (e.g., the
er

Holocaust, Rwanda) are exemplars of the latter. Other forms of aggression do

Am

not fit this dichotomy so well. For example, some “terrorists,” such as Okla-
homa City bomber Timothy McVeigh, do not seem to fit either. Larger and
e

more persistent terrorist groups seem more institutional (e.g., Irish Republican
th

Army, Hamas), but labeling such groups as “terrorist organizations” is to some

by

extent more of a political statement than an objective description. From the

standpoint of understanding how individuals become involved in violence
11

against others, it may be best to avoid the political labels whenever possible,
20

and focus research (and intervention) efforts on understanding the precursors.

A lot is known about the precursors of violence, and in our view the
©

GAM can be used to organize them all. One point to keep in mind, however,
ht

is that the GAM is intended to provide an overarching general view of aggres-

ig

sion. It is not a compendium of specific factors and microprocesses that are

yr

unique to each specific type of aggression and violence.

op

Social psychologists have long been interested in understanding how

“ordinary” people can carry out horrific acts of aggression and violence. Stan-
C

ley Milgram’s (1963) obedience to authority studies and Haney, Banks, and

THE GENERAL AGGRESSION MODEL 25

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 26

Zimbardo’s (1973) simulation of prison conditions shocked researchers and

laypersons by showing how easy it is to create a situation in which people will
behave aggressively toward strangers. The terrorist attacks on the World
Trade Center and the Pentagon on September 11, 2001; the proliferation of

A)
suicide bombers in the Middle East; and recent instances of waterboarding
and other brutal tactics used by members of the U.S. military continue to raise

P
(A
the question: How are terrorists, suicide bombers, and torturers created?
Subsequent chapters in this volume provide many more specific details

n
about the precursors of specific categories of violence. Many can be summarized

io
in a few statements. Many precursors can be seen as factors that promote the

at
development of individuals who are capable of and predisposed to use aggres-

ci
sion and violence to solve conflicts (e.g., Chapters 4, 5, and 7). Other precur-

so
sors are immediate situational factors (in the case of individual violence) or the

As
current social milieu, serving in some cases as triggers (i.e., precipitating causes)
or as factors that support ongoing violence (e.g., Chapters 9 and 16).

al
Beliefs, attitudes, and expectations supportive of violence must come from

ic
somewhere. And they do. If you want to create people who are predisposed to

og
aggression and violence, begin by depriving them of resources necessary to meet
ol
basic needs—physical, emotional, psychological, and social. Provide them with
ch
multiple examples of aggression and violence, examples in which such behav-
y

ior appears to work. Desensitize them to the images, sounds, smells, and, in gen-
Ps

eral, to the horrors of violence by exposing them to these stimuli, both live and
in electronic media form. Then provide them with a belief system that serves
an

to dehumanize potential targets, that justifies on moral grounds any and all
ic

forms of attacks on potential targets, and that minimizes negative consequences

er

to oneself and one’s social group while maximizing positive consequences in the
Am

near future and/or in an afterlife. Finally, if you want specific forms of violence
to emerge, provide training (i.e., the behavioral scripts) in those specific forms
e

of violence. Link these violence scripts to the social support system and the
th

belief systems that you have already provided, and you will have a group of peo-
by

ple who are quite prepared to behave violently. Put these people into the right
situation, and the desired violence will occur (Miller, 2004).
11
20

IMPLICATIONS OF THE GENERAL AGGRESSION MODEL

©

FOR AGGRESSION BETWEEN GROUPS OF PEOPLE

ht
ig

The GAM makes specific predictions about aggression not only between
yr

two people but also between groups of people large (e.g., nations) and small
op

(e.g., two or more people with a defined identity and common goal). The
majority of evidence supporting GAM, however, is derived from correlational,
C

experimental, and longitudinal research on the aggressive behavior of individ-

26 DE WALL AND ANDERSON

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 27

uals. Increasing the number of people involved in an episode changes features

of the situation in the same manner as other situational inputs (e.g., hot ambi-
ent temperatures, violent media), which in turn influences current internal
states and subsequent appraisal and decision processes.

A)
In a recent review, Meier, Hinsz, and Heimerdinger (2007) argued that
groups commit and receive more aggression than individuals. Although the

P
(A
aggression literature is dominated by research on aggression between individ-
uals, the available evidence on aggression between small groups supports this

n
view (Jaffe, Shapir, & Yinon, 1981; Meier & Hinsz, 2004; Wildschut, Pinter,

io
Vevea, Insko, & Schopler, 2003). The findings suggest that group size func-

at
tions as a situational factor that produces increased levels of aggressive behav-

ci
ior from both the initial perpetrator and the initial target.

so
According to the GAM, heightened aggression between groups (rela-

As
tive to individuals) results from increased levels of aggressive affect, arousal,
or cognition. Indeed, expecting to interact with an unfriendly group increases

al
hostile expectations (Hoyle, Pinkley, & Insko, 1989), and the presence of

ic
others increases arousal (Zajonc, 1965). Finally, terror management theory

og
suggests additional ways in which groups will become more embedded in esca-
ol
lating cycles of violence (Chapter 16, this volume).
y ch
Ps

GOVERNMENT ACTIONS AND THEIR IMPACT

ON AGGRESSION AND VIOLENCE
an
ic

Most, if not all, known human societies have governmental structures.

er

Governments have many functions, one of which is to protect citizens against

Am

aggression and violence from perpetrators within and outside the society.
Despite the good intentions behind many government actions, governmental
e

efforts to reduce aggression and violence often fail. Even worse, some govern-
th

mental actions designed to reduce aggression and violence actually increase

by

them. The GAM helps to explain why governmental programs designed to

reduce aggression and violence often fail or even enhance the likelihood that
11

people will respond to the program by behaving more aggressively and violently.
20

War offers an example of how government actions can affect aggression

and violence. Wars begin as a result of one or more events that affect two or
©

more nations, frequently by causing some harm or injustice to one or both

ht

parties. Often, the precipitating events involve disputes over resources,

ig

including not only traditional natural resources (e.g., land, water) but psycho-
yr

logical ones as well (e.g., access to religious sites, traditional homeland bound-
op

aries; Avalos, 2005). One nation responds in a manner that its citizens believe
is justified, whereas the other nation perceives the action to be unjustified and
C

overly harsh. An escalating cycle of violence between nations ensues, with

THE GENERAL AGGRESSION MODEL 27

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 28

each retaliation growing more violent. Over time, these governmental

actions cause citizens of each nation to develop aggression-relevant knowl-
edge structures regarding the enemy. Citizens develop and display more
aggressive personalities, which influence their construal and reaction to situ-

A)
ations and which change the nature of situations they will encounter in the
future. As a result, it is difficult for citizens to understand how their nation’s

P
(A
actions can be considered “evil” by members of the enemy nation and how
the enemy nation can feel justified in its retaliatory actions.

n
Consider how two ongoing conflicts—the U.S. War on Terrorism in

io
Afghanistan and Iraq and the Israeli–Palestinian conflict—fit the GAM’s

at
explanation of how governmental action can influence aggression and vio-

ci
lence. The U.S. War on Terrorism in Afghanistan and the war in Iraq are, in

so
large part, governmental responses to the terrorist attacks on the United

As
States on September 11, 2001, by members of al-Qaeda. Many U.S. citizens
perceive the actions taken by the United States as justified responses to the

al
terrorist attacks, but these same U.S. citizens appear baffled that Arab groups

ic
and other members of the Gulf region perceive U.S. governmental actions as

og
evil or at least unjustified overreactions. There is also evidence that the 9/11
ol
attacks changed aggression-related knowledge structures in a manner consis-
ch
tent with the GAM. After the 9/11 attacks, college students showed increased
y

positive attitudes toward war and more aggressive personalities (Carnagey &
Ps

Anderson, 2007). The increased aggressive attitudes about the war and the
increase in trait physical aggression persisted even a year after the terrorist
an

attacks. These findings provide at least indirect evidence that the U.S. gov-
ic

ernment’s actions in Afghanistan and Iraq affected attitudes toward war and
er

aggressive personality traits. The broader implication is that the U.S. govern-
Am

ment’s actions created not only hostile attitudes toward Arabs and Muslims
among U.S. citizens but also had the unplanned effect of pushing many people
e

who suffered from the U.S. military actions to join terrorist factions to retal-
th

iate against what they perceived as unwarranted and harsh attacks on their
by

nations. The U.S. War on Terrorism may therefore have created more U.S.
enemies than it killed.
11

The Israeli–Palestinian conflict grew out of violent outbreaks between

20

Israeli and Arab residents of the region alternately called Israel or Palestine,
a strip of land that Jews claim as their birthright and Palestinians claim as
©

their own. Persistent fighting between Israelis and Palestinians over the
ht

course of the past 60 years shows little sign of waning (see Chapters 16, 17,
ig

and 21). Although most members of each group perceive their own leaders’
yr

actions as justified, they have difficulty understanding how members of the

op

enemy group perceive their government’s actions as justified. The leaders of

both groups acknowledge that their actions aimed at reducing intergroup vio-
C

28 DE WALL AND ANDERSON

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 29

lence may actually increase the level of violence between the groups. For
example, Israel’s top generals and intelligence officers have admitted that
their military actions in response to Palestinian suicide bombings have had
the effect of creating additional Palestinian terrorist cells (Moore, 2003).

A)
Thus, the GAM provides a useful framework for understanding how govern-
mental actions can produce an escalating cycle of violence between groups

P
(A
and even nations.
These two examples illustrate another key point about escalatory vio-

n
lence. When the two parties in a conflict have vastly different resources, their

io
forms of violence and of escalation will vary. Al-Qaeda cannot launch a con-

at
ventional war against the West, so its attacks include unconventional forms

ci
of violence, such as roadside or suicide bombings. Similarly, the Palestinians

so
cannot win a conventional war with Israel, so they resort to terrorist tactics.

As
al
USING THE GENERAL AGGRESSION MODEL TO REDUCE

ic
AGGRESSION AND VIOLENCE

og
ol
Thus far, we have dwelled on how the GAM can help to describe, pre-
ch
dict, and explain aggressive behavior. The GAM can also help researchers,
y

government officials, and laypersons understand how to control or prevent

Ps

aggression. The most likely points for intervention will vary from case to case,
but several stand out. In the case of groups or individuals already in conflict,
an

the first step should be to stop the violence cycle. Interventions by third par-
ic

ties may be necessary. A second step should be to ensure that people’s basic
er

needs are met. A third step would be to address people’s symbolic needs (e.g.,
Am

by giving them access to religious sites).

Longer term interventions should focus on reducing the risk factors that
e

cause individuals to be predisposed to aggression. This is likely to be most

th

effective in the case of individuals who have not already become aggression-
by

prone, either in general or specifically toward an enemy outgroup. Thus, one

must provide for the healthy development of children, both physical and psy-
11

chological. This includes the learning of basic attitudes, beliefs, and values
20

that foster positive social interactions—even with outgroups—and encourage

nonviolent problem solving.
©

Although retraining people who are already violence prone is more dif-
ht

ficult, research in the violent crime domain has found that intensive inter-
ig

ventions with high-risk youth can be highly successful (U.S. Department of

yr

Health and Human Services, 2001). This is a different conclusion from the
op

one that most scholars and public policy makers in the United States held as
recently as the late 1980s and early 1990s.
C

THE GENERAL AGGRESSION MODEL 29

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 30

CONCLUSIONS

In summary, the GAM integrates several domain-specific theories of

human aggression to form a general understanding of why people behave

A)
aggressively. It identifies a wide range of factors that influence the develop-
ment of aggressive tendencies over time. It explains how highly aggressive

P
(A
contexts are created and maintained through violence escalation cycles. It
clarifies why government actions designed to bring about peace often fail, cre-

n
ating even more violent conflict. And perhaps most important, it offers pos-

io
sible solutions for preventing and reducing aggression and violence both

at
between individuals and groups.

ci
so
As
REFERENCES

al
Anderson, C. A., Buckley, K. E., & Carnagey, N. L. (2008). Creating your own hostile

ic
environment: A laboratory examination of trait aggression and the violence esca-

og
lation cycle. Personality and Social Psychology Bulletin, 34, 462–473. doi:10.1177/
0146167207311282 ol
ch
Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of
y

Psychology, 53, 27–51. doi:10.1146/annurev.psych.53.100901.135231

Ps

Anderson, C. A., Carnagey, N. L., & Eubanks, J. (2003). Exposure to violent media:
The effects of songs with violent lyrics on aggressive thoughts and feelings. Journal
an

of Personality and Social Psychology, 84, 960–971. doi:10.1037/0022-3514.84.5.960

ic

Anderson, C. A., & Huesmann, L. R. (2003). Human aggression: A social–cognitive

er

view. In M. A. Hogg & J. Cooper (Eds.), Handbook of social psychology

Am

(pp. 296–323). London, England: Sage.

Anderson, C. A., Krull, D. S., & Weiner, B. (1996). Explanations: Processes and
e

consequences. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology:

th

Handbook of basic principles (pp. 271–296). New York, NY: Guilford Press.
by

Avalos, H. (2005). Fighting words: The origins of religious violence. Buffalo, NY:
Prometheus Books.
11

Bandura, A. (1977). Social learning theory. New York, NY: Prentice Hall.
20

Bargh, J. A. (1996). Automaticity in social psychology. In E. T. Higgins & A. W.

Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 169–183).
©

New York, NY: Guilford Press.

ht

Bargh, J. A., & Pietromonaco, P. (1982). Automatic information processing and social
ig

perception: The influence of trait information presented outside of conscious

yr

awareness on impression formation. Journal of Personality and Social Psychology,

op

43, 437–449. doi:10.1037/0022-3514.43.3.437

Baron, R. A., & Richardson, D. R. (1994). Human aggression (2nd ed.). New York,
C

NY: Plenum Press.

30 DE WALL AND ANDERSON

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 31

Bartholow, B. D., Bushman, B. J., & Sestir, M. A. (2006). Chronic violent video
game exposure and desensitization: Behavioral and event-related brain poten-
tial data. Journal of Experimental Social Psychology, 42, 532–539. doi:10.1016/
j.jesp.2005.08.006

A)
Berkowitz, L. (1989). Frustration–aggression hypothesis: Examination and reformu-
lation. Psychological Bulletin, 106, 59–73. doi:10.1037/0033-2909.106.1.59

P
(A
Bushman, B. J., & Anderson, C. A. (2001). Is it time to pull the plug on the hostile
versus instrumental aggression dichotomy? Psychological Review, 108, 273–279.

n
doi:10.1037/0033-295X.108.1.273

io
Bushman, B. J., & Anderson, C. A. (2002). Violent video games and hostile expec-

at
tations: A test of the general aggression model. Personality and Social Psychology

ci
Bulletin, 28, 1679–1686. doi:10.1177/014616702237649

so
Bushman, B. J., & Baumeister, R. F. (1998). Threatened egotism, narcissism, self-

As
esteem, and direct and displaced aggression: Does self-love or self-hate lead to
violence? Journal of Personality and Social Psychology, 75, 219–229. doi:10.1037/

al
0022-3514.75.1.219

ic
Carnagey, N. L., & Anderson, C. A. (2005). The effects of reward and punishment

og
in violent video games on aggressive affect, cognition, and behavior. Psycho-
ol
logical Science, 16, 882–889. doi:10.1111/j.1467-9280.2005.01632.x
ch
Carnagey, N. L., & Anderson, C. A. (2007). Changes in attitudes towards war and
y

violence after September 11, 2001. Aggressive Behavior, 33, 118–129. doi:10.1002/
Ps

ab.20173
Carnagey, N. L., Anderson, C. A., & Bushman, B. J. (2007). The effect of video game
an

violence on physiological desensitization to real-life violence. Journal of Exper-

ic

imental Social Psychology, 43, 489–496. doi:10.1016/j.jesp.2006.05.003

er

Cooper, M. (1999, February 5). Officers in Bronx fire 41 shots, and an unarmed man
Am

is killed. The New York Times. Retrieved from http://www.nytimes.com/1999/

02/05/nyregion/officers-in-bronx-fire-41-shots-and-an-unarmed-man-is-
e

killed.html
th

Correll, J., Park, B., Judd, C. M., & Wittenbrink, B. (2002). The police officer’s
by

dilemma: Using ethnicity to disambiguate potentially threatening individu-

als. Journal of Personality and Social Psychology, 83, 1314–1329. doi:10.1037/
11

0022-3514.83.6.1314
20

Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social informa-
tion processing mechanisms in children’s adjustment. Psychological Bulletin, 115,
©

74–101. doi:10.1037/0033-2909.115.1.74
ht

DeWall, C. N., Baumeister, R. F., Stillman, T. F., & Gailliot, M. T. (2007). Violence
ig

restrained: Effects of self-regulatory capacity and its depletion on aggressive

yr

behavior. Journal of Experimental Social Psychology, 43, 62–76. doi:10.1016/

op

j.jesp.2005.12.005
DeWall, C. N., Twenge, J. M., Gitter, S. A., & Baumeister, R. F. (2009). It’s the
C

thought that counts: The role of hostile cognition in shaping aggressive responses

THE GENERAL AGGRESSION MODEL 31

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 32

to social exclusion. Journal of Personality and Social Psychology, 96, 45–59. doi:
10.1037/a0013196
Dodge, K. A. (1980). Social cognition and children’s aggressive behavior. Child
Development, 51, 162–170. doi:10.2307/1129603

A)
Finkel, E. J., DeWall, C. N., Slotter, E. B., Oaten, M., & Foshee, V. A. (2009). Self-
regulatory failure and intimate partner violence perpetration. Journal of Person-

P
ality and Social Psychology, 97, 483–499. doi:10.1037/a0015433

(A
Fiske, S. T., & Taylor, S. E. (1991). Social cognition (2nd ed.). New York, NY:

n
McGraw-Hill.

io
Gailliot, M. T., Baumeister, R. F., DeWall, C. N., Maner, J. K., Plant, E. A., Tice,

at
D. M., . . . Schmeichel, B. J. (2007). Self-control relies on glucose as a limited

ci
energy source: Willpower is more than a metaphor. Journal of Personality and

so
Social Psychology, 92, 325–336. doi:10.1037/0022-3514.92.2.325

As
Ghosh, B. (2005, June 28). Inside the mind of an Iraqi suicide bomber. Time. Retrieved
from http://www.time.com/time/magazine/article/0,9171,1077288,00.html

al
Haney, C., Banks, W. C., & Zimbardo, P. G. (1973). Interpersonal dynamics in a

ic
simulated prison. International Journal of Criminology and Penology, 1, 69–97.

og
Hare, R. D. (1996). Psychopathy: A clinical construct whose time has come. Criminal
ol
Justice and Behavior, 23, 25–54. doi:10.1177/0093854896023001004
ch
Hoyle, R. H., Pinkley, R. L., & Insko, C. A. (1989). Perceptions of social behav-
y

ior: Evidence of differing expectations for interpersonal and intergroup inter-

Ps

action. Personality and Social Psychology Bulletin, 15, 365–376. doi:10.1177/

0146167289153007
an

Huesmann, L. R. (1988). An information processing model for the development of

ic

aggression. Aggressive Behavior, 14, 13–24. doi:10.1002/1098-2337(1988)14:

er

1<13::AID-AB2480140104>3.0.CO;2-J
Am

Jaffe, Y., Shapir, N., & Yinon, Y. (1981). Aggression and its escalation. Journal of
Cross-Cultural Psychology, 12, 21–36. doi:10.1177/0022022181121002
e

Meier, B. P., & Hinsz, V. B. (2004). A comparison of human aggression committed

th

by groups and individuals: An interindividual–intergroup discontinuity. Journal

by

of Experimental Social Psychology, 40, 551–559. doi:10.1016/j.jesp.2003.11.002

Meier, B. P., Hinsz, V. B., & Heimerdinger, S. R. (2007). A framework for explain-
11

ing aggression involving groups. Social and Personality Psychology Compass, 1,

298–312. doi:10.1111/j.1751-9004.2007.00015.x
20

Milgram, S. (1963). Behavioral study of obedience. Journal of Abnormal and Social

©

Psychology, 67, 371–378. doi:10.1037/h0040525

ht

Miller, A. (Ed.). (2004). The social psychology of good and evil (pp. 168–192). New
ig

York, NY: Guilford Press.

yr

Mischel, W. (1973). Toward a cognitive social learning reconceptualization of per-

op

sonality. Psychological Review, 80, 252–283. doi:10.1037/h0035002

Mischel, W., & Shoda, Y. (1995). A cognitive–affective system theory of personal-
C

ity: Reconceptualizing situations, dispositions, dynamics, and invariance in

32 DE WALL AND ANDERSON

12051-02_PT1-CH01-rev2.qxd 5/25/10 11:29 AM Page 33

personality structure. Psychological Review, 102, 246–268. doi:10.1037/0033-

295X.102.2.246
Moore, M. (2003, October 31). Top Israeli officer says tactics are backfiring. The
Washington Post, p. A1.

A)
Neuilly, M.-A. (2007). When murder is not enough: Toward a new definition of com-
munity violence. Aggression and Violent Behavior, 12, 598–610. doi:10.1016/

P
(A
j.avb.2007.02.004
Öhman, A., Lundqvist, D., & Esteves, F. (2001). The face in the crowd effect: An

n
anger superiority effect with schematic stimuli. Journal of Personality and Social

io
Psychology, 80, 381–396.

at
Plant, E. A., & Peruche, B. M. (2005). The consequences of race for police officers’

ci
responses to criminal suspects. Psychological Science, 16, 180–183. doi:10.1111/

so
j.0956-7976.2005.00800.x

As
Smith, C. A., & Lazarus, R. S. (1993). Appraisal components, core relational
themes, and the emotions. Cognition and Emotion, 7, 233–269. doi:10.1080/

al
02699939308409189

ic
og
Strack, F., & Deutsch, R. (2004). Reflective and impulsive determinants of social
behavior. Personality and Social Psychology Review, 8, 220–247. doi:10.1207/
s15327957pspr0803_1 ol
ch
Tooby, J., & Cosmides, L. (1992). The psychological foundations of culture. In
y

J. Barkow, L. Cosmides, & J. Tooby (Eds.), The adapted mind (pp. 19–136). New
Ps

York, NY: Oxford University Press.

U.S. Department of Health and Human Services. (2001). Youth violence: A report of
an

the Surgeon General. Rockville, MD: Author.

ic

Wegner, D. M., & Bargh, J. A. (1998). Control and automaticity in social life. In
er

D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology

Am

(4th ed., Vol. 1, pp. 446–496). New York, NY: McGraw-Hill.

Wildschut, T., Pinter, B., Vevea, J. L., Insko, C. A., & Schopler, J. (2003). Beyond the
e

group mind: A quantitative review of the interindividual–intergroup discontinu-

th

ity effect. Psychological Bulletin, 129, 698–722. doi:10.1037/0033-2909.129.5.698

by

Zajonc, R. B. (1965, July 16). Social facilitation. Science, 149, 269–274. doi:10.1126/
science.149.3681.269
11
20
©
ht
ig
yr
op
C

THE GENERAL AGGRESSION MODEL 33

C
op
yr
ig
ht
©
20
12051-02_PT1-CH01-rev2.qxd

11
by
5/25/10

th
e
Am
11:29 AM

er
ic
an
Page 34

Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 35

PA)
I3 THEORY: INSTIGATING,

(A
IMPELLING, AND INHIBITING

n
io
FACTORS IN AGGRESSION

at
ci
so
ERICA B. SLOTTER AND ELI J. FINKEL

As
al
ic
og
ol
y ch
Ps
an

Interpersonal aggression is prevalent and disturbing. This chapter pres-

ic

ents a metatheoretical perspective, I3 theory, that seeks (a) to impose theoret-

er

ical coherence on the massive number of established risk factors for aggression
Am

and (b) to use the tools of statistical (and conceptual) moderation to gain new
insights into the processes by which a previously nonaggressive interaction
e

escalates into an aggressive one (see Finkel, 2007, 2008). I3 theory (pronounced
th

“I-cubed theory”) does not advance one key variable (or even a specific set of
by

key variables) as the root cause of aggression. Rather, it seeks to present an

organizational structure for understanding both (a) the process by which a
11

given risk factor promotes aggression and (b) how multiple risk factors inter-
20

relate to aggravate or mitigate the aggression-promoting tendencies of each.

As detailed in this chapter, I3 theory suggests that scholars can predict whether
©

an individual will behave aggressively in a given situation by examining the

ht

main and interactive effects of the instigating triggers, impelling forces, and
ig

inhibiting forces at play.

yr

Scholars have advanced a broad range of theories to understand aggres-

op

sion, which in this chapter refers to any behavior carried out with the primary
proximal goal of inflicting physical harm on a target who is motivated to
C

avoid being harmed (Baron & Richardson, 1994). (We do not examine in

35
12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 36

this chapter other forms of aggression, such as verbal, relational, or sexual

aggression.) Craig Anderson and colleagues have sought to integrate many of
these theories into a broad metatheory called the general aggression model
(GAM; Anderson & Bushman, 2002; see Chapter 1, this volume). As with

A)
I3 theory, the GAM focuses less on a particular variable or process than on
general classes of aggression risk factors and processes. The GAM consists

P
(A
of three main foci. The first emphasizes person and situation inputs, or risk
factors, for aggression. Person inputs include personality traits, gender, beliefs,

n
attitudes, values, long-term goals, and scripts; situation inputs include aggressive

io
cues (e.g., presence of guns), provocation, frustration, pain and discomfort,

at
drugs, and incentives (determined by a cost/benefit analysis). The second focus

ci
is the interconnected affective, arousal, and cognitive routes, or mechanisms,

so
through which the inputs influence aggressive behavior. Affective routes include

As
mood and emotion and expressive motor tendencies; arousal routes include
the strengthening of a dominant action tendency or certain misattribution

al
processes; cognitive routes include hostile thoughts and scripts. Finally, the

ic
third focus is the outcomes of the underlying appraisal and decision processes.

og
Individuals are likely to act impulsively when they lack the resources and
ol
motivation to alter their immediate appraisal of the situation. If they possess
ch
the resources and motivation, however, they may reappraise the situation and
y

act in a more thoughtful fashion.

Ps
an

I3 THEORY
ic
er

I3 theory, which is a process-oriented metatheory designed to identify the

Am

circumstances under which a nonaggressive interaction can become an aggres-

sive one, has different emphases from the general aggression model. For exam-
e

ple, although both metatheories seek to integrate extant theories of aggression

th

into a broad, coherent model, I3 theory incorporates recent research on self-

by

regulation as a core emphasis of the model, and it specifies the novel ways in
which aggression risk factors interact to predict aggressive behavior.
11

The theory begins by posing three questions. First, does at least one indi-
20

vidual in the interaction experience strong instigating triggers toward aggres-

sion? Second, does that individual experience strong impelling forces toward
©

aggression? Third, does that individual experience weak forces to inhibit or

ht

override the aggressive impulses? Each affirmative answer increases the like-
ig

lihood of aggressive behavior via both a main effect and interactive effects
yr

with variables relevant to one or both of the other questions. Whereas the
op

strength of impelling forces is determined by the collective power of the vari-

ables that cause the individual to experience an urge to aggress in response to
C

a given instigating trigger, the strength of inhibiting forces is determined by

36 SLOTTER AND FINKEL

12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 37

the collective power of the variables that cause the individual to override this
aggressive urge.
In addition to these three initial questions, I3 theory poses a fourth: How
do effects of variables in one category (i.e., instigating triggers, impelling forces,

A)
or inhibiting forces) interact with effects of one or more variables from the other
categories to predict aggressive behavior? As presented in Table 2.1, answering

P
(A
these four questions enables scholars to identify seven key I3 theory effects. Fig-
ure 2.1 (which builds on work by Fals-Stewart, Leonard, & Birchler, 2005) illus-

n
trates how these seven effects can work together to increase or decrease the

io
likelihood of aggressive behavior.

at
I3 theory diverges from the aggression theories mentioned earlier in its

ci
central emphasis on inhibitory processes. The theory recognizes the impor-

so
tance of instigating triggers and impelling forces, but it argues that such fac-

As
tors cause individuals to enact aggressive behavior only when their collective
power is stronger than the collective power of inhibitory processes. Although

al
other theories address the importance of inhibitory processes in aggression

ic
(e.g., Chapters 1, 6, 9, and 15), the emphasis on such processes gains new

og
prominence with I3 theory.
ol
ch
Instigating Triggers
y
Ps

The first stage of I3 theory concerns the presence of one or more instigat-
ing triggers, which are discrete situational events or circumstances that induce
an

rudimentary action tendencies toward physical aggression. As illustrated at the

ic

lower left of Figure 2.1, impelling and inhibiting forces are irrelevant when
er

instigating triggers are absent. Even the world’s angriest, least controlled per-
Am

son is not aggressive all the time; some situational variable (even if it only
serves to activate a long-standing goal or grievance) is required before the per-
e

son becomes aggressive. Aversive events can trigger (via automatic associa-
th

tive networks or cognitive appraisal processes) hostile cognitive, affective,

by

physiological, and even motor tendencies that prime the individual to aggress
(Berkowitz, 1993; Chapters 1 and 9, this volume). I3 theory suggests that cer-
11

tain events can also trigger aggressive tendencies driven by instrumental goals
20

(e.g., having an individual offer you money to beat up his enemy).

Instigating triggers fall into one of two categories: dyadic and third-
©

party. Dyadic triggers refer to events or circumstances that the potentially

ht

aggressive individual perceives as having originated in the target. Examples

ig

include direct provocation (Bettencourt & Miller, 1996), goal obstruction

yr

(Dollard, Doob, Miller, Mowrer, & Sears, 1939), and social rejection (Leary,
op

Twenge, & Quinlivan, 2006). Third-party triggers refer to events or circum-

stances that the potentially aggressive individual perceives as having origi-
C

nated in somebody other than the target. The same kinds of triggers that lead

I3 THEORY 37
C
op
yr
ig
ht

38
©
12051-03_CH02-rev2.qxd

20
11
5/25/10

by
TABLE 2.1
3 The Seven Effects
th I Theory of Aggression:
I3 Effect
e
11:29 AM

(#) I3 Effect (Stage) I3 Effect (Description) Example Citation for example

SLOTTER AND FINKEL

Am
1 Stage 1 Instigating Trigger Main Effect Social Rejection Twenge et al., 2001
2 Stage 2 Impelling Forces Main Effect Testosterone Dabbs et al., 1987
er
3 Stage 3
ic
Inhibiting Forces Main Effect Self-Regulatory Strength DeWall et al., 2007
Page 38

4 Stage 1 × Stage 2 Instigating Trigger × Impelling Ego Threat × Narcissism Bushman &
an
Forces Interaction Effect Baumeister, 1998
5 Stage 1 × Stage 3 Instigating Trigger × Inhibiting Provocation × Self-Regulatory Finkel et al., 2009
Forces Interaction Effect Strength
Ps
6 Stage 2 × Stage 3 Impelling Forces × Inhibiting
yc Physical Proclivity × Negative Finkel & Foshee,
Forces Interaction Effect Outcome Expectancies 2009
7 Stage 1 × Stage 2 Instigating Trigger × Impelling [No Known Example] [No Known Example]
ho
× Stage 3 Forces × Inhibiting Forces lo
Interaction Effect gi
ca
lA
s so
ci
at
io
n
(A
P A)
12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 39

Strong Impelling Forces

Intensity of Violent Impulses Weak Impelling Forces

Violence Threshold
Given Strong

A)
Inhibiting Forces

P
(A
n
io
Violence Threshold

at
Given Weak
Inhibiting Forces

ci
so
Absent Moderate Extreme

As
Intensity of Instigating Triggers

al
Figure 2.1. How the three components of I3 theory interrelate to predict aggressive
behavior. For ease of illustration, impelling forces and inhibiting forces are depicted

ic
as if they are binary—either weak or strong. In reality, the intensity of each type of

og
force varies continuously from weak to strong.
ol
ch
y

to a rudimentary action tendency to aggress against the provocateur can also

Ps

lead to this tendency with respect to a third party. For example, an individ-
ual who feels provoked or rejected may experience an instigation to aggress
an

not only against the source of the provocation or rejection but also (or alter-
ic

nately) toward another target whom the potential perpetrator believes would
er

make a more acceptable or desirable target (e.g., somebody who is less likely
Am

to fight back).
e

Impelling Forces
th
by

The second stage of I3 theory concerns risk factors that determine the
strength of the aggressive impulse experienced by the individual, through
11

main effects and through interactions with instigating triggers. In some situ-
20

ations, individuals may effortlessly shrug off (or perhaps not even notice; see
Chapter 9) an instigating trigger, experiencing virtually no impulse toward
©

aggression. In others, individuals may react strongly to a trigger, experiencing

ht

a powerful impulse toward aggression. Impelling forces refer to factors that

ig

increase the likelihood that individuals will experience an aggressive impulse

yr

in response to an instigating trigger. Individuals tend to experience more

op

powerful aggressive impulses when impelling forces are strong than when they
are weak (see dashed vs. dotted lines in Figure 2.1), especially to the degree
C

that the instigating trigger is severe.

I3 THEORY 39
12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 40

Impelling forces fall into one of four categories: evolutionary and cul-
tural, personal, dyadic, and situational. Evolutionary and cultural impellors refer
to features of the potentially aggressive individual’s biological or cultural her-
itage, including evolutionary adaptations and social norms (see Chapters 3

A)
and 10). Examples include adaptations resulting from evolutionary pressures
that provided ancestral men and women with a survival advantage for expe-

P
(A
riencing violent impulses in certain situations (Lorenz, 1966) and social
norms delineating the extent to which certain instigating triggers provoke

n
strong aggressive impulses (Nisbett & Cohen, 1996). Personal impellors refer

io
to relatively stable characteristics of a given individual that differ from

at
those of many other individuals, including personality characteristics, atti-

ci
so
tudes, beliefs, interpersonal interaction styles, or biological factors. Exam-
ples include dispositional hostility (Norlander & Eckhardt, 2005; Chapters

As
5 and 8, this volume), narcissism (Twenge & Campbell, 2003; Chapter 11,
this volume), and testosterone (Dabbs, Frady, Carr, & Besch, 1987; Van

al
ic
Goozen, Frijda, & Van de Poll, 1994). Dyadic impellors refer to characteris-

og
tics of the relationship between the potential aggressor and the potential
target. Examples include dissatisfaction with the amount of power one has
ol
in a relationship (Ronfeldt, Kimerling, & Arias, 1998), target-specific jeal-
ch
ousy (Dutton, van Ginkel, & Landolt, 1996; Holtzworth-Munroe, Stuart,
y

& Hutchinson, 1997), and feelings of vulnerability or insecurity in the relation-

Ps

ship (Carney & Buttell, 2005; Holtzworth-Munroe et al., 1997; Chapters 13

an

and 14, this volume). Finally, situational impellors refer to momentarily

activated cognitive, affective, or physiological experiences. Examples include
ic

uncomfortable temperatures (Anderson, Anderson, Dorr, DeNeve, & Flanagan,

er

2000), physical pain (Berkowitz, 1998), and exposure to violent media

Am

(Anderson & Bushman, 2001; Anderson, Carnagey, & Eubanks, 2003).

e

Inhibiting Forces
th
by

The third stage of I3 theory concerns risk factors that determine whether
11

individuals will override the aggressive impulses that emerge from the instigat-
ing triggers, impelling factors, and their interaction. In some situations, indi-
20

viduals succumb to these impulses, engaging in aggressive behavior. In others,

©

individuals override them in favor of nonviolent behavior. Inhibiting forces refer

to factors that increase the likelihood that individuals will override aggressive
ht

impulses rather than acting on them. Inhibiting factors collectively determine

ig

the threshold above which aggressive impulses will manifest themselves in

yr

aggressive behavior. If the inhibiting forces are weak (i.e., the lower horizontal
op

line in Figure 2.1), then aggressive impulses need not be especially strong to
C

result in aggressive behavior. If the inhibiting forces are strong (i.e., the upper

40 SLOTTER AND FINKEL

12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 41

horizontal line in Figure 2.1), then aggressive impulses must be strong to result
in aggressive behavior.
As with impelling forces, inhibiting forces fall into one of four categories:
evolutionary and cultural, personal, dyadic, and situational. Examples of evo-

A)
lutionary and cultural inhibitors include adaptations resulting from evolutionary
pressures that provided ancestral men and women with a survival advantage

P
(A
for overriding aggressive impulses in certain situations (Baumeister, 2005;
Chapter 3, this volume) and social norms or institutions that decrease the like-

n
lihood that individuals will act on aggressive impulses (Guerra, Huesmann, &

io
Spindler, 2003; Sampson, Raudenbush, & Earls, 1997; Chapter 10, this vol-

at
ume). Examples of personal inhibitors include dispositional self-control (Finkel,

ci
DeWall, Slotter, Oaten, & Foshee, 2009), executive functioning (Giancola,

so
2000; Chapter 6, this volume), and beliefs that enacting aggressive behavior

As
will lead to poor outcomes for the self (Slaby & Guerra, 1988). Examples of
dyadic inhibitors include partner empathy or perspective taking (Richardson,

al
Green, & Lago, 1998; Van Baardewijk, Stegge, Bushman, & Vermeiren, in

ic
press), relationship commitment (Gaertner & Foshee, 1999; Slotter, Finkel,

og
& Bodenhausen, 2009), and relative physical size (Archer & Benson, 2008;
ol
Felson, 1996; Chapter 3, this volume). Finally, examples of situational
ch
inhibitors include sobriety (i.e., vs. alcohol intoxication; Bushman & Cooper,
y

1990; Denson et al., 2008), nondepleted self-regulatory resources (DeWall,

Ps

Baumeister, Stillman, & Gailliot, 2007; Finkel et al., 2009; Chapter 6, this
volume), and plentiful cognitive processing time (Finkel et al., 2009).
an
ic

Reviewing the Aggression Literature From the Perspective of I3 Theory

er
Am

One purpose of I3 theory is to provide a coherent framework for catego-

rizing aggression risk factors and examining the interplay among them. To
illustrate how I3 theory can accomplish these goals, we review key findings in the
e
th

aggression literature through its lens, with a particular emphasis on interaction

effects. This review is illustrative rather than exhaustive.
by

I3 theory encompasses seven key effects: three main effects (i.e., insti-
11

gating triggers, impelling forces, and inhibiting forces), three two-way inter-
action effects (i.e., instigating triggers × impelling forces, instigating triggers
20

× inhibiting forces, and impelling forces × inhibiting forces), and one three-
way interaction effect (instigating triggers × impelling forces × inhibiting
©

forces). Table 2.1 lists these seven effects and provides an example of each.
ht

We discuss examples of these effects in turn.

ig
yr

1. An Illustrative Main Effect of Instigating Triggers: Social Rejection

op

As shown in the first row of Table 2.1, an illustrative instigating trigger

C

is social rejection, which refers to a class of interpersonal processes in which

I3 THEORY 41
12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 42

individuals feel rejected, excluded, or ostracized. Diverse lines of evidence

demonstrate that individuals who experience social rejection are more aggres-
sive than individuals who do not (Leary et al., 2006). In one study, partici-
pants who had been unanimously rejected by a group of fellow participants

A)
administered substantially louder, more painful sound blasts to an unknown
stranger than did participants who had been unanimously accepted (Twenge,

P
(A
Baumeister, Tice, & Stucke, 2001, Study 5). In this study, social rejection
served as a displaced instigating trigger because the target of the aggression

n
was not a member of the group who had previously rejected the participant.

io
Another study, which provided an in-depth analysis of all well-documented

at
school shootings in the United States between 1995 and 2001, yielded com-

ci
patible conclusions, with acute or chronic rejection preceding the shootings

so
in 87% of the cases (Leary, Kowalski, Smith, & Phillips, 2003).

As
2. An Illustrative Main Effect of Impelling Forces: Testosterone

al
ic
As shown in the second row of Table 2.1, an illustrative impelling force

og
is the androgen testosterone. Although testosterone is higher in men than in
ol
women, its level predicts aggression in both sexes (Archer, Birring, & Wu,
ch
1998; Dabbs & Hargrove, 1997; see Sapolsky, 1998). In a study of male prison
inmates, testosterone levels correlated positively with crime severity; indeed,
y
Ps

nine of the 11 inmates with the lowest testosterone levels had committed
nonviolent crimes, whereas 10 of the 11 inmates with the highest testos-
an

terone levels had committed violent crimes (Dabbs et al., 1987). In another
study, female-to-male transsexuals became considerably more aggressive in
ic
er

the first 3 months of androgen injections (Van Goozen, Cohen-Kettenis,

Gooren, Frijda, & Van de Poll, 1995).
Am

3. An Illustrative Main Effect of Inhibiting Forces: Self-Regulatory Strength

e
th

As shown in the third row of Table 2.1, an illustrative inhibiting factor

by

is self-regulatory strength, which refers to the psychological resource that

undergirds willful acts of self-regulation. According to the strength model of
11

self-regulation (Baumeister, Gailliot, DeWall, & Oaten, 2006), all such acts
20

depend on a unitary resource that resembles a muscle. The strength of this

resource can be temporarily diminished by self-regulatory exertions (leading
©

to a state of “ego depletion”), just as holding a heavy weight fatigues a muscle

ht

in the short term but it also can be bolstered over time by adherence to a
ig

self-regulatory bolstering “regimen,” just as a consistent weight-lifting regi-

yr

men strengthens a muscle over time. In one study, hungry participants who
op

had, moments earlier, exerted self-regulation by resisting the temptation to

eat an indulgent food (a donut; high ego depletion condition) were more
C

aggressive toward a same-sex provoking interaction partner (forcing this part-

42 SLOTTER AND FINKEL

12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 43

ner to eat a snack with plentiful hot sauce despite his or her distaste for spicy
foods) than were participants who had previously resisted eating a less tempt-
ing food (radishes; low ego depletion condition), even though participants in
the two conditions did not differ in how angry they were in response to the

A)
provocation (DeWall et al., 2007, Study 1).
Complementing this evidence that short-term self-regulatory exertions

P
(A
can deplete self-regulatory resources and thereby predict elevated aggression
is evidence that longer term self-regulatory exertion regimens can bolster self-

n
regulatory strength and thereby predict reduced aggression. A recent study

io
demonstrated that individuals who had adhered to a 2-week self-regulatory

at
strength-bolstering regimen declined significantly from before to after the reg-

ci
imen in their aggressive tendencies toward their romantic partner (Finkel

so
et al., 2009, Study 5). In this study, participants who deliberately regulated

As
either their physical behavior (e.g., brushing their teeth with their non-
dominant hand) or their verbal behavior (e.g., making sure not to begin sen-

al
tences with “I”) reported a reduced likelihood of being physically aggressive

ic
in response to various partner provocations (e.g., “I walk in and catch my part-

og
ner having sex with someone”), whereas participants in a no-intervention
ol
condition exhibited no change from pretest to posttest.
ch
4. An Illustrative Instigating Trigger × Impelling Forces Interaction Effect:
y
Ps

Ego Threat × Narcissism

We now turn from I3 theory’s three main effects to its three two-way
an

interaction effects. As shown in the fourth row of Table 2.1, an illustrative

ic

instigating trigger × impelling forces interaction effect is ego threat × narcissism.

er

Although many scholars have suggested that low self-esteem causes aggres-
Am

sion, others have increasingly argued that a form of high self-esteem is fre-
quently more likely to do so. In particular, individuals whose self-views are
e

not only favorable but also unstable, inflated, or uncertain are especially
th

prone toward aggressive behavior when their favorable self-views are socially
by

threatened (Baumeister, Smart, & Boden, 1996; Chapter 11, this volume).
One series of studies demonstrated that participants who had experienced an
11

ego threat in the form of insulting feedback about an essay they had written
20

(an instigating trigger) were more aggressive toward the same-sex provoca-
teur (subjecting him or her to painfully loud noise blasts) than were partic-
©

ipants who had not experienced an ego threat (Bushman & Baumeister,
ht

1998). The key finding, however, was that this main effect of ego threat was
ig

substantially larger for participants who were high in narcissism (an impelling
yr

factor) than for participants who were low in narcissism.

op

Additional research has examined how self-views moderate the link

between other instigating triggers (aside from insults) and aggression. One
C

study demonstrated that the link between social rejection and displaced

I3 THEORY 43
12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 44

aggression (painful noise blasts) was substantially stronger for individuals who
were high in narcissism than for those who were low in narcissism (Twenge &
Campbell, 2003, Study 4), and another demonstrated that the link between
social rejection and aggression (aversive hot sauce) was substantially stronger

A)
for individuals who were high in rejection sensitivity (those who anxiously
expect, readily perceive, and overreact to rejection) than for those who were

P
(A
low in rejection sensitivity (Ayduk, Gyurak, & Luerssen, 2008).

n
5. An Illustrative Instigating Trigger × Inhibiting Forces Interaction Effect:

io
Provocation × Self-Regulatory Strength

at
ci
As shown in the fifth row of Table 2.1, an illustrative instigating trigger

so
× inhibiting forces interaction effect is provocation × self-regulatory strength. As
discussed previously, both provocation and self-regulatory strength predict

As
aggression via main effects. Recent studies confirm the I3 theory prediction that

al
incorporating their interaction effect yields a richer story (e.g., Chapter 6). Two

ic
recent experiments, one in which the aggression was directed at strangers (aver-

og
sive sound blasts; DeWall et al., 2007, Study 2) and one in which the aggres-
ol
sion was directed at one’s romantic partner (forcing him or her to maintain
ch
body poses for painfully long durations; Finkel et al., 2009, Study 4), demon-
strated that participants were especially aggressive when they experienced both
y
Ps

provocation (in the form of insulting feedback) and ego depletion.

Another example of an instigating trigger × inhibiting forces inter-
an

action effect is provocation severity × frontal lobe functioning (Lau, Pihl, &
ic

Peterson, 1995). In this study, participants were preselected if they were in

er

the upper or lower quartile on frontal-lobe-based cognitive functioning, which

Am

underlies the ability to control one’s impulses (Hecaen & Albert, 1978).
Consistent with I3 theory, participants were more aggressive (administering
painful electric shocks) to the degree that the opponent had previously pro-
e
th

voked them severely rather than mildly (i.e., had administered painful electric
shocks to them), but this provocation main effect was substantially stronger
by

among individuals with weak rather than strong frontal lobe functioning (Lau
11

et al., 1995).
Yet another example of an instigating trigger × inhibiting forces inter-
20

action effect is provocation salience × alcohol consumption (Densen et al., 2008).

©

In this study, participants who had just consumed four alcoholic or placebo
beverages were provoked in either a salient or a subtle way and then had the
ht

opportunity to aggress against their provocateur by determining for how long

ig

the provocateur would have to keep his or her hand immersed in iced water.
yr

Participants were more aggressive in the salient provocation condition than

op

in the subtle provocation condition, and this effect was significantly stronger
C

in the alcohol than in the placebo condition.

44 SLOTTER AND FINKEL

12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 45

A final example of an instigating trigger × inhibiting forces interaction

effect is provocation level × relationship commitment within the context of an
ongoing romantic relationship (Slotter et al., 2009). In these studies, partic-
ipants were more aggressive toward their romantic partner after he or she had

A)
provoked them, and this provocation main effect was especially strong among
participants who were low in relationship commitment. Extending work sug-

P
(A
gesting that commitment promotes prorelationship behaviors in other con-
flictual relationship domains, such as forgiveness (Finkel, Rusbult, Kumashiro,

n
& Hannon, 2002), it appears that individuals who are highly committed to

io
their romantic relationships are able to override aggressive impulses when their

at
partner provokes them.

ci
so
6. An Illustrative Impelling Forces × Inhibiting Forces Interaction Effect:

As
Physical Proclivity × Negative Outcome Expectancies

al
As shown in the sixth row of Table 2.1, an illustrative impelling forces

ic
× inhibiting forces interaction effect is physical proclivity × negative outcome

og
expectancies. Individuals vary in the degree to which they prefer physical ver-
sus cognitive tasks, and a relative preference for the former predicts increased
ol
tendencies toward aggression and criminal behavior (Gottfredson & Hirschi,
ch
1990). Individuals also vary in the degree to which they believe that engaging
y
Ps

in aggressive behavior will cause them to experience negative outcomes, such

as physical harm or social derision, and stronger beliefs in this causal link
an

predict decreased tendencies toward aggression (Slaby & Guerra, 1988). One
recent study examined whether physical proclivity (an impelling factor)
ic
er

interacts with negative outcome expectancies (an inhibiting factor) to pre-

dict aggression toward a romantic partner (Finkel & Foshee, 2009). Results
Am

revealed a strong positive association of physical proclivity with self-reported

aggressive behavior over the previous year, but only when negative outcome
e
th

expectancies were low. It seems that the tendencies to prefer physical to cog-
nitive tasks predicts greater aggression among individuals whose aggression is
by

not restrained by beliefs that being aggressive will bring about negative effects
11

for them, but not among individuals whose aggression is so restrained.

20

7. Instigating Triggers × Impelling Forces × Inhibiting Forces Interaction Effects:

A Null Set
©
ht

Although the three main effects and the three two-way interaction effects
reviewed previously are key components of I3 theory, the instigating triggers ×
ig

impelling forces × inhibiting forces three-way interaction effect represents the

yr

most important component of the theory. Indeed, the theory suggests that all
op

three of the preceding two-way interaction effects are moderated by third vari-
C

ables from whichever category is not represented in that two-way interaction.

I3 THEORY 45
12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 46

Testing for such a three-way interaction would not be difficult. For

example, it would be easy to examine (a) whether the ego threat × narcissism
(instigating trigger × impelling factor) interaction effect is moderated by an
inhibiting factor (e.g., self-regulatory strength, alcohol consumption, strong
relationship commitment), (b) whether the provocation × self-regulation

A)
strength (instigating trigger × inhibiting factor) interaction effect is moderated

P
(A
by an impelling factor (e.g., testosterone, physical proclivity, dispositional
anger), and (c) whether the physical proclivity × negative outcome expectancy

n
(impelling factor × inhibiting factor) interaction effect is moderated by an insti-

io
gating trigger (e.g., social rejection, ego threat, provocation). Testing for such

at
interaction effects is an important direction for future research.

ci
so
As
DISCUSSION

al
I3 theory seeks (a) to impose enhanced theoretical coherence on the

ic
vast array of aggression risk factors by identifying how each of them increases

og
the likelihood of aggression (via instigating triggers, impelling forces, and/or
ol
inhibiting forces) and (b) to examine the manner in which risk factors from
ch
one category interact with those from one or both of the other categories to
y

predict aggressive behavior. In the preceding section (also see Table 2.1 and
Ps

Figure 2.1), we reviewed specific examples of how certain risk factors fit
into I3 theory and how they interface with variables from the other I3 theory
an

categories.
ic

One important direction for future research, aside from providing the
er

first tests of I3 theory’s instigating trigger × impelling factor × inhibiting

Am

factor three-way interaction effect, will be to develop and hone empirical

procedures for classifying a given risk factor into an I3 theory category
e

(instigating trigger, impelling forces, and/or inhibiting forces) or perhaps

th

into more than one category if a given variable both increases aggressive
by

impulses and decreases restraint. In this chapter, we have relied on theory

to determine, for example, (a) that social rejection, ego threat, and provo-
11

cation are instigating triggers; (b) that testosterone, narcissism, and physi-
20

cal proclivity are impelling factors; and (c) that self-regulatory strength, lack
of alcohol consumption, and negative outcome expectancies are inhibiting
©

factors. One limitation of this approach is that existing theory is in many

ht

cases not sufficiently developed vis-à-vis the I3 theory parameters to allow

ig

for definitive classification. For example, we are reasonably confident that

yr

dispositional self-control predicts reduced aggression in large part by raising

op

the inhibition threshold (see the horizontal lines in Figure 2.1), thereby
increasing the likelihood that individuals will override aggressive impulses.
C

That said, perhaps such dispositional self-control also predicts reduced aggres-

46 SLOTTER AND FINKEL

12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 47

sion in part by reducing impelling forces, thereby decreasing the strength of

the aggressive impulse in the first place.
How might scholars use empirical procedures to determine whether a
given variable promotes aggression by increasing aggressive impulses or by

A)
decreasing restraint? One promising approach is to adapt recent developments
in process dissociation paradigms used by social cognition researchers. Schol-

P
(A
ars have recently modeled behavior on laboratory tasks to discern the degree
to which participants exhibit certain automatic tendencies (e.g., toward preju-

n
dice or discrimination) and also controlled tendencies that override these

io
automatic tendencies (Payne, 2001; Sherman et al., 2008; see also Jacoby,

at
1991). After developing empirical procedures for distinguishing impulses

ci
toward aggressive behavior from self-controlled processes that override those

so
impulses, scholars will be able to examine the association of a given risk fac-

As
tor with both (a) individuals’ tendencies to experience impulses to aggress and
(b) their tendencies to override those impulses (Chapter 6, this volume). We

al
predict that variables such as testosterone, narcissism, and physical proclivity

ic
will correlate positively with the automatic aggressive tendencies identified by

og
these process dissociation procedures and negligibly with the controlled ten-
ol
dencies that override these automatic tendencies. In contrast, we predict that
ch
variables such as self-regulatory strength, sobriety, and negative outcome
y

expectancies will correlate negligibly with the automatic aggressive tenden-

Ps

cies identified by these process dissociation procedures and positively with the
controlled tendencies that override these automatic tendencies.
an

Once scholars determine (using theoretical tools, empirical tools, or

ic

both) which risk factors function by strengthening aggressive impulses and

er

which function by weakening behavioral inhibition processes, I3 theory may

Am

hold promise for interventions designed to reduce aggressive behavior. For

example, the theory highlights the importance of inhibitory factors (especially
e

self-regulation) in predicting aggressive behavior, and scholars have suggested

th

that interventions designed to help individuals override their impulses are

by

likely to be more effective than interventions designed to prevent them from

experiencing those impulses in the first place (e.g., Baumeister, 2005). To the
11

degree that such scholars are correct, interventions designed to strengthen

20

inhibitory forces may turn out to be more effective on average than interven-
tions designed to weaken impelling forces. Early research inspired by I3 the-
©

ory suggests that inhibition-relevant interventions at the dispositional level

ht

(e.g., building self-regulatory strength via bolstering regimens), relational

ig

level (e.g., strengthening relationship commitment), and situational level

yr

(e.g., having participants wait 10 seconds before responding to a provocation)

op

can reduce aggressive behavior (Finkel et al., 2009; Slotter et al., 2009; see
Finkel et al., 2002, for evidence that relationship commitment is amenable
C

to experimental manipulation). And, as discussed earlier, distal inhibitory

I3 THEORY 47
12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 48

factors such as prevalent social norms also seem to influence aggressive behav-
ior (Guerra et al., 2003; Sampson, Raudenbush, & Earls, 1997; Chapter 10,
this volume), which hints at the possibility that large-scale social interventions
could potentially reduce aggression at the societal level.

A)
In conclusion, I3 theory is an attempt to categorize aggression risk fac-
tors into instigating triggers, impelling factors, and/or inhibiting factors and

P
(A
to identify the interplay among variables across categories (see Figure 2.1).
To the degree that extant theory provides good reason to believe that partic-

n
ular risk factors fit relatively neatly into one of the I3 theory categories, the

io
theory provides an immediately accessible agenda for future research, ori-

at
ented less toward identifying additional risk factors than toward identifying

ci
(a) the processes by which risk factors, considered in isolation, increase

so
aggression and (b) the manner in which they interact to do so. In the long

As
run, I3 theory can inform interventions designed to help individuals manage
their aggressive impulses in a constructive manner. Indeed, scholarship may

al
well progress to the point where interventions can be tailored to the specific

ic
inhibiting risk factors most relevant to a given person, perhaps reducing one

og
individual’s aggression by bolstering self-regulatory strength and reducing
ol
another individual’s aggression by bolstering empathy. Such tailored inter-
ch
ventions, which would require valid assessment instruments, hold particular
y

promise for reducing the prevalence and severity of interpersonal aggression.

Ps
an

REFERENCES
ic
er

Anderson, C. A., Anderson, K. B., Dorr, N., DeNeve, K. M., & Flanagan, M. (2000).
Am

Temperature and aggression. Advances in Experimental Social Psychology, 32,

63–133. doi:10.1016/S0065-2601(00)80004-0
e

Anderson, C. A., & Bushman, B. J. (2001). Effects of violent video games on aggres-
th

sive behavior, aggressive cognition, aggressive affect, physiological arousal, and

by

prosocial behavior: A meta-analytic review of the scientific literature. Psycho-

logical Science, 12, 353–359. doi:10.1111/1467-9280.00366
11

Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of

20

Psychology, 53, 27–51. doi:10.1146/annurev.psych.53.100901.135231

Anderson, C. A., Carnagey, N. L., & Eubanks, J. (2003). Exposure to violent media: The
©

effects of songs with violent lyrics on aggressive thoughts and feelings. Journal of
ht

Personality and Social Psychology, 84, 960–971. doi:10.1037/0022-3514.84.5.960

ig

Archer, J., & Benson, D. (2008). Physical aggression as a function of perceived fighting
yr

ability and provocation: An experimental investigation. Aggressive Behavior, 34,

9–24. doi:10.1002/ab.20179
op

Archer, J., Birring, S. S., & Wu, F. C. W. (1998). The association between testosterone
C

and aggression among young men: Empirical findings and a meta-analysis.

48 SLOTTER AND FINKEL

12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 49

Aggressive Behavior, 24, 411–420. doi:10.1002/(SICI)1098-2337(1998)24:6<411::

AID-AB2>3.0.CO;2-9
Ayduk, O., Gyurak, A., & Luerssen, A. (2008). Individual differences in the
rejection–aggression link in the hot sauce paradigm: The case of rejection sensi-

A)
tivity. Journal of Experimental Social Psychology, 44, 775–782. doi:10.1016/
j.jesp.2007.07.004

P
(A
Baron, R. A., & Richardson, D. R. (1994). Human aggression (2nd ed.). New York,
NY: Plenum Press.

n
Baumeister, R. F. (2005). The cultural animal: Human nature, meaning, and social life.

io
New York, NY: Oxford University Press.

at
Baumeister, R. F., Gailliot, M., DeWall, C. N., & Oaten, M. (2006). Self-regulation

ci
and personality: How interventions increase regulatory success, and how

so
depletion moderates the effects of trait on behavior. Journal of Personality, 74,

As
1773–1802. doi:10.1111/j.1467-6494.2006.00428.x
Baumeister, R. F., Smart, L., & Boden, J. M. (1996). Relation of threatened egotism

al
to violence and aggression: The dark side of high self-esteem. Psychological

ic
Review, 103, 5–33. doi:10.1037/0033-295X.103.1.5

og
Berkowitz, L. (1993). Aggression: Its causes, consequences, and control. New York, NY:
McGraw-Hill. ol
ch
Berkowitz, L. (1998). Affective aggression: The role of stress, pain, and negative
y

affect. In R. G. Geen & E. Donnerstein (Eds.), Human aggression: Theories,

Ps

research, and implications for social policy (pp. 49–72). San Diego, CA: Academic
Press.
an

Bettencourt, B. A., & Miller, N. (1996). Gender differences in aggression as a func-

ic

tion of provocation: A meta-analysis. Psychological Bulletin, 119, 422–447.

er

doi:10.1037/0033-2909.119.3.422
Am

Bushman, B. J., & Baumeister, R. F. (1998). Threatened egotism, narcissism, self-

esteem, and direct or displaced aggression: Does self-love of self-hate lead to
e

violence? Journal of Personality and Social Psychology, 75, 219–229. doi:10.1037/

th

0022-3514.75.1.219
by

Bushman, B. J., & Cooper, H. M. (1990). Alcohol and human aggression: An

integrative research review. Psychological Bulletin, 107, 341–354. doi:10.1037/
11

0033-2909.107.3.341
20

Carney, M. M., & Buttell, F. P. (2006). Exploring the relevance of attachment

theory as a dependent variable in the treatment of women mandated into treat-
©

ment for domestic violence offenses. Journal of Offender Rehabilitation, 41,

33–61. doi:10.1300/J076v41n04_02
ht
ig

Dabbs, J. M., Jr., Frady, R. L., Carr, T. S., & Besch, N. F. (1987). Saliva testosterone
yr

and criminal violence in young adult prison inmates. Psychosomatic Medicine,

49, 174–182.
op

Dabbs, J. M., Jr., & Hargrove, M. F. (1997). Age, testosterone, and behavior among
C

female prison inmates. Psychosomatic Medicine, 59, 477–480.

I3 THEORY 49
12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 50

Denson, T. F., Aviles, F. E., Pollock, V. E., Earleywince, M., Vasquez, E. A., &
Miller, N. (2008). The effects of alcohol and the salience of aggressive cues
on triggered displaced aggression. Aggressive Behavior, 34, 25–33. doi:10.1002/
ab.20177

A)
DeWall, C. N., Baumeister, R. F., Stillman, R. G., & Gailliot, M. T. (2007). Vio-
lence restrained: Effects of self-regulation and its depletion on aggression. Jour-

P
nal of Experimental Social Psychology, 43, 62–76. doi:10.1016/j.jesp.2005.12.005

(A
Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., & Sears, R. R. (1939). Frustra-

n
tion and aggression. New Haven, CT: Yale University Press. doi:10.1037/10022-000

io
Dutton, D. G., van Ginkel, C., & Landolt, M. (1996). Jealousy, intrusiveness, and

at
intimate abusiveness. Journal of Family Violence, 11, 411–423. doi:10.1007/

ci
BF02333425

so
Fals-Stewart, W., Leonard, K. E., & Birchler, G. R. (2005). The occurrence of male-

As
to-female intimate partner violence on days of men’s drinking: The moderating
effects of antisocial personality disorder. Journal of Consulting and Clinical Psy-

al
chology, 73, 239–248. doi:10.1037/0022-006X.73.2.239

ic
Felson, R. B. (1996). Big people hit little people: Sex differences in physical power

og
and interpersonal violence. Criminology, 34, 433–452. doi:10.1111/j.1745-
9125.1996.tb01214.x ol
ch
Finkel, E. J. (2007). Impelling and inhibiting forces in the perpetration of intimate
y

partner violence. Review of General Psychology, 11, 193–207. doi:10.1037/

Ps

1089-2680.11.2.193
Finkel, E. J. (2008). Intimate partner violence perpetration: Insights from the science
an

of self-regulation. In J. P. Forgas & J. Fitness (Eds.), Social relationships: Cogni-

ic

tive, affective, and motivational processes (pp. 271–288). New York, NY: Psychol-
er

ogy Press.
Am

Finkel, E. J., DeWall, C. N., Slotter, E., B., Oaten, M., & Foshee, V. A. (2009). Self-
regulatory failure and intimate partner violence perpetration. Journal of Person-
e

ality and Social Psychology, 97, 483–499. doi:10.1037/a0015433

th

Finkel, E. J., & Foshee, V. A. (2009). Impelling and inhibiting forces in the perpetration
by

of intimate partner violence. Unpublished manuscript, Northwestern University,

Evanston, IL.
11

Finkel, E. J., Rusbult, C. E., Kumashiro, M., & Hannon, P. A. (2002). Dealing
20

with betrayal in close relationships: Does commitment promote forgiveness?

Journal of Personality and Social Psychology, 82, 956–974. doi:10.1037/0022-
©

3514.82.6.956
ht

Gaertner, L., & Foshee, V. (1999). Commitment and the perpetration of relation-
ig

ship violence. Personal Relationships, 6, 227–239. doi:10.1111/j.1475-6811.1999.

yr

tb00189.x
op

Giancola, P. R. (2000). Executive functioning: A conceptual framework for alcohol-

related aggression. Experimental and Clinical Psychopharmacology, 8, 576–597.
C

doi:10.1037/1064-1297.8.4.576

50 SLOTTER AND FINKEL

12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 51

Gottfredson, M. R., & Hirschi, T. (1990). A general theory of crime. Stanford, CA:
Stanford University Press.
Guerra, N. G., Huesmann, L. R., & Spindler, A. (2003). Community violence expo-
sure, social cognition, and aggression among urban elementary school children.

A)
Child Development, 74, 1561–1576. doi:10.1111/1467-8624.00623

P
Hecaen, H., & Albert, M. L. (1978). Human neuropsychology. New York, NY: Wiley.

(A
Holtzworth-Munroe, A., Stuart, G. L., & Hutchinson, G. (1997). Violent versus non-
violent husbands: Differences in attachment patterns, dependency, and jealousy.

n
Journal of Family Psychology, 11, 314–331. doi:10.1037/0893-3200.11.3.314

io
at
Jacoby, L. L. (1991). A process dissociation framework: Separating automatic from

ci
intentional uses of memory. Journal of Memory and Language, 30, 513–541.

so
doi:10.1016/0749-596X(91)90025-F
Lau, M. A., Pihl, R. O., & Peterson, J. B. (1995). Provocation, acute alcohol intox-

As
ication, cognitive performance, and aggression. Journal of Abnormal Psychology,

al
104, 150–155. doi:10.1037/0021-843X.104.1.150

ic
Leary, M. R., Kowalski, R. M., Smith, L., & Phillips, S. (2003). Teasing, rejection,

og
and violence: Case studies of the school shootings. Aggressive Behavior, 29,
202–214. doi:10.1002/ab.10061
ol
ch
Leary, M. R., Twenge, J. M., & Quinlivan, E. (2006). Interpersonal rejection as a
determinant of anger and aggression. Personality and Social Psychology Review,
y

10, 111–132. doi:10.1207/s15327957pspr1002_2

Ps

Lorenz, K. (1966). On aggression. San Diego, CA: Harcourt, Brace & World.
an

Nisbett, R. E., & Cohen, D. (1996). Culture of honor: The psychology of violence in the
ic

South. Boulder, CO: Westview Press.

er

Norlander, B., & Eckhardt, C. (2005). Anger, hostility, and male perpetrators of inti-
Am

mate partner violence: A meta-analytic review. Clinical Psychology Review, 25,

119–152. doi:10.1016/j.cpr.2004.10.001
e

Payne, B. K. (2001). Prejudice and perception: The role of automatic and controlled
th

processes in misperceiving a weapon. Journal of Personality and Social Psychology,

81, 181–192. doi:10.1037/0022-3514.81.2.181
by

Richardson, D. R., Green, L. R., & Lago, T. (1998). The relationship between
11

perspective-taking and nonaggressive responding in the face of an attack.

Journal of Personality, 66, 235–256. doi:10.1111/1467-6494.00011
20

Ronfeldt, H. M., Kimerling, R., & Arias, I. (1998). Satisfaction with relationship
©

power and the perpetration of dating violence. Journal of Marriage and the Fam-
ily, 60, 70–78. doi:10.2307/353442
ht
ig

Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997, August 15). Neighborhoods
yr

and violent crime: A multilevel study of collective efficacy. Science, 277,

918–924. doi:10.1126/science.277.5328.918
op

Sapolsky, R. M. (1998). The trouble with testosterone and other essays on the biology of
C

the human predicament. New York, NY: Scribner.

I3 THEORY 51
12051-03_CH02-rev2.qxd 5/25/10 11:29 AM Page 52

Sherman, J. W., Gowronski, B., Gonsalkorale, K., Hugenberg, K., Allen, T. J., &
Groom, C. J. (2008). The self-regulation of automatic associations and behavioral
impulses. Psychological Review, 115, 314–335. doi:10.1037/0033-295X.115.2.314
Slaby, R. G., & Guerra, N. G. (1988). Cognitive mediators of aggression in adolescent

A)
offenders: I. Assessment. Developmental Psychology, 24, 580–588. doi:10.1037/
0012-1649.24.4.580

P
(A
Slotter, E. B., Finkel, E. J., & Bodenhausen, G. V. (2009). Provocation and commit-
ment in intimate partner violence. Unpublished manuscript, Northwestern Uni-

n
versity, Evanston, IL.

io
Twenge, J. M., Baumeister, R. F., Tice, D. M., & Stucke, T. J. (2001). If you can’t join

at
them, beat them: Effects of social exclusion on aggressive behavior. Journal of Per-

ci
sonality and Social Psychology, 81, 1058–1069. doi:10.1037/0022-3514.81.6.1058

so
Twenge, J. M., & Campbell, W. K. (2003). Isn’t it fun to get the respect we deserve?

As
Narcissism, social rejection, and aggression. Personality and Social Psychology
Bulletin, 29, 261–272. doi:10.1177/0146167202239051

al
Van Baardewijk, Y., Stegge, H., Bushman, B. J., & Vermeiren, R. (in press). “I feel your

ic
pain” (but only when I am forced to focus on it): Psychopathy, distress, and aggres-

og
sive behavior. Journal of Child Psychology and Psychiatry, and Allied Disciplines.
ol
Van Goozen, S. H. M., Cohen-Kettenis, P. T., Gooren, L. J. G., Frijda, N. H., & Van
ch
de Poll, N. E. (1995). Gender differences in behaviour: Activating effects
y

of cross-sex hormones. Psychoneuroendocrinology, 20, 343–363. doi:10.1016/

Ps

0306-4530(94)00076-X
Van Goozen, S., Frijda, N., & Van de Poll, N. (1994). Anger and aggression in women:
an

Influence of sports choice and testosterone administration. Aggressive Behavior,

ic

20, 213–222. doi:10.1002/1098-2337(1994)20:3<213::AID-AB2480200308>

er

3.0.CO;2-4
Am
e
th
by
11
20
©
ht
ig
yr
op
C

52 SLOTTER AND FINKEL

12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 53

PA)
APPLYING ADAPTATIONISM

(A
TO HUMAN ANGER:

n
io
THE RECALIBRATIONAL THEORY

at
ci
so
AARON SELL

As
al
ic
og
ol
y ch
Ps
an

In the late 1960s and early 1970s, evolutionary biology underwent a sci-
ic

entific revolution in which poorly defined models of evolutionary change were

er

replaced with a theoretically rigorous program of research that integrated the

Am

major findings of evolutionary biology, animal behavior, and genetics. This

integration included a method, known as the adaptationist program, for identi-
e

fying and describing design in organisms, a method that matches the engineer-
th

ing requirements inherent in adaptive problems to features in the organism

by

that are designed to solve those problems.

In this chapter, I apply the methods of the adaptationist program to
11

human anger. The resultant theory, called the recalibrational theory, states that
20

anger is the output of a cognitive mechanism designed by natural selection to

negotiate conflicts of interest. The causes of anger, the behavior it produces,
©

the factors that mitigate it, and its effects on physiology, perception, and cog-
ht

nition can be explained by reference to this adaptive function.

ig

In brief, the recalibrational theory states that anger is a system designed

yr

by natural selection to recalibrate the weight another individual places on the

op

angry person’s interests so that they become less likely to impose costs or deny
benefits to the angry individual. Anger deploys two main strategies to con-
C

vince the target to treat the angry individual better: (a) toward those with

53
12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 54

whom the angry individual has a cooperative relationship, he or she may

withhold benefits; and (b) toward those who have little incentive to value
the cooperation of the angry individual, anger can mobilize aggression. This
latter option is responsible for a great deal of human aggression and is close

A)
in design and function to aggression against conspecifics in other animals.

P
(A
EVOLUTIONARY BIOLOGY OF CONFLICT

n
io
Natural selection is the only process shown capable of creating complex

at
functional design in organisms (Williams, 1966). Evolutionary psychology is

ci
an approach to psychology that uses analyses of selection pressures to discover

so
and understand the cognitive machinery designed by those selection pres-

As
sures. The closer the fit between the logic of the selection pressures and the
known features of the mechanism under study, the more confident one can

al
be about the proposed function of the mechanism, the accuracy of the posited

ic
selection pressures, and any future predictions derived from the model. There-

og
fore, a thorough examination of both the selection pressures and the proposed
ol
cognitive mechanism are necessary for a theoretically sound computational
ch
model of any organic mechanism designed by natural selection. An analysis
y

of the selection pressures inherent in conflicts of interest and the major fea-
Ps

tures of human anger fit together well, making a strong case that human anger
was designed by natural selection to regulate conflicts of interest.
an
ic

Selection Pressures Inherent in Conflicts of Interest

er
Am

By virtue of their common design, organisms of the same species tend

to exploit the same resources in their environment. For example, they tend
e

to mate with members of the same species, eat the same foods, and seek to
th

occupy the same territories. Selection pressures to gain increased access to finite
by

resources will naturally produce organisms that are designed to compete for
those resources. Excluding kin, and assuming there are no reciprocal conse-
11

quences, natural selection will design one organism to attempt to gain resources
20

with no regard for the destructive effects of their activities on competitors.

But animals should be sensitive to the cost–benefit structure of aggres-
©

sive competition. An examination of the selection pressures involved in con-

ht

flicts of interest led evolutionary biologists to two key variables that animals
ig

respond to when deciding whether to engage in aggression to secure a benefit.

yr

The first of these is the relative value of the contested resource to the indi-
op

vidual organisms (i.e., who needs it more). This variable determines an organ-
ism’s fitness budget (i.e., the maximum cost an organism is willing to incur to
C

attain the resource). If a food item is worth 10 fitness units to an organism,

54 AARON SELL
12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 55

then he or she should be willing to accept up to 9 units in cost to attain it.

Therefore, animals that value a contested resource more will fight longer and
harder for it. The second factor animals respond to when deciding whether
to fight for a contested resource is the relative fighting ability of the organ-

A)
isms (Hammerstein & Parker, 1982; Maynard Smith & Parker, 1976). Con-
testants that are better able to impose costs will deplete their competitor’s

P
(A
fitness budgets earlier and win the conflict. If these two selection pressures,
relative formidability and relative valuation, have been active in the design

n
of animal brains, then we would expect to see animals respond in adaptive

io
ways to assessments of these two variables.

at
Five converging lines of evidence demonstrate that animals have been

ci
designed to respond adaptively to the selection pressures inherent in conflicts

so
of interest as modeled by evolutionary biologists.

As
䡲 Animals are designed such that relative formidability partially

al
determines the resolution of resource conflicts. The advantage

ic
of being the more formidable competitor has been noted in

og
species as distantly related as the sea anemone (Brace & Pavey,
ol
1978), African buffalo (Sinclair, 1977), and crayfish (Hazlett,
ch
Rubenstein, & Ritschoff, 1975). A particularly dramatic exam-
ple was found by Petrie (1984), who studied territory size in the
y
Ps

moorhen (Gallinula chloropus). He found that relative male

weight was a perfect predictor (Spearman correlation coefficient
an

of 1.0) of territory size. (For more examples, see Archer, 1988.)

ic

䡲 Animals are designed such that the relative valuation of the con-
er

tested resource partially determines the resolution of resource

Am

conflicts. Animals that have been deprived of food for a longer

time tend to win contests over food. This effect has been repli-
e

cated with chimpanzees (Nowlis, 1941), crayfish (Hazlett et al.,

th

1975), bald eagles (Hansen, 1986), and dark-eyed juncos (Cristol,

1992), among other species. The effect is not limited to differen-
by

tial valuation of food; it has also been found for differential value
11

of mating opportunities, nesting sites, feeding sites, and spider-

webs (for a review, see Enquist & Leimar, 1983; for a particularly
20

clear example, see Austad, 1983).

©

䡲 Animals are designed to assess relative formidability and the

ht

relative value of the conflict. Because the costs of aggression can

ig

be avoided if animals know in advance who will win, animals

yr

will be selected to assess cues that predict who is likely to win

op

and to respond appropriately. A variety of species are designed

not only to calculate their competitor’s fighting ability but also
C

to broadcast their own. Several species of birds and lizards have

APPLYING ADAPTATIONISM TO HUMAN ANGER 55

12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 56

evolved physical signals of fighting ability (i.e., badges) that

track, quite accurately, the organism’s fighting ability (Rohwer
& Rohwer, 1978). These badges would be worthless if competi-
tors did not use the signals to compute the animals’ formidabil-

A)
ity. Furthermore, the existence of dominance hierarchies in many
species indicates the ability of animals to measure their relative

P
(A
formidability. (For a review of the nonprimate animal literature,
see Huntingford & Turner, 1987; for primate examples, see Smuts,

n
Cheney, Seyfarth, Wrangham, & Struhsaker, 1987.)

io
䡲 Animals are designed to demonstrate formidability to lower the

at
costs of conflict. Two organisms can minimize the cost of con-

ci
flict if the eventual loser can recognize that it will lose. If a type

so
of animal aggression is designed to demonstrate relative formi-

As
dability, these conflicts should follow a general pattern of esca-
lation in which lower-cost demonstrations (which are probably

al
less accurate) are exchanged before higher-cost demonstrations

ic
or nonritualistic, “no holds barred” combat ensues. Large dis-

og
crepancies in formidability should be evident even in cases
ol
where demonstrations of formidability are somewhat inaccu-
ch
rate. For example, tail beating in male cichlid fish (i.e., shaking
y

one’s tail at an opponent so that waves hit him) is an indicator

Ps

of body size and strength that is presumed to be less predictive

than mouth-locked wrestling (during which fish lock mouths
an

and vigorously shake each other), but mouth-locked wrestling

ic

is more costly in terms of energy and probability of injury. This

er

pattern of escalation, a model of which is called the sequential

Am

assessment game (Enquist & Leimar, 1983), has been observed

in numerous and distally related species, including cichlid fish
e

(Enquist, Leimar, Ljungberg, Mallner, & Segerdahl, 1990), bowl

th

and doily spiders (Austad, 1983), African buffalo (Sinclair, 1977),

by

red deer (Clutton-Brock & Albon, 1979), and pigs (Jensen &
Yngvesson, 1998).
11

䡲 Species in which animals typically engage in repeated inter-

20

actions will maintain an internal representation of relative for-

midability that governs conflicts of interest without individuals
©

needing to reestablish formidability through competitive inter-

ht

actions. The internal storage of relative formidability is indicated

ig

by what are called dominance hierarchies, which have been noted

yr

in many social species, including dark-eyed juncos (Cristol, 1992),

op

chickens (Guhl, 1956), hyenas (Owens & Owens, 1996), and

every social monkey and ape (Smuts et al., 1987). Although
C

dominance hierarchies are talked about as if they were features

56 AARON SELL
12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 57

of a community, it is important to remember that the information

about one’s position compared with other group members is stored
within individual brains.

A)
Constructing a Cognitive Model of Human Anger

P
(A
Human conflicts of interest often look different from nonhuman animal
resource competition largely because of humans’ ability to mentally represent

n
conflicts of interests. A dung fly can fight with a competitor over a piece of

io
food, but a human can fight over whether someone should have gotten a piece

at
of food or even whether someone should have agreed with someone else who

ci
wanted to give a piece of food to a third party, and so on. Most human con-

so
flicts of interest do not involve tangible material resources but instead involve

As
conflicts over courses of action (retrospective and prospective), exchanges of
information, social alliances, and other abstract cost–benefit tradeoffs between

al
individuals. The logic of the selection pressures and the role of relative valua-

ic
og
tion and relative formidability, however, are the same.
An analysis of human conflicts and the selection pressures surrounding
ol
conflicts of interest argues for the existence of an internal variable similar to
ch
that which underlies dominance hierarchies in nonhuman animals, which
y

regulates decisions about cost–benefit interactions in humans. This has been

Ps

called a welfare tradeoff ratio (Tooby, Cosmides, Sell, Liberman, & Sznycer,
an

2008; Sell, Tooby, & Cosmides, 2009).

ic
er

WELFARE TRADEOFF RATIOS

Am

It is posited that humans internally represent a threshold for acceptable

e

cost–benefit transactions, a welfare tradeoff ratio (WTR), for every individual

th

they interact with. For an agent, X, the WTR with respect to agent Y defines
by

the cost–benefit ratio below which X will give Y the resource and above which
X will attempt to take the resource. This is represented mathematically as
11

v ( X ) > v ( Y ) ⴱ WTRxy
20
©

In other words, the WTR X has toward Y (WTRxy) indicates how much weight
ht

X puts on Y’s interests when making decisions that affect them both. For exam-
ig

ple, if X’s WTR toward Y is .5, then X will give up a resource worth 5 to himself
yr

if Y values it at more than 10. WTRs are theorized to be the computational

op

elements underlying folk notions such as love, respect, and deference.

An individual will have different WTRs for different individuals. For
C

example, a person might be willing to ruin a stranger’s sweater to dress a pet’s

APPLYING ADAPTATIONISM TO HUMAN ANGER 57

12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 58

wound but unwilling to ruin a work supervisor’s sweater for that same pur-
pose, because the WTR he or she has with respect to the supervisor is higher
than toward a stranger.
Given mutual human dependence, the costs of contests, and the nature

A)
of kin and friendship, natural selection is predicted to have designed humans
such that WTRs will be set higher based on numerous factors related to

P
(A
another’s ability to enforce his or her own welfare. One set of factors is related
to the ability to enforce WTRs by threatening to inflict harm; these factors

n
include, for example, greater physical strength and more coalitional support.

io
WTRs set primarily by the threat of force will be consulted, presumably, only

at
when there is some possibility that the individual will be present to defend

ci
his or her interests. This is entirely analogous to the relative formidability that

so
is known to affect conflicts of interest in nonhuman animals. Another set of

As
factors that set WTRs is related to the ability to defend one’s welfare by
threatening to withdraw the benefits of cooperation. These include, for exam-

al
ple, the person’s status as a frequent and dependable reciprocation partner,

ic
his or her status as a friend who has a stake in one’s welfare (Tooby & Cosmides,

og
1996), or his or her possession of special abilities that can be deployed to
benefit others. ol
ch
There are at least two kinds of WTRs that govern cost–benefit transac-
y

tions in different contexts: (a) monitored WTRs, which define the threshold of
Ps

cost–benefit transactions when both parties are present or otherwise capable of

defending their interests; and (b) intrinsic WTRs, which define the threshold
an

of cost–benefit transactions when the other individual is not present or is

ic

unable to defend his or her interests. Presumably, intrinsic WTRs allowed indi-
er

viduals to partition cost–benefit transactions adaptively in a world where the

Am

welfare of other individuals was of adaptive significance for oneself.

Though the full selection pressure analysis of intrinsic WTRs is beyond
e

the scope of this chapter, a quick starting point would be as follows: A subset of
th

individuals in one’s social world can improve one’s welfare as a result of their
by

existence, social power, and well-being (e.g., a devoted friend, a caretaker of

one’s children, a generous acquaintance). Benefits given to them, even with-
11

out their knowledge, will correspond to benefits to oneself.

20
©

RECALIBRATIONAL THEORY OF ANGER: HUMAN ANGER IS AN

ht

ADAPTATION TO RAISE ANOTHER’S WELFARE TRADEOFF

ig

RATIO WITH RESPECT TO ONESELF

yr
op

I propose that a large and well-bounded subset of phenomena that people

refer to when they use the word anger can be understood as the output of a highly
C

sophisticated, complex, reliably developing computational system, instanti-

58 AARON SELL
12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 59

ated in neural tissue and designed by natural selection, that is deployed as a

negotiation tool to resolve present and future conflicts of interest in the
angered individual’s favor. The anger system does this by initiating behaviors
that (a) recalibrate the target’s estimates of the costs and benefits of actions

A)
to the target and to the angered individual and (b) raise the target’s WTR
toward the angered individual, so the target takes that individual’s welfare

P
(A
more into account in the present, the future, or both. The two main negotiation
strategies deployed to recalibrate the target’s WTR are (a) threatening to

n
inflict costs (or actually doing so) and (b) threatening to withdraw cooperation

io
(or actually doing so).

at
ci
Causes of Human Anger

so
As
Anger should be triggered when one individual interprets the actions of
the target as indicating that the target’s WTR toward the individual is lower

al
than the accepted level. WTRs are theorized to be used when making cost–

ic
benefit decisions—indeed, that is their hypothesized function—but they may

og
also be used by other cognitive programs, such as memory storage related to
ol
particular individuals (e.g., we may remember our niece’s birthday but not our
ch
mailman’s birthday), frequency of consideration (e.g., one thinks daily about
y

a spouse but may forget about old friends for weeks at a time), inherent pleasure
Ps

in being around an individual, the weight placed on the veracity of another’s

beliefs, one’s willingness to seek advice or share secrets, and vicarious WTR
an

toward friends of friends. Because WTRs are used by so many cognitive systems,
ic

there is predicted to be a plethora of computationally distinct ways of trigger-

er

ing anger, each with its own blend of behavioral responses dependent on how
Am

the WTR was indicated (see examples that follow).

Theoretically, the clearest indication of a low WTR is the target’s will-
e

ingness to take actions that impose a large cost on the angry individual in
th

order for the target to receive a small benefit. Holding other variables con-
by

stant, anger is more likely to be activated over a cost–benefit transaction as

(a) the cost imposed on the individual increases, (b) the benefit reaped as a
11

result of that cost decreases, and (c) characteristics of the instigator and the
20

angry individual indicate that it is possible for the angry individual to force
the other to use a higher WTR than was indicated by the cost–benefit trans-
©

action imposed.
ht
ig

Behavioral Responses Generated by Anger

yr
op

The primary functions of anger are to raise the magnitude of the WTR
of an individual who has demonstrated a lower WTR than is acceptable to
C

the angry individual and/or to recalibrate that individual’s estimates of the

APPLYING ADAPTATIONISM TO HUMAN ANGER 59

12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 60

magnitude of the costs imposed and benefits received. WTRs in the target
should be open to modification when this will allow the target to avoid
being harmed or having cooperation withdrawn—more precisely, in cir-
cumstances that predicted ancestrally that these two negative outcomes

A)
were likely.
Thus, when the anger system is triggered by evidence that the target’s

P
(A
WTR toward the angry individual is too low, it should motivate him or her
to make credible threats or demonstrate qualities that would make such

n
threats credible, if issued. WTRs are hypothesized to be set partly by relative

io
formidability, as is the case with many other animals. Thus if an individual is

at
showing evidence of a low WTR, it could be the result of an underestimation

ci
of one’s willingness or ability to use force and could be recalibrated by a

so
demonstration of said force. As with nonhuman animals, formidability should

As
be demonstrated starting with low-cost, presumably less accurate, demonstra-
tions of physical strength and escalated as needed to more accurate and

al
dangerous demonstrations of strength.

ic
The theory also predicts that anger should be designed to manipulate

og
the target’s estimates of the magnitude of costs and benefits inherent in the
ol
transaction. To the extent that you can increase another’s perception of a cost
ch
he or she imposed on you, you can decrease the probability that the individual
y

will impose such a cost on you again. The same is true of reducing another’s
Ps

perception of the benefit he or she received.

an
ic

KNOWN FEATURES OF ANGER

er
Am

Given the breadth of data collected on human anger, the first step when
proposing a new theory must be to determine its consistency with empirical
e

findings that have been shown to be both large in effect and robust across
th

studies. It should be noted that although I did my best to choose the following
by

data sets on the basis of their effect sizes, reliability, and cross-cultural docu-
mentation, I am not providing a complete review of the anger literature.
11
20

Feature 1: Anger Frequently Results From the Imposition of Costs

©

Individuals tend to get angry when costs are imposed on them. Most
ht

important, the magnitude of the anger response is positively correlated with

ig

the magnitude of the cost. Empirical studies that varied the magnitude of the
yr

cost have confirmed this relationship across a host of different cost types, such
op

as the voltage of electric shocks (O’Leary & Dengerink, 1973), severity of

insults (Taylor, 1967), seriousness of a crime (Blumstein & Cohen, 1980),
C

and monetary payoffs in economic games (Fehr & Gaechter, 2000).

60 AARON SELL
12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 61

Feature 2: Intentionality Increases the Likelihood of Anger

When a cost is imposed without prior knowledge on behalf of the imposer,

there is much less anger and retaliation than if the cost was knowingly imposed

A)
(see Epstein & Taylor, 1967). Theories that posit intentionality have used
different definitions. For example, Heider (1958) defined intention as a plan

P
(A
that guides action, Kaufmann (1970) defined an aggressive action as one that is
known by the actor to have a nonzero chance of inflicting harm on the target,

n
and Tedeschi and Felson (1994) defined an intentional action as one “performed

io
with the expectation that it will produce a proximate outcome of value to the

at
actor” (p. 164).

ci
Intentionality, not being directly observable, is a category the human

so
mind uses to classify types of actions and thus must be discovered and explored

As
rather than defined as a given, objective feature of the world. A theory of
intentionality will have to specify, at a minimum, what information must be

al
known for something to qualify as intentional. For example, if someone plans

ic
to hit you with a toy ball and expects that you will enjoy this as part of a

og
game but you become angry when hit because you did not want to play, should
ol
the person’s act be viewed as intentional? Heider’s (1958) definition cannot
ch
answer this question. Kaufmann’s (1970) definition would result in the act
y

being judged as not intentional. Tedeschi and Felson’s (1994) definition would
Ps

classify the act as intentional.

an

Feature 3: Apologies Mitigate Anger

ic
er

The most reliable way to reduce anger, according to the empirical liter-
Am

ature, is to apologize (Riordan, Marlin, & Kellogg, 1983). The content of

apologies varies, and a great deal of empirical work remains to be done on dis-
e

tinguishing “real” from “false” apologies and discovering why angry individu-
th

als are so sensitive to the difference (see Holtgraves, 1989).

by

Feature 4: Anger and Aggression Are Often Used

11

by Males to Restore “Face”

20

Violent and homicidal aggressive acts are most common among young
©

men across cultures and time periods (Daly & Wilson, 1988). These acts are
ht

largely the result of insults and attempts to save face or attain status by fighting.
ig

This account of violent aggression among males has been noted, to some extent
yr

independently, by criminologists (Luckenbill, 1977), social psychologists

op

(Berg & Fox, 1947; see also Chapter 10, this volume; Felson, 1982), and evo-
lutionary psychologists (Daly & Wilson, 1988). Most impressively, a host of
C

cultural anthropologists have documented the positive association between

APPLYING ADAPTATIONISM TO HUMAN ANGER 61

12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 62

fighting ability and status in nonpolice societies, including the Yanamamo of

Venezuela (Chagnon, 1983), the Dani of Highland New Guinea (Sargent,
1974), the Montenegrins of Eastern Europe (Boehm, 1984), the Inuit (Balikci,
1970), the Jivaro Indians of the western Amazon (Karsten, 1935), and Amer-

A)
ican gangs (Toch, 1969). In each of those societies, threats to one’s face or
status are often the trigger for violent episodes between young men.

P
(A
Feature 5: Personal Insults Are One of the Most Reliable Causes of Anger

n
io
Though not usually the object of study, personal insults have been used

at
in aggression research for 40 years and have (in most cases) been shown to be

ci
sufficient causes of anger (Geen, 1968). In nonlaboratory cases of aggression, it

so
has been found that personal insults almost always precede homicides (Berg &

As
Fox, 1947; Luckenbill, 1977; Toch, 1969) and assaults (Felson, 1982).

al
Feature 6: Anger Results in an Exchange of Argument

ic
og
The most common response to an anger-inducing event in naturally
ol
occurring situations is to engage in an argument. Averill (1982) established
ch
this fact with an influential study of a large sample of adults.
y
Ps

Feature 7: Anger Has Both Cross-Culturally Universal Features

and Neurophysiological Locality
an
ic

The anger expression is universal across individuals and cultures, includ-

er

ing cultures with no historical contact with the West (Ekman, 1973). It has
Am

been demonstrated in 6-month-old infants (Stenberg & Campos, 1990) and

in congenitally blind children (Galati, Sini, Schmidt, & Tinti, 2003). Physio-
e

logical changes accompanying anger have been found to be similar across cul-
th

tures (Rime & Giovannini, 1986). Subjects from a broad European sample
by

reported that anger felt unpleasant and warm and was frequently associated
with muscular tension. These response patterns were different from those of
11

other emotions.
20

Behavioral responses that result from anger are also similar across European
countries. Specifically, anger often leads to vocal changes involving increased
©

volume and sometimes trembling; it also leads to changes in movement qual-

ht

ity, clenched fists, and increased hand movement (Shaver, Schwartz, Kirson,
ig

& O’Connor, 1987; Wallbott, Ricci-Bitti, & Baenninger-Huber, 1986). Fur-

yr

thermore, Scherer and colleagues categorized vocal expressions of emotion

op

and showed that when a person is angry, fundamental frequency (roughly,

pitch) often increases in mean and variability (Banse & Scherer, 1996). This
C

pattern is also distinct from those of other emotions.

62 AARON SELL
12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 63

The antecedents of anger have also been shown to be cross-culturally

similar. Recall the cross-cultural data on aggression being driven by insults to
one’s face reviewed in feature 4. Wallbott and Scherer (1986) noted similar
antecedents of anger in a broad range of European countries.

A)
Finally, there have been numerous studies of parts of the brain that are
differentially activated by anger, showing that its neural underpinnings are

P
(A
similar across individuals and species and that they are distinct from those of
other emotions (see also Chapter 6). Panksepp (2000) theorized, based on

n
brain imaging and lesion studies, that anger relies mostly on the medial amyg-

io
dala, bed nucleus of stria terminalis, and the medial and perifornical hypo-

at
thalamus. The anger/rage system is moderated primarily by acetylcholine and

ci
glutamate in ways that connect the amygdala and periaqueductal gray with

so
the hypothalamus (Siegel & Schubert, 1995). Finally, a recent meta-analysis

As
confirmed that testosterone tracks individual differences in tendencies toward
anger and aggression with an effect size of .4 (Cohen’s d; Archer, 2006; see

al
also Chapter 2, this volume).

ic
og
ol
HOW DOES THE RECALIBRATIONAL THEORY ACCOUNT
ch
FOR FEATURES OF ANGER?
y
Ps

The recalibrational theory accounts well for the major features of anger
and provides testable predictions that promise to more fully elaborate the
an

computational and functional structure of each feature. These explanations

ic

are discussed briefly next.

er
Am

Feature 1: Cost Imposition

e

The recalibrational theory predicts that the cause of anger is not nega-
th

tive affect but the indication that another person holds a low WTR with
by

respect to oneself. A common indication of such a WTR is the imposition of

a cost that is too large given the benefit the offender received. Holding the
11

magnitude of the benefit constant, the larger the cost one is willing to impose,
20

the more likely anger is to be triggered. Likewise, holding the cost constant
and the more the other person benefits by imposing that cost, the less angry
©

one will be (Sell, 2006).

ht
ig

Feature 2: Intentionality
yr
op

Intentionality judgments, in the context of anger, can be thought of as the

outputs of a cognitive mechanism that determines whether a WTR was engaged
C

when a cost was imposed. This predicts that anger-relevant intentionality

APPLYING ADAPTATIONISM TO HUMAN ANGER 63

12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 64

requires knowledge of the magnitude of the imposed cost, the magnitude of

the benefit received, and the identity of the individual on whom the cost was
imposed. Each of these three components has been shown to affect anger in
the predicted direction (Sell, 2006).

A)
Feature 3: Apologies

P
(A
The recalibrational theory suggests that apologies are explicit acknowl-

n
edgements of either (a) a past discrepant WTR that has been recalibrated

io
upward or (b) a misperception or accident on behalf of the target of anger that

at
led them to commit an act that does not reflect their true WTR toward the

ci
angry individual. The contents of WTR-recalibrated apologies (Type 1) are

so
predicted to contain statements that translate into the following cognitive

As
grammar: “I will demonstrate a more favorable welfare tradeoff ratio with
respect to you, such that I will no longer impose costs of that magnitude on

al
you for benefits of that magnitude.” This kind of claim may be best validated

ic
by restitution or by indicating a willingness to incur a cost to repay the angry

og
individual. The content of Type 2 apologies should contain statements about
ol
the magnitude of perceived costs and benefits; for example, “I didn’t realize
ch
that would hurt you so badly; I thought I had a good reason for doing that, but
y

I was wrong.”
Ps

Feature 4: Role of Face and Status

an
ic

Perceptions of formidability play a part in the setting of WTRs, partic-

er

ularly for males (Sell et al., 2009; Sell, Tooby, & Cosmides, 2009). As such,
Am

the recalibrational theory predicts that demonstrations of physical strength

in humans should involve the same procedures as strength contests in non-
e

human animals, such as signals of challenge, escalating conflicts that start

th

with low levels of violence (e.g., pushing contests, staring contests) and move
by

either to more violent demonstrations (e.g., wrestling, punching, weaponry)

or to signals of surrender that end the violence. These patterns fit the data on
11

homicide quite well (Luckenbill, 1977).

20

Feature 5: Insults
©
ht

The recalibrational theory predicts that anger-inducing insults can be

ig

understood as attempts to directly indicate a low WTR with respect to

yr

another individual and perhaps as attempts to influence others to hold lower

op

WTRs with respect to the insulted individual. Insults should, psychologically,

translate to the form, “I do not value your interests highly.” More proxi-
C

mately, insults can be declarations of a deficit in a variable that is used to

64 AARON SELL
12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 65

determine one’s WTR toward that individual. One such variable, for males
at least, is physical strength (Sell, Tooby, & Cosmides, 2009). Fitting the
theory, many insults applied disproportionately to men target insufficient
masculinity or strength (e.g., wimp, wuss, nerd, girly-man, pussy, weakling;

A)
Harris, 1993; Preston & Stanley, 1987). Although beyond the scope of this
analysis, it seems likely that other insults fit into categories that make a man

P
(A
socially powerful or not, such as his intent to cooperate (e.g., asshole, prick,
bastard [colloquial meaning]), his being unable or unwilling to function as a

n
reliable cooperator or being otherwise unworthy of having others take his

io
interests into account (e.g., punk, white trash, bum), or his competence (e.g.,

at
idiot, fool, loser).

ci
Insulting beliefs are a kind of implied insult that is often mistakenly

so
revealed. For example, imagine a professor who offers to help her student with

As
a simple math problem. The student may be angry and insulted because he
believes the professor thinks he is stupid. This is a different kind of insult, but

al
it causes anger because it reveals that a trait used to set WTRs (in this case

ic
intelligence) is being underestimated and presumably results in the professor

og
holding a lower WTR toward the student.
ol
ch
Feature 6: Arguments
y
Ps

The recalibrational theory predicts two primary functions of arguments:

(a) to recalibrate the target’s WTR, raising it so the target’s decisions will take
an

the angered person’s welfare more fully into account; and (b) to recalibrate
ic

the target’s perception of the cost–benefit transaction that caused anger.

er

When WTRs are largely based on relative fighting ability, one would predict
Am

arguments about aggressive potential, for example, “I could kick your ass” or
“My dad could beat up your dad!” When one has an expectation of inherent
e

value (e.g., with friends, kin, and others with whom long-term cooperation
th

would be mutually beneficial), the theory predicts that statements of relative

by

friendship quality would be used to boost another’s intrinsic WTR toward

oneself. For example, “I wouldn’t do that to you” or “Remember when your
11

mother was sick and I took notes for you in all your classes.”
20

Secondary functions of arguments may include (a) gathering informa-

tion to determine the magnitude of the offender’s WTR, such as the magni-
©

tudes of the values involved in the cost–benefit exchange (e.g., “Why did you
ht

do that?”); the offender’s knowledge of the magnitude of the cost, benefit, and
ig

victim identity (e.g., “Do you know how much that hurt?” “Do you know who
yr

you’re messing with?”); the offender’s perception of his or her WTR with
op

respect to you (e.g., “I thought we were friends”); and other variables that are
used to set WTRs; and (b) testing the boundaries of one’s WTR with respect
C

to the other and vice versa (e.g., verbal bullying).

APPLYING ADAPTATIONISM TO HUMAN ANGER 65

12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 66

Feature 7: Universality

In contrast to many other theories, the recalibrational theory clearly

predicts that anger is an adaptation designed by natural selection and that its

A)
basic computational structure should be universal across cultures and should
share a phylogenetic relationship with structures in closely related nonhuman

P
(A
animals, residing in similarly localized brain areas. Furthermore, like nonhuman
animals, humans should use signals of aggressive intent based on enhanced

n
features of formidability (e.g., facial and vocal expressions), physiological pre-

io
paredness for aggression if necessary, and a structured functional set of causes

at
and behavioral responses that are similar across cultures.

ci
so
As
CONCLUSION: HOW TO ACCOUNT FOR COMPLEX DESIGN

al
Social science has had difficulty accounting for complex features of

ic
human behavior largely because it has ignored the one known cause of com-

og
plex functional design in organisms: natural selection (Tooby & Cosmides,
ol
1992). Without functional theories capable of generating testable predictions
ch
about many different aspects of the domain of study, researchers have had two
y

choices: (a) retreat into smaller data sets that can be predicted and cogently
Ps

summarized by one or two main effects or (b) posit intuitive concepts to

account for a great range of data but remain computationally intractable and
an

pliable enough to account for shifting data sets.

ic

The alternative to computationally vague theories, or theories that are

er

specific but restricted to smaller data sets, are computationally specific, func-
Am

tional theories that posit many testable hypotheses based on a simple model of
selection pressures and the logical extensions of them given what is known
e

about human evolutionary history and the design of other animals. Natural
th

selection is the only process shown to be capable of organizing and designing

by

organisms. It is the only process that could have designed anger, and thus any
functional design in the anger system (including facial expressions, vocaliza-
11

tions, physiological changes, recalibrational learning mechanisms, and behav-

20

ioral responses) can be explained only as a computational system designed to

address the selection pressures that created anger. If one prefers to theorize that
©

anger is the result of learning, natural selection is the only process that could
ht

have designed the learning mechanism responsible for learning anger.

ig

The selection pressures described in this chapter may or may not be the
yr

ones most responsible for the functional structure of human anger, although
op

they have had great early success in making predictions and explaining many
of the most reliable and significant features of anger. Ultimately, the recali-
C

brational theory will be judged by its ability to generate computationally spe-

66 AARON SELL
12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 67

cific explanations capable of organizing and explaining features of anger that

have been intractable to intuitive theories and by its ability to make novel
predictions that lead to large, reliable effects. If it fails, the alternative must be
another computationally specific functional theory that is consistent with

A)
evolutionary biology. The data set on anger is now too large to be described
by simple learning mechanisms and too detailed to be accounted for by theories

P
(A
that do not computationally specify their primary components.

n
io
REFERENCES

at
ci
Archer, J. (1988). The behavioral biology of aggression. Cambridge, England: Cambridge

so
University Press.

As
Archer, J. (2006). Testosterone and human aggression: An evaluation of the challenge
hypothesis. Neuroscience and Biobehavioral Reviews, 30, 319–345. doi:10.1016/

al
j.neubiorev.2004.12.007

ic
Austad, S. (1983). A game theoretical interpretation of male combat in the bowl and

og
doily spider. Animal Behaviour, 31, 59–73. doi:10.1016/S0003-3472(83)80173-0
ol
Averill, J. (1982). Anger and aggression: An essay on emotion. New York, NY: Springer-
ch
Verlag.
y

Balikci, A. (1970). The Netsilik Eskimo. Garden City, NY: The Natural History Press.
Ps

Banse, R., & Scherer, K. (1996). Acoustic profiles in vocal emotion expression. Journal
of Personality and Social Psychology, 70, 614–636. doi:10.1037/0022-3514.70.3.614
an

Berg, I., & Fox, V. (1947). Factors in homicides committed by 200 males. Journal of
ic

Social Psychology, 26, 109–119.

er

Blumstein, A., & Cohen, J. (1980). Sentencing of convicted offenders: An analysis

Am

of the public’s view. Law & Society Review, 14, 223–261. doi:10.2307/3053313
Boehm, C. (1984). Blood revenge: The anthropology of feuding in Montenegro and other
e

tribal societies. Lawrence, KS: University Press of Kansas.

th

Brace, R., & Pavey, J. (1978). Size-dependent dominance hierarchy in the anemone
by

Actinia equine. Nature, 273, 752–753. doi:10.1038/273752a0

Chagnon, N. (1983). Yanomamo: The fierce people (3rd ed.). New York, NY: Holt,
11

Rinehart & Winston.

20

Clutton-Brock, T., & Albon, S. (1979). The roaring of red deer and the evolution of
honest advertisement. Behaviour, 69, 145–170. doi:10.1163/156853979X00449
©

Cristol, D. (1992). Food deprivation influences dominance status in dark-eyed juncos,

ht

Junco hyemalis. Animal Behaviour, 43, 117–124. doi:10.1016/S0003-3472(05)

ig

80077-6
yr

Daly, M., & Wilson, M. (1988). Homicide. New York, NY: Aldine de Gruyter.
op

Ekman, P. (1973). Cross-cultural studies of facial expression. In P. Ekman (Ed.),

Darwin and facial expression: A century of research in review (pp. 169–222). New
C

York, NY: Academic Press.

APPLYING ADAPTATIONISM TO HUMAN ANGER 67

12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 68

Enquist, M., & Leimar, O. (1983). Evolution of fighting behaviour: Decision rules
and assessment of relative strength. Journal of Theoretical Biology, 102, 387–410.
doi:10.1016/0022-5193(83)90376-4
Enquist, M., Leimar, O., Ljungberg, T., Mallner, Y., & Segerdahl, N. (1990). A test

A)
of the sequential assessment game: Fighting in the cichlid fish Nannacara anomala.
Animal Behaviour, 40, 1–14. doi:10.1016/S0003-3472(05)80660-8

P
(A
Epstein, S., & Taylor, S. (1967). Instigation to aggression as a function of degree of
defeat and perceived aggressive intent of the opponent. Journal of Personality,

n
35, 265–-289.

io
Fehr, E., & Gaechter, S. (2000). Cooperation and punishment in public goods experi-

at
ments. The American Economic Review, 90, 980–994.

ci
so
Felson, R. B. (1982). Impression management and the escalation of aggression and
violence. Social Psychology Quarterly, 45, 245–254. doi:10.2307/3033920

As
Galati, D., Sini, B., Schmidt, S., & Tinti, C. (2003). Spontaneous facial expressions

al
in congenitally blind and sighted children aged 8–11. Journal of Visual Impairment

ic
& Blindness, 97, 418–428.

og
Geen, R. (1968). Effects of frustration, attack and prior training in aggressiveness
upon aggressive behavior. Journal of Personality and Social Psychology, 9, 316–321.
doi:10.1037/h0026054
ol
ch
Guhl, A. M. (1956, February). The social order of chickens. Scientific American,
y
Ps

194(2), 42–46.
Hammerstein, P., & Parker, G. A. (1982). The asymmetric war of attrition. Journal
an

of Theoretical Biology, 96, 647–682. doi:10.1016/0022-5193(82)90235-1

ic

Hansen, A. J. (1986). Fighting behavior in bald eagles: A test of game theory. Ecology,
er

67, 787–797. doi:10.2307/1937701

Am

Harris, M. B. (1993). How provoking: What makes men and women angry? Aggres-
sive Behavior, 19, 199–211. doi:10.1002/1098-2337(1993)19:3<199::AID-
AB2480190305>3.0.CO;2-D
e
th

Hazlett, B. A., Rubenstein, D. & Rittschoff, D. (1975). Starvation, aggression and

energy reserves in the crayfish Orconectes virilis. Crustaceana, 28, 11–16.
by

doi:10.1163/156854075X00801
11

Heider, F. (1958). The psychology of interpersonal relations. New York, NY: Wiley.
doi:10.1037/10628-000
20

Holtgraves, T. (1989). The form and function of remedial moves: Reported use,
©

psychological reality, and perceived effectiveness. Journal of Language and Social

ht

Psychology, 8, 1–16. doi:10.1177/0261927X8900800101

ig

Huntingford, F. A., & Turner, A. K. (1987). Animal conflict. New York, NY: Chapman
yr

& Hall.
op

Jensen, P., & Yngvesson, J. (1998). Aggression between unacquainted pigs: Sequen-
tial assessment and effects of familiarity and weight. Applied Animal Behaviour
C

Science, 58, 49–61. doi:10.1016/S0168-1591(97)00097-X

68 AARON SELL
12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 69

Karsten, R. (1935). The head-hunters of western Amazonus. Helsingfors, Finland:

Societas Scientiarum Fennica.
Kaufmann, H. (1970). Aggression and altruism. New York, NY: Holt, Rinehart &
Winston.

A)
Luckenbill, D. (1977). Criminal homicide as a situated transaction. Social Problems,

P
25, 176–186. doi:10.1525/sp.1977.25.2.03a00050

(A
Maynard Smith, J., & Parker, G. A. (1976). The logic of asymmetric contests. Animal
Behaviour, 24, 159–175. doi:10.1016/S0003-3472(76)80110-8

n
io
Nowlis, V. (1941). The relation of degree of hunger to competitive interaction in chim-

at
panzee. Journal of Comparative Psychology, 32, 91–115. doi:10.1037/h0063065

ci
O’Leary, M., & Dengerink, H. (1973). Aggression as a function of the intensity and

so
pattern of attack. Journal of Research in Personality, 7, 61–70. doi:10.1016/0092-
6566(73)90032-9

As
Owens, D., & Owens, M. (1996). Social dominance and reproductive patterns in

al
brown hyaenas, Hyaena brunnea, of the central Kalahari Desert. Animal Behaviour,

ic
51, 535–551. doi:10.1006/anbe.1996.0058

og
Panksepp, J. (2000). Emotions as natural kinds within the mammalian brain. In
ol
M. Lewis & J. Haviland-Jones (Eds.), Handbook of emotions (2nd ed., pp. 137–156).
ch
New York, NY: Guilford Press.
Petrie, M. (1984). Territory size in the moorhen (Gallinula chloropus): An outcome of
y
Ps

RHP asymmetry between neighbours. Animal Behaviour, 32, 861–870. doi:10.1016/

S0003-3472(84)80163-3
an

Preston, K., & Stanley, K. (1987). “What’s the worst thing . . . ?”: Gender-directed
ic

insults. Sex Roles, 17, 209–219. doi:10.1007/BF00287626

er

Rime, B., & Giovannini, D. (1986). The physiological patterns of reported emotional
Am

states. In K. Scherer, H. Wallbott, & A. Summerfield (Eds.), Experiencing emotion:

A cross-cultural study (pp. 84–97). Cambridge, England: Cambridge University
Press.
e
th

Riordan, C., Marlin, N., & Kellogg, R. (1983). The effectiveness of accounts following
transgression. Social Psychology Quarterly, 46, 213–219. doi:10.2307/3033792
by

Rohwer, S., & Rohwer, F. (1978). Status signaling in Harris sparrows: Experimental
11

deceptions achieved. Animal Behaviour, 26, 1012–1022. doi:10.1016/0003-3472

(78)90090-8
20

Sargent, W. (1974). People of the valley. New York, NY: Random House.
©

Sell, A. (2006). Regulating welfare tradeoff ratios: Three tests of an evolutionary-

ht

computational model of human anger. Dissertation Abstracts International: Sec-

ig

tion B. Sciences and Engineering, 66(8-B), 4516.

yr

Sell, A., Cosmides, L., Tooby, J., Sznycer, D., von Rueden, C., & Gurven, M. (2009).
op

Human adaptations for the visual assessment of strength and fighting ability
from the body and face. Proceedings of the Royal Society B. Biological Sciences, 276,
C

575–584. doi:10.1098/rspb.2008.1177

APPLYING ADAPTATIONISM TO HUMAN ANGER 69

12051-04_CH03-rev2.qxd 5/25/10 11:30 AM Page 70

Sell, A., Tooby, J., & Cosmides, L. (2009). Formidability and the logic of human
anger. Proceedings of the National Academy of Science, 106(35), 15073–15078.
Shaver, P. R., Schwartz, J., Kirson, D., & O’Connor, C. (1987). Emotion knowledge:
Further exploration of a prototype approach. Journal of Personality and Social

A)
Psychology, 52, 1061–1086. doi:10.1037/0022-3514.52.6.1061
Siegel, A., & Schubert, K. (1995). Neurotransmitters regulating feline aggressive

P
(A
behavior. Reviews in the Neurosciences, 6, 47–61.
Sinclair, A. (1977). The African buffalo. Chicago, IL: University of Chicago Press.

n
io
Smuts, B., Cheney, D., Seyfarth, R., Wrangham, R., & Struhsaker, T. (1987). Primate

at
societies. Chicago, IL: University of Chicago Press.

ci
Stenberg, C. R., & Campos, J. J. (1990). The development of anger expression in

so
infancy. In N. Stein, B. Leventhal, & T. Trabasso (Eds.), Psychological and bio-
logical approaches to emotion (pp. 247–282). Hillsdale, NJ: Erlbaum.

As
Taylor, S. P. (1967). Aggressive behavior and physiological arousal as a function of

al
provocation and tendency to inhibit aggression. Journal of Personality, 35, 297–310.
doi:10.1111/j.1467-6494.1967.tb01430.x

ic
og
Tedeschi, J., & Felson, R. (1994). Violence, aggression, and coercive actions. Washington,
DC: American Psychological Association. doi:10.1037/10160-000
Toch, H. (1969). Violent men. Chicago, IL: Aldine.
ol
ch
Tooby, J., & Cosmides, L. (1992). The psychological foundations of culture. In
y
Ps

J. Barkow, L. Cosmides, & J. Tooby (Eds.), The adapted mind: Evolutionary

psychology and the generation of culture (pp. 19–136). New York, NY: Oxford
an

University Press.
Tooby, J., & Cosmides, L. (1996). Friendship and the banker’s paradox: Other pathways
ic

to the evolution of adaptations for altruism. Proceedings of the British Academy, 88,
er

119–143.
Am

Tooby, J., Cosmides, L., Sell, A., Liberman, D., & Sznycer, D. (2008). Internal reg-
ulatory variables and the design of human motivation: A computational and
e

evolutionary approach. In A. J. Elliot (Ed.), Handbook of approach and avoidance

th

motivation (pp. 251–271). New York, NY: Psychology Press.

by

Wallbott, H., Ricci-Bitti, P., & Baenninger-Huber, E. (1986). Nonverbal reactions

to emotional experiences. In K. Scherer, H. Wallbott, & A. Summerfield (Eds.),
11

Experiencing emotion: A cross-cultural study (pp. 98–116). Cambridge, England:

20

Cambridge University Press.

Wallbott, H., & Scherer, K. (1986). The antecedents of emotional experiences. In
©

K. Scherer, H. Wallbott, & A. Summerfield (Eds.), Experiencing emotion: A cross-

ht

cultural study (pp. 69–83). Cambridge, England: Cambridge University Press.

ig

Williams, G. (1966). Adaptation and natural selection. Princeton, NJ: Princeton

yr

University Press.
op
C

70 AARON SELL
12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 71

PA)
A BEHAVIORAL SYSTEMS

(A
PERSPECTIVE ON POWER

n
io
AND AGGRESSION

at
ci
so
PHILLIP R. SHAVER, MICHAL SEGEV, AND MARIO MIKULINCER

As
al
ic
og
ol
y ch
Ps
an

Attachment theory (Bowlby, 1982) has proven to be unique among

ic

psychodynamic theories in stimulating a large, coherent body of empirical

er

research (Cassidy & Shaver, 2008). Beginning with a focus on the importance
Am

for personality development of a child’s early relationships with primary care-

givers, Bowlby (1982) developed a theory of mental processes and social
e

behavior that has affected multiple domains of psychological research, includ-

th

ing developmental, social, and clinical psychology. The theory has been suc-
by

cessful partly because Bowlby retained some of the most valuable contributions
of prior psychoanalytic theorists (e.g., unconscious mental processes, self-
11

protective defenses, lasting effects of early social experiences) while adding

20

many important insights and research findings from cognitive psychology,

primate ethology, and community psychiatry. Along the way, Bowlby made
©

several important contributions to psychoanalytic motivation theory, as we

ht

explain in this chapter, but we are especially interested here in something he

ig

largely left out: Freud’s focus on aggression.

yr

For various reasons, Bowlby (1982) wanted to move away from Freud’s
op

(1920/1961) emphasis on sexual and aggressive instincts or drives. It seemed

misleading to conceptualize human infants’ seeking of intimate contact with
C

caregivers as inherently sexual (a mistake Freud may have made because he

71
12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 72

was struck early in his career by the number of his patients who had been sex-
ually abused—or, as he called it, “seduced”—by adults). It also seemed that
infant anger was a reaction to unreliable or frightening parental behavior, not a
primary motivational force in its own right (see Chapters 13 and 14, this volume).

A)
In other respects, Bowlby’s emphasis on what he called behavioral systems—that
is, functional patterns of motivated behavior that evolved over evolutionary

P
time because they contributed to human survival and successful reproduction—

(A
was laudable. It has been one of his ideas that has encouraged behavioral mea-

n
surement of attachment-related phenomena (e.g., Ainsworth, Blehar, Waters,

io
& Wall, 1978), something that earlier psychoanalytic theories did not do. The

at
purpose of the present chapter is to begin exploring the possibility that it would

ci
be worthwhile to supplement attachment theory with a hypothesized behavioral

so
system concerned with power or assertive influence. When this system goes

As
awry, one result might be dysfunctional anger and aggression.
In the following section of the chapter, we briefly explain the behavioral

al
system construct and describe some of the behavioral systems that Bowlby

ic
(1982) proposed (e.g., attachment, exploration, caregiving). We then propose

og
a power or assertion behavioral system and show that individual differences in
ol
the functioning of this system can be measured with self-report scales such as
ch
the ones we have used in the past to measure attachment and caregiving orien-
y

tations. We describe the development of the Power Behavioral System Scale

Ps

(PBSS), which measures hyperactivation and deactivation of the power behav-

ioral system. We then show how these two tendencies relate to other measures
an

of power, anger, and aggression, as well as measures of key personality traits,

ic

attachment insecurities, subjective well-being, and social functioning. We also

er

present preliminary evidence concerning the extent to which the new power
Am

system subscales predict individual differences in actual aggressive behavior.

e
th

THE BEHAVIORAL SYSTEM CONCEPT

by

In his trilogy on attachment theory, Attachment and Loss, Bowlby (1973,

11

1980, 1982), a psychoanalyst, conceptualized personality and social develop-

20

ment in terms of behavioral systems, a concept borrowed from ethology that

Bowlby believed could usefully replace Freud’s (1920/1961) notion of sexual
©

and aggressive, or life and death, instincts. A behavioral system is a species-

ht

universal neural program that governs the choice, activation, and termination
ig

of behavioral sequences in ways that Bowlby (1982) thought had increased

yr

the likelihood of survival and reproductive success in human evolutionary

op

history. He imagined different behavioral systems that served separate func-

tions, systems he called, for example, the attachment system, the exploration
C

system, and the caregiving system.

72 SHAVER, SEGEV, AND MIKULINCER

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 73

Each system was viewed as having a major aim or goal, for example,
attaining a sense of safety and security, curiously exploring and learning about
one’s environment, and promoting others’ (especially loved ones’) safety and
welfare. Each system was thought to include a repertoire of interchangeable,

A)
functionally equivalent behaviors that constitute the primary strategy used by
the system to attain its goal (e.g., maintaining proximity to a protective attach-

P
(A
ment figure in times of need). These sets of behaviors or behavioral tendencies
were thought to be “activated” automatically by stimuli or situations that made

n
a particular goal salient (e.g., loud noises that signaled danger and aroused a

io
need for protection) and were “deactivated” or “terminated” by other stimuli

at
that signaled goal attainment.

ci
Inasmuch as each behavioral system presumably evolved because it

so
increased the likelihood of coping successfully with environmental demands,

As
it is easy to understand why its optimal functioning in today’s human beings
is important for mental health, social adjustment, and a satisfying life.

al
Consider, for example, the attachment behavioral system. It is activated by per-

ic
ceived threats and dangers, which cause a person to seek proximity to another

og
person who is viewed as a “safe haven” and “secure base” (Bowlby, 1982).
ol
Successfully attaining proximity, protection, and emotional comfort from
ch
such an “attachment figure” results in what Sroufe and Waters (1977) called
y

felt security. Moreover, following repeated experiences of successfully attain-

Ps

ing protection and support, a person develops positive mental representations

of self and others that become an inherent part of the attachment system’s
an

operation. Hundreds of empirical studies provide compelling evidence for the

ic

benefits of attachment security in childhood and adulthood (for a review, see

er

Mikulincer & Shaver, 2007).

Am

Bowlby (1982) believed that the strategies associated with each behav-
ioral system undergo experience-based development. People learn to alter the
e

parameters or settings of their behavioral systems to fit pervasive contextual

th

demands, and in the process they form reliable expectations about available
by

access routes and likely barriers to goal attainment. These expectations, which
Bowlby (1973) called internal working models, become part of a behavioral sys-
11

tem’s neural wiring; their systematic and prolonged effects are observable as
20

individual differences in cognitive processes, emotional reactions, emotion-

regulation strategies, and personality traits.
©

Changes in behavioral strategies can be characterized in terms of

ht

hyperactivation or deactivation of the relevant behavioral system (e.g.,

ig

Cassidy & Kobak, 1988; Mikulincer & Shaver, 2007). Hyperactivating strate-
yr

gies intensify the primary strategy of the system to influence other people
op

to respond in accordance with the system’s goals (e.g., to provide adequate

emotional support). These strategies keep a behavioral system chronically
C

and intensely activated until its goal is achieved. They are learned in social

A BEHAVIORAL SYSTEMS PERSPECTIVE 73

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 74

environments that place a person on a partial reinforcement schedule for

noisily and successfully persisting. (The reinforcement schedule is partial
because other people’s responses to it are unreliable and unpredictable.)
Unfortunately, although hyperactivation is sometimes successful, it is also

A)
accompanied by heightened agitation and distress, which often upsets inter-
action partners by seeming overly intrusive, demanding, and controlling.

P
(A
In contrast, deactivating strategies involve down-regulation of a behavioral
system to reduce the frustration and anguish of repeatedly unsuccessful efforts

n
to attain the system’s goal. These strategies develop in the presence of people

io
who disapprove of or punish the system’s primary strategy (e.g., crying, reach-

at
ing, clinging). This disapproval or punishment suggests that one can expect bet-

ci
ter outcomes if the primary strategy of a particular behavioral system is blocked

so
or suppressed, which unfortunately means that the system’s goal is not often

As
fully attained. The problem with deactivating strategies is that they require a
narrowing of experience and the relative absence of many of life’s rewards (e.g.,

al
shared intimacy). By suppressing what would, in other circumstances, be nor-

ic
mal behavior, deactivating strategies prevent a person from realizing that there

og
are other social relationships or social environments in which the system’s pri-
mary strategy would be effective. ol
y ch
Ps

DEFINING THE POWER OR ASSERTION BEHAVIORAL SYSTEM

an

In previous work (e.g., Mikulincer & Shaver, 2007; Shaver, Mikulincer,

ic

& Shemesh-Iron, 2010) we have shown that stable individual differences in

er

hyperactivation and deactivation of the attachment system and the caregiv-

Am

ing system (both of which were proposed by Bowlby, 1982) can be measured
and that they are associated in theoretically predicted ways with other psy-
e

chological and behavioral processes. In the present chapter we propose, more

th

speculatively, that human beings are born with the rudiments of a behavioral
by

system the aim of which is to acquire and control material and social resources
(e.g., food, shelter, social status, sexual mates) that contribute to survival and
11

reproductive success. We propose further that these innate behavioral strate-

20

gies, as they emerge in development, are organized by an evolved and generally

adaptive power or assertion behavioral system, which for the sake of simplicity
©

we will call the power system. This system presumably evolved because it con-
ht

tributes to the propagation of one’s genes in a competitive social environ-

ig

ment. According to this fitness logic, the proliferation of a person’s genes

yr

depends on his or her ability to acquire and control precious resources and to
op

cope effectively with people and events that threaten resource control (see
Chapter 3, this volume).
C

74 SHAVER, SEGEV, AND MIKULINCER

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 75

In personality and social psychology, having a sense of power (analogous

to felt security in the attachment domain) is defined as perceiving that one
has control over valuable resources and outcomes within a particular situa-
tion (e.g., Keltner, Gruenfeld, & Anderson, 2003). Because resources are

A)
finite and people compete to acquire and control them, this definition implies
that people who have a sense of power also have control over others’ access to

P
(A
resources and can influence their behavior (see Chapter 3, this volume). This
reasoning led Keltner et al. (2003) to define power as “an individual’s relative

n
capacity to modify others’ states by providing or withholding resources or

io
administering punishments” (p. 265). Another implication of this definition

at
is that power often involves freedom and independence from others’ influence

ci
when seeking desired resources. Galinsky, Magee, Gruenfeld, Whitson, and

so
Liljenquist (2008) stated, “Power, it could also be said, is the capacity to be

As
uninfluenced by others. Without power, one’s outcomes are constrained by
others” (p. 1451).

al
Before conceptualizing the normative components and operations of the

ic
power behavioral system, we should distinguish power from aggression. From

og
an evolutionary standpoint, human aggression is a “fight” mechanism (e.g., Buss
ol
& Shackelford, 1997). It presumably evolved in many animal species because
ch
it facilitates control over precious resources, which makes it natural to equate
y

aggression with power. We emphasize, however, that there definitely are aggres-
Ps

sive acts whose sole purpose is to damage or destroy someone or something else,
and there are many cases in which one’s sense of power can be restored simply
an

by asserting one’s position and authority (see Chapter 3, this volume). We

ic

would like to consider the possibility that the power system is not “designed”
er

primarily to attack and destroy, but to gain, maintain, or restore one’s sense of
Am

power without necessarily damaging one’s own social ties or equanimity.

e

Normative Parameters of the Power System

th
by

We propose that the main goal of the power system is to remove threats
and obstacles that interfere with a person’s sense of power. In other words, the
11

power system seeks to maintain a stable inner sense of power and to restore
20

this sense when one perceives that others are attempting to constrain one’s
access to valuable resources or influence one’s behavior in a particular situa-
©

tion. This does not imply that people seek power simply for power’s sake.
ht

Rather, following Bowlby’s (1982) contention that attachment security (i.e.,

ig

felt security) provides a solid foundation for exploration, we propose that

yr

power facilitates the smooth functioning of other behavioral systems, such as

op

exploration, affiliation, caregiving, and sex. With a sense of power, people can
more easily explore and master their environment; help and get along with
C

A BEHAVIORAL SYSTEMS PERSPECTIVE 75

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 76

other people without worrying about being influenced, exploited, or con-

strained; and have sex with desirable partners and produce offspring to carry
one’s genes into future generations.
The proposed power behavioral system is likely to be activated in one

A)
of two kinds of situations: (a) when a person competes for access to valu-
able resources and (b) when other people constrain one’s access to resources

P
(A
or attempt to influence one’s attitudes and actions. In either case, people
are motivated to protect or restore their sense of power when they appraise

n
an event or social interaction as a threat to their power, not when they sim-

io
ply encounter someone who has a certain objective status or acts in a par-

at
ticular way. That is, the power system is not typically activated if a person

ci
detects no threat to his or her sense of power. By the same token, a person

so
can inappropriately appraise something as a threat even in the absence of

As
another person’s explicit signaling of competition, provocation, or superi-
ority (see Chapter 9).

al
Once a person’s power system is activated (appropriately or not), he or

ic
she calls on a repertoire of behaviors aimed at protecting or restoring a sense

og
of power. This repertoire, which reflects the system’s primary strategy, includes
ol
behaviors meant to maintain what Parker (1974) called resource-holding power,
ch
behaviors such as asserting one’s dominance, authority, and competence to
y

deal with the situation; expressing confidence in one’s strengths, attitudes, and
Ps

opinions; deterring others from competing for or exerting control over one’s
resources; and verbally or physically attacking (or threatening to attack)
an

others until power is restored (e.g., Gilbert, 1989; see also Chapter 3, this vol-
ic

ume). Beyond these basic strategies that can be observed in most animal
er

species, Gilbert (1989) also proposed that humans can protect or restore their
Am

sense of power by using what he called social attention-holding strategies,

efforts to emphasize one’s attractiveness and social value or display one’s spe-
e

cial talents, skills, and other positive attributes.

th

Activating these strategies is often accompanied by physiological

by

arousal and, often, a feeling of anger, which in our view is an emotional

signature of power-system activation. According to Lazarus (1991), the
11

core relational theme of anger is “a demeaning offense against me and mine”

20

(p. 222), an assault on or threat to one’s identity or other important per-

sonal goals and possessions. Shaver, Schwartz, Kirson, and O’Connor
©

(1987) viewed anger as a signal that control over an important resource is

ht

being threatened illegitimately and that some assertive action needs be

ig

taken to reduce or eliminate the threat, repair the damage, or prevent fur-
yr

ther assaults. In the second volume of his Attachment and Loss trilogy,
op

Bowlby (1973) argued that anger is a protest response against a partner’s

signs of unavailability, detachment, or rejection. Viewed from our concep-
C

tion of power, these signs indicate that one lacks the power to obtain

76 SHAVER, SEGEV, AND MIKULINCER

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 77

needed resources, such as affection and support, and that one is dependent
on others’ unreliable responsiveness.
When the power system, like other behavioral systems, works properly,
it contributes greatly to one’s subjective well-being and social adaptation

A)
(Keltner et al., 2003). Moreover, it encourages what Higgins (1998) called a
promotion focus—a motivational orientation that facilitates goal pursuit and

P
(A
realization of aspirations—because powerful people expect positive outcomes
from their efforts and relatively little interference from others. Research

n
shows that people with a sense of power devote attention to rewards and goal

io
pursuit, have more frequent positive emotions, and experience fewer threat-

at
related thoughts and emotions (for reviews, see Keltner et al., 2003; Chapter

ci
13, this volume).

so
As
Individual Differences in the Activation and Functioning
of the Power System

al
ic
Although we assume that everyone is born with the potential to develop

og
a stable sense of power, the functioning of the power system can be impaired by
ol
experiencing repeated failures to obtain desired outcomes, remove threats, and
ch
overcome obstacles. Such failures may result from physical illnesses that pre-
y

vent the effective use of the power system’s primary strategies. They may also
Ps

result from social arrangements that preclude or constrain competition; frus-

trate attempts to acquire needed resources; severely punish assertiveness, anger,
an

aggression, or other resource-holding power strategies; or demand submission

ic

or self-abasement. Such conditions may arouse anxiety about asserting oneself,

er

serious doubts about one’s power and influence, and a loss of confidence in one’s
Am

abilities to maintain or restore power when power is desirable.

As with the other behavioral systems, failures of the power system can
e

result in one or both of two nonoptimal (i.e., secondary) power strategies:

th

hyperactivation or deactivation of the power system. Hyperactivated power-

by

oriented behavior involves a substantial increase in efforts to restore a sense of

power despite adverse circumstances and doubts and anxieties that arise
11

when one experiences repeated and unpredictable failures. Hyperactivation

20

is fueled by two motives: an excessive urge to gain power and an extreme fear
of failure in the use of resource-holding power strategies. This combination
©

results in chronic activation of the power system, even when there is no

ht

imminent threat or actual damage to one’s power; an indiscriminate urge to

ig

assert power over others; frequent anger and hostility toward others (who are
yr

viewed as potential rivals); and a proclivity to attack others following min-

op

imal or ambiguous signs of competition or provocation. In an extreme form,

it can lead to flagrant vindictiveness and destructive, even murderous,
C

behavior. Hyperactivation is also characterized by a tendency to misinter-

A BEHAVIORAL SYSTEMS PERSPECTIVE 77

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 78

pret social situations as threatening or hostile (i.e., a hostile attribution bias;

see Chapter 9).
Deactivation, in contrast, involves terminating or “shutting off” the
power system, thereby giving up on the possibility of using the system’s pri-

A)
mary strategies to defend against threats and damages to one’s sense of power.
Deactivation is evident in submissiveness, self-abasement, and the absence of

P
(A
resource-holding power strategies, even in the presence of clear-cut, explicit
assault or provocation, to the point of experiencing substantial physical or

n
psychological harm as a result. Deactivation also involves a tendency to avoid

io
situations that call for activation of the power system and assertion of one’s

at
rights and opinions: competitions, arguments, disputes, and interpersonal

ci
conflicts. It is important to note, however, that such deactivation does not

so
necessarily involve reduced sensitivity to threats. In fact, powerless people are

As
often highly sensitive to threat-related cues, prone to ruminate about threats,
and experience negative emotions because of perceived threats and injustices

al
(for a review, see Keltner et al., 2003). In other words, deactivation is not a

ic
peaceful or calm state; it is characterized by a blend of worries, doubts, and

og
defenses against the pain and frustration of “losing” or dodging a “fight.”
ol
A review of self-report measures of beliefs, attitudes, and feelings related
ch
to power and aggression led us to conclude that no instrument had been
y

explicitly designed to assess the two nonoptimal power strategies, activation

Ps

and deactivation. We did gain insights from examining existing scales, how-
ever. We concluded, for example, that Buss and Perry’s (1992) Aggression
an

Questionnaire, which assesses several aspects of trait aggression (e.g., physi-

ic

cal aggression, verbal aggression), taps hyperactivated power, an extreme and

er

chronic proclivity to engage in hostile aggression. This scale, however, does

Am

not distinguish people with an optimally functioning power system from those
who deactivate their power system, because neither kind of person typically
e

relies on hostile aggression as a way to deal with competitions, arguments, or

th

conflicts. Similarly, low scores on Rathus’s (1973) Assertiveness Question-

by

naire might indicate power-system deactivation. However, this scale fails to

distinguish people with an optimally functioning power system from those
11

who hyperactivate the system, because both groups are capable of asserting
20

opinions (although often by different means).

There are also scales that measure a sense of power or dominance (e.g.,
©

Gough, 1964), but they fail to distinguish between people who hyperactivate
ht

or deactivate their power systems because both groups experience doubts and
ig

worries about the extent to which they have power over resources. One measure
yr

that taps constructs similar to the ones we are proposing here is the Inventory of
op

Interpersonal Problems (Horowitz, Rosenberg, Baer, Ureno, & Villasenor,

1988). It includes two subscales that describe forms of hyperactivation—overly
C

autocratic (e.g., “I try to control other people too much”) and overly competitive

78 SHAVER, SEGEV, AND MIKULINCER

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 79

(e.g., “I fight with other people too much”)—and two subscales that describe
forms of deactivation—overly subassertive (e.g., “It is hard for me to be assertive
with another person”) and overly exploitable (e.g., “I let other people take advan-
tage of me too much”). Despite these useful near-approximations to the mea-

A)
sures we were seeking, we decided to create a new measure that, like the
Caregiving System Scale (Shaver et al., 2010) and the Experiences in Close

P
(A
Relationships Inventory (ECR; Brennan, Clark, & Shaver, 1998) that is used
to measure insecure attachment, would be specifically designed to assess hyper-

n
activating and deactivating power strategies. (These scales are all part of an

io
overall effort to create a theory and set of measures to cover all major human

at
behavioral systems related to social behavior.) Here, we present data from the

ci
first stage of this research program: the construction of a self-report Power

so
Behavioral System Scale (PBSS) to assess individual differences in hyper-

As
activation and deactivation of the power system.

al
ic
ASSESSING HYPERACTIVATION AND DEACTIVATION

og
OF THE POWER SYSTEM
ol
ch
In the first stage of scale development, we constructed a pool of 50 items
y

that might index the two secondary power-system strategies. In writing the
Ps

items we attempted to capture the various cognitive, emotional, motivational,

and behavioral aspects of hyperactivated and deactivated power strategies
an

described in the present chapter. For example, the 25 items designed to assess
ic

hyperactivation focused on the urgent and exaggerated need for power and
er

control over resources and other people, frequent bouts of anger and aggres-
Am

sion, and anxieties and worries about being defeated in competitions and dis-
putes. The 25 items designed to assess deactivation of the system focused on
e

attempts to avoid behavioral assertions of power and authority as well as feel-

th

ings of uneasiness with competition and disputes.

by

The instructions asked respondents to think about situations in which

they had a disagreement or conflict with another person or group and to rate
11

the extent to which each item was or was not self-descriptive. Hence, the PBSS
20

measures a person’s general, overall orientation to power rather than the exer-
tion of power in a particular situation or relationship, although the items could
©

easily be adapted to assess domain-specific or partner-specific power strategies,

ht

as is sometimes done with the ECR when measuring attachment insecurities.

ig
yr

Factor Structure
op

Our initial 50-item scale was administered to a sample of 292 Israeli

C

undergraduates (185 women and 88 men). Item and factor analyses indicated

A BEHAVIORAL SYSTEMS PERSPECTIVE 79

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 80

that the items did assess the two secondary power-system strategies: hyper-
activating and deactivating. Based on these analyses, we chose the 14 most
representative items from each factor (i.e., the ones that loaded highest
on the intended factor and lowest on the other factor), keeping in mind

A)
our goal of representing various aspects of hyperactivation and deactiva-
tion. We then administered the new 28-item scale to 362 Israeli under-

P
(A
graduates (211 women and 151 men) and conducted a new factor analysis.
As expected, the analysis yielded the intended two factors, which accounted

n
for 52% of the variance. The 14 deactivation items loaded higher than .40

io
on the first factor (28% of explained variance). Here are examples of these

at
items: “I tend to relinquish important goals if their attainment requires

ci
confronting other people”; “I tend to avoid attacking, even if it’s a matter

so
of self-defense”; “I’d rather let others win an argument, even when I know

As
I’m right”; and “I’d rather not show people I’m angry, even when my anger
is justified.” The 14 hyperactivation items loaded higher than .40 on the

al
second factor (24%). The following are examples: “I feel anxious in situa-

ic
tions where I have little control over other people and their actions”;

og
“In an argument or disagreement, my strong desire to fight back makes it
ol
difficult for me to consider other possible responses”; “It’s hard for me to
ch
stop arguing, even when the other person has conceded”; and “When
y

somebody hurts me, I’m flooded with thoughts of revenge.” Cronbach

Ps

alphas were .85 for the hyperactivation items and .90 for deactivation items.
(Similar results were obtained in a replication study conducted in English
an

at the University of California, Davis.)

ic

As intended, the correlation between the hyperactivation and deacti-

er

vation scores was not statistically significant, r(360) = .07. That is, hyper-
Am

activation and deactivation are orthogonal strategies, and the two scales
form a two-dimensional space in which different power orientations can be
e

represented.
th
by

Stability Over Time and Across Reporters and Measure Type

11

In a new sample of 97 Israeli undergraduates, we administered the PBSS

20

twice, in sessions separated by 4 months, and found adequate test–retest reli-

ability (.74 and .79, respectively, for the hyperactivation and deactivation
©

scales). For another sample of 82 Israeli undergraduates, we asked two rela-

ht

tives, friends, or romantic partners of each participant to use the PBSS items
ig

to describe the participant. Significant correlations were found between self-

yr

reports and partner-reports regarding both the hyperactivation and deactiva-

op

tion dimensions, with rs ranging from .46 to .54. These findings imply that
the PBSS measures, in part, behavioral tendencies that can be observed by
C

relationship partners.

80 SHAVER, SEGEV, AND MIKULINCER

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 81

Convergent Validity

In eight additional samples of Israeli undergraduates (with ns ranging

from 120 to 178), we examined the convergent validity of the PBSS scores.

A)
First, we examined associations between the PBSS and preexisting self-
report measures tapping various aspects of aggression: the Aggression Ques-

P
(A
tionnaire (Buss & Perry, 1992), a measure of violence risk (Plutchik & Van
Praag, 1990), and the Abuse within Intimate Relationships Scale (Borjesson,

n
Aarons, & Dunn, 2003). The hyperactivation score was associated with

io
reports of physical aggression, verbal aggression, anger, and hostility; risk

at
of violent behavior; and abusive behavior in intimate relationship, with rs

ci
ranging from .27 to .46, all ps < .01. The deactivation score was not signif-

so
icantly associated with these measures. As expected, aggressive, violent,

As
and abusive behaviors can be viewed as manifestations of hyperactivated
power, but they do not differentiate between people scoring low or high on

al
the deactivation dimension. Stated differently, aggression and violence are

ic
not default strategies for gaining power but seem to develop from repeated

og
failure to control resources that eventually results in hyperactivation of the
power system. ol
ch
Second, we examined associations between the PBSS and preexisting
y

self-report measures of various aspects of anger arousal and expression: the

Ps

Anger Expression Scale (Spielberger, Jacobs, Russell, & Crane, 1983), the
Trait subscale of the State–Trait Anger Scale (Spielberger, 1983), the Multi-
an

dimensional Anger Inventory (Siegel, 1986), and the Anger Rumination

ic

Scale (Sukhodolsky, Golub, & Cronwell, 2001). In line with our theoretical
er

analysis, hyperactivation was associated with trait anger, anger externalization,

Am

anger arousal, hostile outlook, rumination on anger-related thoughts, and

problems in controlling anger expression, with rs ranging from .31 to .58, all
e

ps < .01. The deactivation score was not significantly associated with most of
th

these signs of anger, with the exception of a positive association with anger
internalization, r(176) = .37, p < .05. This finding suggests that an angry state
by

of mind is still active despite deactivation of the power system and that anger-
11

related feelings are directed toward the self rather than other people.
20

Third, we examined associations between the PBSS and preexisting self-

report measures of power, dominance, and assertiveness: the Dominance and
©

Abasement scales of the Personality Research Form (Jackson, 1984), and the
ht

Dominance and Submissiveness scales of the Revised Interpersonal Adjective

ig

Scales (Wiggins, Trapnell, & Phillips, 1988). As expected, both hyperactiva-

yr

tion and deactivation were associated with lower scores on scales measuring
feelings of dominance and power (rs ranging from −.33 to −.42, all ps < .01),
op

implying that these orientations may be alternative ways of coping with lack of
C

power. In addition, deactivation but not hyperactivation was associated with

A BEHAVIORAL SYSTEMS PERSPECTIVE 81

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 82

measures of submissiveness and self-abasement, with rs ranging from .30 to .65,

all ps < .01. These findings indicate that people who score high on deactiva-
tion of the power system suppress their own needs and desires while defer-
ring to others.

A)
Fourth, we examined associations between the PBSS and preexisting
self-report measures of interpersonal conflicts and reactions to others’

P
(A
transgressions: the Styles of Handling Interpersonal Conflict Scale (Rahim,
1983) and the Transgression-Related Interpersonal Motivations Inventory

n
(McCullough & Hoyt, 2002). As expected, hyperactivation was associated

io
with more aggressive and conflict-escalating behavior during conflicts and

at
greater vengeance and less forgiveness following others’ transgressions, with

ci
rs ranging from .29 to .51, all ps < .01. Correlations for the deactivation scale

so
were also compatible with our theoretical analysis: Deactivation was associ-

As
ated with avoidance and giving up during interpersonal conflicts and a ten-
dency to withdraw in response to interpersonal transgressions, with rs ranging

al
from .22 to .56, all ps < .01.

ic
og
Discriminant and Construct Validity
ol
ch
In the eight participant samples we also considered the discriminant
y

and construct validity of the PBSS scores. Participants completed the Crowne-
Ps

Marlowe Social Desirability Scale (Crowne & Marlowe, 1964), the Big
Five Inventory (John & Srivastava, 1999), and the ECR (Brennan et al.,
an

1998). The two PBSS scores were not associated with social desirability, rs
ic

< .05, but they were correlated in theoretically expected ways with other
er

personality constructs. Hyperactivation was associated with emotional

Am

instability (i.e., neuroticism), social avoidance and unpleasantness (i.e., low

agreeableness), and both attachment anxiety and avoidance, with rs rang-
e

ing from .26 to .54, p < .01. Deactivation was significantly associated with
th

introversion and high levels of neuroticism, agreeableness, and attachment

anxiety, with rs ranging from .22 to .37, all ps < .01. These correlations were
by

only moderate in size, suggesting that the PBSS scores are not simply redun-
11

dant with attachment insecurities or with broad personality traits. That is,
20

we believe that the PBSS measures something unique to the power system
that is not measured precisely by the other scales.
©

Whereas people who deactivate the power system tend to be introverted

ht

but agreeable, those who hyperactivate the system tend to be quarrelsome and
ig

socially unpleasant. However, despite these differences, people scoring high

yr

on either hyperactivation or deactivation of the power system share emo-

op

tional instability (i.e., neuroticism) and worries about being loved, accepted,
and esteemed by others (i.e., anxious attachment). That is, even the external
C

82 SHAVER, SEGEV, AND MIKULINCER

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 83

façade of agreeableness, submissiveness, and passivity that characterizes power

deactivation tends to be accompanied by inner emotionality, including
attachment-related anxieties (for a broader analysis of dispositional influences
on aggression, see Chapter 5).

A)
In the eight samples, we also examined the extent to which hyper- and
deactivated forms of power are associated with regulatory deficits and social

P
(A
skill deficits. Power hyperactivation was associated with deficits in emotion-
regulation, self-regulation, and social skills. The deficits were reflected in lower

n
scores on scales measuring self-control (Tangney, Baumeister, & Boone, 2004),

io
negative mood regulation (Catanzaro & Mearns, 1990), and social skills

at
(Buhrmester, Furman, Wittenberg, & Reis, 1988), with rs ranging from −.38

ci
to −.55, all ps < .01. As expected, hyperactivation was also correlated with two

so
of the interpersonal problems measured by Horowitz et al.’s (1988) Inventory

As
of Interpersonal Problems (IIP): being overly autocratic and overly competi-
tive, with rs of .44 and .47, ps < .01. Although deactivation was not associated

al
with problems in self-control, it was associated with lower scores on negative

ic
mood regulation and social skills, with rs ranging from −.42 to −.50, all ps <

og
.01. As expected, deactivation was also associated with two kinds of IIP inter-
ol
personal problems: being overly subassertive and overly exploitable, with rs of
ch
.58 and .52, ps < .01.
y

Taken together, these correlations imply that people who either hyper-
Ps

activate or deactivate their power system have a difficult time regulating their
negative emotions and lack the social skills that promote effective social
an

interactions. However, whereas those who hyperactivate the power system

ic

have problems with self-control and have difficulties in being overly aggres-
er

sive and competitive (see also Chapters 2 and 6), those who deactivate the
Am

system have problems related to subassertiveness.

e

Beneficial Correlates of Optimal Functioning of the Power System

th
by

In conceptualizing the power system, we assumed that optimal function-

ing of the system enhances a person’s self-esteem and sense of well-being. If
11

hyperactivation and deactivation interfere with optimal system functioning,

20

our scales should be inversely correlated with measures of positive psychologi-

cal states. In assessments of the eight participant samples, we included mea-
©

sures of self-esteem (Rosenberg, 1979), self-mastery (Pearlin, Menaghan,

ht

Lieberman, & Mullan, 1981), optimism (Scheier & Carver, 1985), and psycho-
ig

logical well-being (Veit & Ware, 1983). As expected, both hyperactivation and
yr

deactivation were associated with lower self-esteem, mastery, optimism, and

psychological well-being, with rs ranging from −.23 to −.46 for hyperactivation
op

and from −.22 to −.48 for deactivation, all ps < .01.

C

A BEHAVIORAL SYSTEMS PERSPECTIVE 83

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 84

Predicting Actual Behavior

In an additional study, we gathered preliminary data examining the pre-

dictive validity of the PBSS with respect to actual behavior during a couple

A)
conflict in the laboratory. Both members of 100 young Israeli heterosexual
couples who had been dating for less than 5 months completed the PBSS, the

P
(A
ECR, and the Big Five Inventory, and then were invited to a laboratory ses-
sion in which they were asked to discuss a major unresolved problem in their

n
relationship. Each couple was videotaped while discussing this problem. Two

io
independent judges who were unaware of participants’ scores on the other

at
measures rated the extent to which each member of the couple displayed signs

ci
of anger, hostility, and distress, and the extent to which they attacked their

so
partner, deferred submissively to their partner, and reached a joint solution

As
to the problem.
As expected, participants who scored higher on power hyperactivation

al
were rated by judges as displaying more anger, hostility, and distress, and as exe-

ic
cuting more attacks on their partner, with rs ranging from .33 to .46, ps < .01. In

og
contrast, participants who scored higher on power deactivation were rated by
ol
judges as displaying more distress, but not anger or hostility, and more submis-
ch
sive behavior, with rs of .49 and .42, ps < .01. In addition, both hyperactivation
y

and deactivation scores were associated with problems in finding a solution to

Ps

the relationship problem, with rs of −.32 and −.27, ps < .01. It is important that
these associations were unique to the PBSS and were not explained by attach-
an

ment insecurities or the Big Five personality trait scores. These findings indicate
ic

that conflict resolution between members of dating couples is impeded by either

er

deactivation or hyperactivation of the power system.

Am
e

CONCLUSIONS
th
by

The new concepts of power-system hyperactivation and deactivation fit

well with the rest of attachment theory and make increased sense of many of
11

the power, aggression, and anger measures that were already available in the
20

literature. Our new scales, although still preliminary, seem promising in effi-
ciently assessing power system hyperactivation and deactivation as orthog-
©

onal dimensions. As is the case in the attachment and caregiving domains

ht

outlined by attachment theory, both hyperactivation and deactivation, which

ig

are viewed as secondary strategies that come into play when a system’s pri-
yr

mary strategy fails to work, are associated with nonoptimal outcomes in adult
op

relationships. The new concepts and scales fit well in our overall conception
of social motivation based on the behavioral system construct. Their addition
C

to attachment theory may allow us to recover an important component of

84 SHAVER, SEGEV, AND MIKULINCER

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 85

Freud’s psychoanalytic theory that was deemphasized and almost omitted

from Bowlby’s theory.

A)
REFERENCES

P
(A
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attach-
ment: Assessed in the Strange Situation and at home. Hillsdale, NJ: Erlbaum.

n
Borjesson, W. I., Aarons, G. A., & Dunn, M. E. (2003). Development and confirma-

io
tory factor analysis of the Abuse within Intimate Relationship Scale. Journal of

at
Interpersonal Violence, 18, 295–309. doi:10.1177/0886260502250089

ci
so
Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. New
York, NY: Basic Books.

As
Bowlby, J. (1980). Attachment and loss: Vol. 3. Sadness and depression. New York, NY:
Basic Books.

al
ic
Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). New York, NY:

og
Basic Books.
ol
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of
adult attachment: An integrative overview. In J. A. Simpson & W. S. Rholes
ch
(Eds.), Attachment theory and close relationships (pp. 46–76). New York, NY:
y

Guilford Press.
Ps

Buhrmester, D., Furman, W., Wittenberg, M. T., & Reis, H. T. (1988). Five domains
an

of interpersonal competence in peer relationships. Journal of Personality and

Social Psychology, 55, 991–1008. doi:10.1037/0022-3514.55.6.991
ic
er

Buss, A. H., & Perry, M. (1992). The Aggression Questionnaire. Journal of Personal-
ity and Social Psychology, 63, 452–459. doi:10.1037/0022-3514.63.3.452
Am

Buss, D. M., & Shackelford, T. K. (1997). Human aggression in evolutionary psy-

chological perspective. Clinical Psychology Review, 17, 605–619. doi:10.1016/
e
th

S0272-7358(97)00037-8
Cassidy, J., & Kobak, R. R. (1988). Avoidance and its relationship with other defen-
by

sive processes. In J. Belsky & T. Nezworski (Eds.), Clinical implications of attach-

11

ment (pp. 300–323). Hillsdale, NJ: Erlbaum.

Cassidy, J., & Shaver, P. R. (Eds.). (2008). Handbook of attachment: Theory, research,
20

and clinical applications (2nd ed.). New York, NY: Guilford Press.
©

Catanzaro, S. J., & Mearns, J. (1990). Measuring generalized expectancies for nega-
tive mood regulation: Initial scale development and implications. Journal of Per-
ht

sonality Assessment, 54, 546–563. doi:10.1207/s15327752jpa5403&4_11

ig
yr

Crowne, D. P., & Marlowe, D. (1964). The approval motive: Studies in evaluative
dependence. New York, NY: Wiley.
op

Freud, S. (1961). Beyond the pleasure principle (J. Strachey, Trans.). New York, NY:
C

Norton. (Original work published 1920)

A BEHAVIORAL SYSTEMS PERSPECTIVE 85

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 86

Galinsky, A. D., Magee, J. C., Gruenfeld, D. H., Whitson, J. A., & Liljenquist, K. A.
(2008). Power reduces the press of the situation: Implications for creativity, con-
formity, and dissonance. Journal of Personality and Social Psychology, 95, 1450–1466.
doi:10.1037/a0012633

A)
Gilbert, P. (1989). Human nature and suffering. Hove, England: Erlbaum.

P
Gough, H. G. (1964). California Psychological Inventory manual. Palo Alto, CA: Con-

(A
sulting Psychologists Press.
Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motiva-

n
tional principle. In M. P. Zanna (Ed.), Advances in experimental social psychol-

io
ogy (Vol. 30, pp. 1–46). New York, NY: Academic Press.

at
ci
Horowitz, L. M., Rosenberg, S. E., Baer, B. A., Ureno, G., & Villasenor, L. (1988).

so
Inventory of Interpersonal Problems: Psychometric properties and clinical appli-
cations. Journal of Consulting and Clinical Psychology, 56, 885–892. doi:10.1037/

As
0022-006X.56.6.885
Jackson, D. N. (1984). Personality Research Form manual. Port Huron, MI: Research

al
Psychologists Press.

ic
og
John, O. P., & Srivastava, S. (1999). The Big Five trait taxonomy: History, measure-
ment, and theoretical perspectives. In O. P. John & L. A. Pervin (Eds.), Hand-
ol
book of personality: Theory and research (2nd ed., pp. 102–138). New York, NY:
ch
Guilford Press.
y

Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and inhi-
Ps

bition. Psychological Review, 110, 265–284. doi:10.1037/0033-295X.110.2.265

an

Lazarus, R. S. (1991). Emotion and adaptation. New York, NY: Oxford University
Press.
ic

McCullough, M. E., & Hoyt, W. T. (2002). Transgression-related motivational

er

dispositions: Personality substrates of forgiveness and their links to the Big

Am

Five. Personality and Social Psychology Bulletin, 28, 1556–1573. doi:10.1177/

014616702237583
e

Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynam-

th

ics, and change. New York, NY: Guilford Press.

by

Parker, G. A. (1974). Assessment strategy and the evolution of fighting behavior. Jour-
nal of Theoretical Biology, 47, 223–243. doi:10.1016/0022-5193(74)90111-8
11

Pearlin, L. I., Menaghan, E. G., Lieberman, M. E., & Mullan, J. T. (1981). The stress
20

process. Journal of Health and Social Behavior, 22, 337–356. doi:10.2307/2136676

©

Plutchik, R., & Van Praag, H. M. (1990). A self-report measure of violence risk.
Comprehensive Psychiatry, 31, 450–456. doi:10.1016/0010-440X(90)90031-M
ht

Rahim, M. A. (1983). A measure of styles of handling interpersonal conflict. Acad-

ig

emy of Management Journal, 26, 368–376. doi:10.2307/255985

yr

Rathus, S. A. (1973). A 30-item schedule for assessing assertive behavior. Behavior

op

Therapy, 4, 398–406. doi:10.1016/S0005-7894(73)80120-0

C

Rosenberg, M. (1979). Conceiving the self. New York, NY: Basic Books.

86 SHAVER, SEGEV, AND MIKULINCER

12051-05_CH04-rev2.qxd 5/25/10 11:30 AM Page 87

Scheier, M. F., & Carver, C. S. (1985). Optimism, coping, and health: Assessment
and implications of generalized outcome expectancies. Health Psychology, 4,
219–247. doi:10.1037/0278-6133.4.3.219
Shaver, P. R., Mikulincer, M., & Shemesh-Iron, M. (2010). A behavioral systems per-

A)
spective on prosocial behavior. In M. Mikulincer & P. R. Shaver (Eds.), Prosocial
motives, emotions, and behavior: The better angels of our nature (pp. 73–91). Wash-

P
ington, DC: American Psychological Association. doi:10.1037/12061-004

(A
Shaver, P. R., Schwartz, J., Kirson, D., & O’Connor, C. (1987). Emotion knowledge:

n
Further exploration of a prototype approach. Journal of Personality and Social Psy-

io
chology, 52, 1061–1086. doi:10.1037/0022-3514.52.6.1061

at
Siegel, J. M. (1986). The Multidimensional Anger Inventory. Journal of Personality

ci
and Social Psychology, 51, 191–200. doi:10.1037/0022-3514.51.1.191

so
Sroufe, L. A., & Waters, E. (1977). Attachment as an organizational construct. Child

As
Development, 48, 1184–1199. doi:10.2307/1128475
Sukhodolsky, D. G., Golub, A., & Cronwell, E. N. (2001). Development and vali-

al
dation of the anger rumination scale. Personality and Individual Differences, 31,

ic
689–700. doi:10.1016/S0191-8869(00)00171-9

og
Tangney, J. P., Baumeister, R. F., & Boone, A. L. (2004). High self-control predicts
ol
good adjustment, less pathology, better grades, and interpersonal success. Jour-
ch
nal of Personality, 72, 271–324. doi:10.1111/j.0022-3506.2004.00263.x
y

Veit, C. T., & Ware, J. E. (1983). The structure of psychological stress and well being
Ps

in general populations. Journal of Consulting and Clinical Psychology, 51, 730–742.

doi:10.1037/0022-006X.51.5.730
an

Wiggins, J. S., Trapnell, P., & Phillips, N. (1988). Psychometric and geometric char-
ic

acteristics of the revised interpersonal adjective scales (IAS-R). Multivariate

er

Behavioral Research, 23, 517–530. doi:10.1207/s15327906mbr2304_8

Am
e
th
by
11
20
©
ht
ig
yr
op
C

A BEHAVIORAL SYSTEMS PERSPECTIVE 87

C
op
yr
ig
ht
©
12051-05_CH04-rev2.qxd

20
11
5/25/10

by
th
e
11:30 AM

Am
er
ic
Page 88

an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 89

PA)
DISPOSITIONAL INFLUENCES

(A
ON HUMAN AGGRESSION

n
io
at
ci
JENNIFER L. TACKETT AND ROBERT F. KRUEGER

so
As
al
ic
og
ol
y ch
Ps
an

People differ in their propensity to engage in aggressive behavior. The

ic

primary goal of this chapter is to review recent research on stable individual

er

differences in aggressive tendencies. A secondary goal is to consider how these

Am

tendencies arise, that is, how genetic and environmental factors combine
to create individual differences in aggression. We begin by considering how
e

aggressive tendencies fit with other personality constructs and how aggression
th

is situated within prominent structural models of personality. We then extend

by

this understanding to encompass a broader range of more pathological exter-

nalizing or disinhibitory tendencies, describing how aggressive tendencies fit
11

with other forms of externalizing behavior. We conclude by reviewing specific

20

genetic and environmental factors that shape externalizing and aggressive

tendencies.
©
ht
ig

THE TRAIT APPROACH IN PERSONALITY PSYCHOLOGY

yr
op

The most common approach to examining enduring human dispositions

is personality trait theory. Traits are typically conceptualized as pervasive and
C

enduring characteristics that can be used to predict future behavior. The most

89
12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 90

widely accepted trait model is the five-factor model (FFM; e.g., Goldberg, 1993).
The FFM is a taxonomy of higher order, broadly defined personality traits that
include Neuroticism, Extraversion, Conscientiousness, Agreeableness, and
Openness to Experience. Neuroticism is defined by the experience of negative

A)
moods and emotions, such as anxiety, dysphoria, and low self-esteem. Extra-
version encompasses positive emotions as well as behaviors such as sociability

P
(A
and assertiveness. Conscientiousness includes abilities to organize and inhibit
behavioral tendencies. Agreeableness includes affiliation and altruism concep-

n
tualized as the opposites of interpersonal irritability and antagonism. Openness

io
to Experience includes intellectual curiosity as well as imagination and fantasy.

at
The FFM is replicable across languages and cultures, making it a useful

ci
general framework for characterizing major individual differences. In addition,

so
recent research has demonstrated empirical connections between various

As
higher order personality trait models, which provides a better understanding
of interrelationships among them (Markon, Krueger, & Watson, 2005). There

al
is much less agreement about the structure of lower order personality traits

ic
(e.g., the facets of the five major traits assessed with the commonly used per-

og
sonality instrument, the NEO–PI–R measure of the traits anxiety, excitement-
ol
seeking, trust, openness to feelings, and self-discipline; for details, see Costa &
ch
McCrae, 1992), even though lower order traits are useful for distinguishing
y

among specific behavioral outcomes (e.g., Reynolds & Clark, 2001).

Ps

Researchers have more recently addressed questions regarding the poten-

tial applicability of the FFM across the life span, including childhood (e.g.,
an

Halverson et al., 2003). Structural analyses of trait measures suitable for chil-
ic

dren have found empirical relations among higher order trait models that par-
er

allel those found in adults (Tackett, Krueger, Iacono, & McGue, 2008),
Am

allowing comparison across studies using different trait models. Scores on the
five trait factors change systematically across the adult years (Srivastava, John,
e

Gosling, & Potter, 2003), but individual differences on the factors generally
th

remain quite stable (e.g., McCrae & Costa, 1990).

by
11

THE EXTERNALIZING SPECTRUM: CONNECTING AGGRESSION

20

WITH OTHER FORMS OF MALADAPTIVE DISINHIBITION

©

In parallel with research on trait models of normal personality, such as

ht

the FFM, investigators have attempted to find the empirical structure of the
ig

mental disorders described in standard psychiatric classification systems.

yr

Although mental disorders have traditionally been conceptualized as discrete

op

and categorical, extensive evidence suggests that many disorders are better
conceived of in terms of continuous dimensions (Helzer, Kraemer, & Krueger,
C

2006). In addition, putatively separate disorders tend to blend into each

90 TACKETT AND KRUEGER

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 91

other, a phenomenon traditionally conceptualized in terms of comorbidity

(i.e., the co-occurrence of separate disorders). This phenomenon can be bet-
ter understood in terms of dimensionally organized spectrums of psychopathol-
ogy (Krueger & Markon, 2006).

A)
Psychopathology spectrums bear conceptual similarities to higher order
personality traits, both of which can be organized in a multidimensional space.

P
(A
This resemblance is empirical as well as conceptual. Especially relevant to this
chapter, forms of psychopathology that involve aggression can be viewed in

n
terms of a broad spectrum that has generally been labeled externalizing disorders.

io
Aggressive personality dispositions also fall within this spectrum (Krueger,

at
Markon, Patrick, Benning, & Kramer, 2007). The externalizing spectrum

ci
encompasses disinhibitory personality traits such as impulsivity and aggression,

so
as well as clinical disorders such as antisocial personality disorder and illicit sub-

As
stance dependence. In FFM terms, the externalizing spectrum is closely aligned
with a combination of disagreeableness and low conscientiousness, domains

al
that are associated with a higher order domain of disinhibition (Markon et al.,

ic
2005). That is, FFM disagreeableness and low conscientiousness are correlated

og
with each other, giving rise to a broader personality trait domain of disinhibi-
ol
tion. Stated somewhat differently, disinhibition (a combination of disagree-
ch
ableness and low conscientiousness) appears to be the personality-trait core of
y

aggressive and externalizing behavioral tendencies (see also Chapters 1, 2, and

Ps

6, this volume).
an

The Structure of the Externalizing Spectrum in Adults

ic
er

We recently sought to develop an empirical model of individual differences

Am

in externalizing tendencies, including diverse forms of aggression (Krueger et al.,

2007). To do this, we created 23 novel facet-level measures of various external-
e

izing tendencies. The 23 facets cover a variety of specific individual-difference

th

variables, ranging from those more traditionally characterized as personality

by

(e.g., problematic impulsivity) to those more traditionally characterized as

psychopathology or behavior problems (e.g., drug problems). Research participants
11

completed self-report questionnaires containing the items, and then the items
20

were pruned and refined based on psychometric analyses of the results. Both
factor and cluster analyses of the item-level data were used to isolate specific,
©

narrow-band facets of the broader externalizing spectrum.

ht

With specific facets in hand, we then used item response theory (IRT)
ig

analyses to ensure that the items indexing a specific facet also covered a range
yr

of individual differences within that facet. IRT models are suited to this task
op

because they parameterize the location of items along a specific dimension. The
goal was to ensure that the items were appropriately arrayed to cover a wide
C

range of the dimension. Items were then deleted or revised, and additional

DISPOSITIONAL INFLUENCES ON HUMAN AGGRESSION 91

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 92

items written, to improve our ability to measure facets of externalizing tenden-

cies well, across their entire range (for additional methodological details, see
Krueger et al., 2007). Three specific facets of aggression were identified in the
course of this scale development project: relational, physical, and destructive

A)
aggression. In addition to these three facets of aggression, 20 additional facets
of externalizing were identified (for a complete list, see Krueger et al., 2007,

P
(A
Table 2). These additional facets cover a range of content, including deficient
empathy; a tendency to externalize blame (i.e., to blame others for one’s own

n
problems); alienation from others in the interpersonal environment; problems

io
with alcohol, marijuana, and “harder” substances; difficulties controlling

at
impulses; behavior that would be grounds for arrest (e.g., theft and fraud); dis-

ci
honesty; irresponsibility; rebelliousness; a tendency to prefer exciting but

so
potentially unsafe activities to safer but more boring activities; and proneness

As
to boredom.

al
Structural Analysis of the 23 Facets of the Externalizing

ic
Spectrum in Adults

og
ol
Exploratory analyses of the 23 facets of the externalizing spectrum sug-
ch
gested that three factors could account for the relations among the 23 facets:
y

(a) a broad, general factor (i.e., overall externalizing tendencies, indicated by

Ps

all of the scales to a nontrivial extent but most strongly indicated by irrespon-
sible and impulsive tendencies); (b) a narrower factor encompassing callous
an

aggression (indicated primarily by the three aggression facets and a facet

ic

indexing lack of empathy); and (c) a narrower factor encompassing use of and
er

problems with substances (indicated primarily by the alcohol, marijuana, and

Am

“hard drug” facets).

Following these exploratory analyses, we compared the fit of three confir-
e

matory models: (a) a one-factor model, where all 23 scales were indicators of
th

one and only one general externalizing factor; (b) a higher order model, where
by

the two narrower factors (callous aggression and substance use/problems) load
directly on the 23 facet scales, and the general externalizing factor, in turn,
11

loads on callous aggression and substance use/problems; and (c) a bifactor struc-
20

tural model (also known as a hierarchical model; Yung, Thissen, & McLeod,
1999) of the three factors (i.e., general externalizing, callous aggression, sub-
©

stance use/problems), where all indicators load on the general factor, and indi-
ht

cators of callous aggression and substance use/problems also load on their

ig

respective narrower factors, but the factors are modeled as uncorrelated.

yr

The contrast between the higher order and bifactor models pertains to the
op

ways in which relations among the general, broad externalizing factor and the
narrower callous aggression and substance use/problems factors are modeled.
C

In the higher order model, the general, broad externalizing factor is indicated

92 TACKETT AND KRUEGER

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 93

indirectly via loadings of the narrower callous aggression and substance use/
problems factors, which are at an intermediate level between the 23 primary
facets and the general, broad externalizing factor. By contrast, in the bifactor
model, all 23 primary facets load directly on the general, broad externalizing

A)
factor, and specific subsets of scales also load directly on the narrower factors.
Multiple fit indices converged to indicate a superior fit for the bifactor

P
(A
model over the higher order model (Krueger et al., 2007). This finding had
intriguing implications for the meaning of individual differences in aggressive

n
tendencies, as indexed by the three aggression facet scales (i.e., relational, phys-

io
ical, destructive). These implications were revealed by considering the struc-

at
ture of the bifactor model (i.e., three mutually uncorrelated factors) along with

ci
the relative magnitudes of the loadings of the aggression scales on those factors.

so
Relational aggression was a stronger indicator of the callous–aggressive factor

As
than of the general externalizing factor, physical aggression was more closely
linked to overall externalizing than to callous aggression, and destructive

al
aggression loaded similarly on overall externalizing and callous aggression

ic
(albeit higher on overall externalizing).

og
The general conclusion is that there are multiple pathways to aggressive
ol
outcomes. Specific aggressive outcomes can emerge because a person’s impulses
ch
are unconstrained in general (i.e., via the general externalizing factor) or
y

because a person is unusually callous and aggressive (i.e., via the callous–
Ps

aggressive factor, which is uncorrelated with general externalizing, this path-

way applying in particular to relational aggression). As described earlier, evi-
an

dence in favor of this multiple-pathways conceptualization derives from the

ic

superior fit of the bifactor model, when compared with the two other models
er

(i.e., the one factor model and higher order model). In the bifactor model,
Am

the factors are mutually uncorrelated. Hence, the factors can be interpreted
as independent, multiple pathways to a specific outcome such as relational
e

aggression because relational aggression loaded notably on both factors, that

th

is, variation in relational aggressive tendencies can be traced to at least two,

by

independent, underlying processes.

11

The Externalizing Spectrum in Children and Adolescents

20

Research examining associations between the domains of personality

©

and psychopathology in childhood and adolescence has produced results that

ht

are largely consistent with findings for adults, although these younger age
ig

groups are less often studied in the personality literature. Aggressive and
yr

externalizing behaviors in children are related primarily to disinhibitory dispo-

op

sitions (i.e., disagreeableness and low conscientiousness; Lahey & Waldman,

2003; Nigg, 2006; Tackett, 2006), although much of this research has been
C

correlational in nature, precluding a deeper understanding of the develop-

DISPOSITIONAL INFLUENCES ON HUMAN AGGRESSION 93

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 94

ment of associations between personality and psychopathology over time

(Tackett, 2006).
The longitudinal research that does exist pertaining to connections
between personality and externalizing behaviors has generally been con-

A)
ducted within a vulnerability, or risk, framework, which postulates that per-
sonality or temperament serves as a risk factor for the development of later

P
(A
disorders. An alternative framework is the spectrum or common cause model,
which conceptualizes personality and externalizing psychopathology as

n
dimensionally related and as sharing core etiologic factors. One issue in this

io
literature is that longitudinal studies that fail to measure potentially shared

at
causal factors cannot disentangle evidence for a vulnerability model from evi-

ci
dence for a spectrum model (Tackett, 2006).

so
Externalizing disorders in younger age groups have their own disorder cat-

As
egories in the Diagnostic and Statistical Manual of Mental Disorders (4th ed., text
rev., American Psychiatric Association, 2000). They include oppositional defi-

al
ant disorder, conduct disorder, and attention-deficit/hyperactivity disorder

ic
(ADHD). Other potential externalizing behaviors in early adolescence include

og
precocious sexual behavior and early substance use. Some behaviors, such as
ol
relational aggression (i.e., using social power and social exclusion to aggress
ch
against one’s victims), are still relatively new as forms of pathological aggres-
y

sion and are not yet incorporated into the childhood externalizing disorders
Ps

framework (e.g., Tackett, Waldman, & Lahey, 2009), although relational

aggression was included as a key facet when we developed the aforementioned
an

model of adult externalizing behaviors (Krueger et al., 2007).

ic

One important task for future research is to identify both common and
er

unique personality correlates that provide a more comprehensive picture of

Am

the hierarchical nature of externalizing behavior problems (Tackett, 2010).

For example, relational aggression appears to have personality correlates sim-
e

ilar to those of other externalizing behaviors: lower conscientiousness and

th

agreeableness (Schell & Tackett, 2010). This finding suggests that we should
by

connect relational aggression to a broader externalizing spectrum, as we did

in the studies of adults described earlier (Krueger et al., 2007). This work is
11

currently limited by the lack of strong empirical studies of the structure of the
20

personality, temperament, and disorder measures used in studies of children

and adolescents.
©

As noted previously, lower order personality traits may be especially useful

ht

in predicting aggressive and externalizing behaviors (e.g., Paunonen & Ashton,

ig

2001; Reynolds & Clark, 2001), but the structure of these traits in childhood
yr

and adolescence is also not well studied (Shiner, 1998; Tackett et al., 2008).
op

Recently proposed models of childhood personality traits offer possible lower

order facets for further investigation (Halverson et al., 2003), as do common
C

models of temperament (e.g., Rothbart, Ahadi, Hershey, & Fisher, 2001). As

94 TACKETT AND KRUEGER

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 95

mentioned previously, lower order traits reflect more narrowly defined charac-
teristics (e.g., warmth) than do more broadly defined higher order traits (e.g.,
extraversion). A few theoretical reviews have begun to compare and integrate
proposed lower order trait structures in different models (e.g., Halverson et al.,

A)
2003; Shiner & Caspi, 2003), which provides an important starting point for
future empirical work.

P
(A
An additional place to look for relevant lower order traits is the research
literature on such narrowly defined personality characteristics as callous–

n
unemotional traits, which are related to conduct problems (e.g., Frick,

io
Cornell, Barry, Bodin, & Dane, 2003). Callous–unemotional characteristics

at
reflect remorselessness and a lack of empathy. Frick and colleagues have begun

ci
to integrate callous–unemotional traits into broader personality and tempera-

so
ment models, which have identified consistent connections with the broader

As
externalizing work by highlighting the importance of negative emotionality
and effortful control (Frick & Sheffield-Morris, 2004) and low levels of agree-

al
ableness and conscientiousness (Essau, Sasagawa, & Frick, 2006). One useful

ic
approach might be to use the 23-facet model developed by Krueger et al.

og
(2007) for adults and extend it downward to see whether it applies to children
ol
and adolescents. Some parallels already seem clear. For example, callousness
ch
figures prominently as a narrow-band factor in the Krueger et al. (2007) model,
y

as it does in the thinking of Frick and his colleagues.

Ps
an

INTEGRATING ENVIRONMENTAL
ic

AND SITUATIONAL FACTORS

er
Am

A diathesis–stress perspective (i.e., a perspective in which a vulnerabil-

ity combines with stressful experiences) can be applied to conceptualizing
e

personality’s influence on the development of externalizing problems. In fact,

th

examining person-by-situation interactions is already a flourishing approach

by

to personality research (for a recent review, see Funder, 2008) and is useful
for understanding aggressive and externalizing tendencies (see Chapters 1, 7,
11

8, and 9, this volume). Personality traits are substantially influenced by

20

genetic factors and this general finding appears to be largely consistent across
the life span.
©

In addition to stability, of course, there is evidence for personality change.

ht

It is important to adopt a dynamic perspective on the influences of personality

ig

on behavioral outcomes that includes both stability and change and considers
yr

environmental factors that can impinge on a person at different points in devel-

op

opment. For example, although early research demonstrated connections

between trait impulsivity, negative affect, and increased risk of suicidal behav-
C

iors (Caspi, Moffitt, Newman, & Silva, 1996), recent research suggests that this

DISPOSITIONAL INFLUENCES ON HUMAN AGGRESSION 95

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 96

is particularly important early in life, with the strength of association declining

across the life span (McGirr et al., 2008). Similarly, research suggests that early-
onset conduct problems are more likely to be influenced by dispositional fac-
tors than those that arise later in development (Lahey & Waldman, 2003;

A)
Moffitt, 2003).
Personality can be an important moderator of individual responses to a

P
(A
particular kind of situation, including aggressive responses. For example, indi-
viduals low in Conscientiousness are more likely to engage in aggressive behav-

n
ior in the face of anger-provoking stimuli than individuals who are high in

io
Conscientiousness (Jensen-Campbell, Knack, Waldrip, & Campbell, 2007).

at
That is, the experience of angry affect alone does not determine a person’s

ci
responses, because self-regulatory capacities act on the affective experience to

so
differentiate individual responses. This is consistent with temperament models

As
that differentiate approach/positive emotionality and avoidance/negative emo-
tionality from self-regulatory traits that are considered superordinate in these

al
models (e.g., Ahadi & Rothbart, 1994; Carver, Johnson, & Joormann, 2008;

ic
Clark, 2005).

og
Psychobiological factors such as hormones, genes, and cortical activity
ol
also play an important role in explaining connections between personality traits
ch
and externalizing behaviors (see Chapter 6; see also Chapter 8 and 9). For
y

example, low serotonergic functioning is associated with personality traits such

Ps

as hostility and impulsivity as well as aggressive behavior (e.g., Carver et al.,

2008). This connection has been illuminated by molecular genetic investiga-
an

tions of the short allele of the serotonin 5-HTTLPR polymorphism (Carver

ic

et al., 2008; see also Chapter 8, this volume). Electroencephalograph (EEG)

er

studies of brain processes have also shed light on associations between person-
Am

ality variables and externalizing behaviors. For example, reduced amplitude of

the P300 event-related potential measured by EEG during a visual oddball task
e

has been linked to both externalizing behaviors and trait impulsivity (Iacono,
th

Malone, & McGue, 2003). This research provides a possible endophenotype,

by

or marker, for genetic risk of externalizing behavior (Hicks et al., 2007).

It is also possible that psychobiological risk factors increase response
11

strength to external stimuli, leading to differential expressions of aggression and

20

antisocial behavior (Hay, 2007). Such processes can be properly investigated

only with research designs that include measures of both person and situation
©

variables (for a taxonomy of such interactions, see Chapter 2, this volume). For
ht

example, adolescents with early-onset conduct disorder (compared with con-

ig

trols and adolescents with ADHD) show increased left-sided amygdala activa-
yr

tion when presented with negative pictures (Herpertz et al., 2008). In related
op

work, individuals with high levels of trait anger (compared with individuals
lower in trait anger) show increased left frontal activation in response to anger-
C

provoking pictures (Harmon-Jones, 2007). These findings suggest psycho-

96 TACKETT AND KRUEGER

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 97

biological predispositions that influence reactivity to negative or threaten-

ing situations.

A)
GENETIC RESEARCH ON EXTERNALIZING PHENOMENA

P
(A
In recent years, genetic research on externalizing syndromes and behav-
iors has focused on exploring the coherence of these syndromes as elements

n
within a broader spectrum, following from the phenotypic work described

io
earlier in this chapter. Genetic effects on different externalizing syndromes

at
are mostly common across these syndromes, but there are also specific genetic

ci
effects on substance-dependence syndromes that are not shared with other

so
externalizing syndromes (Kendler, Prescott, Myers, & Neale, 2003; Krueger

As
et al., 2002; Young, Stallings, Corley, Krauter, & Hewitt, 2000). This makes
a great deal of physiological sense: It is reasonable for dependence on sub-

al
stances to be traceable to both genetic effects unique to substances (presum-

ic
ably reflecting substance metabolism) and more general effects, presumably

og
reflecting disinhibited personality traits such as disagreeableness and lack of
ol
conscientiousness. Note the similarity of these findings to the model proposed
ch
by Krueger et al. (2007) and described earlier, where drug problems were
y

affected by a specific factor, beyond the effect of the general externalizing fac-
Ps

tor. The heritability of general externalizing tendencies is quite substantial,

around 80%, suggesting that this general factor would be a good target for
an

gene-hunting studies.
ic

Along these lines, Dick et al. (2008) recently conducted molecular

er

genetic research on alcohol dependence, antisocial personality disorder,

Am

conduct disorder, drug dependence, novelty seeing, sensation seeking, and a

general externalizing tendency linking these syndromes. The strength of
e

genetic linkage was stronger for the externalizing component than for the
th

individual syndromes. In particular, a region on Chromosome 7 appeared to

by

contribute to general risk, transcending specific externalizing syndromes (cf.

Stallings et al., 2005).
11

Going beyond linkage, Dick and her colleagues also studied single
20

nucleotide polymorphisms (SNPs) in the CHRM2 gene. CHRM2 had been

associated with alcohol dependence in previous research and has been associ-
©

ated with risk of alcohol dependence combined with drug dependence, mak-
ht

ing it an appealing potential candidate gene for studies focused on general

ig

externalizing tendencies. In Dick et al.’s (2008) study, general externalizing,

yr

as compared with the other syndromes, was most strongly associated with with
op

SNPs in CHRM2.
With regard to Gene × Environment interactions, both twin studies and
C

molecular genetic studies suggest that environmental effects are critical in

DISPOSITIONAL INFLUENCES ON HUMAN AGGRESSION 97

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 98

shaping the expression of genetic risk of externalizing behavior (see Chap-

ter 8). Statistical indexes such as heritability (the overall magnitude of
genetic influence on a phenotype) are typically estimated for an entire pop-
ulation. For example, the 80% heritability of externalizing behavior is an esti-

A)
mate that applies to an entire population, without regard to various subgroups
within that population. Refining these general estimates with subgroup infor-

P
mation is one way to pursue Gene × Environment interaction effects, because

(A
genetic effects may be moderated by measurable environmental variables

n
associated with population subgroups.

io
An example is a recent study by Legrand, Keyes, McGue, Iacono, and

at
Krueger (2008). A sample of 17-year-old twins assessed on diverse externaliz-

ci
ing syndromes was divided into those living in rural areas and those living in

so
urban areas. In urban areas, genetic influences predominated, but in rural areas,

As
shared environmental effects (environmental effects making people similar
within families) predominated. This suggests that the previously described 80%

al
heritability masks interesting and potentially important subgroup differences.

ic
In the domain of molecular genetic studies, Caspi et al. (2002) examined

og
how childhood maltreatment interacted with a gene coding for MAO-A
ol
(monoamine oxidase-A, an enzyme that metabolizes major neurotransmitters)
ch
in predicting antisocial behavior in a birth cohort of males. These investiga-
y

tors found that the effect of childhood maltreatment on antisocial behavior

Ps

was moderated by MAO-A genotype. The genetic polymorphism that pre-

dicts high MAO-A activity had a protective effect, such that men with this
an

genotype were protected from the measured deleterious effects of childhood

ic

maltreatment. Although there are exceptions, this finding has proven to be

er

generally replicable (Kim-Cohen et al., 2006).

Am
e

SUMMARY AND CONCLUSIONS

th
by

The hierarchical trait model provides a compelling way to conceptualize

stable dispositional characteristics that differentiate individuals. Such hierar-
11

chical models have been integrated with structural dimensions of psycho-

20

pathology, including externalizing behaviors such as aggression and violence

(Krueger et al., 2007). Studies aimed at the higher order trait level have
©

implicated primarily disinhibitory traits such as disagreeableness and a lack of

ht

conscientiousness in connection with aggression measured from early child-

ig

hood through adulthood (Blonigen & Krueger, 2007; Tackett, 2006). Lower
yr

order personality traits may offer increased predictive validity over higher
op

order traits, particularly for specific behavioral outcomes such as physical

aggression and violence. Research linking lower order traits to externalizing
C

behaviors is much less common, but extant research implicates alienation in

98 TACKETT AND KRUEGER

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 99

adults (Blonigen & Krueger, 2007) and callous–unemotional traits in chil-

dren (Frick et al., 2003).
Much of our recent work has been conducted within the externalizing
spectrum framework (Krueger et al., 2007; Tackett, 2010), which posits dimen-

A)
sional relations among correlated personality traits and externalizing behaviors.
This approach has been fruitful in identifying factors common to the broader

P
(A
externalizing domain as well as specific factors differentiating types of external-
izing behaviors (Krueger et al., 2002, 2007; Tackett, Krueger, Sawyer, & Graetz,

n
2003; Tackett, Krueger, Iacono, & McGue, 2005; Tackett et al., 2009). Indeed,

io
as work on molecular genetic linkage and Gene × Environment interactions

at
progresses, it is possible that the spectrum approach will be helpful in the search

ci
for specific genes and relevant environmental stressors (Dick et al., 2008).

so
Nevertheless, limitations to a full understanding of personality connec-

As
tions to externalizing behaviors remain. There is not yet a clear picture of the
life span trajectory of the externalizing spectrum (Tackett, 2010). For example,

al
some disorders typically conceptualized as externalizing syndromes in child-

ic
og
hood do not have officially recognized analogs in adult disorder typologies (e.g.,
ADHD, oppositional defiant disorder). Other behaviors, such as relational
ol
aggression, which are often viewed as externalizing (e.g., Baker, Jacobson,
ch
Raine, Lozano, & Bezdjian, 2007; Krueger et al., 2007), are not yet clearly iden-
y

tified as pathological behaviors in younger age groups (Tackett et al., 2009). In

Ps

addition, pathways from early temperamental traits to adult personality traits

have not been fully articulated, limiting our ability to provide a complete life
an

span perspective.
ic

As psychopathology research moves toward a broad person × environ-

er

ment perspective, it presents an opportunity for integration with developmen-

Am

tal approaches, which have often emphasized environmental influences on

specific behavioral outcomes (Jenkins, 2008). Epigenetic approaches, which
e

focus on the role of gene expression, are also becoming increasingly salient to
th

psychopathology researchers. Methodological advances in epigenetics offer

by

new opportunities for integrating complex personality and psychopathology

phenotypes (Kaminsky et al., 2008; Mill & Petronis, 2007). They highlight
11

the need to focus on particular developmental periods and processes and on

20

better measurement of potential environmental influences.

©
ht

REFERENCES
ig
yr

Ahadi, S. A., & Rothbart, M. K. (1994). Temperament, development, and the

op

Big Five. In C. F. Halverson, Jr., G. A. Kohnstamm, & R. P. Martin (Eds.),

The developing structure of temperament and personality from infancy to adulthood
C

(pp. 189–207). Hillsdale, NJ: Erlbaum.

DISPOSITIONAL INFLUENCES ON HUMAN AGGRESSION 99

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 100

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental

disorders (4th ed., text rev.). Washington, DC: Author.
Baker, L. A., Jacobson, K. C., Raine, A., Lozano, D. I., & Bezdjian, S. (2007). Genetic
and environmental bases of childhood antisocial behavior: A multi-informant

A)
twin study. Journal of Abnormal Psychology, 116, 219–235. doi:10.1037/0021-
843X.116.2.219

P
(A
Blonigen, D. M., & Krueger, R. F. (2007). Personality and violence: The unifying
role of structural models of personality. In I. Waldman, D. J. Flannery, & A. T.

n
Vazsonyi (Eds.), The Cambridge handbook of violent behavior (pp. 288–305). Cam-

io
bridge, England: Cambridge University Press.

at
Carver, C. S., Johnson, S. L., & Joormann, J. (2008). Serotonergic function, two-mode

ci
models of self-regulation, and vulnerability to depression: What depression has

so
in common with impulsive aggression. Psychological Bulletin, 134, 912–943.

As
doi:10.1037/a0013740
Caspi, A., McClay, J., Moffitt, T., E., Mill, J., Martin, J., Craig, I. W., Taylor, A., &

al
Poulton, R. (2002, August). Role of genotype in the cycle of violence in mal-

ic
treated children. Science, 297, 851–854. doi:10.1126/science.1072290

og
Caspi, A., Moffitt, T. E., Newman, D. L., & Silva, P. A. (1996). Behavioral obser-
ol
vations at age 3 years predict adult psychiatric disorders. Archives of General
ch
Psychiatry, 53, 1033–1039.
y

Clark, L. A. (2005). Temperament as a unifying basis for personality and psycho-

Ps

pathology. Journal of Abnormal Psychology, 114, 505–521. doi:10.1037/0021-

843X.114.4.505
an

Costa, P. T., Jr., & McCrae, R. R. (1992). NEO PI-R professional manual. Odessa, FL:
ic

Psychological Assessment Resources.

er

Dick, D. M., Aliev, F., Wang, J. C., Grucza, R. A., Schuckit, M., Kuperman, S.,
Am

. . . Goate, A. (2008). Using dimensional models of externalizing psycho-

pathology to aid in gene identification. Archives of General Psychiatry, 65,
e

310–318.
th

Essau, C. A., Sasagawa, S., & Frick, P. J. (2006). Callous-unemotional traits in

by

a community sample of adolescents. Assessment, 13, 454–469. doi:10.1177/

1073191106287354
11

Frick, P. J., Cornell, A. H., Barry, C. T., Bodin, S. D., & Dane, H. E. (2003).
20

Callous-unemotional traits and conduct problems in the prediction of conduct

problem severity, aggression, and self-report of delinquency. Journal of Abnormal
©

Child Psychology, 31, 457–470. doi:10.1023/A:1023899703866

ht

Frick, P. J., & Sheffield-Morris, A. (2004). Temperament and developmental path-

ig

ways to conduct problems. Journal of Clinical Child and Adolescent Psychology, 33,
yr

54–68. doi:10.1207/S15374424JCCP3301_6
op

Funder, D. C. (2008). Persons, situations, and person–situation interactions. In

O. P. John, R. W. Robins, & L. A. Pervin (Eds.), Handbook of personality: Theory
C

and research (3rd ed., pp. 568–580). New York, NY: Guilford Press.

100 TACKETT AND KRUEGER

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 101

Goldberg, L. R. (1993). The structure of phenotypic personality traits. American Psy-

chologist, 48, 26–34. doi:10.1037/0003-066X.48.1.26
Halverson, C. F., Havill, V. L., Deal, J., Baker, S. R., Victor, J. B., Pavlopoulos, V.,
. . . Wen, L. (2003). Personality structure as derived from parental ratings of free

A)
descriptions of children: The Inventory of Child Individual Differences. Journal
of Personality, 71, 995–1026. doi:10.1111/1467-6494.7106005

P
(A
Harmon-Jones, E. (2007). Trait anger predicts relative left frontal cortical activation
to anger-inducing stimuli. International Journal of Psychophysiology, 66, 154–160.

n
doi:10.1016/j.ijpsycho.2007.03.020

io
Hay, D. F. (2007). The gradual emergence of sex differences in aggression:

at
Alternative hypotheses. Psychological Medicine, 37, 1527–1537. doi:10.1017/

ci
S0033291707000165

so
Helzer, J. E., Kraemer, H. C., & Krueger, R. F. (2006). The feasibility and need for

As
dimensional psychiatric diagnoses. Psychological Medicine, 36, 1671–1680.
doi:10.1017/S003329170600821X

al
ic
Herpertz, S. C., Huebner, T., Marx, I., Vloet, T. D., Fink, G. R., Stoecker, T., . . .

og
Herpertz-Dahlmann, B. (2008). Emotional processing in male adolescents with
childhood-onset conduct disorder. Journal of Child Psychology and Psychiatry, and
ol
Allied Disciplines, 49, 781–791. doi:10.1111/j.1469-7610.2008.01905.x
ch
Hicks, B. M., Bernat, E., Malone, S. M., Iacono, W. G., Patrick, C. J., Krueger, R. F.,
y

& McGue, M. (2007). Genes mediate the association between P3 amplitude

Ps

and externalizing disorders. Psychophysiology, 44, 98–105. doi:10.1111/j.1469-

8986.2006.00471.x
an

Iacono, W. G., Malone, S. M., & McGue, M. (2003). Substance use disorders, exter-
ic

nalizing psychopathology, and event-related potential amplitude. International

er

Journal of Psychophysiology, 48, 147–178. doi:10.1016/S0167-8760(03)00052-7

Am

Jenkins, J. (2008). Psychosocial adversity and resilience. In M. Rutter, D. Bishop,

D. Pine, S. Scott, J. Stevenson, E. Taylor, & A. Thapar (Eds.), Rutter’s handbook
e

of child and adolescent psychiatry (5th ed., pp. 377–391). Oxford, England: Black-
th

well. doi:10.1002/9781444300895.ch25
by

Jensen-Campbell, L. A., Knack, J. M., Waldrip, A. M., & Campbell, S. D. (2007).

Do Big Five personality traits associated with self-control influence the regula-
11

tion of anger and aggression? Journal of Research in Personality, 41, 403–424.

20

doi:10.1016/j.jrp.2006.05.001
Kaminsky, Z., Petronis, A., Wang, S. C., Levine, B., Ghaffar, O., Floden, D., &
©

Feinstein, A. (2008). Epigenetics of personality traits: An illustrative study of

ht

identical twins discordant for risk-taking behavior. Twin Research and Human
ig

Genetics, 11, 1–11. doi:10.1375/twin.11.1.1

yr

Kendler, K. S., Prescott, C. A., Myers, J., & Neale, M. C. (2003). The structure of
op

genetic and environmental risk factors for common psychiatric and substance
use disorders in men and women. Archives of General Psychiatry, 60, 929–937.
C

doi:10.1001/archpsyc.60.9.929

DISPOSITIONAL INFLUENCES ON HUMAN AGGRESSION 101

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 102

Kim-Cohen, J., Caspi, A., Taylor, A., Williams, B., Newcombe, R., Craig, I. W., &
Moffitt, T. E. (2006). MAO-A, maltreatment, and gene–environment inter-
action predicting children’s mental health: New evidence and a meta-analysis.
Molecular Psychiatry, 11, 903–913. doi:10.1038/sj.mp.4001851

A)
Krueger, R. F., Hicks, B. M., Patrick, C. J., Carlson, S. R., Iacono, W. G., & McGue,
M. (2002). Etiologic connections among substance dependence, antisocial

P
behavior, and personality: Modeling the externalizing spectrum. Journal of

(A
Abnormal Psychology, 111, 411–424. doi:10.1037/0021-843X.111.3.411

n
Krueger, R. F., & Markon, K. E. (2006). Reinterpreting comorbidity: A model-based

io
approach to understanding and classifying psychopathology. Annual Review of

at
Clinical Psychology, 2, 111–133. doi:10.1146/annurev.clinpsy.2.022305.095213

ci
Krueger, R. F., Markon, K. E., Patrick, C. J., Benning, S. D., & Kramer, M. (2007).

so
Linking antisocial behavior, substance use, and personality: An integrative

As
quantitative model of the adult externalizing spectrum. Journal of Abnormal
Psychology, 116, 645–666. doi:10.1037/0021-843X.116.4.645

al
Lahey, B. B., & Waldman, I. D. (2003). A developmental propensity model of the

ic
origins of conduct problems during childhood and adolescence. In B. B. Lahey,

og
T. E. Moffitt, & A. Caspi (Eds.), Causes of conduct disorder and juvenile delin-
ol
quency (pp. 76–117). New York, NY: Guilford Press.
ch
Legrand, L. N., Keyes, M., McGue, M., Iacono, W. G., & Krueger, R. F. (2008).
y

Rural residency reduces the genetic influence on adolescent substance-use and

Ps

rule-breaking behavior. Psychological Medicine, 38, 1341–1350. doi:10.1017/

S0033291707001596
an

Markon, K. E., Krueger, R. F., & Watson, D. (2005). Delineating the structure of nor-
ic

mal and abnormal personality: An integrative hierarchical approach. Journal of

er

Personality and Social Psychology, 88, 139–157. doi:10.1037/0022-3514.88.1.139

Am

McCrae, R. R., & Costa, P. T. (1990). Personality in adulthood. New York, NY: Guil-
ford Press.
e

McGirr, A., Renaud, J., Bureau, A., Seguin, M., Lesage, A., & Turecki, G. (2008).
th

Impulsive-aggressive behaviours and completed suicide across the life cycle: A

by

predisposition for younger age of suicide. Psychological Medicine, 38, 407–417.

doi:10.1017/S0033291707001419
11

Mill, J., & Petronis, A. (2007). Molecular studies of major depressive disorder:
20

The epigenetic perspective. Molecular Psychiatry, 12, 799–814. doi:10.1038/

sj.mp.4001992
©

Moffitt, T. E. (2003). Life-course-persistent and adolescence-limited antisocial

ht

behavior: A 10-year research review and a research agenda. In B. B. Lahey,

ig

T. E. Moffitt, & A. Caspi (Eds.), Causes of conduct disorder and juvenile delin-
yr

quency (pp. 49–75). New York, NY: Guilford Press.

op

Nigg, J. T. (2006). Temperament and developmental psychopathology. Journal of

Child Psychology and Psychiatry, and Allied Disciplines, 47, 395–422. doi:10.1111/
C

j.1469-7610.2006.01612.x

102 TACKETT AND KRUEGER

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 103

Paunonen, S. V., & Ashton, M. C. (2001). Big Five factors and facets and the pre-
diction of behavior. Journal of Personality and Social Psychology, 81, 524–539.
doi:10.1037/0022-3514.81.3.524
Reynolds, S. K., & Clark, L. A. (2001). Predicting dimensions of personality disorder

A)
from domains and facets of the five-factor model. Journal of Personality, 69,
199–222. doi:10.1111/1467-6494.00142

P
(A
Rothbart, M. K., Ahadi, S. A., Hershey, K. L., & Fisher, P. (2001). Investigations of
temperament at 3 to 7 years: The Children’s Behavior Questionnaire. Child

n
Development, 72, 1394–1408. doi:10.1111/1467-8624.00355

io
Saudino, K. J., Plomin, R., & DeFries, J. C. (1996). Tester-rated temperament at 14,

at
20, and 24 months: Environmental change and genetic continuity. The British

ci
Journal of Developmental Psychology, 14, 129–144.

so
Schell, G. C., & Tackett, J. L. (2009). Agency and communion as indicators of person-

As
ality in middle childhood. Manuscript submitted for publication.
Shiner, R. L. (1998). How shall we speak of children’s personalities in middle child-

al
hood? A preliminary taxonomy. Psychological Bulletin, 124, 308–332. doi:10.1037/

ic
0033-2909.124.3.308

og
Shiner, R., & Caspi, A. (2003). Personality differences in childhood and adolescence:
ol
Measurement, development, and consequences. Journal of Child Psychology and
ch
Psychiatry, and Allied Disciplines, 44, 2–32. doi:10.1111/1469-7610.00101
y

Srivastava, S., John, O. P., Gosling, S. D., & Potter, J. (2003). Development of per-
Ps

sonality in early and middle adulthood: Set like plaster or persistent change?
Journal of Personality and Social Psychology, 84, 1041–1053. doi:10.1037/0022-
an

3514.84.5.1041
ic

Stallings, M. C., Corley, R. P., Dennehey, B., Hewitt, J. K., Krauter, K. S., Lessem,
er

J. M., . . . Crowley, T. J. (2005). A genome-wide search for quantitative trait

Am

loci that influence antisocial substance dependence in adolescence. Archives of

General Psychiatry, 62, 1042–1051. doi:10.1001/archpsyc.62.9.1042
e

Tackett, J. L. (2006). Evaluating models of the personality–psychopathology rela-

th

tionship in children and adolescents. Clinical Psychology Review, 26, 584–599.

by

doi:10.1016/j.cpr.2006.04.003
Tackett, J. L. (2010). Toward a developmentally integrated externalizing spectrum in
11

DSM–V. Manuscript submitted for publication.

20

Tackett, J. L., Krueger, R. F., Iacono, W. G., & McGue, M. (2005). Symptom-based
subfactors of DSM-defined conduct disorder: Evidence for etiologic distinctions.
©

Journal of Abnormal Psychology, 114, 483–487. doi:10.1037/0021-843X.114.3.483

ht

Tackett, J. L., Krueger, R. F., Iacono, W. G., & McGue, M. (2008). Personality in
ig

middle childhood: A hierarchical structure and longitudinal connections with

yr

personality in late adolescence. Journal of Research in Personality, 42, 1456–1462.

doi:10.1016/j.jrp.2008.06.005
op

Tackett, J. L., Krueger, R. F., Sawyer, M. G., & Graetz, B. W. (2003). Subfactors of
C

DSM–IV conduct disorder: Evidence and connections with syndromes from the

DISPOSITIONAL INFLUENCES ON HUMAN AGGRESSION 103

12051-06_CH05-rev2.qxd 5/25/10 11:30 AM Page 104

Child Behavior Checklist. Journal of Abnormal Child Psychology, 31, 647–654.

doi:10.1023/A:1026214324287
Tackett, J. L., Waldman, I., & Lahey, B. B. (2009). Etiology and measurement of rela-
tional aggression: A multi-informant behavior genetic investigation. Journal of

A)
Abnormal Psychology, 118, 722–733.
Young, S. E., Stallings, M. C., Corley, R. P., Krauter, K. S., & Hewitt, J. K. (2000).

P
(A
Genetic and environmental influences on behavioral disinhibition. American
Journal of Medical Genetics. Part B, Neuropsychiatric Genetics, 96, 684–695.

n
doi:10.1002/1096-8628(20001009)96:5<684::AID-AJMG16>3.0.CO;2-G

io
Yung, Y., Thissen, D., & McLeod, L. D. (1999). On the relationship between the

at
higher order factor model and the hierarchical factor model. Psychometrika, 64,

ci
113–128. doi:10.1007/BF02294531

so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

104 TACKETT AND KRUEGER

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 105

A)
P
A SOCIAL NEUROSCIENCE

(A
PERSPECTIVE ON THE

n
io
NEUROBIOLOGICAL BASES

at
ci
OF AGGRESSION

so
As
THOMAS F. DENSON

al
ic
og
ol
y ch
Ps
an

The discovery of brain regions and mental processes that contribute to

ic

aggressive behavior has long been a significant concern in psychology.

er

Although much progress has been made, identification of the underlying

Am

neural mechanisms remains elusive (Davidson, Putnam, & Larson, 2000). It

is thought that neural functioning mediates aggressive behavior by biasing
e

mental processes toward aggressive responses to social situations (Raine,

th

2008). Relying on advances in the emerging field of social neuroscience, this

by

chapter reviews research that has increased our understanding of the neural
mechanisms associated with human aggression. This review contains a brief
11

discussion of relevant brain anatomy, followed by a review of structural and

20

functional brain abnormalities in highly aggressive populations. Next, I dis-

cuss the application of a recent social neuroscience dual-process model, the
©

X- and C-systems model, to the study of anger and aggression. This is followed
ht

by a discussion of social neuroscience research that is consistent with social

ig

psychological theory. Finally, I discuss the role played by top-down executive

yr

control in determining aggressive behavior.

op
C

This chapter was supported by the Australian Research Council’s Discovery Projects funding scheme
(DP0985182). Thank you to Ajay Satpute for comments on an earlier draft of this chapter.

105
12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 106

ANATOMY

Reviews of anger and aggression converge on the importance of the pre-

frontal cortex (PFC) underlying these phenomena (Davidson et al., 2000; Raine,

A)
2008; Siever, 2008). Indeed, the PFC is broadly involved in the regulation and
control of affect and behavior. Within the PFC, four regions are particularly rel-

P
(A
evant to aggressive behavior. The dorsolateral PFC is involved in planning and
behavioral control. The ventral PFC, which encompasses the ventromedial

n
PFC, ventrolateral PFC, and the orbitofrontal cortex, is involved in emotion reg-

io
ulation. The dorsal region of medial PFC (mPFC) is involved in introspection

at
and the awareness of emotion as well as emotion regulation. Finally, the dorsal

ci
anterior cingulate cortex (dACC) is involved in the detection of conflict and

so
triggering activity in top-down control regions (see Figure 6.1).

As
In terms of neurotransmitter involvement, converging evidence sug-
gests that serotonin has a prominent role in facilitating and inhibiting anger

al
and hostile aggression via 5-HT2 receptors in the PFC (Davidson et al., 2000;

ic
Siever, 2008; see also Chapter 8, this volume). Because a detailed discussion

og
of neurotransmitter modulation is beyond the scope of this chapter, the pur-
ol
pose here is to review the most current research on brain regions implicated
ch
in anger and aggression.
y
Ps

Dorsolateral PFC
an

Dorsal anterior
ic

Medial PFC cingulate cortex

er
Am
e
th
by
11
20
©
ht
ig

Ventromedial PFC
yr
op

Orbitofrontal cortex
C

Figure 6.1. Brain regions implicated in anger and aggression. PFC = prefrontal cortex.

106 THOMAS F. DENSON

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 107

EVIDENCE FROM ABNORMAL POPULATIONS

Early examinations of patients with brain lesions revealed that abnormal-

ities in the PFC were associated with aggressive and antisocial behavior. The

A)
classic case of Phineas Gage, who suffered trauma to his orbitofrontal cortex, is
illustrative of the dramatic within-person change from agreeable and conscien-

P
(A
tious to hostile and antagonistic that is associated with trauma to this region. A
fairly large study of 279 Vietnam War veterans with brain lesions revealed that

n
those who suffered injury to the orbitofrontal cortex and mPFC were more irri-

io
table, hostile, and aggressive than control participants (Grafman et al., 1996).

at
More recent neuroimaging studies have examined structural differences

ci
in groups of individuals known to be highly aggressive compared with less

so
aggressive matched controls. Several studies have identified prefrontal struc-

As
tural differences in the brains of aggressive individuals, such as violent offend-
ers and psychopaths, relative to matched controls (for reviews, see Raine, 2008;

al
Yang, Glenn, & Raine, 2008). For instance, individuals with antisocial person-

ic
ality disorder have an 11% to 14% deficit in prefrontal gray matter relative to

og
normal controls, substance-dependent individuals, and individuals with other
ol
psychiatric disorders (Raine, Lencz, Bihrle, LaCasse, & Colletti, 2000).
ch
These differences are not limited to brain structure. Numerous studies
y

document functional impairment in frontal and limbic regions such as the

Ps

ventral PFC, dorsolateral PFC, mPFC, anterior cingulate cortex (ACC), pos-
terior cingulate cortex (PCC), hippocampus, and amygdala (Raine, 2008;
an

Raine & Yang, 2008). For instance, a positron emission tomography (PET)
ic

study reported reduced glucose metabolism in the PFC in a group of murder-

er

ers compared with a matched control group (Raine, Buchsbaum, & LaCasse,
Am

1997). Another study found that when processing emotional versus neutral
words, criminal psychopaths show less activation in the lateral PFC, ACC,
e

PCC, and amygdala than noncriminal psychopaths and normal controls

th

(Kiehl et al., 2001). In summary, violence-prone groups show deficits in the

by

structure and functioning of key brain regions involved in emotion regulation

and behavioral control. Although these data are correlational, presumably
11

these abnormalities reduce the ability to regulate and control angry feelings and
20

aggressive behavior.
©
ht

A SOCIAL NEUROSCIENCE FRAMEWORK:

ig

THE X- AND C-SYSTEMS

yr
op

Although research on abnormally aggressive individuals is instructive,

most social and personality psychologists are primarily interested in the normal
C

spectrum of human behavior. Indeed, decades of social psychological and

A SOCIAL NEUROSCIENCE PERSPECTIVE 107

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 108

personality research demonstrate that we are all capable of behaving aggres-

sively under certain circumstances and that some individuals within the nor-
mal range tend to be more aggressive than others across a variety of contexts
(e.g., Anderson & Bushman, 2002; see also Chapters 1, 2, and 7, this volume).

A)
Social neuroscience is an emerging field that examines the neural correlates of
social psychological phenomena. Because functional neuroimaging technology,

P
(A
especially functional magnetic resonance imaging (fMRI), is relatively widely
available, we are now able to examine neural processes in normal individuals

n
using experimental methods drawn from social and cognitive psychology. Tools

io
such as fMRI enable us to explore previously unquantifiable aspects of human

at
functioning that are directly relevant to social psychological questions.

ci
In an effort to integrate findings from the rapidly expanding field of

so
social neuroscience, Lieberman and colleagues (Lieberman, 2007; Lieberman,

As
Gaunt, Gilbert, & Trope, 2002; Satpute & Lieberman, 2006) proposed a
dual-process framework known as the X- and C-systems model. The X com-

al
ponent of the model, which we likely share with other animals, represents our

ic
continuous stream of current experience (Lieberman et al., 2002, p. 204). The

og
X stands for the x in reflexive. The X-system is phylogenetically older than
ol
the C-system, operates quickly, supports spontaneous processes, and engages
ch
in parallel processing. Brain regions that make up the X-system are the ventro-
y

medial PFC, dACC, amygdala, basal ganglia, and lateral temporal cortex.
Ps

Social psychological phenomena associated with this region include implicit

prejudice, emotional pain resulting from social rejection, and intuition-based
an

self-knowledge (see Lieberman, 2007).

ic

In contrast with the X-system, the C-system underlies reflective and

er

control processes. The C stands for the c in reflective. The C-system is phylo-
Am

genetically newer than the X-system, operates slowly, is associated with voli-
tional processes, uses serial processing (i.e., not parallel processing), and
e

involves abstract thought (Lieberman, 2007). We experience activity in the

th

C-system during reflective thought. Thus, current experience is the result

by

of the X-system, whereas reflecting on this experience is supported by the

C-system. The C-system often becomes active when the X-system detects a
11

problem. Brain regions that make up the C-system include the lateral and
20

mPFC, lateral and medial parietal cortex, medial temporal lobe, and ventro-
lateral PFC. Social psychological phenomena associated with the C-system
©

include explicit attitudes, reflecting on current experiences, impulse control,

ht

reappraising emotional events, and moral reasoning (see Lane, Fink, Chau,
ig

& Dolan, 1997; Lieberman, 2007; MacDonald, 2008). Although research on

yr

anger and aggression was not considered in Lieberman’s (2007) review, the
op

X- and C-systems provide a valuable framework for discussing the neural bases
of anger and aggression. Specifically, many of the processes associated with
C

the X- and C-systems are relevant for understanding aggressive behavior.

108 THOMAS F. DENSON

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 109

SOCIAL NEUROSCIENCE RESEARCH

ON ANGER AND AGGRESSION

In this section, I review research conducted primarily with normal indi-

A)
viduals. Specifically, I review social neuroscience research relevant to anger,
angry rumination, cognitive neoassociation theory, social learning, media

P
influences, and aggressive behavior.

(A
n
Anger

io
at
Most functional imaging studies of anger in nonclinical samples have

ci
exposed participants to angry faces or asked them to recall and reexperience

so
anger-inducing events. Two recent meta-analyses of nine PET and f MRI

As
studies revealed that some of the most prominent areas of frontal and limbic
brain reactivity were the mPFC, ventromedial PFC, ACC, PCC, lateral PFC,

al
and thalamus (Murphy, Nimmo-Smith, & Lawrence, 2003; Phan, Wager,

ic
Taylor, & Liberzon, 2002). It is interesting that whereas the amygdala has a

og
prominent role in fear, it was not implicated in these studies of angry faces
and memories (Phan et al., 2002). ol
ch
My colleagues and I recently examined the neural correlates of anger more
directly by exposing participants to an interpersonal insult (Denson, Pedersen,
y
Ps

Ronquillo, & Nandy, 2009). Anderson and Bushman (2002) described such a
provocation as “perhaps the most important single cause of human aggression”
an

(p. 37). In this fMRI study, participants were asked to complete difficult ana-
grams and state the answer aloud or say “no answer” if they did not know the
ic

answer. Following two polite prompts to speak louder, the experimenter then
er

insulted participants by saying in an irritated and condescending voice, “Look,

Am

this is the third time I’ve had to say this! Can’t you follow directions?” We found
that, relative to baseline, participants showed increased activation in many
e

of the same regions active during exposure to angry faces and autobiographical
th

recall of anger experiences. Moreover, we found that a component of the

by

X-system, the dACC, played a special role in the subjective experience of

anger. Specifically, self-reported anger was correlated with dACC activation
11

(r = .56). Activity in the dACC was also correlated with scores on the Buss and
20

Perry (1992) Aggression Questionnaire, a measure of general trait anger, hos-

tility, and aggression (r = .61). Consistent with the claim that the dACC is
©

involved in the subjective experience of anger, individuals who have had por-
ht

tions of the ACC removed demonstrate decreased anger (Cohen et al., 2000).
ig
yr

Angry Rumination
op

Reflective processes also play a role in anger and aggression. Regions

C

of the C-system that have been implicated in anger and aggression include

A SOCIAL NEUROSCIENCE PERSPECTIVE 109

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 110

the lateral PFC, mPFC, and medial parietal cortex. As noted earlier,
following provocation, participants demonstrated increased activity
in regions of the C-system as well as the X-system. One important reflec-
tive process that can influence aggressive behavior is the way one

A)
regulates emotions. One particularly pernicious form of emotion regula-
tion is angry rumination. Immersive rumination on anger-inducing expe-

P
(A
riences increases anger, aggression, cardiovascular arousal, and cortisol
levels (Bushman, 2002; Bushman, Bonacci, Pedersen, Vasquez, & Miller,

n
2005; Denson, Fabiansson, Creswell, & Pedersen, 2009; Denson, Pedersen,

io
& Miller, 2006; Ray, Wilhelm, & Gross, 2008; Rusting & Nolen-Hoeksema,

at
1998).

ci
In addition to these negative consequences of angry rumination, our

so
recent research demonstrates that angry rumination also increases activity

As
in parts of the C-system (Denson, Pedersen, et al., 2009). In the second part
of the experiment described earlier, following the provocation, participants

al
were asked to engage in a “memory task,” which served as a rumination

ic
manipulation. Using a modified within-participants rumination task from

og
prior research (Bushman et al., 2005; Denson et al., 2006; Rusting &
ol
Nolen-Hoeksema, 1998), during the rumination task participants were
ch
asked to think about what had occurred in the experiment so far, who they
y

interacted with, and their current mood. During the distraction period, par-
Ps

ticipants were asked to think about neutral events. Relative to distraction,

rumination increased activity in regions of the C-system such as medial and
an

lateral PFC, insula, precuneus, and PCC. Furthermore, self-reported rumi-

ic

nation was correlated with activity in the mPFC (r = .42), as were scores
er

on the Displaced Aggression Questionnaire, an individual difference mea-

Am

sure of trait displaced aggression (Denson et al., 2006). The relationship

between the mPFC and the displaced aggression measure was likely due to
e

the fact that when provoked, individuals high in trait displaced aggression
th

tend to ruminate about the event rather than immediately retaliate against
by

the provocateur.
We expected that the mPFC would be especially relevant to angry rumi-
11

nation because it supports many of the reflective processes at work during

20

rumination. For example, the mPFC is activated during tasks that require the
self-awareness of emotions and self-relevant cognition (Lane et al., 1997;
©

Lieberman, 2007; Ochsner et al., 2004). This region is also active when mon-
ht

itoring one’s emotional state, reflecting on feelings, and reappraising emo-

ig

tional responses to distressing stimuli (Amodio & Frith, 2006; Ochsner et al.,
yr

2004; Ochsner, Bunge, Gross, & Gabrieli, 2002). Furthermore, Ray et al.
op

(2005) reported that when participants were asked to decrease their negative
affective responses to aversive photographs, a composite measure of trait
C

rumination was correlated with mPFC activity.

110 THOMAS F. DENSON

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 111

Cognitive Neoassociation Theory

As previously noted, my colleagues and I found that activity in the

dACC was linearly related to self-reported anger following provocation

A)
(Denson, Pedersen, et al., 2009). The dACC is of social psychological inter-
est because it is involved in at least two additional negative emotional states

P
(A
that have been shown to increase aggression. This is relevant because a core
tenet of Berkowitz’s (1993) cognitive neoassociationistic model of aggression

n
states that any form of negative affect can increase aggression. A large body

io
of evidence supports Berkowitz’s supposition. For example, physical pain and

at
social rejection both increase aggression and activation in the dACC

ci
(Berkowitz, Cochran, & Embree, 1981; Eisenberger, Liberman, & Williams,

so
2003; Rainville, Duncan, Price, Carrier, & Bushnell, 1997; Twenge, Baumeis-

As
ter, Tice, & Stucke, 2001). These and the anger findings converge to suggest
the existence of a common neural mechanism underlying the process Berkowitz

al
proposed, although the mediating role of dACC activation between anger,

ic
pain, social rejection, and actual aggressive behavior remains to be investi-

og
gated. Future neuroimaging research examining additional aversive stimuli
ol
known to increase aggression, such as noxious odors and heat, might demon-
ch
strate increased activity in the dACC as well.
y
Ps

Social Learning, Media Violence, and Script Theory

an

According to social learning perspectives on aggression, individuals

ic

learn how and when to behave aggressively by observing others either in per-
er

son or vicariously in the media (Bandura, 1973). A tremendous amount of

Am

research supports this notion (e.g., Anderson et al., 2003). Expanding on

social learning theory, Huesmann (1998) proposed that individuals learn
e

behavioral scripts from aggressive media exposure. Scripts are closely con-
th

nected concepts in memory that can become strengthened by rehearsal and

by

chronic exposure to violent media. When activated, such scripts can increase
the likelihood of aggressive behavior (see Chapter 7, this volume).
11

Only a handful of social neuroscience studies have examined neural

20

responses during exposure to violent media (for a review, see Carnagey,

Anderson, & Bartholow, 2007). In one study, in an attempt to identify chil-
©

dren most responsive to violent media, children who showed heart rate accel-
ht

eration to violent media on a pretest were exposed to violent (i.e., boxing)

ig

and nonviolent (i.e., animal) scenes while functional images were acquired
yr

(Murray et al., 2006). The strongest activity was located in the right PCC and
op

the right precuneus (in the medial parietal cortex). Because of its role in
memory, these authors speculated that activity in the PCC might correspond
C

to the activation of stored aggressive scripts. This is consistent with our

A SOCIAL NEUROSCIENCE PERSPECTIVE 111

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 112

research showing that the PCC was active during angry rumination, because
the revenge planning that occurs during angry rumination likely involves the
activation of aggressive scripts (Denson, Pedersen, et al., 2009). Indeed, fac-
tor analytic work indicates that angry rumination involves rehearsing acts

A)
of revenge (Caprara, 1986; Denson et al., 2006; Sukhodolsky, Golub, &
Cromwell, 2001).

P
(A
Two additional studies found that violent media influence brain activ-
ity in what is thought to be a maladaptive manner. Specifically, in an inves-

n
tigation of the hypothesis that chronic exposure to violent media desensitizes

io
individuals to aggressive content, one study examined the P300 component

at
of the event-related brain potential in people who played relatively high lev-

ci
els of aggressive video games and those who played aggressive video games less

so
often (Bartholow, Bushman, & Sestir, 2006). When exposed to violent

As
images in the laboratory, chronic exposure to violent video games was asso-
ciated with decreased P300 amplitude to violent images but not to nega-

al
tive or neutral images. These findings support the notion that violent video

ic
game play can desensitize basic neural responses to violent stimuli. Further-

og
more, the P300 deficit predicted increased aggression as assessed by choosing
ol
to deliver loud noise blasts to a fictitious participant. Although electro-
ch
encephalogram methods do not allow for precise localization of brain processes,
y

presumably the automatic response of the P300 reflects activity in the X-system,
Ps

suggesting that chronic exposure to media violence can alter even quite rudi-
mentary information processing such that individuals with high levels of
an

exposure actually experience violent media differently than those low in

ic

exposure.
er

An fMRI study investigated brain activity during actual violent video

Am

game play (Weber, Ritterfeld, & Mathiak, 2006). In a sophisticated frame-

by-frame analysis of violent game play, these authors demonstrated that activ-
e

ity in the dACC preceded suppression in the rostral ACC (rACC), which is
th

involved in affective information processing, and in the amygdala during

by

aggressive “search and destroy” sequences. Recall that the dACC is associ-
ated with the subjective experience of anger (Denson, Pedersen, et al., 2009).
11

This suppression of the rACC by the dACC when committing acts of video
20

game violence suggests that the dACC overrides affective input from the
rACC. Consistent with the suppression function of the dACC, when partic-
©

ipants were in danger, under attack, or using a weapon, the dACC was more
ht

active than when participants were passive or safe.

ig
yr

Aggressive Behavior
op

Only two neuroimaging studies have investigated brain activity during

C

actual acts of aggression. The first such fMRI study examined brain activity in

112 THOMAS F. DENSON

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 113

14 men high and low in psychopathy during their performance in a Taylor

(1967) aggression paradigm that was modified such that participants could see
a fictitious participant receive the allocated bursts of physically painful pres-
sure to their hand (Lotze, Veit, Anders, & Birbaumer, 2007). Of primary inter-

A)
est, activity in the mPFC was positively correlated with the intensity of pain
participants chose to have administered to the confederate. This latter finding

P
(A
is consistent with the role of the mPFC in angry rumination, emotion regula-
tion, attributions, and theory of mind (e.g., Amodio & Frith, 2006; Denson,

n
Pedersen, et al., 2009; Harris, Todorov, & Fiske, 2005; Ochsner et al., 2002).

io
A second study of normal young adults used a modified white-noise

at
Taylor (1967) paradigm in which the decision phase (i.e., deciding what noise

ci
level to choose) and the outcome phase (i.e., the aggressive act) were analyzed

so
separately (Krämer, Jansma, Tempelmann, & Münte, 2007). Furthermore,

As
participants were given the opportunity to aggress against highly provocative
and less provocative bogus participants as well as against a computer. This

al
allowed the authors to identify the neural mechanisms specifically associated

ic
with aggression rather than social interaction per se. Under high provocation,

og
the dACC and mPFC were active during the decision phase, suggesting the
ol
presence of anger and rumination (e.g., Denson, Pedersen, et al., 2009).
ch
Perhaps the most intriguing result of this study is that a component of
y

the reward system—the dorsal striatum—was activated during the decision

Ps

phase in which participants chose the level at which to blast the highly
provocative participant with noise. This finding, which the authors called the
an

“sweetness of revenge” (Krämer, Jansma, Tempelmann, & Münte, 2007, p. 209),

ic

suggests that aggression can be inherently rewarding, and thereby provides a

er

neural basis that might partially explain why it is difficult to reduce retalia-
Am

tory aggressive behavior. During the outcome phase, activation in another

region of the reward system, the ventral striatum, was also observed. How-
e

ever, the authors concluded that this was most likely due to relief derived
th

from the successful avoidance of the noise blast.

by

In summary, the brain regions associated with anger, angry rumination,

and actual aggressive behavior, as well as issues addressed by cognitive theo-
11

ries of aggression and media violence, involve elements of both the X- and
20

C-systems. The concept of a dual system harkens back to the days of Freud,
who posited an innate form of destruction motivation (i.e., Thanatos) ema-
©

nating from the id that resists control by the ego and superego. This notion
ht

of conflict between primitive aggressive urges and control of these urges

ig

remains with us today (see Chapter 2). However, the X- and C-systems frame-
yr

work differs markedly from Freud’s notion of destructive drives. The X- and
op

C- systems are compatible. When the situation calls for it, the C-system inter-
venes. For example, when someone cuts us off on the freeway while making
C

an obscene gesture, the dACC sounds the neural alarm and snaps us out of

A SOCIAL NEUROSCIENCE PERSPECTIVE 113

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 114

our placid stream of experience. Our subsequent behavior, aggressive or other-

wise, will depend on a number of factors, one of which is the activation in
top-down control regions of the brain. I discuss this in more detail later.

A)
ROLE OF TOP-DOWN CONTROL AND EMOTION

P
(A
REGULATION MECHANISMS

n
For the aggressor, aggressive behavior can have positive consequences

io
(e.g., self-defense, achievement of dominance) and negative consequences

at
(e.g., developing a bad reputation, instigating retaliation, and even being

ci
killed). Evolutionary theorists argue that in our ancestral past, aggression was

so
a risky strategy, but when successfully executed, aggression likely increased

As
reproductive success (e.g., MacDonald, 2008; see also Chapters 3 and 4, this
volume). Although he did not use the X- and C-system framework, in a review

al
of the literature on effortful control, MacDonald (2008) argued that a con-

ic
scious system located in the PFC (C-system) allows humans to inhibit prepo-

og
tent impulses toward aggression stemming primarily from limbic structures
ol
(X-system). When the X-system cannot solve a problem, it calls on the
ch
C-system. It is the C-system that allows individuals to make “explicit appraisals
y

of costs and benefits” (MacDonald, p. 1014) that are “only . . . available

Ps

through explicit processing” (MacDonald, p. 1015). These explicit appraisals

play a key role in determining whether aggression will or will not occur, and
an

it is precisely these appraisals that distinguish human aggression from the

ic

purely reflexive aggression observed in other animals.

er

The general aggression model (GAM; Anderson & Bushman, 2002; see
Am

also Chapter 1, this volume) highlights the importance of the explicit decision-
making process. According to the GAM, appraisals and decision-making
e

processes precede thoughtful or impulsive action. One key implication of the

th

GAM and other models of effortful control is that individual differences in

by

impairment of self-control should be related to aggression and impulsivity in

general. Indeed, individual differences in trait aggression, executive function-
11

ing, and impulsivity are interrelated (MacDonald, 2008). Moreover, tempo-

20

rary experimental impairment in self-control increases aggression (DeWall,

Baumeister, Stillman, & Gailliot, 2007; Stucke & Baumeister, 2006; see also
©

Chapter 2, this volume). Acute alcohol intoxication has similar effects, likely
ht

via altered activity in the dorsolateral PFC (Dao-Castellana et al., 1998).

ig

In the modern world, aggression is still risky. One might even argue that
yr

the negative consequences (e.g., imprisonment, legal fees, social rejection)

op

are typically a more likely outcome following aggression than any positive
consequences. Thus, the ability to effectively weigh the costs and benefits of
C

aggression is critical (see Chapter 3). This is not to say that the C-system flaw-

114 THOMAS F. DENSON

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 115

lessly functions in our best interest. Indeed, in modern society we still find
that most aggressive acts are impulsive acts. For example, the majority of
homicides and other aggressive crimes occur when people are provoked and
angry and either explicitly decide to aggress or are simply unable to resist the

A)
motivation to behave aggressively. Thus, the immediate cause of many acts
of aggression is often a loss of self-control (DeWall et al., 2007).

P
(A
Substantial neuropsychological and imaging research indicates that the
PFC is the seat of self-control (Banfield, Wyland, Macrae, Munte, & Heather-

n
ton, 2004), and integral parts of the circuit that underlie self-control are the

io
dACC, dorsolateral PFC, and mPFC (Cohen, Botvinick, & Carter, 2000).

at
The neural substrates of emotion regulation include the medial, ventrolateral

ci
PFC, and ventromedial PFC. The lateral and medial regions of the PFC share

so
rich connectivity with cortical and limbic structures such as the dACC and

As
ventromedial PFC and have been implicated in emotion regulation and
behavioral control (Inzlicht & Gutsell, 2007). Accordingly, Davidson et al.

al
(2000) proposed that impaired functioning of an emotion regulation circuit

ic
involving the dACC, ventromedial PFC, and the dorsolateral PFC predisposes

og
individuals to aggressive behavior. Indeed, engaging in self-control or com-
ol
pleting neuropsychological measures that rely on inhibitory ability recruits the
ch
dACC and dorsolateral PFC (Botvinick, Braver, Barch, Carter, & Cohen,
y

2001; Richeson et al., 2003). As noted previously, both provocation and sub-
Ps

sequent angry rumination recruit neural regions underlying executive control

and emotion regulation mechanisms (Denson, Pedersen, et al., 2009).
an

Part of the anger and aggression circuit, the dACC, has been dubbed a
ic

“neural alarm system” because of its role in detecting conflict (Eisenberger &
er

Lieberman, 2004; Kross, Egner, Ochsner, Hirsch, & Downey, 2007). In the
Am

presence of unjustified wrongdoing, there is likely a conflict between how

people feel they should be treated and how they were actually treated (see
e

Chapter 3, this volume). This is consistent with the associations of the dACC
th

with self-reported social distress and anger following ostracism and provoca-
by

tion (Denson, Pedersen, et al., 2009; Eisenberger et al., 2003). Because there
are costs and benefits to aggression, there is also likely to be a conflict between
11

motivation to aggress and motivation not to aggress. As Krämer et al. (2007)

20

suggested, when a person is unjustly wronged, it is likely that the dACC ini-
tiates regulatory behavior via activity in the dorsolateral PFC.
©

There is some evidence to suggest that exposure to media violence

ht

impairs top-down control mechanisms. Mathews et al. (2005) examined ado-

ig

lescents who had been diagnosed with a disruptive behavior disorder, includ-
yr

ing aggressive features, and matched controls who had either high or low levels
op

of exposure to media violence. During neuroimaging, participants completed

the Stroop task, which typically activates the dACC and dorsolateral PFC.
C

The aggressive group demonstrated decreased activity, and this activation was

A SOCIAL NEUROSCIENCE PERSPECTIVE 115

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 116

not different from that in the normal adolescents who had been exposed to
high levels of media violence. Only the normal adolescents with low levels of
exposure to media violence demonstrated the typical pattern of lateral PFC
and dACC activity during the Stroop task.

A)
P
(A
CONCLUSION AND FUTURE DIRECTIONS

n
The evidence reviewed here implicates a network of neural regions that

io
underlie anger, angry rumination, aggression, and media violence. Activity

at
in these regions and the processes that they support are consistent with social

ci
psychological models of aggression, such as cognitive neoassociation theory,

so
script theory, and the GAM. The hostile aggression circuit described here

As
implicates limbic and top-down prefrontal regions, which support both reflec-
tive and reflexive processes. Anger, pain, and social rejection are mediated by

al
activity in the dACC, whereas angry rumination is mediated by the mPFC.

ic
There is also evidence that high levels of exposure to media violence are asso-

og
ciated with abnormal functioning in the aggression circuit. Furthermore, dur-
ol
ing actual acts of aggression, neural regions involved in reward processing are
ch
active. This finding partially explains the difficulty associated with effectively
y

reducing aggression.
Ps

Much work remains to be done. For instance, because alcohol is involved

in a large number of aggressive acts, neuroimaging research investigating the
an

pathways by which alcohol influences reactions to provocation seems worth-

ic

while. Future research could also investigate genetic markers that might influ-
er

ence neural reactivity to provocation such as the MAOA polymorphism.

Am

Additional work could also investigate the connection between neural activity,
other systems (e.g., cardiovascular, endocrine), and actual aggressive behavior
e

in more detail. Furthermore, one might also examine the effects of inter-
th

ventions known to reduce aggression (e.g., distraction, self-control training)

by

on long-term changes in neural reactivity. By grounding social psychological

theory in brain processes, future research will expand our understanding of
11

situational and personological influences on aggressive behavior.

20
©

REFERENCES
ht
ig

Amodio, D. M., & Frith, C. D. (2006). Meeting of minds: The medial frontal cortex and
yr

social cognition. Nature Reviews. Neuroscience, 7, 268–277. doi:10.1038/nrn1884

op

Anderson, C. A., Berkowitz, L., Donnerstein, E., Huesmann, L. R., Johnson, J. D.,
Linz, D., . . . Wartella, E. (2003). The influence of media violence on youth.
C

Psychological Science in the Public Interest, 4, 81–110.

116 THOMAS F. DENSON

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 117

Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of

Psychology, 53, 27–51. doi:10.1146/annurev.psych.53.100901.135231
Bandura, A. (1973). A social learning analysis. Englewood Cliffs, NJ: Prentice-Hall.
Banfield, J. F., Wyland, C. L., Macrae, C. N., Munte, T. F., & Heatherton, T. F. (2004).

A)
The cognitive neuroscience of self-regulation. In R. F. Baumeister & K. D. Vohs
(Eds.), Handbook of self-regulation (pp. 62–83). New York, NY: Guilford Press.

P
(A
Bartholow, B. D., Bushman, B. J., & Sestir, M. A. (2006). Chronic violent video
game exposure and desensitization to violence: Behavioral and event-related

n
brain potential data. Journal of Experimental Social Psychology, 42, 532–539.

io
doi:10.1016/j.jesp.2005.08.006

at
Berkowitz, L. (1993). Aggression: Its causes, consequences, and control. New York, NY:

ci
McGraw-Hill.

so
Berkowitz, L., Cochran, S., & Embree, M. C. (1981). Physical pain and the goal of

As
aversively stimulated aggression. Journal of Personality and Social Psychology, 40,
687–700. doi:10.1037/0022-3514.40.4.687

al
ic
Botvinick, M. M., Braver, T. S., Barch, D. M., Carter, C. S., & Cohen, J. D. (2001).

og
Conflict monitoring and cognitive control. Psychological Review, 108, 624–652.
doi:10.1037/0033-295X.108.3.624
ol
Bushman, B. J. (2002). Does venting anger feed or extinguish the flame? Catharsis,
ch
rumination, distraction, anger, and aggressive responding. Personality and Social
y

Psychology Bulletin, 28, 724–731. doi:10.1177/0146167202289002

Ps

Bushman, B. J., Bonacci, A. M., Pedersen, W. C., Vasquez, E. A., & Miller, N.
(2005). Chewing on it can chew you up: Effects of rumination on triggered dis-
an

placed aggression. Journal of Personality and Social Psychology, 88, 969–983.

ic

doi:10.1037/0022-3514.88.6.969
er

Buss, A. H., & Perry, M. (1992). The Aggression Questionnaire. Journal of Personality
Am

and Social Psychology, 63, 452–459. doi:10.1037/0022-3514.63.3.452

Caprara, G. V. (1986). Indicators of aggression: The Dissipation–Rumination Scale.
e

Personality and Individual Differences, 7, 763–769. doi:10.1016/0191-8869(86)

th

90074-7
by

Carnagey, N. L., Anderson, C. A., & Bartholow, B. D. (2007). Media violence and
social neuroscience: New questions and new opportunities. Current Directions in
11

Psychological Science, 16, 178–182. doi:10.1111/j.1467-8721.2007.00499.x

20

Cohen, J. D., Botvinick, M., & Carter, C. S. (2000). Anterior cingulate and prefrontal
cortex: Who’s in control? Nature Neuroscience, 3, 421–423. doi:10.1038/74783
©

Cohen, R. A., Paul, R., Zawacki, T. M., Moser, D. J., Sweet, L., & Wilkinson, H.
ht

(2001). Emotional and personality changes following cingulotomy. Emotion, 1,

ig

38–50. doi:10.1037/1528-3542.1.1.38
yr

Dao-Castellana, M. H., Samson, Y., Legault, F., Martinot, J. L., Aubin, H. J., Crouzel,
op

C., . . . Syrota, A. (1998). Frontal dysfunction in neurologically normal chronic

alcoholic subjects: Metabolic and neuropsychological findings. Psychological
C

Medicine, 28, 1039–1048. doi:10.1017/S0033291798006849

A SOCIAL NEUROSCIENCE PERSPECTIVE 117

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 118

Davidson, R. J., Putnam, K. M., & Larson, C. L. (2000, July). Dysfunction in the
neural circuitry of emotion regulation: A possible prelude to violence. Science,
289, 591–594. doi:10.1126/science.289.5479.591
Denson, T. F., Fabiansson, E. C., Creswell, J. D., & Pedersen, W. C. (2009). Experimen-
tal effects of rumination styles on salivary cortisol responses. Motivation and

A)
Emotion, 33, 42–48. doi:10.1007/s11031-008-9114-0

P
Denson, T. F., Pedersen, W. C., & Miller, N. (2006). The Displaced Aggression Ques-

(A
tionnaire. Journal of Personality and Social Psychology, 90, 1032–1051. doi:10.1037/

n
0022-3514.90.6.1032

io
Denson, T. F., Pedersen, W. C., Ronquillo, J., & Nandy, A. S. (2009). The angry

at
brain: Neural correlates of anger, angry rumination, and aggressive personality.

ci
Journal of Cognitive Neuroscience, 21, 734–744. doi:10.1162/jocn.2009.21051

so
DeWall, C. N., Baumeister, R. F., Stillman, T. F., & Gailliot, M. T. (2007). Violence

As
restrained: Effects of self-regulation and its depletion on aggression. Journal of
Experimental Social Psychology, 43, 62–76. doi:10.1016/j.jesp.2005.12.005

al
Eisenberger, N. I., & Lieberman, M. D. (2004). Why rejection hurts: A common

ic
neural alarm system for physical and social pain. Trends in Cognitive Sciences, 8,

og
294–300. doi:10.1016/j.tics.2004.05.010
ol
Eisenberger, N. I., Liberman, M. D., & Williams, K. D. (2003, October). Does rejec-
ch
tion hurt? An fMRI study of social exclusion. Science, 302, 290–292. doi:10.1126/
science.1089134
y
Ps

Grafman, J., Schwab, K., Warden, D., Pridgen, A., Brown, H. R., & Salazar, A. M.
(1996). Frontal lobe injuries, violence, and aggression: A report of the Vietnam
an

Head Injury Study. Neurology, 46, 1231–1238.

ic

Harris, L. T., Todorov, A., & Fiske, S. T. (2005). Attributions on the brain: Neuro-
er

imaging dispositional inferences, beyond theory of mind. NeuroImage, 28,

763–769. doi:10.1016/j.neuroimage.2005.05.021
Am

Huesmann, L. R. (1998). The role of social information processing and cognitive

schema in the acquisition and maintenance of habitual aggressive behavior. In
e

R. G. Geen & E. Donnerstein (Eds.), Human aggression: Theories, research, and

th

implications for social policy (pp. 73–109). San Diego, CA: Academic Press.
by

Inzlicht, M., & Gutsell, J. N. (2007). Running on empty: Neural signals for self-control
failure. Psychological Science, 18, 933–937. doi:10.1111/j.1467-9280.2007.02004.x
11

Kiehl, K. A., Smith, A. M., Hare, R. D., Mendrek, A., Forster, B. B., Brink, J., &
20

Liddle, P. F. (2001). Limbic abnormalities in affective processing by criminal

psychopaths as revealed by functional magnetic resonance imaging. Biological
©

Psychiatry, 50, 677–684. doi:10.1016/S0006-3223(01)01222-7

ht

Krämer, U. M., Jansma, H., Tempelmann, C., & Münte, T. F. (2007). Tit-for-tat:
ig

The neural basis of reactive aggression. NeuroImage, 38, 203–211. doi:10.1016/

yr

j.neuroimage.2007.07.029
op

Kross, E., Egner, T., Ochsner, K., Hirsch, J., & Downey, G. (2007). Neural dynamics of
rejection sensitivity. Journal of Cognitive Neuroscience, 19, 945–956. doi:10.1162/
C

jocn.2007.19.6.945

118 THOMAS F. DENSON

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 119

Lane, R. D., Fink, G. R., Chau, P. M.-L., & Dolan, R. J. (1997). Neural activation
during selective attention to subjective emotional responses. Neuroreport, 8,
3969–3972. doi:10.1097/00001756-199712220-00024
Lieberman, M. D. (2007). Social cognitive neuroscience: A review of core processes.
Annual Review of Psychology, 58, 259–289. doi:10.1146/annurev.psych.58.

A)
110405.085654

P
Lieberman, M. D., Gaunt, R., Gilbert, D. T., & Trope, Y. (2002). Reflexion and

(A
reflection: A social cognitive neuroscience approach to attributional inference.

n
Advances in Experimental Social Psychology, 34, 199–249. doi:10.1016/S0065-

io
2601(02)80006-5

at
Lotze, M., Veit, R., Anders, S., & Birbaumer, N. (2007). Evidence for a different role

ci
of the ventral and dorsal medial PFC for social reactive aggression: An interactive

so
fMRI study. NeuroImage, 34, 470–478. doi:10.1016/j.neuroimage.2006.09.028

As
MacDonald, K. B. (2008). Effortful control, explicit processing, and the regulation
of human evolved predispositions. Psychological Review, 115, 1012–1031.

al
doi:10.1037/a0013327

ic
Mathews, V. P., Kronenberger, W. G., Want, Y., Lurito, J. T., Lowe, M. J., & Dunn,

og
D. W. (2005). Media violence exposure and frontal lobe activation measure by
ol
functional magnetic resonance imaging in aggressive and nonaggressive adoles-
ch
cents. Journal of Computer Assisted Tomography, 29, 287–292. doi:10.1097/
01.rct.0000162822.46958.33
y
Ps

Murphy, F. C., Nimmo-Smith, I., & Lawrence, A. D. (2003). Functional neuroanatomy

of emotions: A meta-analysis. Cognitive, Affective, & Behavioral Neuroscience, 3,
an

207–233. doi:10.3758/CABN.3.3.207
Murray, J. P., Liotti, M., Ingmundson, P. T., Mayberg, H. S., Pu, Y., Zamarripa,
ic

F., . . . Fox, P. T. (2006). Children’s brain activations while viewing tele-

er

vised violence revealed by fMRI. Media Psychology, 8, 25–37. doi:10.1207/

Am

S1532785XMEP0801_3
Ochsner, K. N., Bunge, S. A., Gross, J. J., & Gabrieli, J. D. E. (2002). Rethinking
e

feelings: An fMRI study of the cognitive regulation of emotion. Journal of

th

Cognitive Neuroscience, 14, 1215–1229. doi:10.1162/089892902760807212

by

Ochsner, K. N., Knierim, K., Ludlow, D. H., Hanelin, J., Ramachandran, T., Glover,
G., & Mackey, S. C. (2004). Reflection upon feelings: An fMRI study of neural
11

systems supporting the attribution of emotion to self and other. Journal of

20

Cognitive Neuroscience, 16, 1746–1772. doi:10.1162/0898929042947829

Phan, K. L., Wager, T., Taylor, S. F., & Liberzon, I. (2002). Functional neuro-
©

anatomy of emotion: A meta-analysis of emotion activation studies in PET and

ht

fMRI. NeuroImage, 16, 331–348. doi:10.1006/nimg.2002.1087

ig

Raine, A. (2008). From genes to brain to antisocial behaviour. Current Directions in

yr

Psychological Science, 17, 323–328. doi:10.1111/j.1467-8721.2008.00599.x

op

Raine, A., Buchsbaum, M., & LaCasse, L. (1997). Brain abnormalities in murderers
indicated by positron emission tomography. Biological Psychiatry, 42, 495–508.
C

doi:10.1016/S0006-3223(96)00362-9

A SOCIAL NEUROSCIENCE PERSPECTIVE 119

12051-07_CH06-rev2.qxd 5/25/10 11:34 AM Page 120

Raine, A., Lencz, T., Bihrle, S., LaCasse, L., & Colletti, P. (2000). Reduced prefrontal
gray matter volume and reduced autonomic activity in antisocial personality dis-
order. Archives of General Psychiatry, 57, 119–127. doi:10.1001/archpsyc.57.2.119
Raine, A., & Yang, Y. (2008). Neural foundations to moral reasoning and antisocial

A)
behaviour. Social Cognitive and Affective Neuroscience, 1, 203–213.
Rainville, P., Duncan, G. H., Price, D. D., Carrier, B., & Bushnell, M. C. (1997,

P
(A
August). Pain affect encoded in human anterior cingulate but not somato-
sensory cortex. Science, 277, 968–971. doi:10.1126/science.277.5328.968

n
Ray, R. D., Ochsner, K. N., Cooper, J. C., Roberston, E. R., Gabrieli, J. D. E., &

io
Gross, J. J. (2005). Individual differences in trait rumination and the neural sys-

at
tems supporting cognitive reappraisal. Cognitive, Affective, & Behavioral Neuro-

ci
science, 5, 156–168. doi:10.3758/CABN.5.2.156

so
Ray, R. D., Wilhelm, F. H., & Gross, J. J. (2008). All in the mind’s eye? Anger rumi-

As
nation and reappraisal. Journal of Personality and Social Psychology, 94, 133–145.
doi:10.1037/0022-3514.94.1.133

al
Richeson, J. A., Baird, A. A., Gordon, H. L., Heatherton, T. F., Wyland, C. L.,

ic
Trawalter, S., & Shelton, J. N. (2003). An fMRI investigation of the impact of

og
interracial contact on executive function. Nature Neuroscience, 6, 1323–1328.
doi:10.1038/nn1156 ol
ch
Rusting, C. L., & Nolen-Hoeksema, S. (1998). Regulating responses to anger: Effects
y

of rumination and distraction on angry mood. Journal of Personality and Social

Ps

Psychology, 74, 790–803. doi:10.1037/0022-3514.74.3.790

Satpute, A. B., & Lieberman, M. D. (2006). Integrating automatic and controlled
an

processes into neurocognitive models of social cognition. Brain Research, 1079,

ic

86–97. doi:10.1016/j.brainres.2006.01.005
er

Siever, L. J. (2008). Neurobiology of aggression and violence. The American Journal

Am

of Psychiatry, 165, 429–442. doi:10.1176/appi.ajp.2008.07111774

Stucke, T. S., & Baumeister, R. F. (2006). Ego depletion and aggressive behavior: Is the
e

inhibition of aggression a limited resource? European Journal of Social Psychology,

th

36, 1–13. doi:10.1002/ejsp.285

by

Sukhodolsky, D. G., Golub, A., & Cromwell, E. N. (2001). Development and vali-
dation of the Anger Rumination Scale. Personality and Individual Differences, 31,
11

689–700. doi:10.1016/S0191-8869(00)00171-9
20

Taylor, S. P. (1967). Aggressive behavior and physiological arousal as a function

of provocation and the tendency to inhibit aggression. Journal of Personality, 35,
©

297–310. doi:10.1111/j.1467-6494.1967.tb01430.x
ht

Twenge, J. M., Baumeister, R. F., Tice, D. M., & Stucke, T. S. (2001). If you can’t join
them, beat them: Effects of social exclusion on aggressive behavior. Journal of Per-
ig

sonality and Social Psychology, 81, 1058–1069. doi:10.1037/0022-3514.81.6.1058

yr
op

Weber, R., Ritterfeld, U., & Mathiak, K. (2006). Does playing violent video games
induce aggression? Empirical evidence of a functional magnetic resonance imaging
C

study. Media Psychology, 8, 39–60. doi:10.1207/S1532785XMEP0801_4

120 THOMAS F. DENSON

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 121

II

A)
P
(A
GENETIC AND

n
io
ENVIRONMENTAL

at
ci
DETERMINANTS

so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C
C
op
yr
ig
ht
©
20
12051-08_PT2-CH07-rev2.qxd

11
by
5/25/10

th
e
Am
11:38 AM

er
ic
an
Page 122

Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 123

A)
P
THE TRANSMISSION OF

(A
AGGRESSIVENESS ACROSS

n
io
GENERATIONS: BIOLOGICAL,

at
ci
CONTEXTUAL, AND SOCIAL

so
LEARNING PROCESSES

As
al
L. ROWELL HUESMANN, ERIC F. DUBOW, AND PAUL BOXER

ic
og
ol
y ch
Ps
an

In this chapter, we present a theoretical perspective for examining the

ic

cross-generational transmission of aggressive or nonaggressive behavior. We

er

begin by reviewing the evidence that there is substantial continuity of aggres-

Am

sion within a generation across the life span. We note that this continuity is
due as much to the continuity of unaggressiveness as to the continuity of aggres-
e

siveness. We then turn to examining the empirical evidence concerning the

th

cross-generational continuity of aggression. We note that although a number

by

of studies suggest such continuity, a number of methodological issues have not

been resolved by most of the studies to date. Most notably, the number of
11

prospective cross-generational life-span studies is still small. We then present a

20

social–cognitive model that has evolved to explain aggressive behavior and its
continuity in the life span, and we discuss how this model can be applied to
©

understanding cross-generational continuity within a broader framework that

ht

explains such continuity as a product of four processes: genetic predispositions

ig
yr
op

This research was supported in part by grants from the National Institute of Mental Health and the
National Institute of Child Health and Human Development. The authors wish to acknowledge the
C

contributions of Leonard Eron, Monroe Lefkowitz, and Leopold Walder to the Columbia County
Longitudinal Study.

123
12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 124

interacting with environmental factors, continuities or discontinuities in

environments, observational learning processes with a focus on parents as
models, and conditioning processes with a focus on parents as the condition-
ers. Finally, we present some data from a 40-year, three-generational longitu-

A)
dinal study that shows that children’s aggressiveness after they grow up is
predicted by their parents’ own childhood aggressiveness and their parents’

P
(A
aggressiveness when the children were young or adults but not by their par-
ents’ adolescent aggressiveness. It is shown that continuity of aggression

n
within the life span plus transmission between generations during the criti-

io
cal childhood period of the second generation can adequately explain the

at
cross-generational relations. These findings are consistent with intergenera-

ci
tional transmission through genetic and biological predispositions and learn-

so
ing processes during the childhood years.

As
al
WITHIN-PERSON CONTINUITY OF AGGRESSIVE

ic
BEHAVIOR ACROSS THE LIFE SPAN

og
ol
One of the most consistently asked questions of prospective, longitudinal
ch
data is whether personality traits, behaviors, or other aspects of psychosocial
y

functioning remain stable over the life course. The continuity of intellec-
Ps

tual ability from childhood to adulthood is substantial, with correlations typi-

cally in the moderate range (about .50). Research, including our own, on the
an

development of aggression suggests that adult aggression and antisocial behav-

ic

ior are also statistically predictable from childhood aggression and antisocial
er

behavior (Farrington, 2003; Farrington, Ttofi, & Coid, 2009; Huesmann,

Am

Dubow, & Boxer, 2009; Huesmann, Eron, Lefkowitz, & Walder, 1984; Kokko,
Pulkkinen, Huesmann, Dubow, & Boxer, 2009; Loeber & Dishion, 1983;
e

Moffitt, 1993; Olweus, 1979; Pulkkinen, Lyyra, & Kokko, 2009).

th

Our most recent estimates are based on analyses of data from the Colum-
by

bia County Longitudinal Study, a study of a cohort of 856 eight-year-olds first

evaluated in 1960 and then reevaluated at age 19 in 1970 to 1971, at age 30
11

in 1982, and at age 48 in 2000. Their parents were also interviewed in 1960,
20

and 525 of their children were interviewed in 2000, providing data on three
generations. We call the main sample the Generation 2 (G2) sample, and
©

they have provided 40 years of data on continuity of aggression across the life
ht

span. When the continuity coefficients for 40 years are estimated with struc-
ig

tural equation models that control for measurement error and method vari-
yr

ance (see Huesmann et al., 2009), we obtain coefficients of .50 for males and
op

.42 for females. These approach the continuity of intellectual ability. The
consequences of such continuity can be serious. As Kokko et al. (2009)
C

recently showed with data from Finland and the United States, early aggres-

124 HUESMANN, DUBOW, AND BOXER

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 125

siveness is more predictive of physical aggression and lack of self-control in

later middle adulthood than it is of verbal aggression.
It is worth noting that the continuity correlations over time are due as
much to low aggressive children staying low on aggression throughout life as

A)
they are to high aggressive children staying high on aggression throughout life
(Huesmann et al., 2009). To demonstrate this fact, we grouped participants

P
(A
based on age-8 peer-nominated aggression into “high” and “low” categories
using both median and one-third splits and examined how many stayed in the

n
same category over 40 years. For the full sample, 37% (52 of 141) of individu-

io
als who were low in third grade stayed low through late adolescence (i.e.,

at
age 19), through young adulthood (i.e., age 30), and into middle adulthood

ci
(i.e., age 48) on the composite measure of aggression. Of individuals who were

so
high in third grade, 35% (i.e., 31 of 89) stayed high through adolescence

As
and into middle adulthood. In other words, continuity was as much due to
low aggressives staying low as to high aggressives staying high. When we

al
inspected these patterns by gender, interesting differences became evident.

ic
For males, 38% of those who were low in childhood stayed low through age

og
48. Similarly, 36% of females who were low in childhood stayed low into
ol
middle adulthood. However, the differences were striking with regard to
ch
those classified as high on aggression. Among males, 47% of those who were
y

high in childhood stayed high into middle adulthood, whereas for females,
Ps

only 18% who were high in childhood continued to be high into middle
adulthood (χ [1] = 14.97, p < .001).
an

These results open the possibility that differential socialization of

ic

males and females places more pressure on females to reduce their aggres-
er

sion than it does on males. Early aggressiveness may also have more last-
Am

ing serious consequences for males. For example, Huesmann, Eron, and
Dubow (2002) found that early aggressiveness was the most important pre-
e

dictor of males being arrested by the time they were 30 years old, in com-
th

parison with a large variety of contextual and personal variables assessed

by

in the Columbia County Longitudinal Study. The standardized odds ratio

for age-8 aggression in predicting “ever arrested by age 30” was highly sig-
11

nificant: 1.45, p < .01.

20

Of course, even continuity coefficients as large as .5 and odds ratios

approaching 1.5 still mean that a substantial portion of adult aggressive
©

and criminal behavior is not predictable from childhood aggressive behav-

ht

ior and is probably related to context (Broidy et al., 2003; Sameroff, Seifer,
ig

Baldwin, & Baldwin, 1993). In addition, it seems to be only aggressive and

yr

antisocial behavior that emerges early in life that has lasting negative con-
op

sequences. As Moffitt (1993) suggested and a number of recent studies con-

firmed, some aggressive and antisocial behavior often emerges in adolescence
C

and is relatively short lived. This kind of aggression seems to have few long-

THE TRANSMISSION OF AGGRESSIVENESS 125

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 126

term negative consequences, whereas life course persistent aggression that

begins in early childhood has lasting detrimental consequences (Bergman
& Andershed, 2009; Farrington et al., 2009; Huesmann et al., 2009; Pulkkinen
et al., 2009).

A)
Similar findings regarding continuity have been reported in studies of
other indicators of adjustment over time. For example, Helson, Jones, and

P
(A
Kwan (2002) demonstrated quadratic changes in various indicators of person-
ality functioning from early to late adulthood: Certain attributes such as

n
dominance and independence peaked in middle adulthood. In her influential

io
island of Kauai prospective study, Werner (2002) summarized long-term out-

at
comes for children and adolescents with behavior disorders: By age 40, only

ci
one third of those males and one fifth of those females exhibited continuing

so
difficulties (e.g., financial, marital, substance use). However, far less is

As
known about the degree of continuity of positive psychosocial adjustment
from childhood to adulthood, although this appears to be an emergent con-

al
cern of longitudinal researchers. As an example, with data from the Jyväsklyä

ic
Longitudinal Study, Pulkkinen and her colleagues demonstrated that pro-

og
social behavior in childhood predicted greater self-esteem and subjective
ol
well-being and shorter lived periods of unemployment in adulthood (Kokko
ch
& Pulkkinen, 2000; Pulkkinen, Nygren, & Kokko, 2002). Flouri and Buchanan
y

(2002) showed that good family relationships in childhood led to better

Ps

marital adjustment in adulthood.

an
ic

RECENT STUDIES OF INTERGENERATIONAL

er

CONTINUITY OF AGGRESSION
Am

The discovery of strong continuity of aggression across the life course

e

has stimulated interest in the continuity of aggression and antisocial behav-

th

ior across generations. Although many studies have investigated parent–child

by

relations statically (i.e., cross-sectionally), longitudinal cross-generational

investigations of continuity and discontinuity in personality, behavior, and
11

adjustment are relatively few and are limited primarily to examining contem-
20

poraneous or retrospective links between parent and child behavior. Several

discussions of the intergenerational transmission of aggression have been
©

published in the child development literature over the past decade (e.g.,
ht

Constantino, 1996; MacEwen, 1994). Those reviews indicated that most rel-
ig

evant studies have used self-report, retrospective questionnaire data obtained

yr

from two generations. However, response bias problems cloud the interpretation
op

of such findings. Three studies in a 1998 special issue of Developmental Psychol-

ogy did use prospective methodology and multiple methods of measuring aggres-
C

sion (Cairns, Cairns, Xie, Leung, & Hearne, 1998; Capaldi & Clark, 1998;

126 HUESMANN, DUBOW, AND BOXER

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 127

Serbin et al., 1998), and all reported modest to moderate cross-generational con-
tinuity. But those studies included only two generations.
More recently, four studies on cross-generational consistencies in par-
enting and in aggressive and antisocial behavior were published in a special

A)
issue of the Journal of Abnormal Child Psychology (Capaldi, Pears, Patterson,
& Owen, 2003; Conger, Neppl, Kim, & Scaramella, 2003; Hops, Davis,

P
(A
Leve, & Sheever, 2003; Thornberry, Freeman-Gallant, Lizotte, Krohn, &
Smith, 2003; see also Smith & Farrington, 2004). All four studies measured

n
parenting by two generations (G1 and G2) and aggressive behavior in two

io
generations (G2 and G3). Two of the four studies found significant inter-

at
generational continuity of aggression. The two that did not report such

ci
continuity had the smallest sample sizes. All four studies also reported inter-

so
generational continuity for some parenting factors relevant to aggression

As
(e.g., negative affect in parent–child interactions, consistency of discipline).
The results also suggested that child aggression mediates some of the conti-

al
nuity of parenting as well as parenting mediating some of the continuity of

ic
aggression. For example, Thornberry et al. (2003), using self-report question-

og
naire data, found a chain of relations for males from G1 parenting to G2
ol
aggression to G2 parenting to G3 aggression. For females, the pattern was
ch
similar, although the link from G1 parenting to G2 aggression was not sig-
y

nificant. Conger et al. (2003), using observational data, found both that G1
Ps

observed parenting had direct effects on G2 observed aggression and that G2

observed parenting had direct effects on G3 observed aggression. Hops et al.
an

(2003), also using observations of parenting, obtained a fully mediated path

ic

from G1 parenting, to G2 aggression, to G2 parenting, and finally to G3

er

aggression. Capaldi et al. (2003) reported findings similar to those of the

Am

other studies, with an important methodological distinction: the use of

multiple informants and sources of data, moving beyond the parent and child
e

observational and questionnaire data used in the other studies to include

th

teacher reports and archival records.

by

A plausible conclusion that can be drawn from these similar results in the
four studies is that parenting behavior and aggressive behavior seem to have
11

reciprocal influences on each other. Within generations, aggression in youth is

20

often followed by aggression-promoting parenting. Aggression-promoting

parenting, in turn, seems to contribute to aggression in offspring. However, this
©

conclusion might be attenuated by important considerations related to the the-

ht

ory on which these cross-generational investigations are based. In spite of

ig

increasing empirical attention, the processes by which patterns of positive or

yr

negative adjustment are transmitted from parents to children are not yet well
op

understood or firmly established, and more research is needed to explicate cross-

generational links and the theory explaining such links (Dubow, Huesmann,
C

& Boxer, 2003; Rutter, 1998; Shaw, 2003).

THE TRANSMISSION OF AGGRESSIVENESS 127

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 128

PSYCHOLOGICAL PROCESSES INVOLVED

IN AGGRESSIVE BEHAVIOR

To begin with, to talk theoretically about the cross-generational trans-

A)
mission of aggression, we need a model for the psychological processes
through which predisposing personal factors and precipitating situational fac-

P
(A
tors interact to determine whether a person behaves aggressively. The model
needs to include a representation of the enduring psychological structures that

n
control and influence these processes. Such a model has been provided by

io
Huesmann (1998) in a unified information-processing model for social prob-

at
lem solving. According to this model that integrates previous theorizing of

ci
Huesmann (1988), Bandura (1977), Dodge (1982), and Anderson (Anderson

so
& Huesmann, 2003), an individual’s emotional state and encoded schemas

As
about the world interact with situational cues to lead the individual to make
attributions about the situation. These attributions change the individual’s

al
emotional state and prime the activation of scripts for behaving. The scripts

ic
are filtered through a set of normative beliefs about appropriateness until a

og
script is accepted and followed.
ol
Within this model, the long-term determinants of aggressiveness
ch
include (a) encoded cognitions represented by schemas about the world,
y

(b) scripts for behavior, and (c) normative beliefs for filtering scripts, along
Ps

with (d) the individual’s emotional predispositions. More specifically,

hostile attributional biases (e.g., “People are mean”; Dodge, 1982; see also
an

Chapter 9, this volume) occur when hostility is emphasized by the “world

ic

schema” that an individual has acquired. Attributions influence emotions

er

and the type of script an individual will retrieve to deal with a social situ-
Am

ation. Scripts (Huesmann, 1998) can be viewed as cognitive programs that

have been acquired over time and are stored in a person’s memory and are
e

used as guides for behavior and social problem solving. Not all scripts that
th

occur to the child will be used. Before acting out the script, the child
by

reevaluates the appropriateness of the script in light of existing internalized

social norms—called normative beliefs (Huesmann, 1998; Huesmann &
11

Guerra, 1997)—and examines the likely consequences of the script. These

20

normative beliefs and expectations about outcome must also have been
acquired over time.
©

Our focus on cognitive-information processes does not mean that emo-

ht

tions are unimportant. We view emotion regulation as an important influ-

ig

ence on aggressive behavior that influences attributions, script selection, and

yr

evaluation of scripts. Emotion regulation has held a place of prominence in

op

the study of behavioral development for some time, particularly with regard
to developmental psychopathology (e.g., Cicchetti, Ackerman, & Izard,
C

1995; Frick & Morris, 2004).

128 HUESMANN, DUBOW, AND BOXER

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 129

Processes for the Intergenerational Transmission of Aggression

Given this model, how can a tendency to behave in a characteristic

manner (e.g., aggressively or nonaggressively) be transmitted from one

A)
person (e.g., a parent) to another (e.g., the parent’s child)? Our position is
that there are four major processes by which this may occur that need to

P
(A
be considered: (a) through the transmission of genes that influence social
behavior, (b) by changes parents make in the child’s environment or through

n
continuity in the parent’s and child’s environment, (c) through children

io
observing parents’ behaviors, and (d) through conditioning of the children’s

at
behavior in which the parent participates.

ci
so
Genetic Influences on Aggression

As
Genetic influences are well established empirically (see Chapter 8).

al
Individual differences in emotional arousal, neurotransmitter levels, percep-

ic
tual biases, and other characteristics relevant to aggression seem to be influ-

og
enced by genes. The evidence for a heritable predisposition to aggression from
twin and adoption studies is impressive (Cloninger & Gottesman, 1987;
ol
Deater-Deckard & Plomin, 1999; Mednick, Gabrielli, & Hutchings, 1984).
ch
Miles and Carey (1997) performed a meta-analysis of 24 genetically infor-
y

mative studies that included twin and adoption designs and found significant
Ps

heritability estimates (in the .4 range) for self-report measures of aggression.

In addition, a variety of adoption studies have revealed relations between chil-
an

dren’s aggressiveness or antisocial behavior and the aggressive or antisocial

ic

behavior of both their natural and adoptive parents.

er

However, these seem to be predisposing influences rather than determin-

Am

istic influences. Estimates from behavior genetic analyses of low shared environ-
mental variance depend on unlikely assumptions of genes being uncorrelated,
e

noninteracting, and not influencing the environment of the individual. The

th

weight of evidence suggests, rather, that biosocial interactions between genes

by

and the environment are likely influences.

Recent studies are beginning to go beyond global estimates of the rela-
11

tive influences of genetic differences in accounting for phenotypic differences

20

to look for the specific genes responsible (e.g., Chapter 8). Strategies based
on genome scans have proved disappointing in the search for loci that influ-
©

ence behavioral phenotypes, but investigations of polymorphisms of candi-

ht

date genes have been surprisingly productive. Although methods are still
ig

developing for statistical discrimination of reliable signals from noise when

yr

many candidate genes are studied, effects of monoamine oxidase (MAO)

op

and serotonin promoters have been confirmed and others are under study (for
a review, see Munafò et al., 2003). Studies by Caspi and colleagues (2002,
C

2003) have indicated that there are important childhood contextual–genetic

THE TRANSMISSION OF AGGRESSIVENESS 129

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 130

interactions in the expression of social behavior. For example, those authors

found that a gene that causes slightly lower MAO increases risk of aggression
only in a child who is exposed to high amounts of stress during childhood and
adolescence. Similarly, the perinatal environment is known to affect the risk

A)
of aggressive behavior (Raine, Brennan, & Mednick, 1995). Whatever their
source, a variety of individual differences in neurophysiology, neurotransmit-

P
(A
ters, hormones, and heart rate correlate with individual differences in early
aggressiveness (Knoblich & King, 1992; Olweus, Mattsson, Schaling, &

n
Loew, 1988; Raine & Jones, 1987).

io
at
Intergenerational Environmental Continuity and Change and Their Influences

ci
on Continuity and Change in Aggression

so
Many environmental influences on risk and resilience for aggression and

As
violence have been identified. These include stress, poverty, abuse, parental

al
rejection, peer behaviors, and religion, to name a few. To the extent that such

ic
environments are passed from parent to child, one can expect cross-generational

og
continuity. To the extent that such environments change from parent to child,
one can expect cross-generational discontinuity.
ol
A major question in developmental research concerns how changes
ch
over time in the social contexts people inhabit affect development (Higgins
y

& Parsons, 1983; Huesmann, Dubow, Eron, & Boxer, 2006; Sameroff, 1983).
Ps

One needs to understand whether the degree of continuity in positive and

negative adjustment over time and across generations is related to the degree
an

of continuity in contextual factors. Is there continuity in parenting practices

ic

from what a current parent experienced in interactions with his or her par-
er

ents to what the parent delivers to his or her child? Does continuity or change
Am

in socioeconomic status promote continuity or change in aggression? Many

children have to cope with family changes: Their parents might divorce, get
e

into trouble with the law, or lose their jobs. Evidence suggests that the stress
th

engendered by such changes is a risk factor for children’s socioemotional

by

development (Caspi et al., 2002), thus increasing the likelihood of negative

changes that might have long-term consequences. Our hypothesis is that sim-
11

ilar discontinuities of a positive nature (e.g., a substantial improvement in the

20

family’s financial situation, a significant improvement in the child’s academic

performance) can turn trajectories of psychosocial adjustment upward toward
©

greater success, achievement, and life satisfaction.

ht

A second question in developmental research concerns the timing of

ig

exposure to contextual influences. What contexts at what points during

yr

childhood are most predictive of later outcomes? For example, Duncan

(2002) found that poverty experienced during the early childhood years had
op

the strongest effects on the number of years of education the child attained
C

by early adulthood. This suggests that the environment in which the parent

130 HUESMANN, DUBOW, AND BOXER

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 131

lives and the child develops would be more important than the parent’s own
early environment.
A third issue of great relevance is the extent to which the larger histor-
ical context surrounding individual development influences the trajectories

A)
taken by cohorts embedded in those circumstances. As an example, what is
the differential impact of being an adolescent in a society that is at war as

P
compared with the same society experiencing relative peace, and how might

(A
the institution of selective or compulsory service magnify that impact?

n
These issues lead to a fourth broad contextual concern: major life tran-

io
sitions or turning points in individual development (Rönkä, Oravala, &

at
Pulkkinen, 2002; Rutter, 1996). Turning points are positive or negative events,

ci
over which the individual may or may not have control, which significantly

so
alter the life trajectory. According to Rutter (1996), an event can be a turn-

As
ing point only if it leads to an enduring, long-term modification of the trajec-
tory. Thus, turning points cannot be assessed contemporaneously for their

al
impact. Longitudinal data are necessary because often individuals do not rec-

ic
ognize turning points in their lives until some time has passed and the indi-

og
vidual has understood the importance and meaningfulness of the event.
ol
ch
Children Observing Parents’ Behaviors
y

The observation of parents’ behaviors by children can be expected to

Ps

influence social behavior in general and aggressive behavior in particular, and

this can occur through two quite different processes: short-term stimulating
an

processes and long-term learning processes. These processes operate more

ic

generally to make the child imitate anything he or she observes, including

er

aggressive acts by others (Huesmann & Kirwil, 2007).

Am

In the short run, when children see their parents behave aggressively,
schemas, scripts, and normative beliefs associated with aggression are primed
e

in the children’s minds. In addition, emotions associated with the behavior

th

are aroused in the child (e.g., anger may be stimulated). These processes alone
by

can lead to short-term increases in the risk of aggressive behavior. In addi-

tion, however, because children generally identify with their parents, they are
11

likely to mimic behaviors almost immediately. All of these short-term stim-

20

ulating processes make it likely that a child will behave aggressively after
observing his or her parents behaving aggressively.
©

However, the more important observational processes for the inter-

ht

generational transmission of aggression are probably long-term observational

ig

learning processes involving parent and child. As mentioned earlier, children

yr

generally identify strongly with their parents. Consequently, they tend to

op

encode into their repertoire of scripts the scripts they see their parents using,
they tend to adopt the world schemas they perceive their parents to be hold-
C

ing, and they tend to accept the normative beliefs of their parents about the

THE TRANSMISSION OF AGGRESSIVENESS 131

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 132

appropriateness of social behavior. For a long time, children’s imitation of

parents’ behaviors was thought to be a relatively low-level childish form of
behavior. “But recent work across a variety of sciences argues that imitation
is a rare ability fundamentally linked to characteristically human forms of intel-

A)
ligence, in particular to language, culture, and the ability to understand other
minds” (Hurley & Chater, 2005, p. 1). Imitation of parents’ behaviors appears

P
(A
to be innate and occurs automatically in young primate infants (Meltzoff &
Moore, 1983). Specific mirror neurons seem to organize imitation in primate

n
brains (Gallese, Fadiga, Fogassi, & Rizzolatti, 1996). However, recent work

io
suggests that imitation goes far beyond the copying of specific sequences of

at
behaviors to the encoding of social cognitions fundamental to the control of

ci
social behavior (Meltzoff, 2007).

so
These observational learning processes are also involved in the develop-

As
ment of emotion regulation. Eisenberg, Cumberland, and Spinrad (1998)
described three processes by which emotion socialization leads to social com-

al
petence. First, through everyday family interactions, children display a wide

ic
range of both positive and negative emotions, thus providing parents with

og
numerous opportunities to react in both positive (e.g., encouraging) and neg-
ol
ative (e.g., punishing) ways. Second, when parents discuss various aspects of
ch
emotion (e.g., causes and consequences, emotional experience and regulation)
y

with their children, they serve as models for understanding and coping with
Ps

emotion. A child who receives a high degree of emotion knowledge from his
or her parents should be better equipped to contend with emotionally charged
an

events. Third, family emotional expressivity relates in important ways to chil-

ic

dren’s social competence. Through observation, children can learn positive

er

(e.g., smiling) or negative (e.g., yelling) modes of expressing emotions, as well

Am

as heuristics for interpreting the emotions of others. This model of emotion

socialization suggests individual and cross-generational continuity of emo-
e

tional regulation and, in turn, social behaviors, as well as taking into account
th

potential moderating influences of contextual or individual factors.

by

Conditioning of Children’s Behaviors to Be Similar to or Different from Parents

11

Of course, observational learning is not the only learning process cru-

20

cial to intergenerational transmission of aggressive tendencies. As Patterson

©

(1986) demonstrated, parents may unintentionally condition their children

operantly to behave aggressively. Through coercive family interactions, they
ht

may unintentionally reinforce their children for behaving aggressively. They

ig
yr

may also deliberately and directly reinforce their children for being aggressive
if the parents believe aggression is appropriate.
op

Parents who harshly punish and abuse their children may also classically
C

condition them to experience anger responses to stimuli that might seem

132 HUESMANN, DUBOW, AND BOXER

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 133

benign to others (e.g., persons of authority). They may classically condition

alienation from society in this way and make appropriate socialization of the
child by others in the child’s environment difficult.

A)
Analysis of Intergenerational Continuity in
the Columbia County Longitudinal Study

P
(A
The first analyses of the four waves of data spanning three generations

n
from the Columbia County Longitudinal Study suggest that genetic and dis-

io
positional processes, learning processes operating during the child-rearing

at
years, and environmental continuities may all contribute to cross-generational

ci
continuity of aggressiveness. The sample we used in these analyses consists of

so
349 of the 551 G3 offspring who were interviewed when their G3 parent was

As
48 years old. The 349 represent one child of each G2 subject interviewed—
the youngest child for subjects with more than one child (in 202 families two

al
children were interviewed). The ages of the G3 sample at the time of the inter-

ic
view ranged from 6 to 33, with a median age of 18.

og
In these analyses, we assessed aggression in the G2 generation with the
ol
same composite measure we developed to assess continuity of aggression across
ch
the life span in the G2 generation. The indicators we used in that analysis of
y

four waves of G2 data were peer nominations at age 8 and 19; severe physical
Ps

aggression at ages 19, 30, and 48; and aggressive personality at ages 19, 30,
and 48. The “severe physical aggression” measure assessed how often the per-
an

son punched, choked, beat up, or used a knife or a gun against another person
ic

in the preceding year. The “aggressive personality” measure was the sum of the
er

F, 4, and 9 scales of the Minnesota Multiphasic Personality Inventory (MMPI;

Am

Hathaway & McKinley, 1940). For the G3 data, we used the same two indi-
cators of adult aggression as with G2: aggressive personality and behaving
e

severely physically aggressively. However, we analyzed these indicators sepa-

th

rately because the sample sizes are quite different for the two variables, given
by

that only those G3 offspring who were over 13 received the aggressive person-
ality assessment.
11

In Table 7.1, the intergenerational correlations are shown from G2 to

20

G3 for these variables. One can see substantial cross-generational correlations

between the G2 parent’s aggression at age 8, 19, 30, and 48 and the G3 off-
©

spring’s aggressive personality and severe physical aggression at the time the
ht

G2 parent was 48. For a subsample of 125 of the G3 offspring we had data on
ig

the aggressiveness of the other parent as well. Their aggressiveness correlated

about .20 (p < .05) with the G2 spouses’ aggressiveness at both age 30 and 48
yr
op

and with the G3 offsprings’ aggressiveness. However, the total prediction of

G3s’ aggression was not enhanced by adding in this other parent’s aggression
C

to a prediction equation.

THE TRANSMISSION OF AGGRESSIVENESS 133

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 134

TABLE 7.1
Correlations Between Generation G2’s (Parents’) Aggression at Four Ages
and Generation G3’s Aggression When G2 Parent Was Age 48
Child aggression when parent was age 48

A)
G3 child’s G3 child’s severe

P
aggressive physical

(A
G2 parent’s aggression personality aggression
at different ages (N = 212) (N = 348)

n
Age 8

io
G2’s composite aggression 0.24*** 0.09*

at
Age 19

ci
G2’s composite aggression 0.26*** 0.18***

so
Age 30
G2’s composite aggression 0.46*** 0.17***

As
Age 48
G2’s composite aggression 0.19*** 0.26***

al
* p < .10. ** p < .05. *** p < .01.

ic
og
ol
The correlations with G2s’ age-8 aggression are consistent with genetic
ch
and biological predisposition models of intergenerational transfer, whereas
y

the correlations between G2s’ later aggression and the G3 offsprings’ aggres-
Ps

sion are consistent with learning models that emphasize transmission of

aggressive behavior during the child-rearing years. Although the differences
an

are not large, an inspection of the correlations suggests that the strongest rela-
ic

tions with the offspring’s aggressive personality are from the parent’s aggres-
er

sion years earlier when the parent was a child, adolescent, or young adult.
Am

However, the strongest correlations with the G3 offsprings’ actual physically

aggressive behavior are with the G2 parents’ aggressiveness later in life at age
e

19, 30, and 48 (i.e., during the early child-rearing years). These results are
th

consistent with the concept that characteristically aggressive personalities are

by

influenced by predisposing genetic and biological factors modified by learn-

ing experiences during the early child-rearing years, whereas the emergence
11

of adult aggressive behaviors is influenced more by the child-rearing environ-

20

ment in which the offspring develops. The highest correlations with G3s’
aggressive personality around age 18 is with the parents’ age 30 aggressiveness
©

for both genders 18 years earlier. However, the highest correlations with G3s’
ht

physically aggressive behavior around age 18 is with the parents’ concurrent

ig

age 48 aggressiveness.
yr

These specific relations are illustrated in more detail in Figure 7.1,

op

which displays the mean aggression scores for the offspring of G2 parents who
scored low (i.e., < 25th percentile), medium, or high (i.e., > 75th percentile)
C

on aggression concurrently at age 48 or 18 years earlier at age 30. The effects

134 HUESMANN, DUBOW, AND BOXER

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 135

0.6
G3 offspring's severe physically aggressive behavior

0.5

A)
when G2 parent was age 48

0.4

P
(A
Lo G2 Agg
0.3 Med G2 Agg

n
io
Hi G2 Agg

at
0.2

ci
so
0.1

As
al
0

ic
G2 parent's aggression at age 48

og
210
ol
ch
G3 offspring's aggressive personality when G2

200
Ps

190
an

180
parent was age 48

ic

Lo G2 Agg
er

170
Med G2 Agg
Am

160 Hi G2 Agg

150
e
th

140
by

130
11

120
20

G2 parent's aggression at age 30

©

Figure 7.1. In the upper panel, the relation between a parent’s age-48 aggression and
ht

their offspring’s concurrent severe physical aggression when the offspring averaged
18 years of age, F (2, 312) = 11.7, p < .001. In the lower panel, the relation between
ig

a parent’s age-30 aggression and their offspring’s aggressive personality 18 years

yr

later (i.e., when the offspring averaged 18 years of age), F (2, 113) = 6.24, p < .01.
op

G3 aggression is low if G2 is below the 25th percentile, medium if G2 is between the

25th to 75th percentile, and high if G2 is above the 75th percentile.
C

THE TRANSMISSION OF AGGRESSIVENESS 135

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 136

are significant and illustrate that it is the offspring of parents who were above
the 75th percentile on aggression at age 30 who are most at risk of develop-
ing an aggressive personality, and it is the offspring of parents who are above
the 75th percentile on aggression at age 48 who are most at risk of being seri-

A)
ously physically aggressive around age 18.
Given these relations, we decided to test a model of cross-generational

P
(A
transmission that combines continuity of aggression within the G2 generation
with transmission across generations only during the age 30 to age 48 period

n
when the G3 child is being reared. The model incorporated both the measure

io
of aggressive personality and the measure of severe physical aggression. This

at
two-generational structural model for continuity of aggression is shown in

ci
the left panel of Figure 7.2. The model assumes that the only path from gen-

so
eration G2 to generation G3 is the path from the aggression of the G2 parent

As
at age 30 to the aggression of the child 18 years later. The model fits the data well
with a nonsignificant chi-square statistic and other reasonable goodness-of-fit

al
statistics. The path from G2 aggression at age 30 to G3 aggression 18 years

ic
later is a highly significant and large effect.

og
We next expanded the model to incorporate all three generations we had
ol
studied. The results are shown in the right panel of Figure 7.2. Unfortunately,
ch
our only good measure of G1’s aggression is the individual’s tendency to hit
y

G2 when G2 was age 8. This measure is related to G2s’ aggression, but it prob-
Ps

ably represents G2s’ tendency to behave badly as well; thus, the path relating
it to G2s’ aggression at the same time is bidirectional. Still, the model fits the
an

data well, and the best estimates of the path coefficients for the model suggest
ic

a total cross-generational effect from G1 to G3 of .15. Adding a path from G2s’

er

aggression at age 48 directly to G3s’ concurrent aggression did not significantly

Am

improve the fit of the model, indicating that the strong relation between G2s’
age 48 aggression and G3s’ concurrent severe physical aggression (see Table 7.1
e

and Figure 7.1) is a consequence of the stability of G2s’ aggression through the
th

child-rearing years. Thus, the estimated path coefficients are consistent with
by

the concept that aggression is transmitted across generations primarily dur-

ing the child-rearing years, with genetic and biological influences exerting
11

long-term effects through within-generation continuity.

20
©

SUMMARY
ht
ig

In this chapter, we argued that cross-generational continuity in aggres-

yr

siveness is undoubtedly the product of four factors: biological predispositions

op

that are inherited by offspring, continuity in environment across genera-

tions, and conditioning and observational learning experienced by the child.
C

We elaborated a model that takes into consideration all of these factors, and

136 HUESMANN, DUBOW, AND BOXER

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 137

Continuity of Aggression Over 40 Years Across 2 Generations

Sev Phys Agg MMPI F49 Sev Phys Agg MMPI F49
.63# .51# .63# .51#

A)
G2 SUBJECT

P
G2 Agg 8 G2 Agg 19 G2 Agg 30 G2 Agg 48

(A
.64*** .86*** .80***

n
.67# .67# .63# .51# .65***

io
PrAgg PrAgg Sev Phys Agg MMPI F49

at
G3 Agg

ci
.63# .51#

so
Sev Phys Agg MMPI F49
Chi-Sq(43) = 47.9, p > .28, RMSEA = .018, CFI = .989

As
G3 CHILD

al
ic
Continuity of Aggression Over 40 Years Across 3 Generations

og
ol
ch
G1 Par
Hit G2 G1 PARENT
y
Ps

.80#
Sev Phys Agg MMPI F49 Sev Phys Agg MMPI F49
G1 Agg
.63# .51# .63# .51#
an

.13
.25*
ic

G2 SUBJECT
er

G2 Agg 8 G2 Agg 19 G2 Agg 30 G2 Agg 48

.61*** .86*** .80***
Am

.67# .67# .63# .51# .64***

e

PrAgg PrAgg Sev Phys Agg MMPI F49

th

G3 Agg
.63# .51#
by

Sev Phys Agg MMPI F49

Chi-Sq(52) = 56.1, p > .32, RMSEA = .015, CFI = .991
11

G3 CHILD
20

Figure 7.2. In the upper panel, the continuity of aggression over 40 years across 2 gen-
©

erations (G2 and G3). X2 (43) = 47.9, p > .28, RMSEA = .018, Cumulative Fit Index =
.99. In the lower panel, the continuity of aggression over 40 years across 3 generations
ht

(G1, G2, and G3). X2 (52) = 56.1, p > .32, RMSEA = .015, Goodness of Fit Index =
ig

.991. # indicates that the measurement parameter was fixed at the value estimated in
the within-generation model. Correlated errors were specified for the same measures
yr

across ages. PrAgg = peer nominated aggression; Sev Phys Agg = severe physical
op

aggression; MMPI F49 = the sum of scales F, 4, and 9 on the Minnesota Multiphasic
Personality Inventory; G1 Par Hit G2 = the amount of hitting of G2 child at age 8 by
C

the G1 parents; G2 Agg ‘n’ = G2’s composite aggression score at age ‘n’.
*p<.05 **p<.01 ***p<.001.

THE TRANSMISSION OF AGGRESSIVENESS 137

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 138

we applied it to data from a three-generation longitudinal study. The data

showed significant cross-generational continuity including significant relations
between an offspring’s aggressive behavior as an adult and their parents’ aggres-
sive behaviors in both childhood and during the time the child was being raised.

A)
A model fits the data well in which the cross-generational effects across two
and three generations are mostly mediated by the parent’s aggression during

P
(A
the child-rearing period.

n
io
REFERENCES

at
ci
Anderson, C. A., & Huesmann, L. R. (2003). Human aggression: A social–cognitive

so
view. In M. Hogg & J. Cooper (Eds.), Handbook of social psychology (pp. 296–323).

As
London, England: Sage.
Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice Hall.

al
Bergman, L. R., & Andershed, A. (2009). Predictors and outcomes of persistent or age-

ic
limited registered criminal behavior: A 30-year longitudinal study of a Swedish

og
urban population. Aggressive Behavior, 35, 164–178. doi:10.1002/ab.20298
ol
Broidy, L. M., Nagin, D. S., Tremblay, R. E., Bates, J. E., Brame, B., Dodge, K. A.,
ch
. . . Vitaro, F. (2003). Developmental trajectories of childhood disruptive
behaviors and adolescent delinquency: A six-site, cross-national study. Develop-
y
Ps

mental Psychology, 39, 222–245. doi:10.1037/0012-1649.39.2.222

Cairns, R. B., Cairns, B. D., Xie, H., Leung, M. C., & Hearne, S. (1998). Paths
an

across generations: Academic competence and aggressive behaviors in young

ic

mothers and their children. Developmental Psychology, 34, 1162–1174. doi:10.1037/

er

0012-1649.34.6.1162
Am

Capaldi, D. M., & Clark, S. (1998). Prospective family predictors of aggression

toward female partners for at-risk young men. Developmental Psychology, 34,
1175–1188. doi:10.1037/0012-1649.34.6.1175
e
th

Capaldi, D. M., Pears, K. C., Patterson, G. R., & Owen, L. D. (2003). Continuity of
parenting practices across generations in an at-risk sample: A prospective com-
by

parison of direct and mediated associations. Journal of Abnormal Child Psychol-

ogy, 31, 127–142. doi:10.1023/A:1022518123387
11

Caspi, A., McClay, J., Moffitt, T. E., Mill, J., Martin, J., Craig, I. W., . . . Poulton, R.
20

(2002, August). Role of genotype in the cycle of violence in maltreated children.

Science, 297, 851–854. doi:10.1126/science.1072290
©

Caspi, A., Sugden, K., Moffitt, T. E., Taylor, A., Craig, I. W., Harrington,
ht

H., . . . Poulton, R. (2003, July). Influence of life stress on depression: Modera-

ig

tion by a polymorphism in the 5-HTT gene. Science, 301, 386–389. doi:10.1126/

yr

science.1083968
op

Cicchetti, D., Ackerman, B. P., & Izard, C. E. (1995). Emotions and emotion regu-
lation in developmental psychopathology. Development and Psychopathology, 7,
C

1–10. doi:10.1017/S0954579400006301

138 HUESMANN, DUBOW, AND BOXER

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 139

Cloninger, C. R., & Gottesman, I. I. (1987). Genetic and environmental factors

in antisocial behavior disorders. In S. A. Mednick, T. E. Moffitt, & S. A. Stack
(Eds.), The causes of crime: New biological approaches (pp. 92–109). New York,
NY: Cambridge University Press.

A)
Conger, R. D., Neppl, T., Kim, K. J., & Scaramella, L. (2003). Angry and aggressive
behavior across three generations: A prospective, longitudinal study of parents

P
and children. Journal of Abnormal Child Psychology, 31, 143–160. doi:10.1023/

(A
A:1022570107457

n
Constantino, J. N. (1996). Intergenerational aspects of the development of aggres-

io
sion: A preliminary report. Journal of Developmental and Behavioral Pediatrics,

at
17, 176–182. doi:10.1097/00004703-199606000-00006

ci
Deater-Deckard, K., & Plomin, R. (1999). An adoption study of the etiology of

so
teacher and parent reports of externalizing behavior problems in middle child-

As
hood. Child Development, 70, 144–154. doi:10.1111/1467-8624.00011
Dodge, K. A. (1982). Social cognition and children’s aggressive behavior. Child

al
Development, 53, 620–635. doi:10.2307/1129373

ic
Dubow, E. F., Huesmann, L. R., & Boxer, P. (2003). Theoretical and methodologi-

og
cal considerations in cross-generational research on parenting and child aggres-
ol
sive behavior. Journal of Abnormal Child Psychology, 31, 185–192. doi:10.1023/
ch
A:1022526325204
y

Duncan, G. J. (2002). The PSID and me. In E. Phelps, F. F. Furstenberg, &

Ps

A. Colby (Eds.), Looking at lives: American longitudinal studies of the twentieth

century (pp. 133–166). New York, NY: Russell Sage Foundation.
an

Eisenberg, N., Cumberland, A., & Spinrad, T. L. (1998). Parental socialization of

ic

emotion. Psychological Inquiry, 9, 241–273. doi:10.1207/s15327965pli0904_1

er

Farrington, D. P. (2003). Key results from the first forty years of the Cambridge Study
Am

in delinquent development. In T. P. Thornberry & M. D. Krohn (Eds.), Taking

stock of delinquency: An overview of findings from contemporary longitudinal studies
e

(pp. 137–183). New York, NY: Kluwer/Plenum.

th

Farrington, D. P., Ttofi, M. M., & Coid, J. W. (2009). Development of adolescence-

by

limited, late-onset, and persistent offenders from age 8 to age 48. Aggressive
Behavior, 35, 150–163. doi:10.1002/ab.20296
11

Flouri, E., & Buchanan, A. (2002). What predicts good relationships with parents in
adolescence and partners in adult life: Findings from the 1958 British birth cohort.
20

Journal of Family Psychology, 16, 186–198. doi:10.1037/0893-3200.16.2.186

©

Frick, P. J., & Morris, A. (2004). Temperament and developmental pathways to con-
ht

duct problems. Journal of Clinical Child and Adolescent Psychology, 33, 54–68.
doi:10.1207/S15374424JCCP3301_6
ig
yr

Gallese, V., Fadiga, L., Fogassi, L., & Rizzolatti, G. (1996). Action recognition in the
premotor cortex. Brain, 119, 593–609. doi:10.1093/brain/119.2.593
op

Hathaway, S. R., & McKinley, J. C. (1940). The MMPI manual. New York, NY:
C

Psychological Corporation.

THE TRANSMISSION OF AGGRESSIVENESS 139

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 140

Helson, R., Jones, C., & Kwan, V. S. Y. (2002). Personality change over 40 years
of adulthood: Hierarchical linear modeling analyses of two longitudinal sam-
ples. Journal of Personality and Social Psychology, 83, 752–766. doi:10.1037/
0022-3514.83.3.752
Higgins, E. T., & Parsons, J. E. (1983). Social cognition and the social life of the

A)
child: Stages as subcultures. In E. T. Higgins, D. W. Ruble, & W. W. Hartup

P
(Eds.), Social cognition and social behavior: Developmental issues (pp. 15–62). New

(A
York, NY: Cambridge University Press.

n
Hops, H., Davis, B., Leve, C., & Sheeber, L. (2003). Cross-generational transmission

io
of aggressive parent behavior: A prospective, mediational examination. Journal

at
of Abnormal Child Psychology, 31, 161–169. doi:10.1023/A:1022522224295

ci
Huesmann, L. R. (1988). An information processing model for the development of

so
aggression. Aggressive Behavior, 14, 13–24.
Huesmann, L. R. (1998). The role of social information processing and cognitive

As
schema in the acquisition and maintenance of habitual aggressive behavior. In
R. G. Geen & E. Donnerstein (Eds.), Human aggression: Theories, research, and

al
implications for social policy (pp. 73–109). San Diego, CA: Academic Press.

ic
og
Huesmann, L. R., Dubow, E. F., Eron, L. D., & Boxer, P. (2006). Middle childhood
family-contextual and personal factors as predictors of adult outcomes. In A. C.
ol
Huston & M. N. Ripke (Eds.), Developmental contexts in middle childhood: Bridges
ch
to adolescence and adulthood (pp. 62–86). New York, NY: Cambridge University
y

Press. doi:10.1017/CBO9780511499760.005
Ps

Huesmann, L. R., Dubow, E. F., & Boxer, P. (2009). Continuity of aggression from
childhood to early adulthood as a predictor of life outcomes: Implications for the
an

adolescent-limited and life-course-persistent models. Aggressive Behavior, 35,

ic

136–149. doi:10.1002/ab.20300
er

Huesmann, L. R., Eron, L. D., & Dubow, E. F. (2002). Childhood predictors of adult
Am

criminality: Are all risk factors reflected in childhood aggressiveness? Criminal

Behaviour and Mental Health, 12, 185–208. doi:10.1002/cbm.496
Huesmann, L. R., Eron, L. D., Lefkowitz, M. M., & Walder, L. O. (1984). The sta-
e
th

bility of aggression over time and generations. Developmental Psychology, 20,

1120–1134. doi:10.1037/0012-1649.20.6.1120
by

Huesmann, L. R., & Guerra, N. G. (1997). Children’s normative beliefs about aggres-
sion and aggressive behavior. Journal of Personality and Social Psychology, 72,
11

408–419. doi:10.1037/0022-3514.72.2.408
20

Huesmann, L. R., & Kirwil, L. (2007). Why observing violence increases the risk of vio-
lent behavior in the observer. In D. J. Flannery, A. T. Vazsonyi, & I. D. Waldman
©

(Eds.), The Cambridge handbook of violent behavior and aggression (pp. 545–570).
ht

Cambridge, England: Cambridge University Press.

ig

Hurley, S., & Chater, N. (Eds.). (2005). Perspectives on imitation: From neuroscience
yr

to social science. Cambridge, MA: The MIT Press.

op

Knoblich, G., & King, R. (1992). Biological correlates of criminal behavior. In

J. McCord (Ed.), Advances in criminological theory (Vol. 3, pp. 1–21). New
C

Brunswick, NJ: Transaction.

140 HUESMANN, DUBOW, AND BOXER

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 141

Kokko, K., & Pulkkinen, L. (2000). Aggression in childhood and long-term

unemployment in adulthood: A cycle of maladaptation and some protective fac-
tors. Developmental Psychology, 36, 463–472. doi:10.1037/0012-1649.36.4.463
Kokko, K., Pulkkinen, L., Huesmann, L. R., Dubow, E. F., & Boxer, P. (2009). Inten-

A)
sity of aggression in childhood as a predictor of different forms of adult aggres-
sion: A two-country (Finland and United States) analysis. Journal of Research on

P
Adolescence, 19, 9–34. doi:10.1111/j.1532-7795.2009.00579.x

(A
Loeber, R., & Dishion, T. (1983). Early predictors of male delinquency: A review.

n
Psychological Bulletin, 94, 68–99. doi:10.1037/0033-2909.94.1.68

io
MacEwen, K. E. (1994). Refining the intergenerational transmission hypothesis.

at
Journal of Interpersonal Violence, 9, 350–365. doi:10.1177/088626094009003005

ci
Meltzoff, A. N. (2007). Like me: A foundation for social cognition. Developmental

so
Science, 10, 226–134. doi:10.1111/j.1467-7687.2007.00574.x

As
Mednick, S. A., Gabrielli, W. F., Jr., & Hutchings, B. (1984, May). Genetic influ-
ences in criminal convictions: Evidence from an adoption cohort. Science, 224,

al
891–894. doi:10.1126/science.6719119

ic
Meltzoff, A. N., & Moore, M. K. (1983). Newborn infants imitate adult facial ges-

og
tures. Child Development, 54, 702–709. doi:10.2307/1130058
ol
Miles, D. R., & Carey, G. (1997). Genetic and environmental architecture of human
ch
aggression. Journal of Personality and Social Psychology, 72, 207–217. doi:10.1037/
y

0022-3514.72.1.207
Ps

Moffitt, T. E. (1993). Adolescence-limited and life-course-persistent antisocial behav-

ior: A developmental taxonomy. Psychological Review, 100, 674–701. doi:10.1037/
an

0033-295X.100.4.674
ic

Munafò, M. R., Clark, T. G., Moore, L. R., Payne, E., Walton, R., & Flint, J. (2003).
er

Genetic polymorphisms and personality in healthy adults: A systematic review

Am

and meta-analysis. Molecular Psychiatry, 8, 471–484. doi:10.1038/sj.mp.4001326

Olweus, D. (1979). The stability of aggressive reaction patterns in males: A review.
e

Psychological Bulletin, 86, 852–875. doi:10.1037/0033-2909.86.4.852

th

Olweus, D., Mattsson, A., Schaling, D., & Loew, H. (1988). Circulating testosterone
by

levels and aggression in adolescent males: A causal analysis. Psychosomatic Med-

icine, 50, 261–272.
11

Patterson, G. R. (1986). Performance models for antisocial boys. American Psychologist,

20

41, 432–444. doi:10.1037/0003-066X.41.4.432

Pulkkinen, L., Lyyra, A.-L., & Kokko, K. (2009). Life success of males on nonoffender,
©

adolescence-limited, persistent, and adult-onset antisocial pathways: Follow-up

ht

from age 8 to 42. Aggressive Behavior, 35, 117–135. doi:10.1002/ab.20297

ig

Pulkkinen, L., Nygren, H., & Kokko, K. (2002). Successful development: Childhood
yr

antecedents of adaptive psychosocial functioning in adulthood. Journal of Adult

Development, 9, 251–265. doi:10.1023/A:1020234926608
op

Raine, A., Brennan, P., & Mednick, S. A. (1995). Birth complications combined
C

with early maternal rejection at age 1 year predispose to violent crime at age

THE TRANSMISSION OF AGGRESSIVENESS 141

12051-08_PT2-CH07-rev2.qxd 5/25/10 11:38 AM Page 142

18 years. Obstetrical & Gynecological Survey, 50, 775–776. doi:10.1097/00006254-

199511000-00010
Raine, A., & Jones, F. (1987). Attention, autonomic arousal, and personality in
behaviorally disordered children. Journal of Abnormal Child Psychology, 15,

A)
583–599. doi:10.1007/BF00917243
Rönkä, A., Oravala, S., & Pulkkinen, L. (2002). “I met this wife of mine and things

P
(A
got on a better track”: Turning points in risk development. Journal of Adoles-
cence, 25, 47–63. doi:10.1006/jado.2001.0448

n
Rutter, M. (1996). Transitions and turning points in developmental psychopathology:

io
As applied to the age span between childhood and mid-adulthood. International

at
Journal of Behavioral Development, 19, 603–626. doi:10.1080/016502596385712

ci
Rutter, M. (1998). Some research considerations on intergenerational continuities

so
and discontinuities: Comment on the special section. Developmental Psychology,

As
34, 1269–1273. doi:10.1037/0012-1649.34.6.1269
Sameroff, A. J. (1983). Developmental systems: Contexts and evolution. In

al
W. Kessen (Ed.), Handbook of child psychology: Vol. 1. History, theory, and methods

ic
(pp. 237–294). New York, NY: Wiley.

og
Sameroff, A. J., Seifer, R., Baldwin, A., & Baldwin, C. (1993). Stability of intelli-
ol
gence from preschool to adolescence: The influence of social and family risk
ch
factors. Child Development, 64, 80–97. doi:10.2307/1131438
y

Serbin, L. A., Cooperman, J. M., Peters, P. L., Lehoux, P. M., Stack, D. M., &
Ps

Schwartzman, A. E. (1998). Intergenerational transfer of psychosocial risk

in women with childhood histories of aggression, withdrawal, or aggression
an

and withdrawal. Developmental Psychology, 34, 1246–1262. doi:10.1037/0012-

ic

1649.34.6.1246
er

Shaw, D. S. (2003). Advancing our understanding of intergenerational continu-

Am

ity in antisocial behavior. Journal of Abnormal Child Psychology, 31, 193–199.

doi:10.1023/A:1022578309274
e

Smith, C. A., & Farrington, D. P. (2004). Continuities in antisocial behavior and

th

parenting across three generations. Journal of Child Psychology and Psychiatry, 43,
230–247.
by

Thornberry, T. P., Freeman-Gallant, A., Lizotte, A. J., Krohn, M., & Smith, C. A.
11

(2003). Linked lives: The intergenerational transmission of antisocial behavior.

Journal of Abnormal Child Psychology, 31, 171–184. doi:10.1023/A:1022574208366
20

Werner, E. E. (2002). Looking for trouble in paradise: Some lessons learned from the
©

Kauai longitudinal study. In E. Phelps, F. F. Furstenberg, & A. Colby (Eds.), Look-

ing at lives: American longitudinal studies of the twentieth century (pp. 297–314).
ht

New York, NY: Russell Sage Foundation.

ig
yr
op
C

142 HUESMANN, DUBOW, AND BOXER

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 143

A)
P
GENETIC AND ENVIRONMENTAL

(A
INFLUENCES ON AGGRESSION

n
io
at
ci
SOO HYUN RHEE AND IRWIN D. WALDMAN

so
As
al
ic
og
ol
y ch
Ps
an

Considerable research has been conducted on the etiology of aggression

ic

and antisocial behavior, and it has relied on many different theoretical and
er

empirical approaches. Of these, behavior genetic methods have the advan-

Am

tages of testing competing alternative etiological models and of clearly dis-

tinguishing genetic from environmental influences and estimating their
e

relative magnitudes. More than 100 twin and adoption studies of antisocial
th

behavior have been published, which led us (Rhee & Waldman, 2002) to
by

conduct a meta-analysis of twin and adoption studies of antisocial behavior

in general and several specific operationalizations of antisocial behavior,
11

including aggression.
20

The role of familial influences on antisocial behavior has been studied

extensively. Dysfunctional familial influences, such as parental psychopathol-
©

ogy (e.g., Robins, 1966), coercive parenting styles (e.g., Patterson, Reid, &
ht

Dishion, 1992), physical abuse (Dodge, Bates, & Pettit, 1990), and family
ig
yr
op

This work was supported in part by NIDA DA-13956 to Soo Hyun Rhee and NIMH MH-01818 to
Irwin D. Waldman. Earlier versions of this chapter were presented at the meeting of the American
C

Society of Criminology in 1996 and the meeting of the Behavior Genetics Association in 1997, and a
more extensive version has been published in Psychological Bulletin, 128, pp. 490–529.

143
12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 144

conflict (e.g., Norland, Shover, Thornton, & James, 1979), are significantly
associated with antisocial behavior. Often, these variables are considered to
be environmental influences, and the possibility that they may also reflect
genetic influences is not considered. This is unfortunate because disentan-

A)
gling the influences of nature and nurture is the first step toward reaching the
eventual goal of explaining the specific etiology of antisocial behavior (see

P
(A
Chapter 7, this volume). Also, estimating the relative magnitude of genetic
and environmental influences on antisocial behavior is an important step in

n
the search for specific candidate genes and environmental risk factors under-

io
lying antisocial behavior. Although it is not possible to disentangle genetic

at
from environmental influences in family studies because genetic and environ-

ci
mental influences are confounded in nuclear families, twin and adoption

so
studies have the unique ability to disentangle genetic and environmental

As
influences and to estimate the magnitude of both simultaneously.
Twin studies can disentangle genetic and environmental influences on

al
a trait by comparing the similarity between monozygotic (MZ) twin pairs,

ic
who are genetically identical, with the similarity between dizygotic (DZ) twin

og
pairs, who share 50% of their genes, on average. Traits with genetic influences
ol
will show greater similarity between MZ twins than between DZ twins. Adop-
ch
tion studies demonstrate (a) genetic influences on a trait if there is a significant
y

correlation between the trait in adoptees and their biological relatives and
Ps

(b) environmental influences if there is a significant correlation between the

trait in adoptees and their adoptive parents or siblings.
an

In this chapter, we address the specific question of the magnitude of

ic

genetic and environmental influences on aggression (see also Chapters 7

er

and 9) and explore how the pattern and magnitude of these influences com-
Am

pare with those on antisocial behavior in general. In the twin and adoption
studies reviewed here, aggression is usually studied as a personality character-
e

istic and assessed with such measures as the Adjective Checklist (Gough &
th

Heilbrun, 1972) and the Multidimensional Personality Questionnaire (as cited

by

in Tellegen et al., 1988). The operationalizations of aggression have been

diverse, ranging from reports of negative affect (Partanen, Bruun, & Markkanen,
11

1966) to observations of the number of times a child hits a Bobo doll (Plomin,
20

Foch, & Rowe, 1981). For the present review, the operationalization of aggres-
sion was restricted to behavior that meets the Diagnostic and Statistical Manual
©

of Mental Disorders (4th ed.; American Psychiatric Association, 1994) criteria

ht

for conduct disorder (e.g., bullying, initiating physical fights, using a weapon
ig

that can cause serious physical harm). Toward the end of the chapter, we qual-
yr

itatively review the emerging behavior genetic literature on specific forms of

op

aggression in children, including reactive, proactive, and relational aggression.

We also briefly describe the burgeoning molecular genetic literature examin-
C

ing the association of aggression with relevant candidate genes.

144 RHEE AND WALDMAN

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 145

META-ANALYSIS OF BEHAVIOR GENETIC

STUDIES OF AGGRESSION

We began our search for twin and adoption studies of aggression by

A)
examining the PsycINFO and Medline databases. The search terms used in this
process were aggressive, aggression, crime, criminality, delinquent, or delinquency

P
(A
in combination with the terms twin(s), adoptee(s), adoptive, genetic, genetics,
genes, environmental, and environment. We examined the references cited in

n
research articles and review articles found in the search process, to locate any

io
additional studies that might have been missed or published before the databases

at
were established.

ci
so
Inclusion Criteria

As
After excluding unsuitable studies according to the criteria described

al
later (i.e., construct validity, inability to calculate tetrachoric or intraclass cor-

ic
relations, and assessment of related disorders) and addressing the problem of

og
nonindependence in these studies, 14 studies examining aggression remained.
ol
Table 8.1 lists the behavior genetic studies examining aggression included in
ch
the meta-analysis. Also listed are the method of assessment and method of zygos-
y

ity determination (in twin studies) used in each study, the mean or median age
Ps

of the sample, the sex of the participants, the number of pairs, the relationship
of the pairs, and the effect sizes.
an
ic

Construct Validity
er

Studies examining aggression were included if they examined behavioral

Am

aggression (e.g., physical fighting, cruelty to animals, bullying). Studies that

examined other related variables, such as anger, hostility, or impulsivity, were not
e
th

included because it was not clear whether they examined aggression or some
related but distinct trait. An additional study (Partanen et al., 1966) was excluded
by

because the aggression items used by the authors (e.g., “Are you readily insulted?”
and “Do you easily become unhappy about even small things?”) suggested that
11

negative affect or anger, rather than aggression per se, was being assessed.
20

Inability to Calculate Tetrachoric or Intraclass Correlations

©
ht

The effect sizes used in this meta-analysis were the Pearson product-
ig

moment or intraclass correlations that were reported in the studies, or the

yr

tetrachoric correlations that were estimated from the concordances or per-

op

centages reported in the studies. These effect sizes were analyzed using model-
fitting programs that estimate the relative contribution of genetic and
C

environmental influences and test the fit of alternative etiological models.

GENETIC AND ENVIRONMENTAL INFLUENCES 145

C
op
yr
ig
ht
©

146
12051-09_CH08-rev2.qxd

20
11
5/25/10

by
TABLE 8.1
Effect Sizes for Behavior Genetic Studies Examining Aggression
th
Study
e
Assessment Zygosity Age Sex N Relationship Effect size
11:38 AM

Midwest twins (Cates, Houston, Self-report Blood grouping/ 42.50 F–F 77 MZ .07 (assault)
Am

RHEE AND WALDMAN

Vavak, Crawford, & Uttley, 1993) questionnaire 21 DZ .41 (assault)
77 MZ .41 (verbal)
er
ic 21 DZ .06 (verbal)
77 MZ .40 (indirect)
Page 146

21 DZ .01 (indirect)
an
California twins (Ghodsian-Carpey Parent report Questionnaire 5.20 Both–both 21 MZ .78
& Baker, 1987) 17 DZ .31
Ps
London twins (adults, 1970s; Reaction to stimuli y 30.50 Both–both 49 MZ .59
Wilson, Rust, & Kasriel, 1977) 52 DZ .34
London twins (adults, 1980s; Self-report Blood grouping/ 30.00 M–M 90 MZ .33
ch
Rushton, Fulker, Neale, Nias, questionnaire
ol 46 DZ .16
& Eysenck, 1986) F–F 206 MZ .43
133 DZ .00
og
ic M–F 98 DZ .12
Minnesota twins (reared together, Self-report Blood grouping 21.65 Both–both 217 MZ .43
1970s; Tellegen et al., 1988) 114 DZ .14
al
Minnesota twins (1990s, adults; Self-report Blood grouping/ 37.76 M–M As 220 MZ .37
Finkel & McGue, 1997) questionnaire 165 DZ .12
F–F 406 MZ .39
so
352 DZ .14
M–F 114 DZ .12
ci
at
io
n
(A
P A)
C
op
yr
ig
ht
©
Boston twins (children; Scarr, 1966) Parent report Blood grouping 8.08 F–F 24 MZ .35
12051-09_CH08-rev2.qxd

20 28 DZ −.08
Philadelphia twins (Meininger, Teacher report Blood grouping 8.50 Both–both 61 MZ .67
Hayman, Coates, & Gallagher, 11 34 DZ .11
1988)
Missouri twins (Owen & Sines, 1970) Reaction to stimuli Blood grouping 10.00 M–M 10 MZ .09
5/25/10

11 DZ
by
−.24
th F–F 11 MZ .58
e 13 DZ .22
Colorado twins (1980s; Plomin, Objective test Questionnaire 7.60 Both–both 53 MZ .42
Foch, & Rowe, 1981) 32 DZ .42
11:38 AM

California twins (Rahe, Hervig, & Self-report Blood grouping 48.00 M–M 82 MZ .31
Am
Rosenman, 1978) 79 DZ .21
British Columbia twins (Blanchard, Self-report 36.18 Both–both 96 MZ .59
er
Vernon, & Harris, 1995)
ic 48 DZ .34
Page 147

Dutch twins (van den Oord, Verhulst, Parent report anBlood grouping 3.00 M–M 210 MZ .81
& Boomsma, 1996) (questionnaire 265 DZ .49
F–F 236 MZ .83
238 DZ .49
Ps
y M–F 409 DZ .45
Swedish Twins (adults; Gustavsson, Self-report ch Both–both 15 MZ ra .22 (indirect)
Pedersen, Åsberg, & Schalling, 26 MZ .41 (indirect)
1996) 29 DZ .27 (indirect)
ol
og 15 MZ ra −.03 (verbal)
26 MZ .22 (verbal)
29 DZ .23 (verbal)
ic
Note. M = male; F = female; Both = both male and female; MZ = monozygotic twin pairs; DZ = dyzygotic twin pairs; MZ ra = monzygotic twin pairs reared apart.
al

GENETIC AND ENVIRONMENTAL INFLUENCES

As
so
ci

147
at
io
n
(A
P A)
12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 148

Nonindependent Samples
Another justification for exclusion from the meta-analysis was nonin-
dependent sampling. Several effect sizes from studies in the original reference
list were based on nonindependent samples because researchers examined

A)
more than one dependent measure of antisocial behavior in their sample or

P
published follow-up data from the same sample in separate publications. Sev-

(A
eral suggestions for dealing with nonindependent samples have been offered
in the meta-analytic literature (Mullen, 1989; Rosenthal, 1991). For exam-

n
io
ple, Mullen (1989) gave four options for dealing with this problem: choosing

at
the best dependent measure, averaging the effect sizes of the different depen-

ci
dent measures, conducting separate meta-analyses for each of the dependent

so
measures, or using nonindependent samples as if they were independent sam-
ples (the least recommended approach). We did not follow the option of

As
choosing the best dependent measure unless one of the dependent measures

al
did not fulfill the inclusion criteria described previously, making the decision

ic
easy. Taking this option would have required making subjective choices

og
because we were aware of the effect sizes associated with each of the depen-
dent measures. The option of conducting separate meta-analyses for each of
ol
the dependent measures was not chosen simply because it was impractical:
ch
There were many effect sizes from nonindependent samples. The most prac-
y

tical and prudent option was to average the effect sizes from nonindependent
Ps

samples.
an

Given that model-fitting analyses require specification of the sample

size, we used the option of averaging multiple effect sizes in cases where the sam-
ic

ple size was identical across the nonindependent samples. If the sample size was
er

not identical across the nonindependent samples, we used the effect size from
Am

the largest sample. More specifically, in cases of nonindependence where the

same dependent measure was used in the same sample multiple times (e.g., in
e

follow-up analyses), we chose the effect size estimated from the largest sam-
th

ple. In cases of nonindependence in which different dependent measures were

by

used in the same sample (e.g., when the author of one publication examined
more than one dependent measure or authors of different publications exam-
11

ined different dependent measures in one sample), the effect sizes were aver-
20

aged if the sample size was the same across the nonindependent samples, and
the effect size from the largest sample was used if the sample size differed across
©

the nonindependent samples.

ht
ig

Determination of Effect Sizes

yr
op

Some adoption and twin studies used a continuous variable to measure

antisocial behavior and reported either Pearson product-moment or intraclass
C

correlations, which were the effect sizes used from these studies in the meta-

148 RHEE AND WALDMAN

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 149

analysis. In other studies, a dichotomous variable was used, and concordances,

percentages, or a contingency table (including the number of twin pairs with
both members affected, one member affected, and neither member affected)
were reported. The information from the concordances or percentages was

A)
transformed into a contingency table, which was then used to estimate the
tetrachoric correlation (i.e., the correlation between the latent continuous

P
(A
variables that are assumed to underlie the observed dichotomous variables),
which was the effect size used in the meta-analysis. For some studies, we directly

n
estimated the tetrachoric correlation from the raw data because it had to be esti-

io
mated from contingency tables. For these studies, we were also able to estimate

at
the weight matrix (i.e., the asymptotic covariance matrix of the correlation

ci
matrix). If the weight matrix can be estimated, it is possible to use weighted

so
least squares estimation in the model-fitting analyses, which is more appropri-

As
ate for nonnormally distributed variables (e.g., diagnoses) than maximum
likelihood estimation.

al
ic
Model-Fitting Analyses

og
ol
The magnitude of additive and nonadditive genetic influences (a2 and
ch
2
d ) constitutes the proportion of variance in the liability for aggression that
y

is due to genetic differences among individuals. If genetic influences are addi-

Ps

tive, the effects of alleles from different loci are independent and “add up” to
influence the liability underlying a trait. If genetic influences are nonadditive,
an

the alleles interact with each other to influence the liability for a trait, either
ic

at a single genetic locus (i.e., dominance) or at different loci (i.e., epistasis).

er

Many studies report the magnitude of additive and nonadditive genetic influ-
Am

ences combined, and this estimate of broad-sense heritability is symbolized

by h2. Shared environmental influences (c2) represent the proportion of lia-
e

bility variance that is due to environmental influences that are experienced

th

in common and make family members similar to one another, whereas non-
by

shared environmental influences (e2) represent the proportion of liability

variance that is due to environmental influences that are experienced uniquely
11

and make family members different from one another. It is important to note
20

that estimates of nonshared environmental influences also include measure-

ment error.
©

It is customary in contemporary behavior genetic analyses to compare

ht

alternative models containing different sets of causal influences for their fit to
ig

the observed data (i.e., twin or familial correlations or covariances). These

yr

models posit that aggression is caused by the types of influences described pre-
op

viously: additive genetic influences (A), nonadditive genetic influences (D),

shared environmental influences (C), and nonshared environmental influ-
C

ences (E). In the present meta-analysis, we contrasted the fit of the ACE

GENETIC AND ENVIRONMENTAL INFLUENCES 149

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 150

model, the AE model, the CE model, and the ADE model. We assessed the
fit of each model, as well as of competing models, using both the chi-square sta-
tistic and the Akaike information criterion (AIC), a fit index that reflects both
the fit of the model and its parsimony (Loehlin, 1992). The AIC has been used

A)
extensively in both the structural equation modeling and behavior genetic lit-
eratures. Among competing models, that with the lowest AIC and the lowest

P
χ2 relative to its degrees of freedom is considered to be the best-fitting model.

(A
Unfortunately, it is not possible to estimate c2 and d2 simultaneously or

n
test an ACDE model with data only from twin pairs reared together because

io
the estimation of c2 and d2 both rely on the same information (i.e., the differ-

at
ence between the MZ and DZ twin correlations). If the DZ correlation is

ci
greater than half the MZ correlation, the ACE model is the correct model,

so
and the estimate of d2 in the ADE model is always zero. If the DZ correlation

As
is less than half the MZ correlation, however, the ADE model is the correct
model, and the estimate of c2 in the ACE model is always zero.

al
ic
Meta-Analytic Findings

og
ol
The ACE model was the best-fitting model for aggression (a2 = .44, c2 =
ch
.06, e = .50), although the fit of this model was close to that of the AE model,
2
y

and the magnitude of shared environmental influences on aggression is mod-

Ps

est. Models that included nonadditive genetic influences instead of shared

environmental influences or that omitted additive genetic influences fit less
an

well. By comparison, in our omnibus meta-analysis of behavior genetic stud-

ic

ies of antisocial behavior there were moderate additive genetic (a2 = .32),
er

nonadditive genetic (d2 = .09), shared environmental (c2 = .16), and non-
Am

shared environmental (e2 = .43) influences. These results suggest that the
magnitude of genetic and nonshared environmental influences is slightly
e

higher for aggression than for antisocial behavior in general and that evidence
th

for the role of shared environmental influences on aggression is at best tenta-

by

tive pending future studies.

11

New Directions in Behavior Genetic Studies of Aggression

20

Unfortunately, we were unable to examine a meaningful distinction

©

between overt and relational aggression (Crick, Casa, & Mosher, 1997; Crick
ht

& Grotpeter, 1995) because there are no published twin or adoption studies
ig

of relational aggression. Overt aggression harms others through physical dam-

yr

age or the threat thereof, whereas relational aggression harms others by dam-
op

aging their peer relationships or reputation (e.g., spreading rumors, excluding

them from a peer group). Although relational aggression does not physically
C

harm the victims, it has serious consequences for both the aggressors (e.g.,

150 RHEE AND WALDMAN

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 151

higher levels of loneliness, depression, and negative self-perceptions, as well

as concurrent and future peer rejection; Crick & Grotpeter, 1995) and the
victims (e.g., depression, anxiety; Crick & Grotpeter, 1996). The distinction
between overt and relational aggression is especially important when exam-

A)
ining sex differences in aggression and its causes, given that females are sig-
nificantly more relationally aggressive and less overtly aggressive than males

P
(A
(Crick & Grotpeter, 1995; Crick et al., 1997). Given the evidence that overt
and relational aggression are correlated but distinct (Crick et al., 1997),

n
behavior genetic studies are necessary to determine the degree of genetic and

io
environmental influences that are common to both overt and relational

at
aggression rather than being specific to each.

ci
Similarly, few behavior genetic studies have distinguished between reac-

so
tive and proactive aggression (Dodge, Lochman, Harnish, Bates, & Pettit,

As
1997; Vitaro, Brendgen, & Tremblay, 2002; Waschbusch, Willoughby, &
Pelham, 1998). Reactive aggression is characterized by impulsive “hot-blooded”

al
anger, appears to be a response to frustration or perceived threat, and is asso-

ic
ciated with lack of self-control. In contrast, proactive aggression is premedi-

og
tated and “cold-blooded,” less emotional, and more likely to be driven by
ol
expected rewards (Dodge et al., 1997). Evidence suggests that reactively and
ch
proactively aggressive children differ in developmental histories, adjustment,
y

and social information-processing patterns and that reactive and proactive

Ps

aggression are distinct (Dodge et al., 1997; Vitaro et al., 2002; see also Chap-
ter 9, this volume). Therefore, behavior genetic studies of antisocial behavior
an

should distinguish between reactive and proactive aggression, and multi-

ic

variate studies should examine the etiology of the overlap between different
er

forms of antisocial behavior. Genetically informative studies of aggression

Am

have recently made important strides along these lines. In addition, several
behavior genetic studies examining the development of aggression have been
e

conducted.
th
by

The Overlap Between Different Forms of Aggression

11

Two genetically informative studies have examined the common and

20

unique etiologies of proactive and reactive aggression, with differing results.

Brendgen, Vitaro, Boivin, Dionne, and Pérusse (2006) examined the associ-
©

ation between proactive and reactive aggression in a sample of 172 six-year-

ht

old twin pairs from Quebec. The magnitude of genetic influences was similar
for proactive (h2 = .41, e2 = .59) and reactive aggression (h2 = .39, e2 = .61).
ig
yr

The correlation between proactive and reactive aggression ranged from .51
op

to .60 in this sample, with the correlation between genetic influences on the
two types of aggression being .87, and the correlation between nonshared
C

environmental influences being .34. Brendgen et al. (2006) also found that

GENETIC AND ENVIRONMENTAL INFLUENCES 151

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 152

most of the association between proactive and reactive aggression was due to
common etiological factors influencing physical aggression.
The second study examining proactive and reactive aggression was
Baker, Raine, Liu, and Jacobson’s (2008) study of 1,219 nine- to 10-year-old

A)
twins from southern California. Results were presented separately for child,
mother, and teacher report, because the correlation between these sources

P
(A
was low (.18 to .26). Baker et al. found a significant sex difference in the mag-
nitude of genetic and environmental influences on aggression according to

n
child reports, such that moderate genetic influences were present for boys but

io
not girls, whereas moderate shared environmental influences were present

at
for girls but not boys (boys: h2 = .38 and e2 = .62 for reactive aggression and

ci
h2 = .50 and e2 = .50 for proactive aggression; girls: c2 = .36 and e2 = .64 for

so
reactive aggression and c2 = .14 and e2 = .86 for proactive aggression). In

As
contrast, no differences were found for mother reports (h2 = .26, c2 = .27, and
e2 = .46 for reactive aggression and h2 = .32, c2 = .21, and e2 = .47 for proac-

al
tive aggression) or teacher reports (h2 = .20, c2 = .43, and e2 = .37 for reactive

ic
aggression and h2 = .45, c2 = .14, and e2 = .41 for proactive aggression). The

og
phenotypic correlation between proactive and reactive aggression ranged
ol
from .46 to .80. The authors suggest that both common genetic and environ-
ch
mental influences are responsible for this correlation, but results varied across
y

the three sources (child report: rg = .57 and re = .46 for boys, rc = .53 and re =
Ps

.53 for girls; mother report: rg = .76, rc = .76, re = .43; teacher report: rg = 1.0,
rc = 1.0, re = .53).
an

As stated previously, another meaningful distinction between forms of

ic

aggression is that between overt and relational aggression (Crick & Grotpeter,
er

1995; Crick et al., 1997). Two recent twin studies have examined the associ-
Am

ation between relational and overt aggression. Brendgen et al. (2005) exam-
ined the association between physical aggression (i.e., overt aggression) and
e

social aggression (i.e., relational aggression) in 234 six-year-old twin pairs

th

from Quebec. Teacher and peer reports of physical and social aggression were
by

obtained, and the magnitude of genetic and environmental influences was

similar for the two sources (teacher report: h2 = .63, e2 = .37 for physical
11

aggression; h2 = .20, c2 = .20, e2 = .60 for social aggression; peer report: h2 = .54,
e2 = .46 for physical aggression; h2 = .23, c2 = .23, e2 = .54 for social aggres-
20

sion). The phenotypic correlation between physical and social aggression was
©

.43 for teacher ratings and .41 for peer ratings, and there was evidence of com-
ht

mon genetic and nonshared environmental influences on the two types of

aggression (teacher report: rg = .79, re = .31; peer report: rg = 1.0, re = .12).
ig
yr

Ligthart et al. (2005) examined relational versus direct (i.e., overt) aggres-
op

sion in 7,449 seven-year-old Dutch twin pairs. They found evidence of

genetic, shared environmental, and nonshared environmental influences on
C

both relational and direct aggression and found evidence of significant sex dif-

152 RHEE AND WALDMAN

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 153

ferences in the magnitude of genetic and environmental influences for direct

aggression, such that genetic influences were stronger for girls and shared
environmental influences were stronger for boys (relational aggression, boys
and girls combined: h2 = .66, c2 = .16, e2 = .18; direct aggression: h2 = .53,
c2 = .23, e2 = .24 for boys; h2 = .60, c2 = .13, e2 = .27 for girls). The phenotypic

A)
correlation between relational and direct aggression was .58 for boys and .47

P
(A
for girls, and this correlation was due to both common genetic and shared
environmental influences (55% genetic, 33% shared environmental, and 12%

n
nonshared environmental influences in boys; 58% genetic, 30% shared envi-

io
ronmental, and 12% nonshared environmental influences in girls).

at
ci
The Overlap Between Aggression and

so
Other Types of Antisocial Behavior

As
Several studies have examined differences in the etiology of aggressive

al
and nonaggressive antisocial behavior. As shown in Table 8.2, although the

ic
results from these studies are not uniform, several general conclusions can

og
be drawn from them. First, although most studies reported similar, moder-
ol
ate heritabilities for both types of antisocial behavior, some studies (e.g.,
ch
Eley, Lichtenstein, & Stevenson, 1999) suggest that aggressive behavior is
y

more heritable than nonaggressive antisocial behavior. This conclusion is

Ps

borne out by findings from a recent meta-analysis of 19 studies of aggressive

and 15 studies of nonaggressive antisocial behavior (Burt, 2009). In this
an

meta-analysis, the etiology of aggression included additive genetic and non-

ic

shared environmental influences but no evidence for shared environ-

er

mental influences (h2 = .65, c2 = .00, e2 = .35), whereas the etiology of

Am

nonaggressive antisocial behavior also included shared environmental

influences and showed a lower magnitude of genetic influences (h2 = .48,
e

c2 = .18, e2 = .34).
th

Second, the phenotypic correlation between aggressive and nonaggres-

by

sive antisocial behavior was moderate in most studies (e.g., .32 in Gelhorn
et al., 2006, and .48 to .76 in Bartels et al., 2003). Third, most studies suggest
11

that there are significant common genetic influences on the covariance

20

between aggressive behavior and nonaggressive antisocial behavior. Fourth,

there was phenotypic continuity between aggressive and nonaggressive
©

antisocial behavior from childhood to adolescence (r = .28 to .61 in Eley,

ht

Lichtenstein, & Moffitt, 2003), which also had genetic influences. However,
ig

it is possible that some of this phenotypic continuity is due to shared method

yr

variance, because the correlation between aggressive behavior in child-

op

hood assessed via parent report and nonaggressive antisocial behavior in ado-
lescent assessed via self-report was lower (r = .07 to .15 in Tuvblad, Eley, &
C

Lichtenstein, 2005).

GENETIC AND ENVIRONMENTAL INFLUENCES 153

C
op
yr
ig
ht
©

154
12051-09_CH08-rev2.qxd

20
11
5/25/10

by
th
e
11:38 AM

TABLE 8.2
Am

RHEE AND WALDMAN

Results of Studies Examining Aggressive and Nonaggressive Antisocial Behavior
er
Operationalization/ Etiology of Aggressive
ic Etiology of Nonaggressive Etiology of Phenotypic
Page 154

Study Assessment Behavior an Behavior Covariance

Bartels (2003) Parent report of aggressive/ h2 = .69, c2 = .00, h2 = .79, c2 = .00, h2 = .80, c2 = .02-.06,
rule breaking behavior e2 = .31 (males) e2 = .21 (males) e2 = .12-.14
Ps
h2 = .72, c2 = .00,y h2 = .56, c2 = .23,
e2 = .28 (females) e2 = .21 (females)
Button (2004) Self-report of aggressive h2 = .58-.68
ch h2 = .59-.74 h2 = .87, c2 = .00, e2 = .13
and nonaggressive c2 = .00-.09 ol c2 = .00-.14
antisocial behavior e2 = .32-.33 e2 = .26-.27
Eley (1999) Parent report of aggressive h2 = .55, c2 = .18,
og
h2 = .30, c2 = .44, h2 = .38, c2 = .53,
and nonaggressive e2 = .27 (Swedish males) e2 = .26 (Swedish males)
ic e2 = .09 (Swedish males)
behavior 2
h = .42, c2 = .25, 2 2
h = .13, c2 = .54,
al h = .25, c2 = .56,
e2 = .33 (British males) e2 = .32 (British males) e2 = .19 (British males)
2
h = .76, c2 = .06, 2 2
h = .45, c2 = .34,
As h = .72, c2 = .18,
e2 = .19 (Swedish e2 = .22 (Swedish e2 = .10 (Swedish
females) females) females)
so
h2 = .71, c2 = .04, h2 = .42, c2 = .30, h2 = .88, c2 = .06,
e2 = .25 (British females) e2 = .28 (British females) e2 = .06 (British females)
ci
at
io
n
(A
P A)
C
op
yr
ig
ht
Eley (2003)
©
Parent report of aggressive h2 = .60, c2 = .15, h2 = .49, c2 = .35, h2 = .70, c2 = .28,
nonaggressive antisocial
12051-09_CH08-rev2.qxd

20 e2 = .24 (time 1) e2 = .16 (time 1) e2 = .02 (aggressive at

behavior at two time h2 = .46, c2 = .30, h2 = .44, c2 = .42, time 1 and non-
points (childhood and
11 e2 = .24 (time 2) e2 = .14 (time 2) aggressive at time 2)
adolescence) h2 = .69, c2 = .27,
e2 = .04 (non-aggressive
5/25/10

at time 1 and aggressive

by
th at time 2)
Gelhorn (2006) Self-report of aggressive e h2 = .49, c2 = .00, e2 = .51 h2 = .55, c2 = .00, e2 = .45 h2 = .61, c2 = .00, e2 = .39
and nonaggressive
conduct disorder
11:38 AM

symptoms
Am
Tuvblad (2005) Parent report of aggressive h2 = .59, c2 = .18, h2 = .37, c2 = .30, h2 = 1.00 (girls)
behavior in childhood e2 = .23 (girls) e2 = .33 (girls)
er
and self-report of non- h2 = .67, c2 = .04,
ic h2 = .27, c2 = .43, h2 = .39, c2 = .47,
Page 155

aggressive behavior in e2 = .29 (boys)

an e2 = .30 (boys) e2 = .14 (boys)
adolescence
Note. h2 = heritability, or magnitude of genetic influences; c2 = magnitude of shared environmental influences; e2 = magnitude of nonshared environmental influences.
Ps
y ch
ol
og
ic
al

GENETIC AND ENVIRONMENTAL INFLUENCES

As
so
ci

155
at
io
n
(A
P A)
12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 156

Behavior Genetic Studies of the Development of Aggression

Several researchers have examined the stability of aggressive behavior

in behavior genetic studies. The overall conclusion is that aggressive behav-

A)
ior is moderately stable from childhood to adolescence (for a similar conclu-
sion, see Chapter 7) and that genetic influences explain a larger percentage

P
(A
of the stability of aggressive behavior than do shared or nonshared environ-
mental influences. Van der Valk, Verhulst, Neale, and Boomsma (1998) con-

n
ducted a longitudinal study of aggressive behavior in 111 pairs of adopted

io
biological siblings, 221 pairs of adopted nonbiological siblings, and 1,484

at
adopted singletons from the Netherlands. Aggressive behavior was assessed

ci
via parent questionnaires at age 10 to 15 years, then again 3 years later. The

so
correlation between the two time points ranged from .51 to .70, and the

As
covariance between the two time points was due 69% to genetic influences,
14% to shared environmental influences, and 17% to nonshared environ-

al
mental influences. Van Beijsterveldt, Bartels, Hudziak, and Boomsma (2003)

ic
examined the stability of aggression at ages 3, 7, 10, and 12 in a large sample

og
of Dutch twin pairs (ranging from 1,509 pairs at age 12 to 6,488 pairs at age 3).
ol
Aggressive behavior was assessed via parent questionnaires. The correlations
ch
for aggression assessed at different ages ranged from .41 to .77. Genetic influ-
y

ences explained approximately 65% of the total stability of aggression, whereas

Ps

shared environmental influences accounted for approximately 25% of the

total stability of aggression. A simplex model, wherein a dynamic developmen-
an

tal process consisting of transmission of existing influences interacting with

ic

new influences, fit best for genetic influences. In contrast, there was a stable
er

set of the same shared environmental influences and age-specific nonshared

Am

environmental influences.
In a longitudinal twin study, Haberstick, Schmitz, Young, and Hewitt
e

(2006) examined the development of aggressive behavior from childhood to

th

early adolescence (through parent report at ages 7, 9, 10, 11, and 12 and
by

teacher report at ages 7, 8, 9, 10, 11, and 12) and reached conclusions slightly
different from those of van Beijsterveldt et al. (2003). The contribution of
11

common genetic influences to the stability of aggression ranged from 66% to

20

87% for parent report and 59% to 95% for teacher report, and the rest was
attributable to nonshared environmental influences that were common across
©

age. There was no evidence of significant shared environmental influences

ht

that were common across age. The authors noted that there is consistent evi-
ig

dence of common genetic influences on aggression assessed at different ages

yr

despite the fact that different teachers rated the children at each age. Eley,
op

Lichtenstein, and Moffitt (2003) examined over 1,000 twin pairs from the
Swedish Twin Registry assessed at ages 8 to 9 and 13 to 14 years. Aggressive
C

behavior was assessed via parent report at both time points. Continuity in

156 RHEE AND WALDMAN

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 157

aggression between childhood and adolescence ranged from .53 to .69 and
was due 84% to genetic influences, 8% to shared environmental influences,
and 8% to nonshared environmental influences.

A)
Candidate Genes for Aggression and Antisocial Behavior

P
(A
Based on the evidence that aggression is heritable, researchers have ini-
tiated attempts to find specific genes that contribute to its etiology using a

n
candidate gene approach. In well-designed candidate gene studies, genes are

io
selected based on the known or hypothesized involvement of their gene prod-

at
uct in the etiology of a trait or disorder (i.e., its pathophysiological function

ci
and etiological relevance). Whereas genome scans may be thought of as

so
exploratory searches for putative genes that underlie a disorder or trait, well-

As
conducted candidate gene studies are targeted tests of the role of specific
genes in the etiology of a disorder or trait, as the location, function, and eti-

al
ological relevance of candidate genes are most often known a priori.

ic
Genes underlying various aspects of the dopaminergic, noradrenergic,

og
and serotonergic neurotransmitter pathways represent viable candidates
ol
given the role of these neurotransmitter systems in the etiology and patho-
ch
physiology of aggression. For example, there is considerable overlap between
y

antisocial behavior and childhood attention-deficit hyperactivity disorder

Ps

(ADHD; e.g., Lilienfeld & Waldman, 1990); thus, candidate genes for
ADHD may also be relevant candidates for aggression and antisocial behav-
an

ior. Several genes within the dopamine system appear to be risk factors for
ic

ADHD (Waldman & Gizer, 2006). Dopamine genes are plausible candi-
er

dates for ADHD, given that the stimulant medications that are the most
Am

frequent and effective treatments for ADHD appear to act primarily by reg-
ulating dopamine levels in the brain (Seeman & Madras, 1998; Solanto,
e

1984), although they also affect noradrenergic and serotonergic function

th

(Solanto, 1998). In addition, knock-out gene studies in mice, which exam-

by

ine the behavioral effects of the deactivation of specific genes, have further
demonstrated the potential relevance of genes within these neurotransmit-
11

ter systems. Results of such studies have markedly strengthened the consid-
20

eration as candidate genes for ADHD of dopaminergic genes, such as the

dopamine transporter gene (DAT1; Giros, Jaber, Jones, Wightman, & Caron,
©

1996) and the dopamine receptor D3 and D4 genes (DRD3 and DRD4;
ht

Accili et al., 1996; Dulawa, Grandy, Low, Paulus, & Geyer, 1999; Rubin-
ig

stein et al., 1997), as well as genes within the serotonergic system, such as
yr

the serotonin 1β receptor gene (HTR1β; Saudou et al., 1994). Serotoner-

op

gic genes also are plausible candidates for aggression, given the demon-
strated relations between serotonergic function and aggression (Berman,
C

Kavoussi, & Coccaro, 1997).

GENETIC AND ENVIRONMENTAL INFLUENCES 157

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 158

Although a comprehensive review of molecular genetic studies of aggres-

sion and antisocial behavior is beyond the scope of this chapter, several lines
of research suggest an association between serotonin and aggression. Several
researchers have found lower cerebrospinal fluid levels of 5-hydroxyindoleacetic

A)
acid, a serotonin metabolite, in aggressive or violent individuals (e.g., Brown,
Goodwin, Ballenger, Goyer, & Major,1979; Linnoila et al., 1983). Mice lack-

P
(A
ing the HTR1β gene show enhanced aggressive behavior (Saudou et al.,
1994), and a serotonin transporter (5HTT) polymorphism is associated with

n
aggression in nonhuman primates who experienced insecure early attach-

io
ment relationships (Suomi, 2003).

at
Candidate genes for neurotransmitter systems may include (a) precursor

ci
genes that affect the rate at which neurotransmitters are produced from pre-

so
cursor amino acids (e.g., tyrosine hydroxylase for dopamine, tryptophan

As
hydroxylase for serotonin), (b) receptor genes that are involved in receiving
neurotransmitter signals (e.g., genes corresponding to the five dopamine recep-

al
tors, DRD1, D2, D3, D4, and D5, and to the serotonin receptors, such as

ic
HTR1β and HTR2A), (c) transporter genes that are involved in the reuptake

og
of neurotransmitters back into the presynaptic terminal (e.g., the dopamine
ol
and serotonin transporter genes, DAT1 and 5HTT), (d) metabolite genes that
ch
are involved in the metabolism or degradation of these neurotransmitters
y

(e.g., the genes for catechol-o-methyl-transferase and for monoamine oxidase

Ps

A and B), and (e) genes that are responsible for the conversion of one neuro-
transmitter into another (e.g., dopamine beta hydroxylase, or DβH, which
an

converts dopamine into norepinephrine). We anticipate that there will be a

ic

steep increase in the number of studies of the involvement of such candidate

er

genes in aggression and antisocial behavior over the next decade.

Am
e

CONCLUSIONS
th
by

In conclusion, the results of a meta-analysis suggest that there are mod-

erate additive genetic and nonshared environmental influences and modest
11

shared environmental influences on aggression. Behavior genetic studies exam-

20

ining relational versus overt aggression and reactive versus proactive aggression
have recently been undertaken. The first set of these studies demonstrates
©

moderate heritabilities and nonshared environmental influences on these

ht

aggression dimensions, with genetic influences contributing substantially to

ig

their overlap. Shared environmental influences were also found to underlie

yr

some of these dimensions of aggression, particularly social or relational aggres-

op

sion. An association between dopamine and serotonin and aggression has been
implicated in some early candidate gene studies, although human studies
C

examining the association between the serotonin transporter gene and vio-

158 RHEE AND WALDMAN

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 159

lence or aggression have yielded conflicting results. Future molecular genetic

studies will illuminate the specific genetic underpinnings of aggression and
antisocial behavior.

A)
REFERENCES

P
(A
Note: References marked with an asterisk indicate studies included in

n
the meta-analysis.

io
at
Accili, D., Fishburn, C. S., Drago, J., Steiner, H., Lachowicz, J. E., Park, B. H.,

ci
. . . Fuchs, S. (1996). A targeted mutation of the D3 dopamine receptor gene is

so
associated with hyperactivity in mice. Proceedings of the National Academy of Sci-

As
ences of the United States of America, 93, 1945–1949. doi:10.1073/pnas.93.5.1945
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental

al
disorders (4th ed.). Washington, DC: Author.

ic
Baker, L. A., Raine, A., Liu, J., & Jacobson, K. C. (2008). Differential genetic and envi-

og
ronmental influences on reactive and proactive aggression in children. Journal of
Abnormal Child Psychology, 36, 1265–1278. doi:10.1007/s10802-008-9249-1
ol
ch
Bartels, M., Hudziak, J. J., van den Oord, E. J., van Beijsterveldt, C. E., Rietveld, M. J.,
& Boomsma, D. I. (2003). Co-occurrence of aggressive behavior and rule-
y

breaking behavior at age 12: Multi-rater analyses. Behavior Genetics, 33, 607–621.
Ps

doi:10.1023/A:1025787019702
an

Berman, M. E., Kavoussi, R. J., & Coccarro, E. F. (1997). Neurotransmitter correlates

of human aggression. In D. M. Stoff, J. Breiling, & J. D. Maser (Eds.), Handbook
ic

of antisocial behavior (pp. 305–313). New York, NY: Wiley.

er

*Blanchard, J. M., Vernon, P. A., & Harris, J. A. (1995). A behavior genetic inves-
Am

tigation of multiple dimensions of aggression. Behavior Genetics, 25, 256.

Brendgen, M., Dionne, G., Girard, A., Boivin, M., Vitaro, F., & Pérusse, D. (2005).
e

Examining genetic and environmental effects on social aggression: A study of

th

6-year-old twins. Child Development, 76, 930–946. doi:10.1111/j.1467-8624.

by

2005.00887.x
Brendgen, M., Vitaro, F., Boivin, M., Dionne, G., & Pérusse, D. (2006). Examining
11

genetic and environmental effects on reactive versus proactive aggression.

20

Developmental Psychology, 42, 1299–1312. doi:10.1037/0012-1649.42.6.1299

Brown, G. L., Goodwin, F. K., Ballenger, J. C., Goyer, P. F., & Major, L. F. (1979).
©

Aggression in humans correlates with cerebrospinal fluid amine metabolites.

ht

Psychiatry Research, 1, 131–139. doi:10.1016/0165-1781(79)90053-2

ig

Burt, S. A. (2009). Are there meaningful etiological differences within antisocial

yr

behavior? Results of a meta-analysis. Clinical Psychology Review, 29. 163–178.

doi:10.1016/j.cpr.2008.12.004
op

Button, T. M., Scourfield, J., Martin, N., & McGuffin, P. (2004). Do aggressive and
C

nonaggressive antisocial behavior in adolescents result from the same genetic

GENETIC AND ENVIRONMENTAL INFLUENCES 159

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 160

and environmental effects? American Journal of Medical Genetics, 129B, 59–63.

doi:10.1002/ajmg.b.30045
*Cates, D. S., Houston, B. K., Vavak, C. R., Crawford, M. H., & Uttley, M. (1993).
Heritability of hostility-related emotions, attitudes, and behaviors. Journal of

A)
Behavioral Medicine, 16, 237–256. doi:10.1007/BF00844758
Crick, N. R., Casa, J. F., & Mosher, M. (1997). Relational and overt aggression in pre-

P
(A
school. Developmental Psychology, 33, 579–588. doi:10.1037/0012-1649.33.4.579
Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social–

n
psychological adjustment. Child Development, 66, 710–722. doi:10.2307/1131945

io
at
Crick, N. R., & Grotpeter, J. K. (1996). Children’s treatment by peers: Victims of
relational and overt aggression. Development and Psychopathology, 8, 367–380.

ci
doi:10.1017/S0954579400007148

so
Dodge, K. A., Bates, J., & Pettit, G. S. (1990, December). Mechanisms in the cycle

As
of violence. Science, 250, 1678–1683. doi:10.1126/science.2270481

al
Dodge, K. A., Lochman, J. E., Harnish, J. D., Bates, J. E., & Pettit, G. S. (1997). Reac-
tive and proactive aggression in school children and psychiatrically impaired

ic
og
chronically assaultive youth. Journal of Abnormal Psychology, 106, 37–51.
doi:10.1037/0021-843X.106.1.37
ol
Dulawa, S. C., Grandy, D. K., Low, M. J., Paulus, M. P., & Geyer, M. A. (1999).
ch
Dopamine D4 receptor-knock-out mice exhibit reduced exploration of novel
y

stimuli. The Journal of Neuroscience, 19, 9550–9556.

Ps

Eley, T. C., Lichtenstein, P., & Moffitt, T. E. (2003). A longitudinal behavioral genetic
analysis of the etiology of aggressive and nonaggressive antisocial behavior. Devel-
an

opment and Psychopathology, 15, 383–402. doi:10.1017/S095457940300021X

ic

Eley, T. C., Lichtenstein, P., & Stevenson, J. (1999). Sex differences in the etiology
er

of aggressive and nonaggressive antisocial behavior: Results from two twin stud-
Am

ies. Child Development, 70, 155–168. doi:10.1111/1467-8624.00012

*Finkel, D., & McGue, M. (1997). Sex differences and nonadditivity in heritability
e

of the multidimensional personality questionnaire scales. Journal of Personality

th

and Social Psychology, 72, 929–938. doi:10.1037/0022-3514.72.4.929

by

Gelhorn, H., Stallings, M., Young, S., Corley, R., Rhee, S. H., Christian, H., & Hewitt,
J. (2006). Common and specific genetic influences on aggressive and nonaggres-
11

sive conduct disorder domains. Journal of the American Academy of Child and Ado-
20

lescent Psychiatry, 45, 570–577. doi:10.1097/01.chi.0000198596.76443.b0

*Ghodsian-Carpey, J., & Baker, L. A. (1987). Genetic and environmental influences
©

on aggression in 4- to 7-year-old twins. Aggressive Behavior, 13, 173–186.

ht

doi:10.1002/1098-2337(1987)13:4<173::AID-AB2480130402>3.0.CO;2-Y
ig

Giros, B., Jaber, M., Jones, S. R., Wightman, R. M., & Caron, M. G. (1996, Feb-
yr

ruary). Hyperlocomotion and indifference to cocaine and amphetamine in mice

op

lacking the dopamine transporter. Nature, 379, 606–612. doi:10.1038/379606a0

Gough, H. G., & Heilbrun, A. B. (1972). The Adjective Checklist manual. Palo Alto,
C

CA: Consulting Psychologists Press.

160 RHEE AND WALDMAN

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 161

*Gustavsson, J. P., Pedersen, N. L., Åsberg, M., & Schalling, D. (1996). Exploration
into the sources of individual differences in aggression-, hostility-, and anger-
related (AHA) personality traits. Personality and Individual Differences, 21,
1067–1071. doi:10.1016/S0191-8869(96)00146-8

A)
Haberstick, B. C., Schmitz, S., Young, S. E., & Hewitt, J. K. (2006). Genes and devel-
opmental stability of aggressive behavior problems at home and school in a com-

P
munity sample of twins aged 7–12. Behavior Genetics, 36, 809–819. doi:10.1007/

(A
s10519-006-9092-5

n
Ligthart, L., Bartels, M., Hoekstra, R. A., Hudziak, J. J., & Boomsma, D. I. (2005).

io
Genetic contributions to subtypes of aggression. Twin Research and Human

at
Genetics, 8, 483–491. doi:10.1375/twin.8.5.483

ci
Lilienfeld, S. O., & Waldman, I. D. (1990). The relation between childhood attention-

so
deficit hyperactivity disorder and adult antisocial behavior reexamined: The

As
problem of heterogeneity. Clinical Psychology Review, 10, 699–725. doi:10.1016/
0272-7358(90)90076-M

al
Linnoila, M., Virkkunen, M., Scheinin, M., Nuutila, A., Rimon, R., & Goodwin,

ic
F. K. (1983). Low cerebrospinal fluid 5-hydroxyindoleacetic acid concentration

og
differentiates impulsive from nonimpulsive violence behavior. Life Sciences, 33,
2609–2614. doi:10.1016/0024-3205(83)90344-2 ol
ch
Loehlin, J. C. (1992). Latent variable models: An introduction to factor, path, and structural
analysis (2nd ed.). Hillsdale, NJ: Erlbaum.
y
Ps

*Meininger, J. C., Hayman, L. L., Coates, P. M., & Gallagher, P. (1988). Genetics
or environment? Type A behavior and cardiovascular risk factors in twin chil-
an

dren. Nursing Research, 37, 341–346. doi:10.1097/00006199-198811000-00006

ic

Mullen, B. (1989). Advanced BASIC meta-analysis. Hillsdale, NJ: Erlbaum.

er

Norland, S., Shover, N., Thornton, W., & James, J. (1979). Intrafamily conflict and
Am

delinquency. Pacific Sociological Review, 22, 233–237.

*Owen, D., & Sines, J. O. (1970). Heritability of personality in children. Behavior
e

Genetics, 1, 235–248. doi:10.1007/BF01074655

th

Partanen, J., Bruun, K., & Markkanen, T. (1966). Inheritance of drinking behavior: A
by

study on intelligence, personality, and use of alcohol of adult twins. Helsinki, Finland:
The Finnish Foundation for Alcohol Studies.
11

Patterson, G. R., Reid, J. B., & Dishion, T. J. (1992). Antisocial boys. Eugene, OR:
20

Castalia Publishing.
*Plomin, R., Foch, T. T., & Rowe, D. C. (1981). Bobo clown aggression in child-
©

hood: Environment, not genes. Journal of Research in Personality, 15, 331–342.

ht

doi:10.1016/0092-6566(81)90031-3
ig

*Rahe, R. H., Hervig, L., & Rosenman, R. H. (1978). Heritability of type A behavior.
yr

Psychosomatic Medicine, 40, 478–486.

op

Rhee, S. H., & Waldman, I. D. (2002). Genetic and environmental influences on

antisocial behavior: A meta-analysis of twin and adoption studies. Psychological
C

Bulletin, 128, 490–529. doi:10.1037/0033-2909.128.3.490

GENETIC AND ENVIRONMENTAL INFLUENCES 161

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 162

Robins, L. N. (1966). Deviant children grown up. Baltimore, MD: Williams & Wilkins.
Rosenthal, R. (1991). Meta-analytic procedures for social research. Newbury Park, CA:
Sage.
Rubinstein, M., Phillips, T. J., Bunzow, J. R., Falzone, T. L., Dziewczapolski, G.,

A)
Zhang, G., . . . Chester, J. (1997). Mice lacking dopamine D4 receptors are
supersensitive to ethanol, cocaine, and methamphetamine. Cell, 90, 991–1001.

P
(A
doi:10.1016/S0092-8674(00)80365-7
*Rushton, J. P., Fulker, D. W., Neale, M. C., Nias, D. K. B., & Eysenck, H. J. (1986).

n
Altruism and aggression: The heritability of individual differences. Journal of Per-

io
sonality and Social Psychology, 50, 1192–1198. doi:10.1037/0022-3514.50.6.1192

at
Saudou, F., Amara, D. A., Dierich, A., LeMeur, M., Ramboz, S., Segu, L., . . . Hen, R.

ci
(1994, September). Enhanced aggressive behavior in mice lacking 5-HT1B recep-

so
tor. Science, 265, 1875–1878. doi:10.1126/science.8091214

As
*Scarr, S. (1966). Genetic factors in activity motivation. Child Development, 37,
663–673. doi:10.2307/1126688

al
Seeman, P., & Madras, B. K. (1998). Anti-hyperactivity medication: Methylphenidate

ic
og
and amphetamine. Molecular Psychiatry, 3, 386–396. doi:10.1038/sj.mp.4000421
Solanto, M. V. (1984). Neuropharmacological basis of stimulant drug action in
ol
attention deficit disorder with hyperactivity: A review and synthesis. Psycholog-
ch
ical Bulletin, 95, 387–409. doi:10.1037/0033-2909.95.3.387
y

Solanto, M. V. (1998). Neuropsychopharmacological mechanisms of stimulant drug

Ps

action in attention-deficit hyperactivity disorder: A review and integration.

Behavioural Brain Research, 94, 127–152. doi:10.1016/S0166-4328(97)00175-7
an

Suomi, S. J. (2003). Gene-environment interactions and the neurobiology of social

ic

conflict. Annals of the New York Academy of Sciences, 1008, 132–139. doi:10.1196/
er

annals.1301.014
Am

*Tellegen, A., Lykken, D. T., Bouchard, T. J., Wilcox, K., Segal, N., & Rich, S.
(1988). Personality similarity in twins reared apart and together. Journal of Per-
e

sonality and Social Psychology, 54, 1031–1039. doi:10.1037/0022-3514.54.6.1031

th

Tuvblad, C., Eley, T. C., & Lichtenstein, P. (2005). The development of antisocial
by

behaviour from childhood to adolescence. A longitudinal twin study. European

Child & Adolescent Psychiatry, 14, 216–225. doi:10.1007/s00787-005-0458-7
11

van Beijsterveldt, C. E., Bartels, M., Hudziak, J. J., & Boomsma, D. I. (2003). Causes
20

of stability of aggression from early childhood to adolescence: A longitudinal

genetic analysis in Dutch twins. Behavior Genetics, 33, 591–605. doi:10.1023/
©

A:1025735002864
ht

*van den Oord, E. J. C. G., Verhulst, F. C., & Boomsma, D. I. (1996). A genetic study
ig

of maternal and paternal ratings of problem behaviors in 3-year-old twins. Journal

yr

of Abnormal Psychology, 105, 349–357. doi:10.1037/0021-843X.105.3.349

op

van der Valk, J. C., Verhulst, F. C., Neale, M. C., & Boomsma, D. I. (1998). Longitudi-
nal genetic analysis of problem behaviors in biologically related and unrelated
C

adoptees. Behavior Genetics, 28, 365–380. doi:10.1023/A:1021621719059

162 RHEE AND WALDMAN

12051-09_CH08-rev2.qxd 5/25/10 11:38 AM Page 163

Vitaro, F., Brendgen, M., & Tremblay, R. E. (2002). Reactively and proactively
aggressive children: Antecedent and subsequent characteristics. Journal of Child
Psychology and Psychiatry, and Allied Disciplines, 43, 495–505. doi:10.1111/1469-
7610.00040

A)
Waldman, I. D., & Gizer, I. (2006). The genetics of attention deficit hyperactivity
disorder. Clinical Psychology Review, 26, 396–432. doi:10.1016/j.cpr.2006.01.007

P
(A
Waschbusch, D. A., Willoughby, M. T., & Pelham, W. E. (1998). Criterion validity
and utility of reactive and proactive aggression: Comparisons to attention deficit

n
hyperactivity disorder, oppositional defiant disorder, conduct disorder, and other

io
measures of functioning. Journal of Clinical Child Psychology, 27, 396–405.

at
doi:10.1207/s15374424jccp2704_3

ci
*Wilson, G. D., Rust, J., & Kasriel, J. (1977). Genetic and family origins of humor

so
preferences: A twin study. Psychological Reports, 41, 659–660.

As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

GENETIC AND ENVIRONMENTAL INFLUENCES 163

C
op
yr
ig
ht
©
12051-09_CH08-rev2.qxd

20
11
5/25/10

by
th
e
11:38 AM

Am
er
ic
Page 164

an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 165

PA)
SOCIAL INFORMATION PROCESSING

(A
PATTERNS AS MEDIATORS OF THE

n
io
INTERACTION BETWEEN GENETIC

at
ci
FACTORS AND LIFE EXPERIENCES

so
IN THE DEVELOPMENT OF

As
AGGRESSIVE BEHAVIOR

al
ic
og
KENNETH A. DODGE
ol
y ch
Ps
an

It is well established that social–cognitive processes correlate with

ic

aggressive behavioral acts (see Chapters 1, 2, and 7). For example, in response
er

to an ambiguous provocation by another person, when a respondent infers

Am

that the act was committed with hostile intent (a hostile attribution), the prob-
ability that the respondent will react aggressively is high (about .76; Dodge,
e

1980), whereas when that same respondent infers that the act was commit-
th

ted benignly, the probability of an aggressive behavioral reaction is low

by

(about .25). Likewise, if during another stage of processing information the

respondent evaluates an aggressive response as being likely to lead to desired
11

outcomes, the probability of engaging in aggression is high (Fontaine, Yang,

20

Dodge, Bates, & Pettit, 2008). Although the evidence is less clear that these
social–cognitive processes cause the aggressive behavioral response during the
©

microseconds of interpersonal interaction, the correlation has been found

ht

over and over, suggesting that social–cognitive processes, and the develop-
ig

mental processes that support them, are closely intertwined with the matrix
yr

of causes of aggression.
op

It has also been found that individuals develop characteristic styles of

processing social information within specific social situations. These styles act
C

as acquired personality characteristics. They correlate significantly with and

165
12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 166

predict individual differences in aggressive behavior in particular situations.

In this chapter, the empirical evidence on this topic will be reviewed briefly
and integrated with recent discoveries regarding psychophysiology and neu-
ral processes. A general model will be advanced that proposes that (a) pro-

A)
cessing patterns provide the proximal mechanism through which aggressive
behavior occurs; (b) these patterns correlate with neural and psychophysio-

P
(A
logical processes; (c) these patterns are acquired through genetic and envi-
ronmental processes, especially in interaction; and (d) acquired processing

n
patterns account for the effects of genetic and environmental factors in

io
behavioral development and provide the mechanism through which these

at
factors exert their impact. Findings will be presented from the Child Devel-

ci
opment Project (CDP), an ongoing longitudinal study of 585 boys and girls

so
followed from age 4 through young adulthood.

As
al
SOCIAL INFORMATION PROCESSING MECHANISMS

ic
IN AGGRESSIVE BEHAVIORS

og
ol
Models of the processing of information in response to social stimuli posit
ch
a sequence of steps that lead to behavioral responding including aggression
y

toward others (Crick & Dodge, 1994; see also Chapter 7, this volume). These
Ps

steps are logically ordered and assumed to flow temporally, although evidence
for the sequential ordering is scant (see Chapter 1, this volume). Methods to
an

assess an individual’s self-report of her or his processing have been developed

ic

using hypothetical situational vignettes as stimuli. In each vignette, the per-

er

son is asked to contemplate a hypothetical scenario in which a social event

Am

occurs and then to answer questions about attributions and possible responses
(Dodge, 1980). The stimuli have been presented orally, in the form of car-
e

toons, or by video.
th

The first several steps of processing describe the sensation and interpreta-
by

tion of cues, and the latter steps describe behavioral decision making. The first
step is attention to and sensation of the stimulus. Because the stimulus array is
11

so large, selective attention to some cues over others is inevitable. In a situa-

20

tion involving provocation by another—for example, being pushed to the

ground—one child might attend to the provocation itself and the pain it causes,
©

whereas a second child might attend to the teacher watching in the back-
ht

ground, and a third child might attend to the peers who are laughing nearby.
ig

Attention to cues can obviously influence downstream processing and ultimate

yr

behavior, for example, when attention to the provocateur’s look of surprise and
op

regret might mitigate a retaliatory response. Numerous factors affect selective

attention to hostile versus other cues, such as recent threat, fatigue, and stress.
C

A pattern of habitual selective attention to hostile cues (e.g., the provocateur’s

166 KENNETH A. DODGE

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 167

angry voice, repeated negative behavior, lack of sensitivity to the child’s

pain) has been associated with chronic aggressive behavior (Dodge, Pettit,
McClaskey, & Brown, 1986).
Closely following the sensation of cues is a mental representation of

A)
those cues, often involving an interpretation of the other person’s intention.
As noted earlier, when a hostile intent is inferred, aggressive behavior likely

P
(A
ensues, in contrast with an inference of a benign intention. The process of
mental representation occurs in microseconds and may be updated across

n
time during a social interaction. It is not usually a conscious process, although

io
it can become so if prompted. The process undoubtedly involves neural activ-

at
ity that is conditioned by experience (for evidence regarding neural activity

ci
underlying mental representations of social intent, see Chapter 6). Inferences

so
of hostile intent have also been found to correlate with heightened auto-

As
nomic reactivity (Crozier et al., 2008). Numerous studies have shown a robust
pattern in which hostile attributional biases are associated with aggressive

al
behavior, especially reactive aggression (Dodge, 1980). A review by Orobio

ic
de Castro, Veerman, Koops, Bosch, and Monshouwer (2002) indicates that

og
this pattern holds across ages, demographic and cultural groups, and contexts.
ol
The third step is goal selection, in which the mentally represented stim-
ch
ulus is associated with an emotional reaction and the narrowing of a goal.
y

Again, the respondent is not usually aware of this process but might reflect after-
Ps

ward on the cognitive processes involved. Children who experience anger and
regularly select instrumental and self-defensive goals are likely to behave aggres-
an

sively, whereas children who select social goals are likely to behave nonaggres-
ic

sively (Crick & Dodge, 1994).

er

Steps of response generation, response evaluation, and enactment con-

Am

stitute the response–decision phase of processing. Mental representation and

goal selection trigger one or more possible behavioral responses such as aggres-
e

sion, withdrawal, and social deflection. The trigger from mental representa-
th

tion of hostile intent to aggressive response generation is a neural association

by

that is probably both “ready” at birth (because of evolutionary adaptation) and

conditioned from experience and observation (see Chapter 6). One of the
11

most frequently replicated findings in this area is the empirical association

20

between the generation of aggressive responses to hypothetical social stimuli

and actual chronic aggressive behavior, beginning at about age 4 and contin-
©

uing through adolescence (Dodge, Coie, & Lynam, 2006).

ht

Generation of an aggressive response does not lead inevitably to aggres-

ig

sive behavior (see Chapters 3 and 6, this volume). Processes of response eval-
yr

uation and decision (called RED by Fontaine & Dodge, 2006) follow. During
op

RED, the respondent immediately decides (nonconsciously, in microseconds)

to accept the generated response without any consideration of consequences
C

or to consider multiple domains of evaluative judgment, including (a) response

SOCIAL INFORMATION PROCESSING PATTERNS 167

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 168

efficacy, the estimation of how likely the respondent is to be successful if the

considered response were to be carried out; (b) response valuation, which is the
assignment of value to the response in terms of its social and moral qualities;
(c) outcome expectancy, which is the estimation of the likelihood of various

A)
consequences of a behavior; and (d) outcome valuation, in which the estimated
outcome is given value. Fontaine and Dodge (2006) hypothesized that differ-

P
(A
ent possible responses are compared (response comparison) before the most
appropriate response is selected (response selection). Measurement of response

n
decision during hypothetical social stimuli has yielded robust correlations

io
between all of these subprocesses and chronic aggressive behavior, especially

at
proactive aggression (see the review by Fontaine & Dodge, 2006).

ci
so
As
SOCIAL INFORMATION PROCESSING PATTERNS
AS PREDICTORS OF INDIVIDUAL DIFFERENCES

al
IN AGGRESSIVE BEHAVIOR

ic
og
Although these processing patterns are empirically correlated with indi-
ol
vidual differences in aggressive behavior, the correlational findings often suffer
ch
from two problems. First, because the measurement of processing is typically
y

based on individuals’ self-report, the data depend on self-observation of cogni-

Ps

tive processing and not processing itself, which occurs at the neural level and
nonconsciously. Even self-reports that are collected “online” are immediate
an

self-observations. Recent evidence suggests that individuals actually begin to

ic

respond with neural activity in microseconds prior to self-awareness of respond-

er

ing. The second problem is that the empirical correlation between patterns of
Am

processing and patterns of behavior might reflect an opposite causal direction

than proposed in the model or might be attributable to an unmeasured third
e

variable. Two kinds of evidence have been mounted to test the hypothesis that
th

chronic patterns of processing improve the prediction of aggressive behavior.

by

Controlling for Prior Aggression

11
20

The first evidence comes from prospective studies in which early levels
of problematic aggressive conduct are controlled statistically. In the CDP of a
©

community sample of 585 boys and girls followed for 20 years, we assessed both
ht

aggressive behavior and processing patterns recurrently. We know that during

ig

the early elementary school years, children develop patterns of processing

yr

social information that become stable when measured annually for 4 years, as
op

assessed by cross-time internal consistency coefficients of .70 to .79 (Dodge,

Pettit, Bates, & Valente, 1995). These patterns function as acquired person-
C

ality characteristics. During this period, the continuity in aggressive behavior

168 KENNETH A. DODGE

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 169

also becomes strong. Aggressive behavior measured by teacher assessments at

age 5 was found to predict patterns of social information processing in kinder-
garten through Grade 3, which in turn predicted aggressive behavior at age 10
and improved the prediction of aggressive behavior even when early levels of

A)
aggression were statistically controlled (Dodge et al., 2003). This pattern sup-
ports the hypothesis that social information processing plays a causal role in

P
(A
generating aggressive behavior.
We also found that we could predict adolescent conduct problems from

n
kindergarten processing patterns. Here, we scored children as displaying prob-

io
lems in social information processing at the early steps (i.e., hypervigilance

at
and hostile attributional biases) or later steps (i.e., response generation or

ci
evaluation) or at both steps or neither step. We found, while controlling for

so
kindergarten externalizing problems as assessed by both mothers and teach-

As
ers, that the four groups differed in mother- and teacher-rated externalizing
behavior problems at the end of Grade 11 (Lansford et al., 2006). Further-

al
more, the effect was cumulative or synergistic, in that the group of children

ic
with kindergarten problems at both stages of processing was especially likely

og
to show conduct problems in high school.
ol
The prediction was even stronger for processing patterns in Grade 8. Con-
ch
trolling for conduct problems before and during Grade 8, the four groups of chil-
y

dren as assessed by processing patterns during Grade 8 differed in mother- and

Ps

teacher-rated conduct problems in Grade 11. Again, a cumulative or inter-

active effect was found, with the group displaying the highest levels of problem-
an

atic conduct being the one with problems at both early and later stages of
ic

processing.
er

The relation between processing patterns and aggressive behavior is

Am

iterative and reciprocal (Fontaine & Dodge, 2006). That is, aggressive behav-
ior at age 14 predicted processing patterns the next year, which in turn pre-
e

dicted growth in aggression the following year, even controlling for prior
th

aggression. Likewise, processing patterns in one year predicted aggression in

by

the following year, which altered processing patterns in the subsequent year.
Across adolescence, this reciprocal effect continues.
11
20

Intervention as Experimental Evidence

©

The second kind of evidence comes from intervention experiments in

ht

which explicit attempts to alter processing patterns are manipulated by ran-

ig

dom assignment and the effect on aggressive behavior is assessed. Graham and
yr

Hudley (1993) developed a brief intervention to help young African Amer-

op

ican boys process information in more benevolent ways (i.e., with a lessened
hostile attribution bias). They found that random assignment to this inter-
C

vention led to lower scores on measures of hostile attribution bias, and this

SOCIAL INFORMATION PROCESSING PATTERNS 169

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 170

impact mediated a change in aggressive behavior. Guerra and Slaby (1990)

also randomly assigned aggressive adolescents to intervention or control con-
ditions. Their intervention involved multiple components of social informa-
tion processing. They also found that random assignment to the intervention

A)
condition was associated with improvements in processing patterns and
reduced aggressive behavior.

P
(A
Situation and Relationship Specificity

n
io
A pivotal issue in constructing models of social information processing

at
is situational specificity in the link between processing patterns and behavior.

ci
Processing patterns within a certain kind of situation, such as being provoked

so
or attempting to enter a peer group or handling conflict in a romantic relation-

As
ship or with a coworker, predict behavior within that type of situation more
strongly than behavior in other situations. For example, Dodge et al. (1986)

al
assessed processing patterns in provocative peer interactions and peer-group

ic
entry situations. We then placed children in a laboratory setting and exposed

og
them to a provocation by a confederate peer and an entry situation in which
ol
they were asked to initiate entry into a strange peer group. We found that pro-
ch
cessing patterns predicted behavior, and the predictions were stronger within
y

kinds of situations than across kinds of situations.

Ps

More recently, we found similar evidence for young adults. Two kinds of
situations are important in young adulthood, defined by relationships. The
an

establishment of successful romantic relationships is a key developmental task

ic

in that age period (Collins & van Dulmen, 2006). Unfortunately, violence is
er

fairly common in these relationships, as indicated by surveys showing that 20%

Am

to 50% of intimate relationships during adulthood involve violence toward a

partner (Silverman, Raj, Mucci, & Hathaway, 2001). Likewise, relationships
e

with adult peers are essential to work and community success (Arnett, 2006).
th

Although some individuals display violence across different kinds of relation-

by

ships, many adults behave violently only in romantic relationships (Archer,

2000), and some research indicates that peer violence and partner violence have
11

distinguishable antecedents (Brendgen, Vitaro, Tremblay, & Lavoie, 2001).

20

We were able to interview 85% of the original CDP sample at age 22 and
follow them through age 24. We assessed processing patterns in situations
©

involving conflict with a romantic partner (e.g., “You are at a gathering with
ht

a group of friends and your girl/boyfriend and learn that your girl/boyfriend and
ig

one of the people at the gathering used to be a couple; they spend most of the
yr

night talking with each other”) and conflict with a coworker or peer (e.g., “You
op

tell a friend something personal and ask your friend not to discuss it with any-
one else. However, a couple of weeks later, you find out that a lot of people
C

know about it”).

170 KENNETH A. DODGE

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 171

We (Pettit, Lansford, Dodge, & Bates, 2009) were able to predict violent
behavior in each of these types of relationships as well as evidence of relation-
ship specificity. Processing patterns in a hypothetical romantic relationship
predicted violent behavior in actual romantic relationships 2 years later, as

A)
reported by both the participant and his or her romantic partner. Also, process-
ing patterns within the peer relationship predicted violent behavior toward

P
(A
peers 2 years later. Furthermore, the predictions were stronger within kinds of
relationships than across kinds of relationships.

n
io
at
NEURAL AND PSYCHOPHYSIOLOGICAL PROCESSES

ci
AND SOCIAL INFORMATION PROCESSING

so
As
A misconception about information processing is that it occurs inde-
pendently of biological processes in real time. Accumulating evidence indi-

al
cates that psychophysiological and neural processes co-occur with information

ic
processing (see Chapter 6, this volume). Most likely, these processes occur in

og
real time outside a person’s awareness, and measures of information processing
ol
represent the individual’s postbehavioral reflection on her or his thoughts and
ch
actions. This suggests that the measures of information processing may be
y

imprecise and subject to self-presentation and other biases.

Ps

Psychophysiological Processes
an
ic

Ortiz and Raine (2004) reviewed evidence indicating that resting heart
er

rate is inversely related to individual differences in aggressive behavior (i.e.,

Am

low resting heart rate predicts higher aggressive behavior), especially proactive
and life-persistent aggression but not situational or adolescence-limited aggres-
e

sion (Moffit & Caspi, 2001). Raine, Venables, and Mednick (1997) found that
th

low resting heart rate at age 3 predicted aggressive behavior 8 years later at
by

age 11. Raine (2002) hypothesized that deficits in volume and function of
prefrontal cortex may be responsible for low autonomic activity as well as
11

aggressive behavior.
20

However, during the processing of social cues, the autonomic nervous

system reacts with rapid changes in heart rate (Crozier et al., 2008). While
©

the individual is attending to cues, heart rate decreases. When a provoca-

ht

tion occurs, heart rate increases and then slowly returns to baseline when
ig

the threat subsides. Lorber (2004) summarized evidence from numerous

yr

studies showing that aggressive children display higher heart rate reactivity
op

to provocative cues than do nonaggressive children. Furthermore, Crozier

et al. (2008) found that high heart rate reactivity to provocation was cor-
C

related with several stages of social information processing. Furthermore,

SOCIAL INFORMATION PROCESSING PATTERNS 171

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 172

processing responses mediate the link between heart rate changes and aggres-
sive behavior.
Thus, two separate psychophysiological processes may be related to
aggressive behavior. Both low resting heart rate and high heart rate reactiv-

A)
ity in response to threatening cues appear to predict aggressive behavior.
These processes may have their antecedents in both life events and heritable

P
(A
biological processes, and their impact on aggressive behavior may be medi-
ated by the manner in which the individual processes social information in

n
response to threat.

io
at
Neural Processes

ci
so
Neuroimaging studies suggest that regions of the prefrontal cortex (e.g.,

As
the dorsolateral prefrontal cortex and ventral prefrontal cortex) and the lim-
bic system (e.g., amygdala) are activated during information processing in

al
response to interpersonal provocations such as unfair allocation of resources

ic
by a peer (Meyer-Lindenberg et al., 2006; see also Chapter 6, this volume).

og
Meyer-Lindenberg et al. (2006) found significant activation of the amygdala
ol
in response to experimental presentation of stimuli similar to those used to
ch
assess social information processing, such as angry and fearful faces and aver-
y

sively valenced cues. Presentation of threatening faces (Pezawas et al., 2005),

Ps

the perception of anger in others, and the experience of anger in oneself

(Murphy, Nimmo-Smith, & Lawrence, 2003) all reliably activate the amyg-
an

dala. The prefrontal cortex is activated during executive function tasks

ic

involved in processing information, such as planning, inhibitory control, and

er

decision making (Raine, Buchsbaum, & LaCasse, 1997). Raine (2008) stated
Am

that “the prefrontal cortex acts as an ‘emergency brake’ on runaway emotions

generated by limbic structures” (p. 324).
e

The activation of these brain regions, in turn, appears to be mediated by

th

the release of neurotransmitters. Monoamine oxidase-A (MAO-A) appears

by

to correlate with amygdala volume and probably with its activation (Meyer-
Lindenberg et al., 2006). Release of 5-HT in the prefrontal cortex is thought
11

to regulate an individual’s impulsive desire to retaliate aggressively to per-

20

ceived hostile treatment (Evers et al., 2006), although the evidence was
only correlational until recently. Crockett, Clark, Tabibnia, Lieberman,
©

and Robbins (2008) experimentally manipulated serotonin levels in each of

ht

20 human subjects through an acute tryptophan depletion procedure that

ig

temporarily lowered 5-HT levels. In contrast with a placebo condition, when

yr

subjects had lower serotonin levels they responded to unfair treatment by a

op

peer with greater retaliatory behavior (for more evidence concerning the role
of serotonin in aggressive behavior, see Chapters 7 and 8, this volume).
C

172 KENNETH A. DODGE

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 173

Thus, in the same way that social information processing involves multi-
ple, relatively independent steps, functional brain imaging studies have identi-
fied multiple brain regions that are implicated in different aspects of responding
to social stimuli.

PA)
(A
ANTECEDENTS OF SOCIAL INFORMATION
PROCESSING PATTERNS AND NEURAL PROCESSES

n
io
at
I turn now to the antecedents of processing patterns and neural processes

ci
that mediate aggressive behavior. Given the empirical findings of both general

so
prediction and situation-specific prediction, I hypothesize that some antecedent

As
factors apply generally and some apply to specific domains, situations, or
relationships.

al
ic
Child Maltreatment

og
ol
The experience of abuse by one’s parent during the first 5 years of life is
ch
devastating and has been hypothesized to alter one’s central working models of
how human relationships work. Children develop basic trust through inter-
y
Ps

action with caring adults, and violation of that trust through extreme or ongo-
ing maltreatment is hypothesized to lead to schemas, scripts, knowledge
an

structures, and working models that others will act maliciously (see Chapters 4,
ic

13, 14, and 19). Thus, the child develops hypervigilant and selective attentional
er

patterns, becomes quick to attribute hostile motives to others, adopts self-

Am

defensive rather than social goals, develops a repertoire of self-defensive behav-

ioral responses, and comes to evaluate self-protective behaviors as effective and
desirable. Given the centrality of these working models, it is hypothesized that
e
th

this experience will lead to long-term and pervasive patterns of processing

information across a range of social situations and relationships.
by

Evidence supporting this hypothesis comes from several sources, includ-

11

ing Pollack’s work on attention patterns in maltreated children and findings

from the CDP study. Dodge, Bates, and Pettit (1990) and Weiss, Dodge, Bates,
20

and Pettit (1992) found that maltreatment in the first 5 years of life predicted
©

processing patterns at school entry that included hypervigilance, hostile attri-

butional biases, aggressive response generation, and favorable evaluation of
ht

aggressive behaviors. In turn, these patterns predicted aggressive behavior

ig
yr

and mediated the impact of maltreatment on aggression. Dodge et al. (2003)

extended the measurement of aggressive outcomes through adolescence and
op

found a similar pattern of prediction and mediation. Edwards et al. (2009)

C

SOCIAL INFORMATION PROCESSING PATTERNS 173

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 174

found that 30.4% of the maltreated children had an official court record of
arrest by age 24, in contrast with 16.5% of the nonmaltreated children.
Lansford et al. (2007) found that the outcomes that accrue from early-life
maltreatment are broad and include not only aggressive behavior but also early

A)
pregnancy, anxiety, depressive symptoms, school dropout, substance-use prob-
lems, and arrests for a variety of crimes (for a review of these long-term effects

P
(A
of child maltreatment, see Chapter 19, this volume). The processing patterns
that were assessed in early elementary school mediated some but not all of these

n
outcomes, perhaps because the stimuli that were used to assess processing pat-

io
terns were restricted to only selected domains. It is plausible that maltreatment

at
alters processing in multiple domains, which then mediate behavioral out-

ci
comes. More comprehensive assessments of processing patterns are necessary

so
to test this hypothesis.

As
Peer Social Rejection

al
ic
Other life experiences appear to have more circumscribed effects. In

og
American elementary schools, chronic peer social rejection is a fairly com-
ol
mon painful experience. We hypothesized, and found empirically, that this
ch
experience would exacerbate children’s aggressive behavior problems
y

beyond whatever aggressive behavior led to the peer response of rejection

Ps

(Dodge et al., 2003). That is, rejection during kindergarten through Grade 2
predicted aggressive behavior in Grade 4, even controlling for early aggres-
an

sive behavior. Furthermore, patterns of social information processing about

ic

peer events partially mediated the growth in aggressive behavior during this
er

period.
Am

More recently, Pettit et al. (2009) found that the impact of early peer
rejection lasts through young adulthood. Recall the CDP findings presented
e

earlier about the significant relation between processing patterns in peer rela-
th

tionships at age 22 and aggressive behavior toward peers at age 24. It turns out
by

that the experience of peer rejection in kindergarten significantly predicts these

processing patterns at age 22, which in turn mediate the impact of kindergarten
11

peer rejection on aggressive behavior toward peers in young adulthood.

20

To describe these relations in person-centered terms, we dichotomized

participants into those who had been rejected by peers in elementary school
©

and those who had not, those who displayed problematic processing about peer
ht

relationships at age 22 and those who did not, and those who displayed vio-
ig

lence toward adult peers versus those who did not. Among those who were not
yr

rejected by peers and who displayed nonproblematic processing about peers,

op

the probability of becoming violent toward a peer was relatively low (.43),
whereas among rejected children who displayed problematic processing, the
C

probability was high (.70). In between were nonrejected children who dis-

174 KENNETH A. DODGE

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 175

played problematic processing (.47) and rejected children who did not display
problematic processing (.51).

Exposure to Romantic Relationship Violence

PA)
The antecedents of romantic relationship violence may differ from those

(A
for peer-directed violence. Exposure to parents’ domestic violence in early life
predicts later violence in romantic relationships (Nay, Dodge, Lansford,

n
io
Pettit, & Bates, 2009), but disentangling the effect of this particular exposure

at
from other family violence may prove difficult. We explored the effect of vic-

ci
timization during adolescent romantic relationships on young adult violent

so
behavior toward a romantic partner (for a review of consequences of aggres-
sion within romantic relationships, see Chapter 20, this volume).

As
We found that being victimized violently during an adolescent roman-

al
tic relationship did indeed predict violent behavior toward a romantic partner

ic
in young adulthood. Also, this adolescent experience predicted problematic

og
processing patterns in romantic relationships, which partially mediated the
effect of victimization on later violent behavior toward a romantic partner.
ol
Among non-victims who display non-problematic processing, the probability
ch
of later becoming violent toward a romantic partner was low (.29), whereas
y
Ps

among victims who displayed problematic processing, the probability was high
(.62). In between were nonvictims who displayed problematic processing (.42)
an

and nonvictims who did not display problematic processing (.43).

Because of the unique paths from problematic processing within a rela-
ic
er

tionship type to violence in that kind of relationship, it follows that media-

tion by processing patterns was found only within relationship types and not
Am

across relationship types.

e
th

Genetic Factors
by

Individual differences in genetic variation in the serotonin transporter

11

5-HT gene have been associated with increased amygdala activation (Pezawas
et al., 2005) and prefrontal cortex functions (Raine, 2008), as assessed through
20

functional magnetic resonance imaging (fMRI). Also, a common variation in

©

the X-linked MAO-A gene has been correlated with amygdala volume and
activation in response to threatening stimuli (Meyer-Lindenberg et al., 2006).
ht

That is, individuals, especially males, with a polymorphism in the gene for
ig

MAO-A demonstrate greater amygdala activation when presented with

yr

threatening faces and aversive emotional situations. It follows, then, that vari-
op

ation in these genes would correlate with individual differences in processing

C

patterns in response to threatening social situations.

SOCIAL INFORMATION PROCESSING PATTERNS 175

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 176

Environmental and Genetic Factors

A comprehensive model of the development of chronic aggressive behav-

ior must account for both genetic and life-experience factors (see Chapters 7

A)
and 8). The major environmental factors in early life have been reviewed
elsewhere (Dodge et al., 2006; Dodge & Pettit, 2003); they include the kinds

P
(A
of personally threatening experiences addressed earlier, such as early physical
maltreatment, peer social rejection, and victimization, as well as exposure to

n
stressful contexts such as poverty and disadvantage. The latter factors are likely

io
to operate through their effects on life experiences such as parenting quality

at
and success in life tasks such as getting an education (Dodge, Pettit, & Bates,

ci
1994). Major environmental variables in mid-childhood and adolescence

so
involve exposure to deviant peers who influence high-risk youth to engage in

As
aggressive behavior and also involve parental failure to supervise and monitor
youths’ activities, thus increasing their exposure to violence and precipitators

al
of violence (Dodge et al., 2006).

ic
Environmental variables that have an enduring effect are likely to be

og
ones that alter brain processes. Evidence has been growing that life stressors
ol
and early trauma have enduring effects on both prefrontal cortex and amyg-
ch
dala processes measured through fMRI. Liston, McEwan, and Casey (2009)
y

reported that the natural experiment of a potent stressor (an upcoming aca-
Ps

demic examination) had observable effects on decreasing dorsolateral pre-

frontal cortex activity during laboratory tasks. Ganzel, Casey, Glover, Voss,
an

and Temple (2007) found that individuals who had been exposed to major
ic

trauma during the September 11, 2001, attack on the World Trade Center
er

displayed heightened amygdala activation when presented with emotionally

Am

aversive and fearful faces.

The heritability of criminal and aggressive behavior patterns has been
e

evident for some time (see Chapter 8), although adoption studies have iden-
th

tified a heritability-by-environment interaction effect. Cloninger, Sigvards-

by

son, Bohman, and von Knorring (1982) found that under conditions of low
heritable risk (i.e., having a biological parent who was not a criminal), the
11

impact of the environment (i.e., having an adoptive parent who was or was
20

not a criminal) on later criminality was rather small (a rate of .03 for non-
criminal adoptive parents vs. .07 for criminal adoptive parents). However,
©

under conditions of high heritable risk, the effect of the environment was
ht

strong (.12 for noncriminal adoptive parents vs. .40 for criminal adoptive par-
ig

ents). Jaffee et al. (2005) studied monozygotic and dizygotic twins from the
yr

British E-Risk study and identified four rank-ordered groups of increasing her-
op

itable risk, with the lowest heritable-risk group being children whose mono-
zygotic twin was not conduct disordered (CD), the next lowest group being
C

176 KENNETH A. DODGE

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 177

children whose dizygotic twin was not CD, the next highest group being chil-
dren whose dizygotic twin was CD, and the highest group being children
whose monozygotic twin was CD. The experience of child physical maltreat-
ment was determined by clinical interview with the mother, following pro-

A)
cedures used by Dodge, Bates, and Pettit (1990). Among the group at lowest
heritable risk, the experience of physical maltreatment had little effect on

P
(A
conduct disorder outcomes (for the nonmaltreated group, the rate was .02; for
maltreated group it was .04). Among those at the next highest level of heri-

n
table risk, the effect of maltreatment was small (.06 vs. .13). Among the group

io
at the next highest level, the effect of maltreatment grew larger (.19 vs. .37).

at
Finally, among those at the highest level, the effect of maltreatment was

ci
largest (.46 vs. .70).

so
The search for specific genetic variables that predict aggressive behav-

As
ior has been plagued by both political pressures and empirical failures to repli-
cate published findings, perhaps because of atheoretical approaches to gene

al
analyses that capitalize on chance (Kim-Cohen et al., 2006). Several theo-

ic
rists argue that genetic factors are likely to operate indirectly through their

og
effects on neurally and biologically mediated dispositions to act impulsively
ol
without planning, without empathy, or without consideration of extenuating
ch
circumstances (Caspi et al., 2002). These dispositions have been called exec-
y

utive functions (Morgan & Lilienfeld, 2000), and they likely affect processing
Ps

responses (or are measured as processing responses) during social interactions.

For example, a deficit in considering another’s feelings would lead to inaccu-
an

rate interpretations of stimulus events, and impulsivity would lead to premature

ic

decision making that failed to consider long-term consequences.

er

In line with this hypothesis, the genetic polymorphisms that have been
Am

most frequently correlated with aggressive behavior are ones that relate to neuro-
transmitter functions, especially dopamine (Dick et al., 2006; Moffitt et al.,
e

2008; see also Chapter 8, this volume). Three genes are discussed here. First,
th

MAO-A is an enzyme that selectively metabolizes serotonin, norepinephrine,

by

and dopamine (Shih, Chen, & Ridd, 1999), which are involved in brain
actions associated with stress regulation (Charney, 2004) and biological sensi-
11

tivity to adverse social contexts (Boyce & Ellis, 2005). A polymorphism in the
20

X-chromosome-linked gene that encodes MAO-A has been identified in about

one third of all human males, suggesting that it might have adaptive value
©

across evolution but is still nonnormative. In rodents, this polymorphism has

ht

been correlated with both aggressive behavior and lower brain serotonin and
ig

norepinephrine levels (Cases et al., 1995). In humans, a modest main effect of

yr

a polymorphism in MAO-A on aggressive behavior has been found (Brunner,

op

Nelen, Breakefield, Ropers, & van Oost, 1993), but other studies have found
contradictory patterns (Kim-Cohen et al., 2006).
C

SOCIAL INFORMATION PROCESSING PATTERNS 177

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 178

More likely in my estimation is a gene-by-environment interaction

effect. Eisenberger, Way, Taylor, Welch, and Lieberman (2007) found that
individuals with the low-expression allele form of MAO-A demonstrated
heightened dorsal anterior cingulate cortex activity but only in response to an

A)
environmental stressor of peer social rejection and not to peer inclusion. They
suggested that MAO-A may dispose individuals to become “hypersensitive” to

P
(A
interpersonal rejection.
If the MAO-A enzyme is involved in regulating stress, especially in

n
reaction to trauma and threat, then it might play a role in moderating the

io
effect of early physical maltreatment on the development of aggressive behav-

at
ior. Caspi et al. (2002) found this interaction effect in their Dunedin longi-

ci
tudinal sample, and Kim-Cohen et al. (2006) replicated the pattern in the

so
British E-Risk study. In the CDP, we recently replicated this interaction effect

As
on externalizing behavior patterns (Edwards et al., 2009). Among children
without the MAO-A polymorphism, those who had been maltreated did not

al
differ from those who had not been maltreated in the proportion who had

ic
been arrested by age 22 (.28 vs. .25), but among children with the polymor-

og
phism, those who had been maltreated had a much higher probability of crim-
ol
inal arrest by age 22 than those who had not been maltreated (.71 vs. .26).
ch
A second gene that has been implicated in a variety of externalizing and
y

addiction disorders is the gamma-aminobutyric acid (GABA) A receptor,

Ps

alpha 2, also known as GABRA2 (Dick et al., 2006). GABA is the major
inhibitory neurotransmitter in the mammalian brain. A polymorphism in
an

GABRA2 has been associated with conduct disorder, but the more powerful
ic

pattern again is a gene-by-environment interaction in which high parental

er

supervision during early adolescence has been found to buffer children from
Am

the adverse effect of the GABRA2 polymorphism (Dick et al., 2006). Among
children whose parents engaged in low rates of supervision, the likelihood of
e

being in a persistently high trajectory of conduct problems across adolescence

th

increases dramatically with the number of copies of the risk allele of GABRA2,
by

from .07 to .19 to .28 for 0, 1, or 2 copies, respectively, whereas among children
whose parents engaged in high rates of supervision, the likelihood of being in
11

the persistently high trajectory of conduct problems increases modestly from

20

.10 to .12 to .13 for 0, 1, or 2 copies, respectively.

A third gene that has been found to correlate with aggressive behavior
©

is 5-HT. Recall that polymorphisms in 5-HT have been associated with

ht

impaired limbic structures and increased amygdala function. Waldman

ig

(2008; see also Chapter 8, this volume) recently took this work a step fur-
yr

ther by finding a significant relation between a polymorphism in 5-HT and

op

individual differences in reactive but not proactive aggressive behavior.

Even more striking was that, using the social information processing instru-
C

ments described earlier, it was possible to find a significant association between

178 KENNETH A. DODGE

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 179

5-HT and hostile attributional biases. Finally, the effect of 5-HT on reactive
aggression was mediated by hostile attributional biases. This work is the first
known attempt to identify molecular genetic bases for social information
processing patterns.

A)
In sum, a comprehensive model of the development of chronic aggres-
sive behavior must take into account environmental main effects, gene main

P
effects, and gene-by-environment interaction effects. Furthermore, the envi-

(A
ronmental variables are likely to include both broad factors that influence

n
aggression across many situations as well as factors that influence aggression

io
within particular kinds of situations.

at
ci
so
A PROPOSED MODEL OF SOCIAL INFORMATION

As
PROCESSING MECHANISMS IN GENE-BY-ENVIRONMENT
INTERACTION EFFECTS

al
ic
Integration of the distal genetic and environmental factors with social

og
information processing proximal mechanisms requires a final leap of theorizing.
ol
The model proposed here builds on models by van Goozen, Fairchild, and
ch
Harold (2008) and Raine (2008) but is unique in positing that within-situation
processing patterns mediate the effects of genes, environments, and their inter-
y
Ps

actions. The overall model is depicted in Figure 9.1, which posits that specific
(albeit as yet unidentified) genes and an early environment characterized by
an

threat, trauma, or adversity pose risks for the long-term development of chron-
ically violent behavior. These distal factors operate as main effects and in inter-
ic

action with each other. Throughout development, particular environments

er

also lead to the development of situation-specific processing patterns that medi-

Am

ate aggressive behavior within those situations. When these situations present
themselves, they pose proximal risk of violent behavior.
e

These distal and proximal risk factors are mediated by brain processes
th

that operate in response to proximal situational stimuli. Three aspects of func-

by

tioning are hypothesized to co-occur during responding: neural activity in

synapses (most likely involving neurotransmitters such as dopamine), social
11

information processing as described earlier, and psychophysiological activity

20

in the autonomic nervous system. These brain processes, in turn, mediate

aggressive behavioral responses within these situations.
©
ht
ig

A RESEARCH AGENDA
yr
op

This model suggests numerous studies that have yet to be completed. To

understand the intricate interactions among specific genes and specific envi-
C

ronments, existing large-sample longitudinal studies need to be mined to test

SOCIAL INFORMATION PROCESSING PATTERNS 179

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 180

Situational
stimulus

A)
P
(A
Genes Neural
activity

n
io
at
ci
Social

so
information Violence
processing

As
al
Psycho-

ic
physiological

og
Early Situational activity
environment environment
ol
ch
Figure 9.1. A proposed model of genetic, environmental, and processing mechanisms
y

in the development of aggressive behavior.

Ps
an

gene–environment hypotheses. Subgroups of individuals who fit profiles of

ic

gene–environmental risk need to be exposed to situational stimuli that elicit

er

processing responses while the individuals are being observed via fMRI and
Am

psychophysiological recording.
Such studies are likely to produce both support and disconfirmation of
e

different components of the proposed model, hopefully with iterative refine-

th

ment of the model over time. Although empirical studies have already yielded
by

promising results, it is highly unlikely that any single gene or group of genes,
even in interaction with environmental histories, will account for much of
11

the variance in aggressive behavior.

20

The implications of these empirical findings and the model in Figure 9.1
for the design of preventive interventions suggest an additional research
©

agenda. Environmental engineering in light of personalized genetic informa-

ht

tion has the exciting but unrealized potential to reduce or prevent the devel-
ig

opment of chronic violence. So, too, environments and training interventions

yr

that alter social information processing patterns offer the hope of secondary
op

prevention among individuals at high risk in terms of gene–environment risk

profiles. Thus, despite an enormous amount of progress so far, research on the
C

determinants of aggressive behavior still has a long way to go.

180 KENNETH A. DODGE

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 181

REFERENCES

Archer, J. (2000). Sex differences in aggression between heterosexual partners: A

meta-analytic review. Psychological Bulletin, 126, 651–680. doi:10.1037/0033-

A)
2909.126.5.651
Arnett, J. J. (2006). The psychology of emerging adulthood: What is known, and

P
(A
what remains to be known? In J. J. Arnett & J. L. Tanner (Eds.), Emerging adults
in America: Coming of age in the 21st century (pp. 303–330). Washington, DC:

n
American Psychological Association. doi:10.1037/11381-013

io
Brendgen, M., Vitaro, F., Tremblay, R. E., & Lavoie, F. (2001). Reactive and pro-

at
active aggression: Predictions to physical violence in different contexts and mod-

ci
erating effects of parental monitoring and caregiving behavior. Journal of

so
Abnormal Child Psychology, 29, 293–304. doi:10.1023/A:1010305828208

As
Boyce, W. T., & Ellis, B. (2005). Biological sensitivity to context: I. An evolutionary–
developmental theory of the origins and functions of stress reactivity. Develop-

al
ment and Psychopathology, 17, 271–301. doi:10.1017/S0954579405050145

ic
Brunner, H. G., Nelen, M., Breakefield, X., Ropers, H., & van Oost, B. (1993, Octo-

og
ber). Abnormal behavior associated with a point mutation in the structural gene
ol
for monoamine oxidase A. Science, 262, 578–580. doi:10.1126/science.8211186
ch
Cases, O., Seif, I., Grimsby, J., Gaspar, P., Chen, K., Pournin, S., . . . De Maeyer, E.
y

(1995, June). Aggressive behavior and altered amounts of brain serotonin and
Ps

norepinephrine in mice lacking MAOA. Science, 268, 1763–1766. doi:10.1126/

science.7792602
an

Caspi, A., McClay, J., Moffitt, T. E., Mill, J., Martin, J., Craig, I.,W., . . . Poulton, R.
ic

(2002, August). Role of genotype in the cycle of violence in maltreated chil-

er

dren. Science, 297, 851–854. doi:10.1126/science.1072290

Am

Charney, D. S. (2004). Psychobiological mechanisms of resilience and vulnerability:

Implications for successful adaptation to extreme stress. The American Journal of
e

Psychiatry, 161, 195–216. doi:10.1176/appi.ajp.161.2.195

th

Cloninger, C. R., Sigvardsson, S., Bohman, M., & von Knorring, A. (1982). Predis-
by

position to petty criminality in Swedish adoptees. Archives of General Psychia-

try, 39, 1242–1247.
11

Collins, W. A., & van Dulmen, M. (2006). “The course of true love(s) . . . ”: Origins
20

and pathways in the development of romantic relationships. In A. C. Crouter

& A. Booth (Eds.), Romance and sex in adolescence and emerging adulthood: Risks
©

and opportunities (pp. 63–86). Mahwah, NJ: Erlbaum.

ht

Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social information-
ig

processing mechanisms in children’s social adjustment. Psychological Bulletin,

yr

115, 74–101. doi:10.1037/0033-2909.115.1.74

op

Crockett, M. J., Clark, L., Tabibnia, G., Lieberman, M. D., & Robbins, T. W. (2008,
June). Serotonin Modulates Behavioral Reactions to Unfairness. Science, 320,
C

1739. doi:10.1126/science.1155577

SOCIAL INFORMATION PROCESSING PATTERNS 181

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 182

Crozier, J. C., Dodge, K. A., Fontaine, R. G., Lansford, J. E., Bates, J. E., Pettit, G. S.,
& Levenson, R. W. (2008). Social information processing and cardiac predic-
tors of adolescent antisocial behavior. Journal of Abnormal Psychology, 117,
253–267. doi:10.1037/0021-843X.117.2.253
Dick, D. M., Bierut, L., Hinrichs, A., Fox, L., Bucholz, K. K., Kramer, J., . . . Foroud,

A)
T. (2006). The role of GABRA2 in risk for conduct disorder and alcohol and

P
drug dependence across developmental stages. Behavior Genetics, 36, 577–590.

(A
doi:10.1007/s10519-005-9041-8

n
Dodge, K. A. (1980). Social cognition and children’s aggressive behavior. Child

io
Development, 51, 162–170. doi:10.2307/1129603

at
Dodge, K. A., Bates, J. E., & Pettit, G. S. (1990, December). Mechanisms in the cycle

ci
of violence. Science, 250, 1678–1683. doi:10.1126/science.2270481

so
Dodge, K. A., Coie, J. D., & Lynam, D. (2006). Aggression and antisocial behavior

As
in youth. In N. Eisenberg, W. Damon, & R. M. Lerner (Eds.), Handbook of child
psychology: Vol. 3. Social, emotional, and personality development (6th ed.,

al
pp. 719–788). New York, NY: Wiley.

ic
Dodge, K. A., Lansford, J. E., Burks, V. S., Bates, J. E., Pettit, G. S., Fontaine, R., &

og
Price, J. M. (2003). Peer rejection and social information-processing factors in
ol
the development of aggressive behavior problems in children. Child Develop-
ment, 74, 374–393. doi:10.1111/1467-8624.7402004
ch
Dodge, K. A., & Pettit, G. S. (2003). A biopsychosocial model of the development
y
Ps

of chronic conduct problems in adolescence. Developmental Psychology, 39,

349–371. doi:10.1037/0012-1649.39.2.349
an

Dodge, K. A., Pettit, G. S., & Bates, J. E. (1994). Socialization mediators of the rela-
tion between socioeconomic status and child conduct problems. Child Develop-
ic

ment, 65, 649–665. doi:10.2307/1131407

er

Dodge, K. A., Pettit, G. S., Bates, J. E., & Valente, E. (1995). Social information-
Am

processing patterns partially mediate the effect of early physical abuse on later
conduct problems. Journal of Abnormal Psychology, 104, 632–643. doi:10.1037/
e

0021-843X.104.4.632
th

Dodge, K. A., Pettit, G. S., McClaskey, C. L., & Brown, M. (1986). Social compe-
by

tence in children. Monographs of the Society for Research in Child Development,

51(2), 1–85.
11

Edwards, A. C., Dodge, K. A., Latendresse, S. J., Lansford, J. E., Bates, J. E., Pettit,
20

G. S., . . . Dick, D. M. (2009). MAOA-uVNTR and early physical discipline

interact to influence delinquent behavior. Journal of Child Psychology and Psy-
©

chiatry. doi:10.1111/j.1469-7610.2009.02196.x
ht

Eisenberger, N. I., Way, B. M., Taylor, S. E., Welch, W. T., & Lieberman, M. D.
ig

(2007). Understanding genetic risk for aggression: Clues from the brain’s
yr

response to social exclusion. Biological Psychiatry, 61, 1100–1108. doi:10.1016/

j.biopsych.2006.08.007
op

Evers, E. A. T., Van der Veen, F. M., Van Deursen, J. A., Schmitt, J. A. J., Deutz,
C

N. E. P., & Jolles, J. (2006). The effect of acute tryptophan depletion on the BOLD

182 KENNETH A. DODGE

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 183

response during performance monitoring and response inhibition in healthy male

volunteers. Psychopharmacology, 187, 200–208. doi:10.1007/s00213-006-0411-6
Fontaine, R. G., & Dodge, K. A. (2006). Real-time decision making and aggressive
behavior in youth: A heuristic model of response evaluation and decision
(RED). Aggressive Behavior, 32, 604–624. doi:10.1002/ab.20150

A)
Fontaine, R. G., Yang, C., Dodge, K. A., Bates, J. E., & Pettit, G. S. (2008). Testing an

P
individual systems model of Response Evaluation and Decision (RED) and anti-

(A
social behavior across adolescence. Child Development, 79, 462–475. doi:10.1111/

n
j.1467-8624.2007.01136.x

io
Ganzel, B., Casey, B. J., Glover, G., Voss, H. U., & Temple, E. (2007). The after-

at
math of 9/11: Effect of intensity and recency of trauma on outcome. Emotion, 7,

ci
227–238. doi:10.1037/1528-3542.7.2.227

so
Graham, S., & Hudley, C. (1993). An attributional intervention to reduce peer-

As
directed aggression among African-American boys. Child Development, 64,
124–138.

al
Guerra, N. G., & Slaby, R. G. (1989). Evaluative factors in social problem solving by

ic
aggressive boys. Journal of Abnormal Child Psychology, 17, 277–289.

og
Jaffee, S. R., Caspi, A., Moffitt, T. E., Dodge, K. A., Rutter, M., Taylor, A., & Tully,
ol
L. A. (2005). Nature × nurture: Genetic vulnerabilities interact with physical
ch
maltreatment to promote conduct problems. Development and Psychopathology,
17, 67–84. doi:10.1017/S0954579405050042
y
Ps

Kim-Cohen, J., Caspi, A., Taylor, A., Williams, B., Newcombe, R., Craig, I. W., &
Moffit, T. E. (2006). MAOA, maltreatment, and gene–environment interaction
an

predicting children’s mental health: New evidence and a meta-analysis. Molec-

ular Psychiatry, 11, 903–913. doi:10.1038/sj.mp.4001851
ic
er

Lansford, J. E., Malone, P. S., Dodge, K. A., Crozier, J. C., Pettit, G. S., & Bates, J. E.
(2006). A 12-year prospective study of patterns of social information processing
Am

problems and externalizing behaviors. Journal of Abnormal Child Psychology, 34,

709–718. doi:10.1007/s10802-006-9057-4
e

Lansford, J. E., Miller-Johnson, S., Berlin, L. J., Dodge, K. A., Bates, J. E., &
th

Pettit, G. S. (2007). Early physical abuse and later violent delinquency: A

by

prospective longitudinal study. Child Maltreatment, 12, 233–245. doi:10.1177/

1077559507301841
11

Liston, C., McEwan, B. S., & Casey, B. J. (2009). Psychosocial stress reversibly dis-
20

rupts prefrontal processing and attentional control. Proceedings of the National

Academy of Sciences of the United States of America, 106, 912–917. doi:10.1073/
©

pnas.0807041106
ht

Lorber, M. F. (2004). Psychophysiology of aggression, psychopathy, and conduct

ig

problems: A meta-analysis. Psychological Bulletin, 130, 531–552.

yr

Meyer-Lindenberg, A., Buckholtz, J. W., Kolachana, B., Hariri, A., Pezawas, L., Blasi,
op

G., . . . Weinberger, D. R. (2006). Neural mechanisms of genetic risk for impul-

sivity and violence in humans. Proceedings of the National Academy of Science
C

Early Edition, 103, 6269–6274. doi: 10.1073/pnas.0511311103

SOCIAL INFORMATION PROCESSING PATTERNS 183

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 184

Moffitt, T. E., Arseneault, L., Jaffee, S. R., Kim-Cohen, J., Koenen, K. C., Odgers,
C. L., . . . Viding, E. (2008). DSM-V conduct disorder: Research needs for an
evidence base. Journal of Child Psychology and Psychiatry, 49, 3–33.
Moffitt, T. E., & Caspi, A. (2001). Childhood predictors differentiate life-course per-

A)
sistent and adolescence-limited antisocial pathways among males and females.
Development and Psychopathology, 13, 355–375. doi:10.1017/ S0954579401002097

P
(A
Morgan, A. B., & Lilienfeld, S. O. (2000). A meta-analytic review of the relation
between antisocial behavior and neuropsychological measures of executive

n
function. Clinical Psychology Review, 20, 113–136. doi:10.1016/S0272-

io
7358(98)00096-8

at
Murphy, F. C., Nimmo-Smith, I., & Lawrence, A. D. (2003). Functional neuro-

ci
anatomy of emotions: A meta-analysis. Cognitive, Affective, & Behavioral Neuro-

so
science, 3, 207–233. doi:10.3758/CABN.3.3.207

As
Nay, S., Dodge, K. A., Lansford, J., Pettit, G. S., & Bates, J. E. (2009). Rejection sen-
sitivity as a mediator of the effect of romantic relationship history on violent behavior

al
in romantic relationships. Unpublished manuscript.

ic
Orobio de Castro, B., Veerman, J. W., Koops, W., Bosch, J. D., & Monshouwer, H. J.

og
(2002). Hostile attribution of intent and aggressive behavior: A meta-analysis.
ol
Child Development, 73, 916–934. doi:10.1111/1467-8624.00447
ch
Ortiz, J., & Raine, A. (2004). Heart rate level and antisocial behavior in children and
y

adolescents: A meta-analysis. Journal of the American Academy of Child and Ado-

Ps

lescent Psychiatry, 43, 154–162. doi:10.1097/00004583-200402000-00010

Pettit, G. S., Lansford, J. E., Dodge, K. A., & Bates, J. E. (2009). Domain specificity in
an

social information processing and violent behavior in early adulthood. Manuscript sub-
ic

mitted for publication.

er

Pezawas, L., Meyer-Lindenberg, A., Drabani, E. M., Verchinski, B. A., Munoz, K. E.,
Am

Kolachana, B. S., . . . Weinberger, D. R. (2005). 5-HTTLPR polymorphism

impacts human cingulate-amygdala interactions: a genetic susceptibility mech-
e

anism for depression. Nature Neuroscience, 8, 828–834. doi:10.1038/nn1463

th

Raine, A. (2002). Annotation: The role of prefrontal deficits, low autonomic arousal
by

and early health factors in the development of antisocial and aggressive behav-
ior in children. Journal of Child Psychology and Psychiatry, and Allied Disciplines,
11

43, 417–434. doi:10.1111/1469-7610.00034

20

Raine, A. (2008). From genes to brain to antisocial behavior. Current Directions in

Psychological Science, 17, 323–328. doi:10.1111/j.1467-8721.2008.00599.x
©

Raine, A., Buchsbaum, M., & LaCasse, L. (1997). Brain abnormalities in murderers
ht

indicated by positron emission tomography. Biological Psychiatry, 42, 495–508.

ig

doi:10.1016/S0006-3223(96)00362-9
yr

Raine, A., Venables, P. H., & Mednick, S. A. (1997). Low resting heart rate at age
op

3 years predisposes to aggression at age 11 years: Evidence from the Mauritius

Child Health Project. Journal of the American Academy of Child and Adolescent
C

Psychiatry, 36, 1457–1464.

184 KENNETH A. DODGE

12051-10_CH09-rev2.qxd 5/25/10 11:53 AM Page 185

Shih, J. C., Chen, K., & Ridd, M. (1999). Monoamine oxidase: From genes to
behavior. Annual Review of Neuroscience, 22, 197–217. doi:10.1146/annurev.
neuro.22.1.197
Silverman, J. G., Raj, A., Mucci, L. A., & Hathaway, J. E. (2001). Dating violence

A)
against adolescent girls and associated substance use, unhealthy weight con-
trol, sexual risk behavior, pregnancy, and suicidality. JAMA, 286, 572–579.

P
doi:10.1001/jama.286.5.572

(A
van Goozen, S. H. M., Fairchild, G., & Harold, G. T. (2008). The role of neurobiolog-

n
ical deficits in childhood antisocial behavior. Current Directions in Psychological

io
Science, 17, 224–228. doi:10.1111/j.1467-8721.2008.00579.x

at
Waldman, I. D. (2008, June). The etiology of hostile perceptual biases and their relation

ci
with children’s aggression. Paper presented at the annual meeting of the Behavior

so
Genetics Association, Louisville, KY.

As
Weiss, B., Dodge, K. A., Bates, J. E., & Pettit, G. S. (1992). Some consequences of
early harsh discipline: Child aggression and a maladaptive social information

al
processing style. Child Development, 63, 1321–1335. doi:10.2307/1131558

ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

SOCIAL INFORMATION PROCESSING PATTERNS 185

C
op
yr
ig
ht
©
12051-10_CH09-rev2.qxd

20
11
5/25/10

by
th
e
11:53 AM

Am
er
ic
Page 186

an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 187

10

PA)
VIOLENCE AND CHARACTER: A CUPS

(A
(CULTURE × PERSON × SITUATION)

n
io
PERSPECTIVE

at
ci
so
DOV COHEN AND ANGELA K.-Y. LEUNG

As
al
ic
og
ol
y ch
Ps
an

Individual-differences and cultural approaches to the study of violence

ic

have both contributed greatly to our understanding, but both approaches face
er

some inherent difficulties. A cultural approach that ignores individual differ-

Am

ences can explain differences between cultures in rates of violence, but it has
difficulty predicting who will be violent and what this means for the violent
e

individuals’ behavior in other contexts. Two problems with an individual-

th

difference approach that ignores culture are that (a) the approach may fail to
by

explain differences in rates of violence between societies and (b), perhaps

more problematically, it may discover “truths” about violence that hold in
11

only one cultural context but do not hold, or even reverse, in others.
20

In this chapter, we argue for the value of combining an individual-

differences approach with a cultural-differences approach in a way that treats
©

both individual differences and cultural differences seriously. Specifically, we

ht

argue for taking culture seriously by examining differences between cultural

ig

logics. Such logics structure behaviors, situations, scripts, and values in ways
yr

that make sense to people within a culture, even if they do not make sense to
op

people outside that culture. We also argue for taking individual differences
seriously by treating people as more than cultural robots who mechanically
C

follow cultural dictates. In the combined approach advocated here, the key

187
12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 188

notion is that individuals are always in a culture, though they are not always
of it. We outline this combined approach, the CuPS approach (culture ×
person × situation) and demonstrate its value by describing experiments
conducted with three different cultural groups.

A)
Before doing so, it is important to say what the CuPS approach is not.
CuPS is not a mediational approach. It does not assume that there is some

P
(A
underlying individual difference that plays itself out similarly across cultures,
and it does not assume that cultural differences would simply derive from such

n
an underlying difference. As such, it does not follow the common procedures

io
for studying cultural and individual differences, namely, (a) showing that

at
there are cultural differences between some dependent variables; (b) measur-

ci
ing an individual-difference variable that also predicts the dependent variable

so
and differs in mean levels across cultures; and (c) showing that the cultural dif-

As
ference is entirely a product of this individual difference, disappearing when
the individual difference is controlled for.

al
We have no doubt that this mediational approach is important and that

ic
it is extremely useful for understanding various phenomena. However, our

og
approach differs in that it treats culture seriously by considering the way cul-
ol
tural logics structure and give psychological meaning to behaviors and situa-
ch
tions, and it allows for different cultures to have different cultural logics.
y

Triandis’s (1994) notion of a cultural syndrome and Mischel and colleagues’

Ps

(Mendoza-Denton & Mischel, 2007) conception of personality as defined by

a behavioral signature are both essential for the CuPS approach.
an
ic
er

SYNDROMES AND SIGNATURES

Am

A cultural syndrome is a central organizing theme of a culture (e.g., “honor,”

e

“dignity”) structured by a cultural logic that clusters behaviors, situations,

th

scripts, and values in a coherent, meaningful way. This coherence is internal

by

in that it is coherent according to a distinct cultural logic for a given culture

or subculture, and it may be completely incoherent for outsiders whose own
11

cultures are structured by a different cultural logic. Thus, clusters of behavior

20

that make perfect sense in one culture may seem puzzling to people outside
that culture: “How come he would risk his life to help me out of a jam, but if
©

I insult his mother, he’d kill me?” “How can this person be so trustworthy
ht

when he has no backbone and won’t stand up for himself?” “How can this per-
ig

son be so arrogant that he ignores other people’s opinions, and yet he is also
yr

so kind and dependable?”

op

Individual differences arise because people are not automata following the
dictates of their culture. People can follow the ideals of their culture, or they
C

can reject them. However, their behavior should not be considered random,

188 COHEN AND LEUNG

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 189

because it is in fact patterned. People choose their behavior, but these choices
are often influenced by ready-made cultural templates that structure certain
sorts of behaviors as belonging together. Thus, within an honor culture, an indi-
vidual who endorses retributive violence may be embracing the honor ideal and

A)
thus may be more honest and trustworthy, these virtues also being elements of
the honor ideal. Conversely, within an honor culture, an individual who rejects

P
(A
retributive violence may be rejecting the honor ideals, including those related
to being an honest and trustworthy person. Outside of an honor culture, the

n
pattern is likely quite different. In a nonhonor culture, a person who endorses

io
retributive violence may be rejecting his culture’s ideals, including ideals of

at
honesty and trustworthiness as well as nonviolence. The particular cultural log-

ci
ics have to be understood in greater detail, but the point is that the person ×

so
situation behavioral signatures in one culture may be very different from the

As
person × situation behavioral signatures in another culture.

al
ic
ORDER AND VALUE IN HONOR, DIGNITY,

og
AND FACE CULTURES
ol
ch
Cultures are defined by how they solve certain problems. Two of the
y

most salient of these are the problem of order and the problem of value. The
Ps

first relates to how cooperation and order emerge in a society, with a central
question being: Who has legitimate authority to use violence and when can
an

they use it? The second relates to how worth is provisioned: Is personal worth
ic

something that is inherent in the individual and inalienable, or is worth

er

something that must be socially conferred?

Am

Our research has been mostly concerned with cultural syndromes of

honor, dignity, and face, and we provide brief sketches of these as “ideal
e

types,” centering on the two issues mentioned previously. In a culture of dig-

th

nity, a defining idea is that “each individual at birth possess[es] an intrinsic

by

value at least theoretically equal to that of every other person” (Ayers, 1984,
p. 19). Dignity is inalienable in that others cannot take it away. Thus, a per-
11

son of dignity is relatively impervious to insults or affronts by others. A per-

20

son with a sense of dignity has an integrity that comes from a sturdy moral
core, centered on a belief in the inherent worth of individual human beings.
©

This sturdy moral core—rather than the threat of being shamed by others—
ht

is presumed to keep the individual behaving correctly, regardless of the whims

ig

of the situation, the temptations of expediency, or the desire to curry favor

yr

with others. When individual conscience fails, however, an effective system

op

of law enforcement is there to back it up. Dignity cultures tend to be found

in societies such as market economies, where individuals act as relatively
C

autonomous agents who work within a system of law that protects property

VIOLENCE AND CHARACTER 189

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 190

rights and punishes predation. Violations of the rules are offenses against the
state, and it is the state that punishes, rather than vigilantes.
In a culture of honor, honor is both internal and external; it must be
claimed from and paid by others. Honor is defined by Pitt-Rivers (1966) as

A)
follows:

P
The value of a person in his own eyes, but also in the eyes of his society.

(A
It is his estimation of his own worth, his claim to pride, but it is also the
acknowledgement of that claim . . . his right to pride. (p. 21)

n
io
An individual must assertively strive for honor, but if that honor is not ulti-

at
mately conferred by other people, it is a big problem. Thus, honor requires for-

ci
titude but is itself fragile and precarious. In Ayers’s (1984) metaphor, it is

so
unlike dignity, which resembles

As
an internal skeleton, . . . a hard structure at the center of the self; honor,

al
on the other hand, resembles a . . . suit of armor that, once pierced, leaves

ic
the self . . . no alternative except to strike back in desperation. (p. 20)

og
The fragility of honor derives from the context in which cultures of
ol
honor typically develop. They arise in contexts where the state is weak, where
ch
there is no effective law enforcement, no mechanism to guarantee contracts,
no police to prevent predation, and no justice system to punish the guilty (Pitt-
y
Ps

Rivers, 1966). In such an environment, “every man should be sheriff on his

own hearth” (as an old North Carolina proverb put it; Fischer, 1988, p. 765).
an

A reputation for reciprocity is key here because one must be known as some-
ic

one who will pay back, who can be relied on to pay back a good turn, who will
er

deliver on promises (and threats), and who will avenge wrongs done to him
Am

(see Chapter 12, this volume). Insults and trivial conflicts take on great impor-
tance in such cultures because they are probes or tests of who can do what to
e

whom. A man who will not let himself be trifled with on small matters surely
th

will not let himself be trifled with on big matters either. In this context, it pays
by

to have a reputation as a man of honor who will show positive reciprocity and,
if wronged, a thirst for vengeance (see Chapters 3 and 12, this volume).
11

Cultures of face exist within strong hierarchical structures. Face, as

20

defined by Ho (1976) and by Heine, Lehman, Markus, and Kitayama (1999),

is “the respectability and/or deference which a person can claim . . . by virtue
©

of [his or her] relative position” in a hierarchy and is the proper fulfillment of

ht

his or her role. A person’s performance in his or her role is not for him or
ig

her to judge but rather must be judged by others; thus, face has primarily an
yr

external quality (Kim & Cohen, 2010). Face can be gained, but the focus is
op

mostly on not losing it—for oneself and for others one interacts with. Indeed,
unlike honor cultures, where one person can often increase his honor by tak-
C

ing someone else’s, in face cultures people often work together to save each

190 COHEN AND LEUNG

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 191

other’s face, avoiding the direct conflicts that become such important con-
tests for status in competitive honor cultures.
Bad behavior in a face culture is not supposed to be punished by the vic-
tim. Doing so would only further upset the harmony that is supposed to pre-

A)
vail in the stable hierarchy of a face culture. Instead, the group or someone
further up the hierarchy will do the punishing. Generally, the three Hs that

P
(A
characterize face cultures are harmony, hierarchy, and humility. The latter,
humility, is quite functional, because in a system where face must be socially

n
conferred, it is not good to claim too much for oneself; otherwise, one may

io
receive a painful and humiliating lesson about where one really falls in the

at
status hierarchy.

ci
These sketches of three kinds of cultures are ideal types (Weber, 1920/

so
1997), and ideal types do not actually exist in the world. However, the sketches

As
are useful because they lead us to predict different behavioral signatures among
people from honor, dignity, and face cultures. It is not the case that individ-

al
uals all follow the ideals of their culture. However, when individuals depart

ic
from the ideals, their departures tend to be systematic. Individuals who depart

og
from the ideals are not simply people who are oblivious to cultural norms, and
ol
they are not random noise. Rather, their behavior—even when they reject
ch
their culture’s dominant ideals—tends to fit preexisting cultural templates,
y

patterned by the culture’s logic even as its ideals are rejected. Again, individ-
Ps

uals are always in a culture, even if they are not always of it.
an
ic

TWO EXPERIMENTS
er
Am

We attempted to demonstrate how this works with two laboratory

experiments showing the different person × situation behavioral signatures
e

that occur in different cultures. In our first study we examined positive reci-
th

procity, and in our second we examined the virtue of trustworthiness, as

by

shown by whether participants cheated on a memory test to earn money. In

both studies, University of Illinois students were participants. Our dignity cul-
11

ture group comprised northern Anglo American students, our honor culture
20

group comprised Latino and southern Anglo American students, and our face
culture group comprised Asian Americans whose ancestors were from East
©

Asia (for supporting evidence, see Ayers, 1984; Cohen, Hoshino-Browne, &
ht

Leung, 2007; Hamamura & Heine, 2008; Heine et al., 1999; Ho, 1976; Kim,
ig

Cohen, & Au, in press; Triandis, 1994; Vandello & Cohen, 2003; Vandello,
yr

Cohen, Grandon, & Franiuk, 2009). Obviously, all of the participants were
op

American, so we speak of our groups not as coming from different societies

but as people whose relative familiarity with dignity, face, and honor differs
C

as a function of the dominant theme within their subculture.

VIOLENCE AND CHARACTER 191

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 192

In both experiments, we obtained individual-difference measures of the

participant’s (a) endorsement of honor-related violence and (b) belief that a per-
son’s worth is either inalienable or must be socially conferred. The measure of
(a) was obtained by showing video clips in which someone responded violently

A)
to an insult, affront, or threat, and the participants had to indicate how justified
the violence was, how much the insulter deserved to be beaten, how reasonable

P
(A
the violent person’s actions were, and so on. The measure of (b) was obtained
with simple Likert scale items: “How others treat me is irrelevant to my worth

n
as a person,” “How much I respect myself is far, far more important than how

io
much others respect me,” “No one (except me) can make me feel diminished,”

at
and “No one can take a person’s self-respect away from him or her.” Participants’

ci
endorsements of (a) or (b) were person-level variables in the design.

so
In Experiment 1, participants were given a cover story, according to

As
which we were examining people’s attitudes toward violence in the media.
This was the putative reason for showing film clips and asking about violence.

al
The cover story was followed by a questionnaire that asked about demograph-

ic
ics and also contained the questions described earlier concerning inalienable

og
versus socially conferred worth. As far as the participant knew, that was the
ol
entire experiment. However, there was, of course, more to it than this. The
ch
participant came into the study with two fellow confederates. On learning that
y

the experiment involved watching movies, one of the confederates pulled out
Ps

a small bag of candies and offered some to the participant and the other con-
federate. In half the cases, the candy offerer turned out to be the “disk confed-
an

erate” and in the other half, the offerer turned out to be the “distraction
ic

confederate” (the reason for the names will become obvious in a moment).
er

After filling out their questionnaires, participants and confederates

Am

received a false debriefing and were then dismissed. After they left the lab room,
the distraction confederate buttonholed the participant and started talking to
e

him as the disk confederate rummaged through his own backpack at a table
th

15 feet down the hall. The disk confederate pulled a crumpled piece of paper
by

out of his backpack and asked the participant and confederate, “Do either of
you know where room 25 is? I’ve got to meet a study group there in 5 minutes.”
11

Either the participant or the distraction confederate would reply that room 25
20

was in the basement, and after learning this, the disk confederate packed his
bag up and left. In doing so, however, he “inadvertently” left behind and forgot
©

his bright, neon-colored computer disk that now lay on the desk.
ht

After allowing the disk confederate to make his getaway, the distraction
ig

confederate ended his conversation with the participant. Both of them then
yr

walked down the hall, having to walk past the “lost” disk. If the participant
op

did not notice the disk, the confederate drew attention to it and gave the par-
ticipant a few subtle opportunities to volunteer to find the disk confederate.
C

If the participant did not take up the offer, the study ended. If the participant

192 COHEN AND LEUNG

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 193

did take up the offer, he needed to find the disk confederate, not an easy task
given that (a) the directional signs in the University of Illinois psychology
building are confusing, contradictory, and occasionally completely mislead-
ing (pointing to the wrong floor) and (b) room 25 is actually a locked broom

A)
closet, tucked behind a set of doors at the bottom of a ramp.
Though a few participants could not or did not find room 25, the major-

P
(A
ity of those who attempted to do so succeeded. (Participant progress was mon-
itored by confederates hidden throughout the building and by the exits,

n
communicating via walkie-talkies.) If a participant reached room 25, however,

io
his journey might continue, because room 25 had a sign on the door, “Meet-

at
ing moved to room 841 (8th floor).” The participant then had to decide either

ci
to walk up the nine floors, take the elevator (if he had the time—the elevators

so
were slow), or abandon the search altogether (the building’s exit was located

As
tantalizingly nearby). For those who chose to go on to the eighth floor and
found room 841 (a broom closet again), there was another sign on the door,

al
“Went to get TV and VCR. Will be back soon.” A confederate hidden in the

ic
hallway watched to see whether the participant waited for at least 1 minute

og
for someone to return. The participant’s degree of helping could thus be mea-
ol
sured, with higher scores indicating that the participant had gone further in
ch
trying to find the disk confederate. Because we manipulated whether it was the
y

disk confederate or the distraction confederate who offered the participant

Ps

candy, this situational manipulation dictated whether finding the disk confed-
erate was or was not an act of reciprocity for his offer of the candy.
an

The key prediction involved a three-way interaction: culture (honor vs.

ic

nonhonor culture) × person (high vs. low endorsement of honor-related vio-

er

lence) × situation (helping is reciprocity vs. not reciprocity). The interaction

Am

was significant in a regression analysis. In analyzing the interaction, we found

two very different patterns for our honor and nonhonor respondents. Among
e

the honor group in the reciprocity condition, there were people who paid
th

back and people who did not. That is, those who paid back insults and affronts
by

(“negative gifts,” as Miller, 1993, called them) were those who also went the
furthest to pay back a favor, traipsing all over the psychology building to help
11

out the person who offered them a piece of candy. Thus, the people who paid
20

back the negative were also the people who paid back the positive. However,
in the nonhonor groups, the pattern was the opposite: Those who rejected
©

retributive violence were most likely to return the favor to the disk confed-
ht

erate who had offered them candy. “Good” people eschewed violence and good
ig

people repaid the confederate.

yr

The moral obligation to repay the confederate was binding for both
op

honor culture participants who embraced the ideal of honor and for face and
dignity participants who embraced the ideals of face and dignity, respectively.
C

(With regard to this latter distinction, we further found that the dignity culture

VIOLENCE AND CHARACTER 193

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 194

participants who were most likely to pay back a favor were also the ones who
endorsed the notion that personal worth is inalienable; conversely, among
nondignity participants, this relationship between paying back favors and
endorsing the principle of inalienable worth did not hold.) Finding the disk

A)
confederate was a matter of fulfilling a duty to reciprocate rather than a sim-
ple act of altruism. The effects described earlier held for the situation where

P
(A
the disk confederate had offered the participant candy; they did not hold
when the disk confederate had not offered candy.

n
The experiment illustrates in static form the argument that (a) cultural

io
syndromes differ in the way they cluster behaviors together and (b) individ-

at
uals position themselves toward or against the dominant syndromes of their

ci
culture or subculture. The experiment we describe next demonstrates this

so
dynamically, illustrating the way this process works in microcosm. That is,

As
our bicultural honor culture participants (i.e., Latinos and southern Anglos
at the University of Illinois) may live in two worlds and may structure their

al
behavior according to different cultural logics, depending on which one hap-

ic
pens to be salient for them in a given situation.

og
The University of Illinois at Urbana–Champaign is an overwhelmingly
ol
White campus in a northern part of the state. Most psychology experiments
ch
there take place in a laboratory room in the Department of Psychology. The
y

default ideal in such a context is dignity. In the experiment we just described,

Ps

honor was “primed” in that all participants watched the film clips before hav-
ing to find the confederate. In the next experiment, half the participants were
an

unprimed, whereas the other half were primed with honor, watching the film
ic

clips before engaging in the crucial behavior (again, as in the experiment

er

already described).
Am

In this study, we also examined a different behavior, one related to trust-

worthiness and integrity (an ideal embraced under the mantle of honor in
e

honor cultures, the mantle of dignity in dignity cultures, and the mantle of
th

face in face cultures). The trustworthiness ideal is a universal good, but

by

depending on the cultural syndrome from which it derives, it can be clustered

with different sorts of behaviors.
11

For this study, we adapted a procedure from Houston and Ziff (1976),
20

giving participants an opportunity to win money by cheating on a memory

test. Briefly, the participant was instructed to remember words from two lists;
©

the experimenter “accidentally” left one of these two lists exposed, and the
ht

dependent variable was (roughly) the likelihood that the participant cheated
ig

(as opposed to remembering words from the exposed list due to chance
yr

alone). Our person-level variable for the endorsement of retributive violence

op

was measured in the same way as in the previous experiment; however, for
half the participants, the honor–violence film clips were shown before the
C

memory test was given and for the other half, the film clips came after.

194 COHEN AND LEUNG

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 195

Again, the key prediction concerned a three-way interaction: culture

(honor vs. nonhonor culture) × person-level endorsement of violence (high
vs. low) × situation (honor is primed vs. honor is not primed). The prediction
can be illustrated by imagining two hypothetical people. Imagine person A

A)
from an honor culture who endorses the notion that insults and affronts need
to be paid back with violence if necessary. Imagine person B from an honor

P
(A
culture, a person who lives by the mantra “Sticks and stones may break my
bones but names will never hurt me” and thus rejects honor-related violence.

n
In the unprimed condition, in the context of a psychology laboratory on a

io
northern, predominately White campus, the cultural ideal that is salient and

at
that one reacts toward or against is dignity. In this case, person A, who

ci
embraces the idea of paying back insults, is a rejecter of the ideal of dignity,

so
and as a rejecter of dignity’s ideal, should be less likely to act in an upright

As
and trustworthy fashion. However, person B is an endorser of the dignity
ideal, and as such, should be more likely to act with integrity.

al
Now, change the salient cultural syndrome to which A and B are react-

ic
ing by administering the honor prime. When honor is the salient ideal, per-

og
son A is an endorser of honor and, as such, should follow honor’s strictures
ol
against cheating and be more honest. In contrast, person B goes from being
ch
an endorser of dignity to a rejecter of honor and, as a rejecter of honor, should
y

be more likely to cheat. This is what happened in our experiment. The honor
Ps

prime made endorsers of retributive violence less likely to cheat whereas mak-
ing rejecters of retributive violence more likely to cheat.
an

It interesting that the priming effect was most pronounced among those
ic

most steeped in the culture of honor tradition. In the unprimed condition, for
er

both the honor and the nonhonor groups, more endorsement of violence was
Am

correlated with more cheating. In the primed condition, among those who
grew up entirely in the South, the correlation flipped dramatically: For those
e

who grew up entirely in the South, the correlation between honor–violence

th

endorsement and cheating was r = −.87. For these Southerners, endorsing

by

honor violence meant embracing all that honor entails, including honesty.
This did not occur for people from nonhonor groups or for people who spent
11

much less of their life in the South. For these groups, one cannot prime—or
20

perhaps more correctly, cannot easily prime—honor as the salient syndrome

to which one reacts.
©

A few caveats are needed, and they derive from a more general point
ht

about priming. The effectiveness of a prime depends on both (a) the “strength”
ig

of the prime and (b) the participant’s susceptibility to being primed by a given
yr

stimulus. Regarding susceptibility, the strong results with the lifelong South-
op

erners are suggestive but must be regarded with caution due to the small n (we
had only 16 lifelong Southerners in our sample). Although, regarding the
C

strength of the prime, we could not effectively activate honor as an ideal for

VIOLENCE AND CHARACTER 195

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 196

our nonhonor groups, this is not to say that the ideal can never be activated.
Our participants watched five film clips. Perhaps stronger primes could induce
honor as a salient ideal, but in the current studies we have no evidence for
this (see also Cohen, 2007).

A)
Cultural Rejectionism

P
(A
As illustrated previously, being an endorser of honor (when honor is the

n
salient ideal) is not the same thing as being a rejecter of dignity (when dig-

io
nity is the salient ideal). And conversely, being a rejecter of honor is not the

at
same thing as being an endorser of dignity. The same inclination toward or

ci
against retributive violence positions one as either a rejecter or endorser of a

so
cultural system (of honor, dignity, or face), and positioning oneself with

As
respect to this cultural system has implications for a wide variety of behav-
iors, including positive reciprocity, honesty, and (as illustrated in our other

al
work not discussed here) financial as well as political actions (see Leung &

ic
Cohen, 2008).

og
A point that these studies reiterate is that cultural rejecters are not sim-
ol
ply oblivious folks who just don’t “get it.” They are not simply undersocial-
ch
ized. Sometimes they understand the cultural ideals perfectly well and just
y

choose to reject them. In our studies, cultural rejectionists behaved badly in

Ps

ways that were striking. In the disk study, Southerners and Latinos who
rejected honor-related violence actually helped someone less after that per-
an

son offered them candy than they did when the person did not offer them
ic

candy. The same was true of northern Anglo and Asian Americans who
er

endorsed honor-related violence. And the same was true of northern Anglo
Am

Americans who rejected the ideal of inalienable worth. These participants all
helped the confederate less after the confederate had done something nice for
e

them. They behaved most badly when social obligations pressed most heav-
th

ily on them.
by

It would be worthwhile to study further this seeming desire to violate

norms—to act badly when one is most obligated not to—perhaps because of
11

a motivation to assert one’s own autonomy, perhaps because of some other

20

motive, or perhaps simply for the sake of spite. There is a Yiddish word—
aftselakhis—translated “very literally, ‘in order to provoke anger,’ the impulse
©

to do things only because someone else doesn’t want you to” (Wex, 2005,
ht

p. 2). In the case of the experiments described here, the word may not apply
ig

because the participants in our studies were not angering the confederate by
yr

not returning his disk. However, the word perhaps captures some of the spirit
op

of how our cultural rejectionists behaved (or misbehaved) when they faced
social obligations. Again, the deeper underlying motives for their behavior,
C

the feelings they experience when they violate social norms (e.g., glee?

196 COHEN AND LEUNG

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 197

shame? dejection? mastery? freedom?), and the circumstances that provoke

such reactions seem worth studying if we want to avoid what Dennis Wrong
(1961) famously called “the oversocialized conception of man” (p. 183).
There may be individual differences in the desire to act spitefully, but the

A)
results from the second experiment showing that participants from honor
cultures can go from endorsers of dignity to rejecters of honor (or from

P
(A
rejecters of dignity to endorsers of honor) by changing the salience of a cul-
tural syndrome suggests that we should consider not just person-level effects

n
but also the interaction of person-level variables with situational and cul-

io
tural variables as well. (For more on these effects, see Cohen, 2009; Kim &

at
Cohen, 2010.)

ci
so
Rejecting the Logic

As
We have argued that individuals are always in a cultural context,

al
although they are not always of it. That is, people react toward or against the

ic
salient ideals of their culture—embracing them or rejecting them. Individ-

og
uals are “free” to go against the ideals of their culture. Our most important
ol
caveat is that there is another level of freedom beyond this. That is, individ-
ch
uals are free to reject not just the ideals of their culture but also the cultural
y

logic itself that binds together and organizes sets of behaviors, values, and
Ps

scripts. Thus, it is one thing to reject the ideals of honor as a package, to

reject retributive violence, reject prosocial reciprocity, and reject honest
an

behavior. However, it is another thing altogether to reject the logic that

ic

weaves together honor by, for example, embracing the notion that one is
er

obligated to pay back good things but not obligated to pay back bad ones or
Am

by embracing the notion that one can show integrity without having to
stand up for him- or herself if challenged. There are obviously people who
e

do this (though in the first experiment there were not many). It seems worth-
th

while to study how, when, and why people challenge not just the ideals but
by

also the underlying logic of their culture. Again, our second study provides
a clue: At least for our bicultural respondents, one might be able to replace
11

one cultural logic (e.g., honor) with that of another (e.g., dignity), or vice
20

versa (see Leung & Cohen, 2009). Particularly in a rather loose social sys-
tem such as that of the United States, people can operate in not just one cul-
©

ture but also in subcultures and sub-subcultures. Research exploring how

ht

people pick the niches they occupy in a culture or subculture may be partic-
ig

ularly useful (Morris, 2009).

yr

Our caveat does not modify the claim that individuals are always in a
op

cultural context even if they are not always aware of it. However, it does make
the claim more complex by noting that sometimes people can choose the
C

salient cultural or subcultural system they are responding to. “Choice” does

VIOLENCE AND CHARACTER 197

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 198

not apply to our experiments because participants were randomly assigned to

a condition that made honor salient or not. But the point about choice
applies to the larger sense in which people have some freedom to decide which
subcultural systems they will be part of (see also Oishi, 2004; Zou, Morris, &

A)
Benet-Martinez, 2008).

P
(A
CONCLUSION

n
io
In sum, all societies must solve the problem of order. How they solve it

at
and how they conceive of individual worth help to define whether a society

ci
resembles the prototype of an honor, dignity, face, or some other sort of culture.

so
Different cultural logics operate in different cultures, meaning that different

As
sorts of behaviors, scripts, and values get bundled together in different ways.
These bundles are coherent to people in the culture even if they do not always

al
seem coherent to those outside the culture.

ic
Individuals react toward or against the salient ideals of their culture, and

og
even when they choose to reject the culture’s ideals, they are still often guided
ol
by its logic and fit their behavior to preexisting cultural templates. The CuPS
ch
approach helps to explain the distinct patterns of within-culture, as well as
y

between-culture, variation, and it takes both cultural logics and individual

Ps

differences seriously. The acceptance or rejection of salient cultural ideals of

honor, dignity, and face entail consequences for patterning a range of behav-
an

iors beyond violence, including whether one is an honest, trustworthy person

ic

who can be counted on to pay back favors. Individuals are always in a cultural
er

context even if they are not always of it. Further research into the way peo-
Am

ple position themselves with respect to cultural ideals, their motives for doing
so, and the forces and choices that influence the cultural syndrome to which
e

they are responding seems likely to be very productive.

th
by

REFERENCES
11
20

Ayers, E. (1984). Vengeance and justice. New York, NY: Oxford University Press.
Cohen, D. (2007). Methods of cultural psychology. In S. Kitayama & D. Cohen (Eds.),
©

Handbook of cultural psychology (pp. 196–236). New York, NY: Guilford Press.
ht

Cohen, D. (2009). How do personal experience and cultural traditions interact to

ig

give rise to intracultural/regional variations within a national culture? In

yr

R. Wyer, C. Y. Chiu, & Y. Y. Hong (Eds.), Understanding culture (pp. 487–489).

op

New York, NY: Taylor & Francis.

Cohen, D., Hoshino-Browne, E., & Leung, A. K.-y. (2007). Culture and the struc-
C

ture of personal experience: Insider and outsider phenomenologies of the self

198 COHEN AND LEUNG

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 199

and social world. In M. Zanna (Ed.), Advances in experimental social psychology

(Vol. 39, pp. 1–67). San Diego, CA: Academic Press.
Fischer, D. H. (1988). Albion’s seed. New York, NY: Oxford University Press.
Hamamura, T., & Heine, S. (2008). The role of self-criticism in self-improvement and

A)
face maintenance among Japanese. In E. C. Chang (Ed.), Self-criticism and self-

P
enhancement (pp. 105–122). Washington, DC: American Psychological Associ-

(A
ation. doi:10.1037/11624-007
Heine, S. J., Lehman, D., Markus, H., & Kitayama, S. (1999). Is there a universal

n
io
need for positive self-regard? Psychological Review, 106, 766–794. doi:10.1037/

at
0033-295X.106.4.766

ci
Ho, D. (1976). On the concept of face. American Journal of Sociology, 81, 867–884.

so
doi:10.1086/226145

As
Houston, J. P., & Ziff, T. (1976). Effects of success and failure on cheating behavior.
Journal of Educational Psychology, 68, 371–376. doi:10.1037/0022-0663.68.3.371

al
Kim, Y. H., & Cohen, D. (2010). Information, perspective, and the self in face

ic
and dignity cultures. Personality and Social Psychology Bulletin, 36, 537–550.

og
doi:10.1177/0146167210362398
ol
Kim, Y. H., Cohen, D., & Au, W. T. (in press). The jury and abjury of my peers. Jour-
ch
nal of Personality and Social Psychology.
y

Leung, A. K.-y., & Cohen, D. (2008). Within- and between-culture variation: Individ-
Ps

ual differences and the cultural logics of honor, face, and dignity cultures. Unpub-
lished manuscript, Singapore Management University.
an

Leung, A. K.-y., & Cohen, D. (2009). The embodiment of moral systems. Unpublished
ic

manuscript, Singapore Management University.

er

Mendoza-Denton, R., & Mischel, W. (2007). Integrating system approaches to cul-

Am

ture and personality. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural

psychology (pp. 175–195). New York, NY: Guilford Press.
e

Miller, W. (1993). Humiliation. Ithaca, NY: Cornell University Press.

th

Morris, M. (2009). How do personal experience and cultural traditions interact to

by

give rise to intracultural/regional variations within a national culture? In

R. Wyer, C. Y. Chiu, & Y. Y. Hong (Eds.), Understanding culture (pp. 484–487).
11

New York, NY: Taylor & Francis.

20

Oishi, S. (2004). Personality in culture. Journal of Research in Personality, 38, 68–74.

doi:10.1016/j.jrp.2003.09.012
©

Pitt-Rivers, J. (1966). Honour and social status. In J. Peristiany (Ed.), Honour and
ht

shame (pp. 19–78). London, England: Weidenfeld & Nicholson.

ig
yr

Triandis, H. C. (1994). Culture and social behavior. New York, NY: McGraw-Hill.
op

Vandello, J. A., & Cohen, D. (2003). Male honor and female fidelity: Implicit cul-
tural scripts that perpetuate violence. Journal of Personality and Social Psychology,
C

84, 997–1010. doi:10.1037/0022-3514.84.5.997

VIOLENCE AND CHARACTER 199

12051-11_CH10-rev2.qxd 5/25/10 11:54 AM Page 200

Vandello, J., Cohen, D., Grandon, R., & Franiuk, R. (2009). Stand by your man. Jour-
nal of Cross-Cultural Psychology, 40, 81–104. doi:10.1177/0022022108326194
Weber, M. (1997). The theory of social and economic organization. New York, NY:
Free Press. (original work in German published 1920)

A)
Wex, M. (2005). Born to kvetch. New York, NY: St. Martin’s Press.

P
Wrong, D. (1961). The oversocialized conception of man in modern sociology.

(A
American Sociological Review, 26, 183–193. doi:10.2307/2089854
Zou, X., Morris, M., & Benet-Martinez, V. (2008). Identity motives and cultural

n
io
priming. Journal of Experimental Social Psychology, 44, 1151–1159. doi:10.1016/

at
j.jesp.2008.02.001

ci
so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

200 COHEN AND LEUNG

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 201

III

A)
P
(A
PSYCHOLOGICAL AND

n
io
RELATIONAL PROCESSES

at
ci
so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C
C
op
yr
ig
ht
©
20
12051-12_PT3-CH11-rev2.qxd

11
by
5/25/10

th
e
Am
11:54 AM

er
ic
an
Page 202

Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 203

11

A)
P
MIRROR, MIRROR, ON THE WALL,

(A
WHO’S THE MOST AGGRESSIVE

n
io
OF THEM ALL? NARCISSISM,

at
ci
SELF-ESTEEM, AND AGGRESSION

so
As
SANDER THOMAES AND BRAD J. BUSHMAN

al
ic
What we have to fight for is the necessary security for the existence and

og
increase of our race and people, the subsistence of its children and the
ol
maintenance of our racial stock unmixed, the freedom and independence
ch
of the Fatherland; so that our people may be enabled to fulfill the mission
y

assigned to it by the Creator.

Ps

—Adolf Hitler
an

Do aggressive people typically think poorly of themselves? Are they full

ic

of self-doubt and feelings of inadequacy? Do they feel inferior to others? Or,

er

instead, are aggressive people typically absorbed with themselves, thinking

Am

they are better than others? Do they overestimate rather than underestimate
their qualities? Although controversies still exist, research has provided sev-
e

eral important insights into the link between self-views and aggression. This
th

chapter provides an overview of historical perspectives and more recent

by

research findings; it addresses several existing controversies (and suggests

ways to solve them), discusses possible interventions for reducing aggression,
11

and recommends directions for future research.

20
©

HISTORICAL PERSPECTIVES
ht
ig

One long-standing view in psychology has held that low self-esteem

yr

causes aggression. Although the origins of this view are difficult to establish
op

(there is no landmark work that made the original observation or theoretical

statement that aggressive people suffer from low self-esteem), a few reasons may
C

explain why this notion sneaked into conventional wisdom. Most important,

203
12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 204

we believe, is that it is intuitively compelling to believe that low self-esteem is

a cause of aggression. Having low self-esteem feels bad, whereas having high
self-esteem feels good. It may therefore seem logical to infer that having low
self-esteem is associated with bad things (e.g., behaving aggressively) and hav-

A)
ing high self-esteem is associated with good things (e.g., behaving prosocially).
Moreover, it is well established that unpleasant events, which make people

P
(A
feel bad, increase aggression (e.g., Berkowitz, 1983). As Nathaniel Branden
(1984), one of the most fervent advocates of the idea that self-esteem is

n
the key determinant of good adjustment, said: “I cannot think of a single

io
psychological problem—from anxiety and depression, to fear of intimacy or

at
of success, to spouse battery or child molestation—that is not traceable to

ci
the problem of low self-esteem” (p.12).

so
A few scholars have gone beyond correlations and tried to explain why

As
low self-esteem should cause aggression. According to one explanation, peo-
ple generally want to experience high self-esteem, so individuals with low

al
self-esteem should suffer from distress and frustration and may behave aggres-

ic
sively in an attempt to feel better about themselves. Aggression is viewed as

og
a strategy to regulate the pain and distress associated with one’s negative self-
ol
feelings (Horney, 1950; Toch, 1969/1993). According to a second explana-
ch
tion, individuals with low self-esteem often reject societal norms, which
y

include norms restricting aggressive behavior (Rosenberg, 1965). In the absence

Ps

of firm empirical data, it was understandable for early psychologists to believe

that low self-esteem is a cause of aggression. More recently, however, several
an

well-designed, rigorous studies have examined the relationship between self-

ic

views and aggression. Is it still valid to believe that low self-esteem is a cause of
er

aggression?
Am

Does Low Self-Esteem Influence Aggressive Behavior?

e
th

In 1996, Roy Baumeister and his colleagues reviewed the literature and
by

found little support for the view that low self-esteem increases aggression
(Baumeister, Smart, & Boden, 1996). In fact, they concluded that inflated
11

self-views increase aggression, especially when people suffer a blow to their

20

ego. It is possible, however, that more recent findings have yielded a different
picture, especially because the review sparked reevaluations and much new
©

research (i.e., it has been cited hundreds of times). Here we provide an

ht

overview of research conducted since 1996.

ig
yr

Research on the Low Self-Esteem Hypothesis as it Applies to Adults

op

Research involving adult participants (mostly undergraduate college

C

students) has typically found a negative association between self-esteem and

204 THOMAES AND BUSHMAN

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 205

self-reported aggression (e.g., Bradshaw & Hazan, 2006; Donnellan,

Trzesniewski, Robins, Moffitt, & Caspi, 2005, Study 3; D’zurilla, Chang, &
Sanna, 2003). Negative correlations in the range of −.20 to −.50 have typically
been found between self-esteem and self-reported hostility and anger. More

A)
pertinent to the present purposes, weak but significant negative correlations
have also been found between self-esteem and self-reported physical aggres-

P
sion. For example, one study found a correlation of −.11 in a sample of more

(A
than 3,000 undergraduate college students (Donnellan et al., 2005, Study 3).

n
The main exception to this pattern of findings is the absence of a correlation

io
between low self-esteem and self-reported verbal aggression. Thus, although

at
low self-esteem individuals do not claim to be more verbally aggressive, they

ci
do claim to be more prone to anger, more likely to harbor hostile feelings

so
toward others, and more likely to be physically aggressive than their high

As
self-esteem counterparts.
Do these findings justify the claim that low self-esteem is an important

al
cause of aggression? We believe not. Self-report measures are often problem-

ic
atic (e.g., Nisbett & Wilson, 1977), especially when examining aggression.

og
Self-report studies typically find that high self-esteem individuals overstate
ol
their good qualities and understate their negative traits (Baumeister, Camp-
ch
bell, Krueger, & Vohs, 2003). Aggression is generally considered to be socially
y

unacceptable, so people with high self-esteem should report low aggression.

Ps

Thus, studies that relate self-esteem to self-reports of aggression are relevant

only to the extent that their findings generalize to more objective measures
an

of aggression. We know of one study that found a negative link between

ic

self-esteem and a behavioral measure of provoked aggression (i.e., prepar-

er

ing samples of hot sauce for a provocative confederate to consume; Webster

Am

& Kirkpatrick, 2006). However, this effect became significant only after a
range of closely related self-regard variables (i.e., superiority, mate value,
e

social inclusion) were statistically controlled. Moreover, another study in

th

the same lab found no link between self-esteem and the same behavioral
by

measure of aggression (Kirkpatrick, Waugh, Valencia, & Webster, 2002). It is

this latter finding that is consistent with the majority of findings based on
11

objective aggression measures. Whether these measures involved blasting

20

opponents with painful noise (e.g., Bushman & Baumeister, 1998; Twenge &
Campbell, 2003), giving a confederate less money (Bushman, Bonacci, van Dijk,
©

& Baumeister, 2003), or third-party observations of real-world aggression

ht

(Goldberg et al., 2007), the bulk of studies found no correlations, or even

ig

small positive correlations between self-esteem and objectively observed

yr

aggression. In summary, the research conducted after Baumeister and col-

op

leagues published their 1996 review has provided no compelling reason to

challenge their conclusion that low self-esteem does not cause aggression and
C

violence.

NARCISSISM, SELF-ESTEEM, AND AGGRESSION 205

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 206

Research on the Low Self-Esteem Hypothesis as It Applies to Children

It is possible that low self-esteem does not cause aggression in adulthood

but does cause aggression in children. Indeed, aggression is much more com-

A)
mon in late childhood and adolescence than it is in adulthood, perhaps
increasing the likelihood of establishing a link with low self-esteem. In addi-

P
(A
tion, late childhood and adolescence are developmental periods marked by
profound concern with maintaining desired self-images (Harter, 2006). If it

n
is true that aggression is a regulatory strategy people use to protect their feel-

io
ings of self-worth (Morf & Rhodewalt, 2001), then children and adolescents

at
who are low in self-esteem may be more likely than their adult counterparts

ci
to behave aggressively.

so
A pair of studies that used the same longitudinal sample (i.e., a complete

As
birth cohort of over 1,000 participants from Dunedin, New Zealand) provided
evidence that appears to be consistent with this reasoning (Donnellan et al.,

al
2005, Study 2; Trzesniewski et al., 2006). One study found a small but signif-

ic
icant negative relationship between self-esteem at age 11 and teacher- and

og
parent-reported externalizing problems (including some items assessing
ol
aggression) at age 13 (Donnellan et al., 2005). The other study found a nega-
ch
tive relationship between young adolescents’ self-esteem and court convictions
y

for violent offenses in adulthood (Trzesniewski et al., 2006). It is important

Ps

that these findings were based on observational measures of aggression and

externalizing problems, so they cannot be explained by self-report biases.
an

Unfortunately, however, these studies were limited in one important

ic

respect. They did not rule out the possibility that the link between low self-
er

esteem and aggression was actually due to social or contextual factors related to
Am

both self-esteem and (later) violent and aggressive behavior. This is important
because third variables often inflate the relationship between self-esteem and
e

its supposed correlates (Baumeister et al., 2003). A different group of researchers

th

tried to replicate the findings from these two studies using a similar New Zealand
by

birth cohort of over 1,000 participants (Boden, Fergusson, & Horwood,

2007). These researchers also found that low self-esteem at ages 10 and 15 pre-
11

dicted aggressive and violent behavior at ages 18, 21, and 25. However, the rela-
20

tionship became nonsignificant when the researchers controlled for Maori

ethnicity, parent education, and other family background factors that were
©

potentially confounded with self-esteem. They concluded the following:

ht

While it may be possible to observe bivariate relationships between self-

ig

esteem and later violent and aggressive behavior, these associations are
yr

in fact very modest in nature, and can be explained by the effects of fam-
op

ily background and contextual factors that are confounded with self-
esteem, rather than the direct effects of self-esteem per se. (Boden et al.,
C

2007, p. 888)

206 THOMAES AND BUSHMAN

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 207

Taken together, many studies have appeared since Baumeister and col-
leagues (1996) challenged the notion that low self-esteem causes aggression. A
few studies have sought to revive the low self-esteem hypothesis. However, these
studies were methodologically flawed in important respects and consistently

A)
failed to replicate when more rigorous research designs were used. Regardless of
whether one studies children, adolescents, or adults, low self-esteem generally

P
(A
fails to predict objective measures of aggressive behavior. Then we might ask
another question: What kind of self-views, if any, predispose people to behave

n
aggressively?

io
at
ci
DOES NARCISSISM INFLUENCE AGGRESSIVE BEHAVIOR?

so
As
Baumeister and his colleagues (1996) proposed that aggression most
commonly stems from threatened egotism. In other words, people with big

al
egos become aggressive when others threaten their inflated egos. Thus,

ic
“grandiose” and “inflated” self-views, rather than simply “positive” self-

og
views, were predicted to lead to aggressive and violent behavior. Such forms
ol
of exaggerated self-love are characteristic of narcissism. In its extreme form,
ch
narcissism is a personality disorder defined by grandiose self-views and an
y

inflated sense of entitlement and personal superiority (American Psychiatric

Ps

Association, 1994). Most current psychological research focuses on normal

narcissism, operationalized as a trait on which people in the general popu-
an

lation vary (Morf & Rhodewalt, 2001). Normal narcissism is typically

ic

measured using self-report questionnaires such as the Narcissistic Person-

er

ality Inventory (NPI). Recent research has confirmed the view that NPI-
Am

measured narcissism is distributed dimensionally in the general population

(Foster & Campbell, 2007).
e

The threatened egotism hypothesis has gained abundant empirical sup-

th

port. Bushman and Baumeister (1998) conducted two laboratory experiments

by

in which participants were given the opportunity to aggress against individ-

uals who insulted or praised them or against an innocent third person. The
11

results showed that people with low self-esteem were not more aggressive than
20

others. Narcissists who aggressed directly against the person who insulted
them showed the highest aggression levels. The finding that narcissism is pos-
©

itively related to aggressive behavior in adults has been replicated by other

ht

researchers using self-report aggression questionnaires (Donnellan et al.,

ig

2005, Study 3; Lawrence, 2006), laboratory aggression measures (e.g., Bushman

yr

et al., 2003; Konrath, Bushman, & Campbell, 2006; Reidy, Zeichner, Foster,
op

& Martinez, 2008; Stucke & Sporer, 2002; Twenge & Campbell, 2003), and
real-world aggression measures (Bushman & Baumeister, 2002; Goldberg
C

et al., 2007).

NARCISSISM, SELF-ESTEEM, AND AGGRESSION 207

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 208

It is possible that self-esteem could have an indirect effect on aggression,

so we recently examined the joint and interactive effects of self-esteem and
narcissism on aggression (Bushman et al., 2009). Interactions between self-
esteem and narcissism are relevant because clinical theories and empirical

A)
research have suggested that there are various kinds of narcissists who differ
in their level of self-esteem (Dickinson & Pincus, 2003; Kernberg, 1975;

P
(A
Kohut, 1977). Covert narcissists have relatively low self-esteem and are described
as socially avoidant individuals who are self-absorbed yet shy and introverted.

n
In contrast, overt narcissists have much higher self-esteem and are described

io
as self-assured extraverts who have a dominant interpersonal orientation. We

at
reanalyzed data from a previous experiment (Bushman & Baumeister, 1998)

ci
and conducted a new experiment, finding that aggression was highest in

so
threatened narcissists who also had high levels of self-esteem (i.e., in overt

As
narcissists). To the extent that threatened narcissists harbored somewhat
lower levels of self-esteem (characteristic of covert narcissists), they were

al
much less aggressive.

ic
Some scholars, however, have argued that laboratory findings may not

og
generalize to the real world because the setting and measures are artificial
ol
(e.g., Donnellan et al., 2005). We therefore conducted a follow-up study as a
ch
naturalistic extension of our laboratory work (Bushman et al., 2009, Study 3).
y

We used a nonlaboratory, naturally occurring situation in which students

Ps

from a class were able to evaluate each other’s work and influence each other’s
grade. Ego threat was not manipulated or artificially induced. Instead, it was
an

measured in the form of naturally occurring feelings of humiliation experi-

ic

enced after receiving negative feedback from a fellow student. Similarly, the
er

target of aggression was not an unseen confederate but a fellow student with
Am

whom participants actually interacted in the classroom. These features of the

study made the events seem real and consequential: People were genuinely
e

affected and sometimes humiliated by the criticisms they received of their

th

actual class work, and they believed they could lower the grades of their eval-
by

uator by giving the evaluator negative feedback.

The results of this field study were the same as those from the laboratory
11

experiments. The most aggressive people were those who experienced feel-
20

ings of humiliation and had high levels of both narcissism and self-esteem.
Humiliated narcissists who held lower levels of self-esteem were the least
©

aggressive. Taken together, this set of studies provides new evidence against
ht

the view that low self-esteem causes aggression. There were indirect (rather
ig

than direct) effects of low self-esteem on aggression, but they were in the
yr

direction opposite of the low self-esteem hypothesis. Low self-esteem reduced

op

or eliminated the independent effect of narcissism on aggression.

Although early work on narcissism focused on adults, narcissism can
C

also be reliably identified and distinguished from related personality con-

208 THOMAES AND BUSHMAN

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 209

structs in children and adolescents (Thomaes, Stegge, Bushman, Olthof, &

Denissen, 2008). In a recent study, we examined whether narcissism and self-
esteem also jointly predict aggression in children who undergo shameful expe-
riences (Thomaes, Bushman, Stegge, & Olthof, 2008). Shameful experiences

A)
are fairly common in late childhood and adolescence, due in part to develop-
mental increases in self-consciousness (Ryan & Kuczkowski, 1994). In our

P
(A
experiment, participants lost to an ostensible opponent in a competitive task.
In the shame condition, they were told their opponent was really bad on the

n
task and that they should easily win. After losing, they saw their own name at

io
the bottom of a ranking list on a bogus web page, below their opponent’s name.

at
In the control condition, they were told nothing about their opponent and did

ci
not see any rankings. Next, participants could blast their opponent with loud

so
noise through headphones (a measure of aggression). Consistent with previ-

As
ous studies of adults, narcissists were more aggressive than others but only
when they were shamed. No support was found for the traditional view that

al
low self-esteem underlies aggression. In fact, that view was contradicted by the

ic
finding that high self-esteem increased narcissistic shame-induced aggression.

og
Once again, low self-esteem eliminated the aggressive behavior characteristic
of shamed narcissists. ol
ch
Taken together, these studies indicate that narcissistic self-views predis-
y

pose people to behave aggressively. The link between narcissism and aggres-
Ps

sion generalizes across research settings and methodologies and seems to be

independent of such factors as age, gender, and type of aggression measure.
an

Narcissists are especially aggressive when their grandiose self-images are chal-
ic

lenged. Self-esteem affects narcissists’ aggressive inclinations by heightening

er

aggression when self-esteem is high and reducing or inhibiting aggression

Am

when self-esteem is low.

e

Controversies
th
by

Although we have tried to explain and reconcile many of the conflict-

ing findings that mark the literature on self-views and aggression, a number
11

of unanswered questions and controversies remain. The first controversy con-

20

cerns the shape of the relation between self-views and aggression. One might
wonder whether the relationship is U-shaped, such that people holding neg-
©

ative, deflated self-views and people holding positive, inflated self-views are
ht

aggression prone. Such a U-shaped relationship would be supported by theo-

ig

ries that argue that extremes of self-esteem (both low and high) are maladap-
yr

tive because they typically reflect distorted perceptions (Jahoda, 1958; Kernis,
op

2003). Prior research has mainly tested linear relationships between self-views
and aggression. We know of only two studies that examined the possibility of
C

a curvilinear relationship (Perez, Vohs, & Joiner, 2005; Webster, 2007). The

NARCISSISM, SELF-ESTEEM, AND AGGRESSION 209

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 210

results from both studies provided little support for a U-shaped relationship, but
both studies used self-report measures of aggression, which as we noted earlier
are problematic. Thus, the issue remains open and awaits further scientific
scrutiny.

A)
Another controversy—one that we already touched on briefly—concerns
the best way to measure aggressive behavior (see Chapters 1, 8, and 9). Some

P
(A
scholars have faulted laboratory aggression studies because they are artificial.
Indeed, participants in laboratory studies are likely aware that researchers will

n
not let them seriously injure or abuse anyone. However, research consistently

io
finds that the same factors that increase aggression in the real world also

at
increase aggression in the laboratory and that the same individual differences

ci
in aggression that exist in the real world also exist in the laboratory (Ander-

so
son & Bushman, 1997; Anderson, Lindsay, & Bushman, 1999). Clearly, one

As
advantage of field studies of aggression is that they do not suffer from lack of
ecological validity. However, field studies typically use informant-based mea-

al
sures of aggression (e.g., teacher, parent, or peer report measures) that also

ic
have limitations (see Chapter 7, this volume). Most important, informants

og
may not validly report on targets’ aggression in response to specific situational
ol
contexts, which makes it difficult to use informant-based measures to test
ch
situation-dependent theories of aggression (e.g., threatened egotism theory).
y

A third controversy concerns what we think of as the “doughnut theory”

Ps

of narcissistic aggression. This theory recognizes that aggressive people typi-

cally hold inflated, narcissistic self-views, but it argues that hidden behind this
an

veneer of grandiosity are deeper feelings of insecurity and insufficiency (much

ic

like a doughnut with an empty hole in the middle). Is there any evidence to
er

support this notion? No and yes.

Am

When self-esteem is defined in the usual way—as one’s conscious appre-

ciation of one’s worth as a person—there is no support for the doughnut theory.
e

In fact, the theory is directly contradicted by the finding that low self-esteem
th

decreases (rather than increases) narcissistic aggression. Moreover, if low self-

by

esteem does not cause aggression, how could hidden low self-esteem cause
aggression? However, recent developments indicate that the distinction
11

between explicit (i.e., conscious, controlled) and implicit (i.e., unconscious,

20

uncontrolled) self-esteem may be important. Some people hold “balanced”

explicit and implicit self-esteem (i.e., they hold similar levels of relatively
©

high or low explicit and implicit self-esteem), but other people hold “dis-
ht

crepant” explicit and implicit self-esteem. The doughnut theory argues that
ig

people holding a combination of high explicit self-esteem and low implicit

yr

self-esteem are exceptionally aggressive. Perhaps these people are aggressive

op

in an attempt to ward off threatening information that would otherwise prime

their implicit feelings of insecurity. Initial research findings provide some sup-
C

port for this notion. One study involving middle school students found an

210 THOMAES AND BUSHMAN

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 211

interaction between explicit and implicit self-esteem in predicting teacher-

reported aggression, such that the most aggressive children tended to have
high explicit and low implicit self-esteem (with implicit self-esteem measured
using a categorization task that assesses the ease with which participants asso-

A)
ciate positive and negative words with themselves; Sandstrom & Jordan,
2008). We know of no research designed to replicate this finding with adults.

P
(A
However, there is evidence that narcissistic adults can have high explicit and
low implicit self-esteem (Jordan, Spencer, Zanna, Hoshino-Browne, & Correll,

n
2003). Future research is needed to determine whether these kinds of adults

io
are especially prone to aggression.

at
A final controversy concerns how aggression is conceptualized. Aggres-

ci
sion is generally defined as any behavior intended to harm another person who

so
does not want to be harmed (Bushman & Thomaes, 2008). Some researchers

As
have focused on more broadly defined categories of behavior that include not
only aggression but also other antisocial behaviors such as lying, stealing, van-

al
dalism, and being disruptive in class (e.g., Donnellan et al., 2005; Trzesniewski

ic
et al., 2006). It is entirely plausible that the relationship between self-views and

og
aggression is different from the relationship between self-views and other
ol
antisocial behaviors (for a review of different genetic markers for aggression and
ch
other antisocial behaviors, see Chapter 8, this volume). Low self-esteem may
y

foster a tendency to break society’s rules because people with low self-esteem
Ps

regard themselves as unsuccessful or marginalized members of society. Thus,

low self-esteem may make people willing to violate social norms, such as by
an

lying, cheating, or stealing. Indeed, most research suggests that low self-esteem
ic

increases the chance that individuals will engage in antisocial or delinquent

er

behavior (Baumeister et al., 2003; Donnellan et al., 2005; Trzesniewski et al.,

Am

2006). There is abundant evidence, however, that low self-esteem does not
make people more likely to behave aggressively. We believe the field will
e

benefit from studies that focus on well-defined, homogeneous categories

th

of behavior (i.e., either aggressive behavior or other antisocial, delinquent,

by

norm-violating behavior). Research is also needed on the different motiva-

tional processes that may underlie various categories of antisocial behavior.
11
20

Clinical Implications
©

What do the research findings reviewed in this chapter imply for pre-
ht

vention and intervention strategies designed to reduce aggression? Aggres-

ig

sion and violence are complex phenomena that have multiple causes (e.g.,
yr

Chapters 1, 2, and 7). Individual differences in aggression are stable over time
op

and typically emerge by early childhood (see Chapter 7). This means that
effective prevention and intervention strategies should be (a) broadly targeted
C

at a constellation of determinants of aggressive behavior and (b) initiated early

NARCISSISM, SELF-ESTEEM, AND AGGRESSION 211

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 212

in children’s development (e.g., Dodge, Coie, & Lynam, 2006; see also
Chapter 9, this volume).
Among the many factors that influence aggression, perpetrators’ self-
regard is important. Until now, many prevention and intervention strate-

A)
gies designed to reduce aggression in children and adolescents have been
aimed at boosting their self-esteem (e.g., Kusché & Greenberg, 1994). For

P
(A
example, boosting self-esteem is a central goal of the Promoting Alterna-
tive Thinking Strategies program, one of the most popular classroom-based

n
aggression prevention programs worldwide. According to the author of the

io
program, “high self-esteem makes it more likely that children will develop

at
internal motivation to eschew antisocial behavior and violence,” whereas

ci
“low self-esteem is emotionally painful and frequently results in . . . angry

so
outbursts including violent and delinquent behavior” (Kusché, 2002,

As
pp. 294–295). The empirical evidence we have reviewed in this chapter
contradicts these claims. In fact, if well-intended efforts to boost self-

al
esteem cultivate the inflated self-views characteristic of narcissism, they

ic
may inadvertently increase (rather than decrease) the aggressive behavior

og
of at-risk youth.
ol
The existing evidence suggests that prevention and intervention strate-
ch
gies should target inflated (rather than deflated) self-views to reduce aggression.
y

The important work by Michael Kernis (2003) on the nature of “optimal”

Ps

self-esteem promises to be useful as a framework for developing such strate-

gies. From Kernis’s perspective, it is important to assist aggressive narcissists
an

in (a) being aware of their strengths and weaknesses, (b) being able to present
ic

themselves in an open and truthful way to others rather than rigidly seeking
er

to promote their self-image, and (c) being able to process self-relevant infor-
Am

mation in an objective and accepting way rather than in a defensive and

biased way. Because self-views may be most amenable to intervention in child-
e

hood and the foundation for lifelong aggressive behavior styles is laid in child-
th

hood, these interventions are likely to be most effective when implemented

by

in childhood.
One other perhaps complementary way of reducing narcissistic aggres-
11

sion is based on the rationale that narcissists are aggressive only when their
20

egos are threatened and that intervention techniques that reduce the psy-
chological impact of ego-threatening experiences may thus lessen narcissis-
©

tic aggression. One such technique is to allow individuals to reaffirm their

ht

sense of self (Sherman & Cohen, 2006; Steele, 1988). Self-affirmation theory
ig

holds that individuals’ overall sense of self is based on multiple domains

yr

of functioning, and a threat to one domain of functioning can be compen-

op

sated for by reflecting on the personal importance of a different domain

(e.g., a self-defining skill or interest). Previous research has shown that self-
C

affirmations reduce the psychological impact of threatening feedback and

212 THOMAES AND BUSHMAN

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 213

social-evaluative stress (e.g., Cohen, Garcia, Apfel, & Master, 2006; Sherman
& Cohen, 2006).
We recently conducted a field experiment to examine whether self-
affirmation can reduce narcissistic aggression (Thomaes, Bushman, Orobio

A)
de Castro, Cohen, & Denissen, 2009). Participants were 405 middle-school
students who completed either a self-affirmation or a control writing exercise

P
(A
in their classes. In the affirmation condition, participants wrote about their
most important values and why these are important to them. In the control

n
condition, participants wrote about their least important values and why

io
these may be important to others (Cohen et al., 2006). The intervention

at
was administered on a Monday morning so that we could examine its possi-

ci
ble effect on aggressive incidents in class during the following week. These

so
aggressive incidents were measured on Friday afternoon using peer reports.

As
We also obtained a measure of ego threats that had been experienced during
the past week. In the control condition, the now familiar pattern of results

al
emerged, with narcissists being more aggressive than others but only when

ic
they experienced relatively high levels of ego threat. In the intervention con-

og
dition, however, that standard pattern of results was eliminated by the
ol
self-affirmation writing exercise. Thus, as predicted, the self-affirmation inter-
ch
vention reduced narcissistic aggression for a period of 1 week, presumably by
y

attenuating the ego protective motives that normally drive narcissists’ aggres-
Ps

sion. We hope this result will encourage the development of theory and evidence-
based aggression interventions aimed at buttressing people’s self-views against
an

ego threats.
ic
er
Am

FUTURE RESEARCH
e

Three main issues need to be addressed to improve our understanding

th

of how self-views relate to aggression. First, we need to know more about

by

developmental processes (see Chapter 7). From what point in development

do children’s narcissistic traits predispose them to behave aggressively? Young
11

children typically hold unrealistically positive and grandiose self-views, but

20

these self-views typically become increasingly realistic later in childhood

(Harter, 2006). It may be that from late childhood, individual differences in
©

narcissism come to influence children’s aggressive behavior. Research is needed

ht

to test this hypothesis and to help clinicians identify critical age periods in
ig

which to influence the self-views underlying children’s aggressive behavior.

yr

It is also important to know more about the developmental origins of narcissism.

op

Why do some people grow up to be narcissists, whereas others do not? Two

theories attribute the development of narcissism to dysfunctional parenting.
C

From a social learning perspective, it is parental overvaluation (i.e., excessive

NARCISSISM, SELF-ESTEEM, AND AGGRESSION 213

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 214

praise, the tendency to rigidly link children’s efforts and achievements to

their worth as a person) that leads to narcissism (Damon, 1995; Millon, 1981;
Twenge, 2006). From a psychodynamic perspective, it is parental coldness
and rejection that leads to narcissism (Kernberg, 1975; Kohut, 1977). Accord-

A)
ing to this perspective, children may learn to seek continuous attention and
admiration to compensate for a lack of parental warmth. The few empirical

P
(A
studies available provide some evidence for both theories (Horton, Bleau, &
Drwecki, 2006; Otway & Vignoles, 2006). These data, however, were based

n
on retrospective adult reports of childhood experiences. Longitudinal research

io
that includes early assessments of parenting behavior and children’s experience

at
with parents is needed. This research should also examine possible interactions

ci
with children’s genetically based temperamental traits.

so
Second, virtually all research on self-views and aggression has been con-

As
ducted in individualistic countries (e.g., United States, Canada, Great Britain,
Germany, the Netherlands, Finland, Australia, New Zealand). Research sug-

al
gests that self-views may play a different motivational role in the lives of peo-

ic
ple from collectivistic cultures. Whereas people from individualistic cultures

og
are typically motivated to enhance their self-views, people from collectivis-
ol
tic cultures tend to be less concerned about sustaining favorable self-views, at
ch
least as a goal in itself (Heine, Lehman, Markus, & Kitayama, 1999). People
y

from collectivistic cultures tend to view themselves in terms of the social roles
Ps

they play and the relationships they have with others (see Chapter 10, this
volume). In fact, it has been argued that the concept of self-esteem is itself a
an

Western phenomenon and that positive self-views in collectivistic cultures

ic

result from a sense of satisfaction with the social role one plays rather than
er

from standing out as an exceptionally admirable individual (Markus &

Am

Kitayama, 1991). It is possible that these different self-construals have differ-

ent effects on aggressive behavior. Moreover, the types of experiences that
e

are perceived as ego threatening are likely to be different in different cultures.

th

Researchers typically focus on threats to people’s personal worth as situational

by

triggers of narcissistic aggression, but such experiences may not be experi-

enced as threatening in collectivistic cultures. A key aim for future research
11

is to explore cultural differences and similarities in the links between self-

20

views and aggression.

Third, whereas research thus far has focused on self-views that increase
©

people’s aggressive inclinations, little is known about self-views that may

ht

decrease aggressive inclinations. If narcissism predisposes people to behave

ig

aggressively, what self-views influence people not to behave aggressively?

yr

Recent research has characterized self-compassion as a healthy cousin of self-

op

esteem (Leary, Tate, Adams, Allen, & Hancock, 2007). Whereas self-esteem
essentially involves a judgment of oneself, self-compassion involves an orien-
C

tation to be kind, caring, and supportive of oneself when things go badly

214 THOMAES AND BUSHMAN

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 215

(Neff, 2003a, 2003b). Self-compassion may be a key self-trait that disengages

people’s inclinations to aggress. People often aggress when they encounter
adversity and experience high levels of emotional distress or frustration. Self-
compassion mitigates the emotional impact of adverse events (Leary et al.,

A)
2007), and so it may also mitigate the aggressive behaviors that often follow
from such events.

P
(A
We recommend that researchers broaden their scope and begin to study
self-compassion or related adaptive self-traits that may reduce aggressive

n
behavior. It would be particularly informative if researchers could show that

io
self-compassion plays a causal role in determining people’s inclinations to

at
aggress. Research has shown that it is possible to induce feelings of self-

ci
compassion, allowing for such a causal test (Leary et al., 2006). This would

so
not only increase our knowledge of how the self is involved in aggressive

As
behavior, but it might also suggest further ways to use self-views as a remedy
for violence and aggression.

al
ic
og
REFERENCES ol
ch
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental
y
Ps

disorders (4th ed.). Washington, DC: Author.

Anderson, C. A., & Bushman, B. J. (1997). External validity of “trivial” experiments:
an

The case of laboratory aggression. Review of General Psychology, 1, 19–41.

ic

doi:10.1037/1089-2680.1.1.19
er

Anderson, C. A., Lindsay, J. J., & Bushman, B. J. (1999). Research in the psycholog-
Am

ical laboratory: Truth or triviality? Current Directions in Psychological Science, 8,

3–9. doi:10.1111/1467-8721.00002
e

Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. (2003). Does high self-
th

esteem cause better performance, interpersonal success, happiness, or healthier

by

lifestyles? Psychological Science in the Public Interest, 4, 1–44. doi:10.1111/1529-

1006.01431
11

Baumeister, R. F., Smart, L., & Boden, J. M. (1996). Relation of threatened egotism
20

to violence and aggression: The dark side of high self-esteem. Psychological

Review, 103, 5–33. doi:10.1037/0033-295X.103.1.5
©

Berkowitz, L. (1983). Aversively stimulated aggression: Some parallels and differences

ht

in research with animals and humans. American Psychologist, 38, 1135–1144.

ig

doi:10.1037/0003-066X.38.11.1135
yr

Boden, J. M., Fergusson, D. M., & Horwood, L. (2007). Self-esteem and violence:
op

Testing links between adolescent self-esteem and later hostility and violent
behavior. Social Psychiatry and Psychiatric Epidemiology, 42, 881–891. doi:10.1007/
C

s00127-007-0251-7

NARCISSISM, SELF-ESTEEM, AND AGGRESSION 215

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 216

Bradshaw, C. P., & Hazan, C. (2006). Examining views of self in relation to views of
others: Implications for research on aggression and self-esteem. Journal of Research
in Personality, 40, 1209–1218. doi:10.1016/j.jrp.2005.11.004
Branden, N. (1984, August–September). In defense of self. Association for Humanistic

A)
Psychology, pp. 12–13.
Bushman, B. J., & Baumeister, R. F. (1998). Threatened egotism, narcissism, self-

P
(A
esteem, and direct and displaced aggression: Does self-love or self-hate lead to
violence? Journal of Personality and Social Psychology, 75, 219–229. doi:10.1037/

n
0022-3514.75.1.219

io
Bushman, B. J., & Baumeister, R. F. (2002). Does self-love or self-hate lead to vio-

at
lence. Journal of Research in Personality, 36, 543–545. doi:10.1016/S0092-

ci
6566(02)00502-0

so
Bushman, B. J., Baumeister, R. F., Thomaes, S., Ryu, E., Begeer, S., & West, S. G.

As
(2009). Looking again, and harder, for a link between low self-esteem and aggres-
sion. Journal of Personality, 77, 427–446. doi:10.1111/j.1467-6494.2008.00553.x

al
Bushman, B. J., Bonacci, A. M., van Dijk, M., & Baumeister, R. F. (2003). Narcis-

ic
sism, sexual refusal, and aggression: Testing a narcissistic reactance model of

og
sexual coercion. Journal of Personality and Social Psychology, 84, 1027–1040.
doi:10.1037/0022-3514.84.5.1027 ol
ch
Bushman, B. J., & Thomaes, S. (2008). Aggression. In R. F. Baumeister & K. D. Vohs
y

(Eds.), Encyclopedia of social psychology (Vol. 1, pp. 20–25). Thousand Oaks, CA:
Ps

Sage.
Cohen, G. L., Garcia, J., Apfel, N., & Master, A. (2006, September). Reducing the
an

racial achievement gap: A social–psychological intervention. Science, 313,

ic

1307–1310. doi:10.1126/science.1128317
er

Damon, W. (1995). Greater expectations: Overcoming the culture of indulgence in our

Am

homes and schools. New York, NY: Free Press.

Dickinson, K. A., & Pincus, A. L. (2003). Interpersonal analysis of grandiose and vul-
e

nerable narcissism. Journal of Personality Disorders, 17, 188–207. doi:10.1521/

th

pedi.17.3.188.22146
by

Dodge, K. A., Coie, J. D., & Lynam, D. (2006). Aggression and antisocial behavior
in youth. In W. Damon, R. Lerner, & N. Eisenberg (Eds.), Handbook of child psy-
11

chology: Vol. 3. Social, emotional, and personality development (pp. 719–788). New
20

York, NY: Wiley.

Donnellan, M. B., Trzesniewski, K. H., Robins, R. W., Moffitt, T. E., & Caspi, A. (2005).
©

Low self-esteem is related to aggression, antisocial behavior, and delinquency.

ht

Psychological Science, 16, 328–335. doi:10.1111/j.0956-7976.2005.01535.x

ig

D’zurilla, T. J., Chang, E. C., & Sanna, L. J. (2003). Self-esteem and social problem
yr

solving as predictors of aggression in college students. Journal of Social and Clin-

op

ical Psychology, 22, 424–440. doi:10.1521/jscp.22.4.424.22897

Foster, J. D., & Campbell, W. K. (2007). Are there such things as “narcissists” in
C

social psychology? A taxometric analysis of the Narcissistic Personality Inven-

216 THOMAES AND BUSHMAN

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 217

tory. Personality and Individual Differences, 43, 1321–1332. doi:10.1016/j.paid.

2007.04.003
Goldberg, B. R., Serper, M. R., Sheets, M. M. A., Beech, D., Dill, C., & Duffy, K. G.
(2007). Predictors of aggression on the psychiatric inpatient service: Self-

A)
esteem, narcissism, and theory of mind deficits. Journal of Nervous and Mental
Disease, 195, 436–442.

P
(A
Harter, S. (2006). The self. In W. Damon, R. M. Lerner, & N. Eisenberg (Eds.),
Handbook of child psychology: Vol. 3. Social, emotional, and personality development

n
(pp. 505–570). New York, NY: Wiley.

io
Heine, S. J., Lehman, D. R., Markus, H. R., & Kitayama, S. (1999). Is there a univer-

at
sal need for positive self-regard? Psychological Review, 106, 766–794. doi:10.1037/

ci
0033-295X.106.4.766

so
Horney, K. (1950). Neurosis and human growth. New York, NY: Norton.

As
Horton, R. S., Bleau, G., & Drwecki, B. (2006). Parenting narcissus: Does parenting
contribute to the development of narcissism? Journal of Personality, 74, 345–376.

al
doi:10.1111/j.1467-6494.2005.00378.x

ic
Jahoda, M. (1958). Current concepts of positive mental health. New York, NY: Basic

og
Books. doi:10.1037/11258-000
ol
Jordan, C. H., Spencer, S. J., Zanna, M. P., Hoshino-Browne, E., & Correll, J. (2003).
ch
Secure and defensive high self-esteem. Journal of Personality and Social Psychol-
y

ogy, 85, 969–978. doi:10.1037/0022-3514.85.5.969

Ps

Kernberg, O. (1975). Borderline conditions and pathological narcissism. New York, NY:
Jason Aronson.
an

Kernis, M. H. (2003). Toward a conceptualization of optimal self-esteem. Psychological

ic

Inquiry, 14, 1–26. doi:10.1207/S15327965PLI1401_01

er

Kirkpatrick, L. A., Waugh, C. E., Valencia, A., & Webster, G. D. (2002). The func-
Am

tional domain specificity of self-esteem and the differential prediction of aggres-

sion. Journal of Personality and Social Psychology, 82, 756–767. doi:10.1037/
e

0022-3514.82.5.756
th

Kohut, H. (1977). The restoration of the self. Madison, WI: International Universities
by

Press.
Konrath, S., Bushman, B. J., & Campbell, W. K. (2006). Attenuating the link
11

between threatened egotism and aggression. Psychological Science, 17, 995–1001.

20

doi:10.1111/j.1467-9280.2006.01818.x
Kusché, C. (2002). Psychoanalysis as prevention. Using PATHS to enhance ego-
©

development, object relationships, and cortical integration in children. Journal

ht

of Applied Psychoanalytic Studies, 4, 283–301. doi:10.1023/A:1015773327131

ig

Kusché, C., & Greenberg, M. (1994). PATHS: Promoting alternative thinking strate-
yr

gies. South Deerfield, MA: Developmental Research Programs.

op

Lawrence, C. (2006). Measuring individual responses to aggression-triggering events:

Development of the Situational Triggers of Aggressive Responses (STAR) scale.
C

Aggressive Behavior, 32, 241–252. doi:10.1002/ab.20122

NARCISSISM, SELF-ESTEEM, AND AGGRESSION 217

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 218

Leary, M. R., Tate, E. B., Adams, C. E., Allen, A. B., & Hancock, J. (2007). Self-
compassion and reactions to unpleasant self-relevant events: The implications
of treating oneself kindly. Journal of Personality and Social Psychology, 92,
887–904. doi:10.1037/0022-3514.92.5.887

A)
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cogni-
tion, emotion, and motivation. Psychological Review, 98, 224–253.

P
(A
Millon, T. (1981). Disorders of personality. DSM-III: Axis II. New York, NY: Wiley.
Morf, C. C., & Rhodewalt, F. (2001). Unraveling the paradoxes of narcissism: A

n
dynamic self-regulatory processing model. Psychological Inquiry, 12, 177–196.

io
doi:10.1207/S15327965PLI1204_1

at
Neff, K. D. (2003a). The development and validation of a scale to measure self-

ci
compassion. Self and Identity, 2, 223–250. doi:10.1080/15298860309027

so
Neff, K. D. (2003b). Self-compassion: An alternative conceptualization of a healthy

As
attitude toward oneself. Self and Identity, 2, 85–101. doi:10.1080/15298860309032
Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal

al
reports on mental processes. Psychological Review, 84, 231–259. doi:10.1037/

ic
0033-295X.84.3.231

og
Otway, L. J., & Vignoles, V. L. (2006). Narcissism and childhood recollections: A
ol
quantitative test of psychoanalytic predictions. Personality and Social Psychology
ch
Bulletin, 32, 104–116. doi:10.1177/0146167205279907
y

Perez, M., Vohs, K. D., & Joiner, T. E., Jr. (2005). Discrepancies between self- and
Ps

other-esteem as correlates of aggression. Journal of Social and Clinical Psychology,

24, 607–620. doi:10.1521/jscp.2005.24.5.607
an

Reidy, D. E., Zeichner, A., Foster, J. D., & Martinez, M. A. (2008). Effects of narcis-
ic

sistic entitlement and exploitativeness on human physical aggression. Personal-

er

ity and Individual Differences, 44, 865–875. doi:10.1016/j.paid.2007.10.015

Am

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton
University Press.
e

Ryan, R. M., & Kuczkowski, R. (1994). The imaginary audience, self-consciousness,

th

and public individuation in adolescence. Journal of Personality, 62, 219–238.

by

doi:10.1111/j.1467-6494.1994.tb00292.x
Sandstrom, M. J., & Jordan, R. (2008). Defensive self-esteem and aggression in child-
11

hood. Journal of Research in Personality, 42, 506–514. doi:10.1016/j.jrp.2007.07.008

20

Sherman, D. K., & Cohen, G. L. (2006). The psychology of self-defense: Self-

affirmation theory. In M. P. Zanna (Ed.), Advances in experimental social
©

psychology (Vol. 38, pp. 183–242). San Diego, CA: Academic Press.doi:10.1016/
ht

S0065-2601(06)38004-5
ig

Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of

yr

the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21,
op

pp. 261–302). New York, NY: Academic Press.

Stucke, T. S., & Sporer, S. L. (2002). When a grandiose self-image is threatened:
C

Narcissism and self-concept clarity as predictors of negative emotions and

218 THOMAES AND BUSHMAN

12051-12_PT3-CH11-rev2.qxd 5/25/10 11:54 AM Page 219

aggression following ego-threat. Journal of Personality, 70, 509–532. doi:10.1111/

1467-6494.05015
Thomaes, S., Bushman, B., Orobio de Castro, B., Cohen, G., Denissen, J. J. A.
(2009). Reducing narcissistic aggression by buttressing self-esteem: An experi-

A)
mental field study. Psychological Science, 20, 1536–1542.
Thomaes, S., Bushman, B. J., Stegge, H., & Olthof, T. (2008). Trumping shame by

P
(A
blasts of noise: Narcissism, self-esteem, shame, and aggression in young adoles-
cents. Child Development, 79, 1792–1801. doi:10.1111/j.1467-8624.2008.01226.x

n
Thomaes, S., Stegge, H., Bushman, B. J., Olthof, T., & Denissen, J. (2008). Develop-

io
ment and validation of the Childhood Narcissism Scale. Journal of Personality

at
Assessment, 90, 382–391. doi:10.1080/00223890802108162

ci
Toch, H. (1993). Violent men: An inquiry into the psychology of violence. Washington,

so
DC: American Psychological Association. (Original work published 1969)

As
Trzesniewski, K. H., Donnellan, M. B., Moffitt, T. E., Robins, R. W., Poulton, R., &
Caspi, A. (2006). Low self-esteem during adolescence predicts poor health,

al
criminal behavior, and limited economic prospects during adulthood. Develop-

ic
mental Psychology, 42, 381–390. doi:10.1037/0012-1649.42.2.381

og
Twenge, J. M. (2006). Generation me: Why today’s young Americans are more confident,
ol
assertive, entitled—and more miserable than ever before. New York, NY: Free Press.
ch
Twenge, J. M., & Campbell, W. K. (2003). “Isn’t it fun to get the respect that we’re
y

going to deserve?” Narcissism, social rejection, and aggression. Personality and

Ps

Social Psychology Bulletin, 29, 261–272. doi:10.1177/0146167202239051

Webster, G. D. (2007). Is the relationships between self-esteem and aggression nec-
an

essarily U-shaped? Journal of Research in Personality, 41, 977–982. doi:10.1016/

ic

j.jrp.2007.01.001
er

Webster, G. D., & Kirkpatrick, L. A. (2006). Behavioral and self-reported aggression

Am

as a function of domain-specific self-esteem. Aggressive Behavior, 32, 17–27.

doi:10.1002/ab.20102
e
th
by
11
20
©
ht
ig
yr
op
C

NARCISSISM, SELF-ESTEEM, AND AGGRESSION 219

C
op
yr
ig
ht
©
20
12051-12_PT3-CH11-rev2.qxd

11
by
5/25/10

th
e
Am
11:54 AM

er
ic
an
Page 220

Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 221

12

PA)
EVOLVED MECHANISMS FOR

(A
REVENGE AND FORGIVENESS

n
io
at
ci
MICHAEL E. MCCULLOUGH, ROBERT KURZBAN,

so
AND BENJAMIN A. TABAK

As
al
ic
og
ol
y ch
Ps
an

In this chapter, we describe our efforts to understand the functions of the

ic

cognitive systems that underlie humans’ capacities for revenge and forgiveness.
er

A better understanding of these concepts is not only scientifically interesting

Am

but socially important as well. In developed nations, the desire for revenge is cited
as a causal factor in as many as 20% of homicides (Kubrin & Weitzer, 2003).
e

Roughly 20% of the perpetrators of violent assault and criminal property

th

damage in the United Kingdom cite the desire for revenge as a motive
by

(Home Office, 2003), and 61% of U.S. school shootings between 1974 and
June 2000 were vengeance-motivated (Vossekuil, Fein, Reddy, Borum, &
11

Modzeleski, 2002). The desire for revenge also motivates people to enlist in
20

terrorist organizations (Speckhard & Ahkmedova, 2006).

Perhaps because revenge is so closely linked to aggression and violence, it
©

has been fashionable in Western thought since the Stoic (and, later, Christian)
ht

philosophers to view revenge as immoral, irrational, or both (Murphy, 2003).

ig

As if seeking to restate this dim view of revenge in modern, therapeutic terms,

yr
op

We gratefully acknowledge the support of the National Institute of Mental Health Grant 5R01MH071258,
C

a grant from the Fetzer Institute, and support from the Center for the Study of Law and Religion at
Emory University to the first author.

221
12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 222

social scientists in the past century have also promulgated the idea that the
desire for revenge is indicative of psychological dysfunction (Murphy, 2003).
The “revenge as disease” conceit had a predictable effect on how forgive-
ness came to be studied empirically as well: If the desire for revenge is a disease,

A)
then perhaps forgiveness is the cure. Indeed, many of the earliest empirical
studies on forgiveness were related to the use of interventions for promoting

P
(A
forgiveness in therapeutic settings (DiBlasio & Benda, 1991; Hebl & Enright,
1993). These treatments do promote forgiveness—and reduce psychological

n
symptoms of anxiety and depression and boost self-esteem (Lundahl, Taylor,

io
Stevenson, & Roberts, 2008)—but such facts do not even come close to

at
proving that forgiveness is a “cure” for revenge.

ci
Evolutionary research and scholarship cast considerable doubt on “disease”

so
and “cure” conceits for conceptualizing the human capacities for revenge and

As
forgiveness. In this chapter, we propose that revenge and forgiveness are the
results of distinct psychological adaptations that evolved to solve specific

al
adaptive problems. We posit that one or more revenge mechanisms evolved

ic
because of their efficacy in deterring interpersonal harms and that one or

og
more forgiveness mechanisms evolved because of their efficacy in preserving
ol
valuable relationships despite those harms. Here, we attempt to define revenge
ch
and forgiveness in functional terms that will make them more amenable to
y

evolutionary analysis (Williams, 1966), to describe the selection pressures

Ps

that gave rise to them, and to outline what we think are the proximate causes
and the computations involved when people make choices to forgive or to
an

avenge a wrong.
ic
er
Am

EVOLUTION OF REVENGE
e

In the section that follows, we attempt to outline a functional approach to

th

understanding revenge. In particular, we attempt to define revenge functionally

by

rather than behaviorally and demonstrate the value of such a definitional

approach. Second, we attempt to outline the evolutionary selection pressures
11

that might have given rise to psychological mechanisms that produce revenge
20

as a functional output.
©

Revenge: A Functional Definition

ht
ig

A great deal of research and writing has been devoted to revenge, and
yr

people have powerful intuitions about it. Still, we believe it is worthwhile to

op

take a step back and reflect on the evolved function of putative revenge systems.
At its heart, revenge solves a problem that is faced to varying degrees by many
C

species: how to change other organisms’ incentives to induce them to emit

222 MC CULLOUGH, KURZBAN, AND TABAK

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 223

benefits and refrain from imposing costs on oneself (see Chapter 3). To the
extent that other organisms can learn that a target organism will retaliate
(or conditionally benefit) as a function of their behavior, it is beneficial for
the target organism to signal that it will do so. One (albeit imperfect) way

A)
to signal that one will retaliate if harmed (or benefit if helped) is to actually
do so. If neural tissue is assembled that reliably motivates these sorts of

P
(A
contingent punishments and contingent rewards, it may boost lifetime repro-
ductive fitness of its bearer and therefore evolve precisely because of these

n
functions. Our analysis begins with this simple, but crucial, idea.

io
This notion contrasts with the way some philosophers have defined

at
revenge, but we think some of the previous definitions create as many problems

ci
as they solve. For example, Govier (2002) wrote, “When we seek revenge, we

so
seek satisfaction by attempting to harm the other (or associated persons)

As
as a retaliatory measure” (p. 2). Elster (1990) likewise defined revenge as “the
attempt, at some cost or risk to oneself, to impose suffering upon those who

al
have made one suffer, because they have made one suffer” (p. 862). Uniacke

ic
(2000) also claimed that “revenge is personal and noninstrumental: With

og
revenge we seek to make people suffer because they have made us suffer, not
ol
because their actions or values require us to bring them down” (p. 62).
ch
These definitions, because they are proximate and do not commit to
y

any function, make no obvious predictions about the design features of

Ps

the psychology of revenge. “Enjoyment” and other proximate explanations

(see Govier’s, 2002, definition) leave a promissory note for an ultimate
an

explanation that must be paid. Why should revenge produce pleasure? For no
ic

organism except humans would we accept that an explanation for a behavior

er

is that it brings enjoyment.

Am

In short, functional thinking about cognition and behavior reminds us

that there is no free lunch. Why would a species such as Homo sapiens engage in
e

costly behavior such as revenge unless the mechanism that creates revenge was
th

designed to produce benefits in the currency of fitness or is a by-product of a

by

structure that does yield fitness payoffs (Andrews, Gangestad, & Matthews,
2002)? What could maintain revenge in humans’ behavioral repertoire?
11

The definitions cited earlier also introduce problems related to inten-

20

tionality. What does it mean that revenge involves an attempt to impose

retaliatory harm on an aggressor? Does the word attempt imply a conscious and
©

deliberate effort to make another individual suffer? Is consciousness necessary?

ht

Is deliberation necessary? Or can this striving to harm one’s provoker be auto-

ig

matic and/or unconscious? And is this distinction critical in any case?

yr

We think a functional definition of revenge can clarify some of these

op

points. Biologists regularly define behavior functionally, as when Maynard

Smith and Harper (2003) defined a signal as “any act or structure which alters
C

the behaviour of other organisms, which evolved because of that effect, and

EVOLVED MECHANISMS FOR REVENGE AND FORGIVENESS 223

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 224

which is effective because the receiver’s response has also evolved” (p. 3).
By designating a function, it becomes possible to search for evidence of the
features—behavioral or physiological—that contribute to accomplishing the
putative function.

A)
In similar fashion, we define revenge functionally as behavior resulting from
a mechanism designed to deter the imposition of costs on (or the withholding

P
(A
of benefits from) oneself or one’s allies by the imposition of costs following
a target’s imposition of costs (or withholding of benefits), where costs and

n
benefits are defined in terms of their effects on lifetime reproductive fitness.

io
That is, revenge is a deterrence system designed to change others’ incentives

at
regarding the self and one’s kin or allies (see Chapter 4 for a similar functional

ci
analysis of what the chapter authors call the power behavioral system). By impos-

so
ing costs after harm (or withheld benefits), revenge signals that subsequent

As
acts will be subject to the same contingent response, thereby altering others’
incentives. We hypothesize that humans possess psychological adaptations

al
designed specifically to produce revenge.

ic
This functional definition has several important features. First, it replaces

og
considerations of intentionality (e.g., whether the organism is deliberately
ol
or consciously attempting to do something) with considerations of design
ch
(e.g., what the system that motivates revenge was designed to do). Moreover,
y

the concept of design makes powerful empirical commitments; adaptation is a

Ps

strong claim (Williams, 1966), and to the extent that the psychological mech-
anisms do not show features that support a deterrence function, the hypothesis
an

that humans possess an innate psychology of revenge is undermined.

ic

Our definition of revenge incorporates as instances of revenge all retal-

er

iatory impositions of costs that are caused by a mechanism designed for this
Am

purpose, even acts that are not based on deliberation or awareness and even
those that do not actually manage to deter anything (as when people behave
e

aggressively toward a driver whom they perceive to have mistreated them

th

on the road). Such a definition also permits a distinction between costs to the
by

provoker that arise from design for that function versus costs that arise as a
by-product. Harming a provoker is only revenge when the system that motivated
11

the harmful behavior was crafted for that purpose. Avoiding a provoker to
20

avert a second harm is not revenge, but avoiding a provoker to limit his or her
access to benefits might be. Likewise, the phenomenon of displaced aggression,
©

in which a victim of aggression proceeds later to harm a third party (Miller,

ht

Pedersen, Earleywine, & Pollock, 2003; see also Chapter 6, this volume), may
ig

not be revenge, even if the third party is a genetic relative or ally of the original
yr

aggressor. If displaced aggression of this nature is not produced by a system

op

designed for deterrence but rather is produced by the psychological processes

that Miller et al. (2003) implicated (e.g., residual arousal and postaggression
C

rumination that lead to what are, essentially, cognitive errors)—that is to say,

224 MC CULLOUGH, KURZBAN, AND TABAK

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 225

if displaced aggression is a mere by-product of other psychological processes—

then it is not revenge. As a side note, to us it is an open question whether some
instances of triggered displaced aggression might actually reflect the operation
of a revenge system. What we wish to point out here is that this “triggered

A)
displaced aggression as revenge” hypothesis—though it might be wrong—
would likely never have been generated solely by relying on the standard,

P
(A
nonfunctional framework that researchers commonly use to understand dis-
placed aggression.

n
io
at
Selection Pressures That Gave Rise to Mechanisms for Revenge

ci
so
In an influential review article, Clutton-Brock and Parker (1995) noted
that retaliation (which they called punishment) is common among nonhuman

As
animals (for a more recent example, see Jensen, Call, & Tomasello, 2007).

al
They speculated that retaliation yields fitness gains by reducing the probability

ic
that the targets of retaliation will repeat their injurious actions against the

og
retaliator in the future. Consistent with Clutton-Brock and Parker’s analysis,
we hypothesize that natural selection gave rise to one or more comparable
ol
deterrence systems in humans. In this sense, the adaptive consequences of
ch
revenge come not from what revenge causes per se but from what it prevents.
y
Ps

For illustrative purposes here, we distinguish among three types of deterrence.

The first two, direct and indirect deterrence of aggression, involve deter-
an

ring the imposition of costs. The third involves deterring the withholding
of benefits.
ic
er
Am

Direct Deterrence
By direct deterrence, we mean that revenge discourages aggressors from
e

harming the avenger a second time. The logic of direct deterrence is straight-
th

forward: If a potential aggressor must make a decision in which he or she can

by

take an action that imposes costs on a potential victim to acquire some benefit,
then the potential victim is better off if he or she can change the potential
11

aggressor’s incentives so that the expected value of the cost-imposing action on

20

the potential victim is negative. Revenge can accomplish this transformation

of expected value by conveying to an aggressor that the retaliatory infliction of
©

fitness costs will exceed the potential benefits to be gained by aggressing

ht

against the potential victim a second time (see Chapter 3). Nevertheless, direct
ig

deterrence gives rise to strategic complications. For example, although revenge

yr

at Time 1 might predict revenge at Time 2, there is nothing that forces this to
op

be true. An organism could be, for example, intermittently vengeful. This leads
to well-known problems of signaling that one’s vengeful dispositions are stable
C

over time (Frank, 1988).

EVOLVED MECHANISMS FOR REVENGE AND FORGIVENESS 225

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 226

Experimental evidence in support of revenge’s effectiveness as a direct

deterrent comes from experiments involving economic games such as the
sequential and iterated prisoner’s dilemma (Axelrod, 1984). In the sequential
prisoner’s dilemma game, there is one round of play, but the second mover

A)
chooses only after seeing the first player’s choice. In such games, the second
player is much more likely to cooperate after a cooperative move than after

P
(A
a defecting move. More relevant to our present point, defection is almost
always met with retaliatory defection (Clark & Sefton, 2001, Table 6), an

n
observation that holds not only in the United States (Hayashi, Ostrom,

io
Walker, & Yamagishi, 1999). However, because noncooperation and punish-

at
ment are the same in the prisoner’s dilemma, such findings must be interpreted

ci
with care.

so
In the iterated prisoner’s dilemma, subjects play multiple rounds of the

As
game with either the same partner or different ones. For the present pur-
pose, key issues are whether people respond to defection with defection—

al
moves plausibly interpretable as revenge—and whether such moves elicit

ic
subsequent cooperation from one’s partner. Experiments using large num-

og
bers of trials in prisoner’s dilemma games suggest that people do respond to
ol
defection with defection (Bixenstine & Wilson, 1963), though the details
ch
vary across studies (Rapoport & Chammah, 1965). Reciprocal strategies
y

such as “tit for tat” or variants of it tend to elicit cooperation from experimen-
Ps

tal subjects (e.g., Wilson, 1971), hinting at their effectiveness in deterring

defection.
an

Moreover, in an analysis of data from five different laboratory studies of

ic

dyadic negotiation in which partners played 250 consecutive trials during which
er

they could either punish, reward, or withhold reward (and punishment) from
Am

each other, Molm (1997) found that the frequency with which retaliatory
punishment was used (i.e., the infliction of punishment after one’s negotiation
e

partner had previously punished the actor) was positively associated with the
th

frequency with which partners rewarded each other. Likewise, the use of
by

punishment following nonreward (i.e., the withholding of benefits) was asso-

ciated with higher rates of rewarding. These findings suggest that retaliatory
11

infliction of punishments in response to punishments and the withholding of

20

rewards creates a relational climate in which the exchange of reward is more

frequent. In contrast, Molm reported that the frequency with which dyads
©

punished noncontingently (i.e., independently of whether the punishment was

ht

a retaliatory response to punishment or the withholding of benefits) was asso-

ig

ciated with lower rates of rewarding: It is only when punishment is contingent

yr

on previous punishment or nonreward that it promotes cooperation.

op

In some situations, one can benefit from revenge’s efficacy as a deterrent

simply by possessing the ability to retaliate against one’s interaction partners;
C

it is not always necessary to retaliate directly. Work in behavioral economics

226 MC CULLOUGH, KURZBAN, AND TABAK

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 227

also illustrates this basic point. Consider the difference in play in the dictator
game (DG) as opposed to the ultimatum game (UG). In both games, some
amount of money, say $10, is to be divided between two people. In the DG,
one person unilaterally decides how to split the money. In the UG, one person,

A)
the “proposer,” proposes a split, and the other person, the “responder,” can
either accept that split or reject it, in which case both players receive nothing.

P
(A
Rejection in the UG is revenge; the cost imposed is the amount that the pro-
poser allocated to him- or herself. It is not surprising that typical proposals in

n
the UG (roughly 40% of the stake), in which revenge is possible, are larger

io
than in the DG (roughly 20% of the stake), in which revenge is not possible

at
(Forsythe, Horowitz, Savin, & Sefton, 1994).

ci
Social psychology experiments also show how the prospect of suffering

so
revenge can deter aggressors from harming the prospective avenger. In one study

As
(Diamond, 1977), undergraduate men wrote an essay, which a confederate
then derogated. Participants were then brought back to the laboratory 24 hours

al
later and were given the opportunity to give 10 (bogus) shocks of varying

ic
intensities to the person who wrote the insulting reviews. Half were led to

og
believe that after they administered shocks, they would then switch roles and
ol
receive the shocks themselves. People who believed that they could harm their
ch
insulting evaluators without the threat of retaliation gave stronger shocks to
y

the evaluators.
Ps

The lessons of empirical studies on the direct deterrent effects of punish-

ment are not always straightforward, however. For example, Fehr and List (2004)
an

used the trust game, a two-step dyadic game in which an “investor” first entrusts
ic

a sum of money to a second person called the “trustee”; the money is then
er

multiplied by some constant (often tripled) by the researchers. In the second

Am

step, the trustee is given the opportunity to return some amount of money to
the investor. Fehr and List permitted investors to indicate a minimum amount
e

of money they required from their trustees in return. If trustees failed to return
th

that minimum amount, that amount was automatically deducted from the
by

trustees’ payoffs. Return transfers were highest when this punishment option
was available but left unused by investors. Nonetheless, the majority of investors
11

used the punishment option. Houser et al. (2008), using a similar design, found
20

that the threat of punishment reduced the fraction of money trustees returned
to investors even if the threat of punishment was applied as a result of a random
©

process rather than as a decision on the part of the investor. Likewise, revenge
ht

in some experimental settings increases, rather than deters, noncooperation.

ig

Dreber, Rand, Fudenberg, and Nowak (2008) found that using punishment
yr

against noncooperators reduced players’ gains in an iterated prisoner’s dilemma,

op

possibly because punishment in this experiment imposed large costs: The cost
of punishment and the size of the damage it inflicts clearly influence revenge’s
C

deterrent effects.

EVOLVED MECHANISMS FOR REVENGE AND FORGIVENESS 227

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 228

Deterring Third Parties

Mechanisms for revenge may also have been naturally selected for their
efficacy in deterring would-be aggressors by virtue of revenge’s ability to signal
the avenger’s aggressive potential. Reputation is important for understanding

A)
how third-party deterrence might work. Ancestral humans lived in small, close-

P
knit groups (Boehm, 2008) without the benefit of institutions for protecting

(A
individual rights, so a readiness to retaliate against interpersonal harms might
have been an important component of people’s social reputations. Researchers

n
io
have documented the importance of defense of honor (i.e., more or less, the

at
perceived ability to defend one’s interests with violent force when necessary)

ci
and the revenge that it stimulates as a major cause of violence among indi-

so
viduals from many societies (for a review, see McCullough, 2008; see also
Chapter 10, this volume).

As
Consistent with the idea that revenge is enacted partly out of reputational

al
concerns, laboratory studies show that the psychological mechanisms that cause

ic
revenge are sensitive to the presence of third parties. Victims retaliate more

og
strongly against their provokers when an audience has witnessed the provo-
cation, especially if the audience communicates to the victim that he or she
ol
looks weak because of the harm suffered or if the victim knows that the audience
ch
is aware that he or she has suffered particularly unjust treatment (Brown, 1968;
y

Kim, Smith, & Brigham, 1998). Also, when two men get into an argument,
Ps

the mere presence of a third person doubles the likelihood that the argument
an

will escalate to a violent encounter (Felson, 1982).

ic
er

Deterring the Withholding of Benefits

Am

Finally, we think mechanisms for revenge might have been naturally

selected because of their efficacy in changing others’ behavior to increase the
e

delivery of benefits (as opposed to only reducing harm). Public goods games
th

are useful for illustrating how revenge can deter the withholding of benefits.
by

In these games, a few (often four to six) participants receive an initial endow-
ment of money and are instructed to choose how to split that endowment
11

between two different pools. One pool is private, and participants simply keep
20

any money they place in it. The other pool is shared; money placed into this
pool is multiplied by some amount greater than one, and the resultant total
©

is subsequently divided evenly among all the players in the group. Money
ht

maximizers keep everything in their private pools; aggregate group wealth

ig

is maximized when everyone contributes to the public pool. These games

yr

are social dilemmas because they create a tension between individual and
op

group outcomes. (The fact that they involve groups rather than dyads is
incidental.)
C

228 MC CULLOUGH, KURZBAN, AND TABAK

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 229

Yamagishi (1986) had subjects play public goods games in groups of four,
repeated over 12 trials. He varied whether participants could punish other
members of the group and varied the price of punishment, that is, the cost one
had to spend to reduce another player’s payoff by one unit. Players used the sanc-

A)
tioning system when it was provided, and, in its presence, players contributed
greater amounts to the public good. However, these results do not distinguish

P
(A
the proximate motive, that is, whether sanctioning is instrumental (i.e., the
result of a motive to increase one’s benefits through the use of incentives) or

n
vengeful (i.e., the result of a motive to impose costs on individuals who had

io
an opportunity to deliver benefits but chose not to do so).

at
Fehr and Gächter’s (2002) results help to clarify the proximate motive

ci
for punishment in this context. Fehr and Gächter ran a similar game with a

so
few modifications, the most important of which was that players changed

As
groups from round to round, so punishment could not be used to induce group
members who were uncooperative in round r to benefit the subject in round

al
r + 1. Nevertheless, their results were similar to Yamagishi’s (1986): Participants

ic
sanctioned uncooperative group members, and group members cooperated more

og
when the punishment option was available (see also Anderson & Putterman,
ol
2006). These results imply the operation of the revenge system, given that
ch
instrumental motives were ruled out. Fehr and Gächter would not agree. They
y

coined a new term, altruistic punishment, to describe their findings.

Ps

Carpenter and Matthews (in press) conducted an experiment that con-

tained an important control condition that helps to identify the limits of any
an

altruism that might be present in so-called altruistic punishment. They ran

ic

noniterated public goods games and varied whether participants could punish
er

members of their own groups or members of other people’s groups. In the key
Am

treatment, the “one-way TPP” (third-party punishment) condition, almost

no one punished. The fact that one-way third-party punishment was so min-
e

imal when directed toward noncooperators in groups to which the subjects

th

themselves did not belong strongly suggests that without the possibility of
by

revenge, people tend not to punish.

11
20

EVOLUTION OF FORGIVENESS
©

In the section to follow, we attempt to outline the basics of a functional

ht

approach to understanding forgiveness. As we did in outlining the basics of a

ig

functional approach to revenge, we offer a functional definition of forgive-

yr

ness and illustrate its advantages over previous definitional approaches. We

op

also outline the evolutionary selection pressures that might have given rise to
psychological mechanisms that produce forgiveness as a functional output.
C

EVOLVED MECHANISMS FOR REVENGE AND FORGIVENESS 229

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 230

Forgiveness: A Functional Definition

Natural selection gave rise to one or more psychological systems that

produce revenge, we posit, by virtue of the fitness payoffs associated with

A)
direct deterrence, third-party deterrence, and, possibly, deterrence of benefit-
withholding. However, avengers trade off the potential benefits lost by virtue

P
(A
of any damage that revenge does to relations with the harm doer, and they
incur the (probabilistic) costs associated with any counterrevenge that might

n
ensue as the result of their revenge. We therefore presume that the revenge

io
system is designed to adjust its operation in response to the potential costs

at
and benefits associated with revenge in any particular instance. When the

ci
costs of revenge are too high relative to its expected deterrence benefits, an

so
organism might pursue an alternative course of behavior—forgiveness being

As
one of the more likely ones.
Over the past decade, the first author’s research group has defined

al
forgiveness as a set of motivational changes whereby an organism becomes

ic
(a) decreasingly motivated to retaliate against an offending relationship partner;

og
(b) decreasingly motivated to avoid the offender; and (c) increasingly moti-
ol
vated by good will for, and a desire to reconcile with, the offender, despite the
ch
offender’s harmful actions (McCullough, 2008; McCullough, Worthington, &
y

Rachal, 1997). Here, we refine this definition by adding a functional addendum:

Ps

that one or more “forgiveness systems” produce these motivational changes

because of their efficacy during evolution in promoting the restoration of
an

beneficial relationships in the aftermath of interpersonal harms.

ic

This newly “functionalized” definition of forgiveness permits all of

er

the important conceptual distinctions that other theorists (e.g., Enright &
Am

Coyle, 1998; Worthington, 2005) consider important (e.g., that forgiveness

is different from forgetting an offense, denying its reality, condoning it, or
e

attempting to minimize its significance), and it enables a tighter conceptual link

th

between forgiveness and reconciliation than has previously been recognized.

by

Many theorists have been careful to distinguish forgiveness from reconciliation,

with the latter concept indicating a restoration of the relationship between
11

offender and victim (Worthington, 2005). In light of the functional defini-

20

tion of forgiveness that we propose, it might be possible to forgive a harm doer

(i.e., to experience motivational changes by which one becomes less vengeful,
©

less avoidant, and more benevolently disposed toward him or her) without
ht

reconciling (i.e., restoring the relationship). Nevertheless, we reason that

ig

modern humans are capable of forgiving because ancestral humans who deployed
yr

this strategy enjoyed the fitness benefits that came from restoring potentially
op

valuable relationships.
Nevertheless, forgiveness, like revenge, involves costs. Forgiveness
C

prepares a victim to reenter constructive relations with a harm doer based on

230 MC CULLOUGH, KURZBAN, AND TABAK

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 231

the prospect of capturing benefits from that relationship, but forgiveness entails
foregoing revenge and its deterrent effects. Forgiveness, therefore, involves a
loss of gains from changing the harm doer’s incentives, potentially inviting
recidivism (e.g., see Gordon, Burton, & Porter, 2004) and attacks from

A)
those who see the opportunity to exploit the forgiver. In short, forgiveness
undermines the function of the revenge system by undermining deterrence.

P
(A
Thus, a forgiveness system, like a revenge system, should be sensitive to costs
and benefits, and these costs and benefits should have shaped the suite of

n
proximate social–psychological factors that turn the system on and off.

io
at
Forgiveness: Selection Pressures

ci
so
As noted earlier, we hypothesize that putative forgiveness systems evolved

As
in response to selection pressures for restoring relationships that, on average,
would have boosted lifetime reproductive fitness, a quality that researchers

al
have called relationship value (de Waal, 2000). The role of relationship value

ic
in determining animals’ propensity to forgive and/or reconcile after conflict

og
has been demonstrated in many simulations of the evolution of cooperation
ol
among dyads and networks of individuals (e.g., Axelrod, 1984; Hruschka &
ch
Henrich, 2006). Similar findings (Koski, Koops, & Sterck, 2007; Watts, 2006)
y

have emerged from behavioral studies of many mammalian species’ postconflict

Ps

conciliatory behaviors. It is in relationships in which substantial potential fitness

gains are possible (e.g., kin, mates, allies, exchange partners) that forgiveness
an

and/or reconciliation appear to be most common in nonhuman animals.

ic

The benefits to lifetime reproductive fitness differ by relationship type.

er

They might entail, of course, inclusive fitness benefits (Hamilton, 1964).

Am

After all, imposing costs on one’s close genetic relatives directly impairs one’s
own inclusive fitness. Also, kin are most likely, all else being equal, to be the
e

source of direct and reciprocal benefits for reasons associated with kin altruism.
th

Therefore, one might expect forgiveness to be more likely in the context of

by

kin relationships, with closer relatives being more easily forgiven than more
distant ones.
11

Social organisms will also undergo selection pressure for forgiveness in

20

the context of cooperation between nonrelatives when repeated encounters

are likely (Axelrod, 1984; Trivers, 1971). Individuals who could forgive in
©

such contexts would acquire two fitness benefits. First, forgiving isolated
ht

transgressions would have inhibited the echo effect (Axelrod, 1984), whereby
ig

individuals who are cooperatively disposed nevertheless become locked in

yr

costly cycles of retaliation when initial unintended defections occur due to

op

noise. Second, individuals who can forgive their reciprocal altruism partners
following defections would have been able to preserve their access to benefits
C

that their partners would have been able to provide them and would have

EVOLVED MECHANISMS FOR REVENGE AND FORGIVENESS 231

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 232

spared themselves the costs associated with establishing new relationships

with new individuals whose social dispositions would be unknown (Hruschka
& Henrich, 2006). On average, it may simply be less costly to forgive some
number of defections from a well-established relationship partner than to

A)
retaliate, or to withdraw from the relationship, following an isolated defection.
Indeed, in computer simulations of the evolution of reciprocal altruism—

P
(A
especially when the possibility of noise is assumed—evolutionarily stable
strategies tend to be more forgiving than tit for tat, which responds to defection

n
with defection and to cooperation with cooperation (Frean, 1994; Hauert &

io
Schuster, 1998; Wu & Axelrod, 1995). This is especially true when one

at
models reciprocal altruism as occurring largely among small networks of

ci
individuals (e.g., friendship groups, individuals within small living groups)

so
who focus their cooperative efforts on other individuals within the net-

As
work and limit their cooperation with individuals outside of the network
(Levine & Kurzban, 2006). Under such circumstances, agents are expected

al
to forgive up to 80% of other network members’ defections (Hruschka &

ic
Henrich, 2006).

og
Other types of relationships generate still other types of benefits that
ol
redound to lifetime reproductive fitness. The benefits that might accrue from
ch
forgiving a mate are different from the benefits that might accrue by forgiving
y

a friend, which in turn are different from the benefits that a forgiver might
Ps

receive by forgiving an ally. Because the fitness-enhancing properties come

in different currencies, the psychological systems that produce forgiveness are
an

likely set up to identify the types of benefits that a particular type of relation-
ic

ship is likely to confer (and to weight them appropriately with respect to the
er

probability of capturing those benefits, the time horizon at which they will be
Am

realized, etc.) and then weigh those benefits against the deterrent value of
revenge, which the organism would trade off if it chose to forgive instead of
e

seeking revenge (McCullough, Luna, Berry, Tabak, & Bono, in press).

th
by

CHOOSING FORGIVENESS OVER VENGEANCE:

11

PROXIMATE CAUSATION
20

If, as we hypothesize, forgiveness systems are sensitive to tradeoffs asso-

©

ciated with sacrificing the deterrence benefits of revenge for the relationship-
ht

restoration benefits of forgiveness, then such systems should be acutely

ig

sensitive to variables that influence the value of each option. These variables
yr

include, but are not necessarily limited to, characteristics of the offender, the
op

transgression itself, and cues that predict the probabilities of future attacks
and/or the potential future value of the restored relationship. In other words,
C

we predict that forgiveness is generated by systems designed to compute and

232 MC CULLOUGH, KURZBAN, AND TABAK

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 233

compare the cost of forgone revenge and the benefits that are expected to
accrue from a restored relationship.

Value of Deterrence

A)
The value of revenge diminishes to the extent that it does not change

P
(A
behavior that would otherwise occur. In the limiting case, suppose that after
an offense, the transgressor could persuasively signal that he or she would

n
never—or could never—again inflict costs. In such a case, revenge would

io
yield no benefit (except through third-party deterrence).

at
Information relevant to inferring intent can come from various sources.

ci
For instance, a transgressor’s apology, expression of sympathy for a victim’s

so
suffering, and declaration of his or her intention to behave better in the future

As
could indicate a low likelihood of trying to harm the victim in the future
(McCullough et al., 1997). Verbal declarations such as these are susceptible

al
to strategic manipulation, of course. Nonverbal displays such as blushing,

ic
which facilitate forgiveness after some transgressions (de Jong, Peters, &

og
de Cremer, 2003), also contain information about changed intent and a
ol
transgressor’s eagerness to distance himself or herself from a transgression, and
ch
their reliability may come from their unfakeability (Frank, 1988).
y

Other situational features might reduce the perceived deterrent value of

Ps

revenge by convincing a victim that the transgressor’s harmful actions were

unintentional in the first place. People more readily forgive transgressors whose
an

behavior was unintentional, unavoidable, or committed without awareness

ic

of the potential negative consequences (McCullough et al., in press). Also,

er

it is unnecessary to engage in deterrence when additional transgressions

Am

are impossible. When the aggressor’s capacity for violence is removed, for
instance, vengeance yields little additional deterrent value. In some ethno-
e

graphic accounts, reconciliation rituals involve the surrender of weapons

th

(e.g., Boehm, 1987), which seems well suited to conveying an unwillingness

by

to commit future aggressive acts. Trust may be a key psychological process by

which the aforementioned factors that cue benevolent intentions lower the
11

likelihood of revenge and raise the likelihood of forgiveness (Kurzban, 2003).

20

People more readily forgive people whom they trust (Hewstone, Cairns, Voci,
Hamberger, & Niens, 2006) and people who are reputed to be trustworthy
©

(Vasalou, Hopfensitz, & Pitt, 2008) despite their recent bad behavior.
ht
ig

Expected Value of the Relationship

yr
op

Against the costs of forgone revenge is the expected value of future ben-
efits in a relationship in which intentions are positive rather than negative.
C

The expected future value of a relationship is computed, we hypothesize, in

EVOLVED MECHANISMS FOR REVENGE AND FORGIVENESS 233

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 234

much the same way that it would be in contexts other than the aftermath of
a transgression. Because of the well-known principles of kin selection, close
relatives are likely to be a source of benefits, and thus, we expect that cues of
kinship will facilitate forgiveness, just as they evidently facilitate the restraint

A)
of vengeance (Lieberman & Linke, 2007).
In similar fashion, those with whom one has a close history of association,

P
(A
shared interests, and many opportunities for mutually beneficial transactions
are good candidates for forgiveness because of the possibility of continued gains.

n
Indeed, priming people with the names of close others (e.g., via subliminal

io
presentation) leads to increased judgments of forgiveness, increased accessi-

at
bility of the concept of forgiveness, and reduced deliberation about whether

ci
forgiveness is an appropriate course of action (Karremans & Aarts, 2007).

so
Karremans and Aarts’s (2007) results complement those from several previous

As
studies showing that people are more inclined to forgive individuals with whom
they feel close and committed (Finkel, Rusbult, Kumashiro, & Hannon, 2002;

al
McCullough et al., 1998; see also Chapter 2, this volume). We would argue

ic
that the reason for these associations of closeness and/or commitment with

og
forgiveness is that relationship closeness and commitment act as cues of
ol
relationship value in many types of relationships. We think the importance
ch
of relationship value can also explain why people tend to want some form of
y

compensation prior to forgiving (Boehm, 1987; Bottom, Gibson, Daniels, &

Ps

Murnighan, 2002): Compensation may serve as a cue of (among other things)

a transgressor’s ability or willingness to be a valuable relationship partner in
an

the future.
ic

In computations of expected value, empathy may play a special role.

er

Empathy for transgressors, which is a sympathy- or pity-like response to the

Am

plight of another person, appears to be a reliable facilitator of forgiveness

(McCullough et al., 1997), perhaps as a result of empathy’s long phylogenetic
e

history as a motivator of care for valuable relationship partners (Preston &

th

de Waal, 2002). Whether the empathy–forgiveness link is part of the design

by

of the forgiveness system or merely an incidental effect that empathy can exert
within the existing forgiveness system, however, is currently difficult to know.
11
20

SUMMARY
©
ht

The desire for revenge and the ability to forgive seem to be universal
ig

psychological endowments of humans (Boehm, 2008; Daly & Wilson, 1988;

yr

McCullough, 2008). Species-typical traits call out for explanations in terms of

op

the mind’s evolved mental structures, either as direct products or as by-products

of what those structures were designed to do (Andrews et al., 2002). Here, we
C

have taken an adaptationist stance and posited that revenge and forgiveness

234 MC CULLOUGH, KURZBAN, AND TABAK

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 235

result from computational mechanisms designed to produce them. Once one

has moved into a functional framework, we think it becomes easier to see what
should qualify as revenge and forgiveness—and what should not—and what
the important questions are if one wants to understand what revenge and

A)
forgiveness are really all about. By outlining the selection pressures that likely
gave rise to humans’ penchants for revenge and forgiveness, we have also tried

P
(A
here to identify the types of information that the structures that produce
revenge and forgiveness should be designed to process. We hope that intro-

n
ducing this sort of thinking can help investigators prioritize their research

io
efforts in the future.

at
ci
so
REFERENCES

As
Anderson, C., & Putterman, L. (2006). Do non-strategic sanctions obey the law of

al
demand? The demand for punishment in the voluntary contribution mechanism.

ic
Games and Economic Behavior, 54, 1–24. doi:10.1016/j.geb.2004.08.007

og
Andrews, P. W., Gangestad, S. W., & Matthews, D. (2002). Adaptationism—How to
ol
carry out an exaptationist program. Behavioral and Brain Sciences, 25, 489–504.
ch
doi:10.1017/S0140525X02000092
y

Axelrod, R. (1984). The evolution of cooperation. New York, NY: Basic Books.
Ps

Bixenstine, V. E., & Wilson, K. V. (1963). Effects of level of cooperative choice by

an

the other player on choices in a prisoner’s dilemma game. Part II. Journal of
Abnormal and Social Psychology, 67, 139–147. doi:10.1037/h0044242
ic

Boehm, C. (1987). Blood revenge: The enactment and management of conflict in Montenegro
er

and other tribal societies (2nd ed.). Philadelphia: University of Pennsylvania Press.
Am

Boehm, C. (2008). Purposive social selection and the evolution of human altruism.
Cross-Cultural Research, 42, 319–352. doi:10.1177/1069397108320422
e
th

Bottom, W. P., Gibson, K., Daniels, S. E., & Murnighan, J. K. (2002). When talk is not
cheap: Substantive penance and expressions of intent in rebuilding cooperation.
by

Organization Science, 13, 497–513. doi:10.1287/orsc.13.5.497.7816

11

Brown, B. R. (1968). The effects of need to maintain face on interpersonal bargaining.

Journal of Experimental Social Psychology, 4, 107–122. doi:10.1016/0022-1031
20

(68)90053-X
©

Carpenter, J. P., & Matthews, P. H. (in press). Norm enforcement: Anger, indignation,
or reciprocity? Journal of the European Economic Association.
ht

Clark, K., & Sefton, M. (2001). The sequential prisoner’s dilemma: Evidence on
ig

reciprocation. The Economic Journal, 111, 51–68. doi:10.1111/1468-0297.00588

yr

Clutton-Brock, T. H., & Parker, G. A. (1995). Punishment in animal societies. Nature,

op

373, 209–216. doi:10.1038/373209a0

C

Daly, M., & Wilson, M. (1988). Homicide. New York, NY: Aldine de Gruyter.

EVOLVED MECHANISMS FOR REVENGE AND FORGIVENESS 235

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 236

de Jong, P. J., Peters, M. L., & de Cremer, D. (2003). Blushing may signify guilt:
Revealing effects of blushing in ambiguous social situations. Motivation and
Emotion, 27, 225–249. doi:10.1023/A:1025059631708
de Waal, F. B. M. (2000, July). Primates: A natural heritage of conflict resolution.
Science, 289, 586–590. doi:10.1126/science.289.5479.586

A)
Diamond, S. R. (1977). The effect of fear on the aggressive responses of anger aroused

P
and revenge motivated subjects. Journal of Psychology, 95, 185–188.

(A
DiBlasio, F., & Benda, B. B. (1991). Practitioners, religion and the use of forgiveness

n
in the clinical setting. Journal of Psychology and Christianity, 10, 166–172.

io
Dreber, A., Rand, D. G., Fudenberg, D., & Nowak, M. A. (2008, March 20). Winners

at
don’t punish. Nature, 452, 348–351. doi:10.1038/nature06723

ci
Elster, J. (1990). Norms of revenge. Ethics, 100, 862–885. doi:10.1086/293238

so
Enright, R. D., & Coyle, C. T. (1998). Researching the process model of forgiveness

As
within psychological interventions. In E. L. Worthington (Ed.), Dimensions
of forgiveness: Psychological research and theological perspectives (pp. 139–161).

al
Philadelphia, PA: Templeton Foundation Press.

ic
og
Fehr, E., & Gächter, S. (2002). Altruistic punishment in humans. Nature, 415,
137–140. doi:10.1038/415137a
ol
Fehr, E., & List, J. A. (2004). The hidden costs and returns of incentives: Trust and
ch
trustworthiness among CEOs. Journal of the European Economic Association, 2,
y

743–771. doi:10.2139/ssrn.364480
Ps

Felson, R. B. (1982). Impression management and the escalation of aggression and

violence. Social Psychology Quarterly, 45, 245–254. doi:10.2307/3033920
an

Finkel, E. J., Rusbult, C. E., Kumashiro, M., & Hannon, P. A. (2002). Dealing with a
ic

betrayal in close relationships: Does commitment promote forgiveness? Journal of

er

Personality and Social Psychology, 82, 956–974. doi:10.1037/0022-3514.82.6.956

Am

Forsythe, R., Horowitz, J. L., Savin, N. E., & Sefton, M. (1994). Fairness in simple
bargaining experiments. Games and Economic Behavior, 6, 347–369. doi:10.1006/
e

game.1994.1021
th

Frank, R. H. (1988). Passions within reason: The strategic role of the emotions. New York,
by

NY: Norton.
Frean, M. R. (1994). The prisoner’s dilemma without synchrony. Proceedings of the
11

Royal Society of London B, 257, 75–79. doi:10.1098/rspb.1994.0096

20

Gordon, K. C., Burton, S., & Porter, L. (2004). Predicting the intentions of women
in domestic violence shelters to return to partners: Does forgiveness play a role?
©

Journal of Family Psychology, 18, 331–338. doi:10.1037/0893-3200.18.2.331

ht

Govier, T. (2002). Forgiveness and revenge. New York, NY: Routledge.

ig

Hamilton, W. D. (1964). The genetical evolution of social behaviour. Parts I & II.
yr

Journal of Theoretical Biology, 7, 1–52.

op

Hauert, C., & Schuster, H. G. (1998). Extending the iterated prisoner’s dilemma
without synchrony. Journal of Theoretical Biology, 192, 155–166. doi:10.1006/
C

jtbi.1997.0590

236 MC CULLOUGH, KURZBAN, AND TABAK

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 237

Hayashi, N., Ostrom, E., Walker, J., & Yamagishi, T. (1999). Reciprocity, trust, and
the sense of control: A cross-societal study. Rationality and Society, 11, 27–46.
doi:10.1177/104346399011001002
Hebl, J., & Enright, R. D. (1993). Forgiveness as a psychotherapeutic goal with elderly
females. Psychotherapy, 30, 658–667. doi:10.1037/0033-3204.30.4.658

A)
Hewstone, M., Cairns, E., Voci, A., Hamberger, J., & Niens, U. (2006). Intergroup

P
contact, forgiveness, and experience of “The Troubles” in Northern Ireland.

(A
Journal of Social Issues, 62, 99–120. doi:10.1111/j.1540-4560.2006.00441.x

n
Home Office. Research, Development and Statistics Directorate. Offending Surveys

io
and Research, National Centre for Social Research and BMRB. (2003). Social

at
Research, Offending, Crime and Justice Survey, 2003 [computer file] (Report

ci
No. SN: 5248, 3rd ed.). Retrieved from http://www.data-archive.ac.uk/doc/5248%

so
5Cmrdoc%5CUKDA%5CUKDA_Study_5248_Information.htm

As
Houser, D., Xiao, E., McCabe, K., & Smith, V. (2008). When punishment fails:
Research on sanctions, intentions and noncooperation. Games and Economic

al
Behavior, 62, 509–532.

ic
Hruschka, D. J., & Henrich, J. (2006). Friendship, cliquishness, and the emergence

og
of cooperation. Journal of Theoretical Biology, 239, 1–15. doi:10.1016/j.jtbi.
2005.07.006 ol
ch
Jensen, K., Call, J., & Tomasello, M. (2007). Chimpanzees are vengeful but not spite-
ful. Proceedings of the National Academy of Sciences of the United States of Amer-
y
Ps

ica, 104, 13046–13050. doi:10.1073/pnas.0705555104

Karremans, J. C., & Aarts, H. (2007). The role of automaticity in determining the
an

inclination to forgive close others. Journal of Experimental Social Psychology, 43,

902–917. doi:10.1016/j.jesp.2006.10.012
ic
er

Kim, S. H., Smith, R. H., & Brigham, N. L. (1998). Effects of power imbalance and
the presence of third parties on reactions to harm: Upward and downward
Am

revenge. Personality and Social Psychology Bulletin, 24, 353–361. doi:10.1177/

0146167298244002
e

Koski, S. E., Koops, K., & Sterck, E. H. M. (2007). Reconciliation, relationship quality,
th

and postconflict anxiety: Testing the integrated hypothesis in captive chimpanzees.

by

American Journal of Primatology, 69, 158–172. doi:10.1002/ajp.20338

Kubrin, C. E., & Weitzer, R. (2003). Retaliatory homicide: Concentrated disadvantage
11

and neighborhood culture. Social Problems, 50, 157–180. doi:10.1525/sp.2003.

20

50.2.157
Kurzban, R. (2003). Trust, reciprocity, and gains from association: Interdisciplinary
©

lessons from experimental research. In E. Ostrom & J. Walker (Eds.), Biological

ht

foundations of reciprocity (pp. 105–127). New York, NY: Sage.

ig

Levine, S. S., & Kurzban, R. (2006). Explaining clustering in social networks: Towards
yr

an evolutionary theory of cascading benefits. Managerial and Decision Economics,

op

27, 173–187. doi:10.1002/mde.1291

Lieberman, D., & Linke, L. (2007). The effect of social category on third party
C

punishment. Evolutionary Psychology, 5, 289–305.

EVOLVED MECHANISMS FOR REVENGE AND FORGIVENESS 237

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 238

Lundahl, B. W., Taylor, M. J., Stevenson, R., & Roberts, K. D. (2008). Process-based
forgiveness interventions: A meta-analytic review. Research on Social Work
Practice, 18, 465–478. doi:10.1177/1049731507313979
Maynard Smith, J., & Harper, D. (2003). Animal signals. Oxford, England: Oxford.

A)
McCullough, M. E. (2008). Beyond revenge: The evolution of the forgiveness instinct.

P
San Francisco, CA: Jossey-Bass.

(A
McCullough, M. E., Luna, L. R., Berry, J. W., Tabak, B. A., & Bono, G. (in press).
On the form of forgiving: Modeling the time-forgiveness relationship and testing

n
the valuable relationships hypothesis. Emotion.

io
at
McCullough, M. E., Rachal, K. C., Sandage, S. J., Worthington, E. L., Brown, S. W., &

ci
Hight, T. L. (1998). Interpersonal forgiving in close relationships. II: Theoretical

so
elaboration and measurement. Journal of Personality and Social Psychology, 75,
1586–1603. doi:10.1037/0022-3514.75.6.1586

As
McCullough, M. E., Worthington, E. L., & Rachal, K. C. (1997). Interpersonal

al
forgiving in close relationships. Journal of Personality and Social Psychology, 73,

ic
321–336. doi:10.1037/0022-3514.73.2.321

og
Miller, N., Pedersen, W. C., Earleywine, M., & Pollock, V. E. (2003). A theoretical
model of triggered displaced aggression. Personality and Social Psychology Review,
7, 75–97. doi:10.1207/S15327957PSPR0701_5
ol
ch
Molm, L. D. (1997). Coercive power in social exchange. New York, NY: Cambridge
y
Ps

University Press.
Murphy, J. G. (2003). Getting even: Forgiveness and its limits. New York, NY: Oxford.
an

Preston, S. D., & de Waal, F. B. M. (2002). Empathy: Its ultimate and proximate bases.
ic

Behavioral and Brain Sciences, 25, 1–72.

er

Rapoport, A., & Chammah, A. M. (1965). The prisoner’s dilemma. Ann Arbor:
Am

University of Michigan Press.

Speckhard, A., & Ahkmedova, K. (2006). The making of a martyr: Chechen suicide
e

terrorism. Studies in Conflict and Terrorism, 29, 429–492. doi:10.1080/10576100

th

600698550
by

Trivers, R. L. (1971). The evolution of reciprocal altruism. The Quarterly Review of

Biology, 46, 35–57. doi:10.1086/406755
11

Uniacke, S. (2000). Why is revenge wrong? The Journal of Value Inquiry, 34, 61–69.
20

doi:10.1023/A:1004778229751
Vasalou, A., Hopfensitz, A., & Pitt, J. V. (2008). In praise of forgiveness: Ways for
©

repairing trust breakdowns in one-off online interactions. International Journal

ht

of Human-Computer Studies, 66, 466–480. doi:10.1016/j.ijhcs.2008.02.001

ig

Vossekuil, B., Fein, R. A., Reddy, M., Borum, R., & Modzeleski, W. (2002). The final
yr

report and findings of the Safe School Initiative: Implications for the prevention of
op

school attacks in the United States. Washington, DC: U.S. Department of Education,
Office of Elementary and Secondary Education, Safe and Drug-Free Schools
C

Program and U.S. Secret Service, National Threat Assessment Center.

238 MC CULLOUGH, KURZBAN, AND TABAK

12051-13_CH12-rev2.qxd 5/25/10 11:55 AM Page 239

Watts, D. (2006). Conflict resolution in chimpanzees and the valuable-relationships

hypothesis. International Journal of Primatology, 27, 1337–1364. doi:10.1007/
s10764-006-9081-9
Williams, G. C. (1966). Adaptation and natural selection. A critique of some current

A)
evolutionary thought. Princeton, NJ: Princeton University Press.
Wilson, W. (1971). Reciprocation and other techniques for inducing cooperation.

P
(A
The Journal of Conflict Resolution, 15, 167–195. doi:10.1177/002200277101500205
Worthington, E. L. (2005). More questions about forgiveness: Research agenda for

n
2005–2015. In E. L. Worthington (Ed.), Handbook of forgiveness (pp. 557–573).

io
New York, NY: Routledge.

at
Wu, J., & Axelrod, R. (1995). How to cope with noise in the iterated prisoner’s dilemma.

ci
The Journal of Conflict Resolution, 39, 183–189. doi:10.1177/0022002795039001008

so
Yamagishi, T. (1986). The provision of a sanctioning system as a public good. Journal

As
of Personality and Social Psychology, 51, 110–116. doi:10.1037/0022-3514.51.1.110

al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

EVOLVED MECHANISMS FOR REVENGE AND FORGIVENESS 239

C
op
yr
ig
ht
©
12051-13_CH12-rev2.qxd

20
11
5/25/10

by
th
e
11:55 AM

Am
er
ic
Page 240

an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 241

13

PA)
ATTACHMENT, ANGER,

(A
AND AGGRESSION

n
io
at
ci
MARIO MIKULINCER AND PHILLIP R. SHAVER

so
As
al
ic
og
ol
y ch
Ps
an

In recent years, attachment theory (Bowlby, 1973, 1980, 1982), designed

ic

originally to characterize infant–parent emotional bonding, has been applied

er

first to the study of adolescent and adult romantic relationships (e.g., Hazan
Am

& Shaver, 1987) and then more broadly to the study of emotion regulation,
social motives, and diverse forms of social behavior (Mikulincer & Shaver,
e

2003; see also Chapter 4, this volume). In this chapter, we explore the theory’s
th

relevance to understanding both normative and individual-difference aspects

by

of human power and aggression. We begin with a brief summary of attachment

theory and an account of the two major dimensions of attachment insecurity
11

in adulthood. We then present a model of attachment-related processes in

20

adulthood, based on an extensive review of the attachment research literature

(Mikulincer & Shaver, 2007a). Next, we focus on the experience of anger
©

and on destructive forms of aggression (see Chapter 14, this volume). We

ht

review research on ways in which attachment processes included in our model

ig

affect (a) functional and dysfunctional forms of anger, (b) domestic violence,
yr

(c) antisocial criminal behavior, and (d) intergroup hostility and aggression.
op

Finally, we consider the main adaptive goal of human aggression: to main-

tain power and dominance (see Chapter 4, this volume). We consider how
C

241
12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 242

attachment orientations and the experience and exercise of power are related.
We also present some new exploratory research concerning the influence of
attachment-related processes on cognition and action when a person’s sense
of power is experimentally enhanced.

A)
P
(A
ATTACHMENT THEORY AND ATTACHMENT STYLE

n
According to Bowlby (1982), human beings are born with an innate

io
psychobiological system (the attachment behavioral system) that motivates

at
them to seek proximity to supportive others (attachment figures) in times of

ci
need. This system accomplishes basic regulatory functions (protection from

so
threats and alleviation of distress) in humans of all ages, but it is most directly

As
observable during infancy and early childhood (Bowlby, 1988). Bowlby (1973)
described important individual differences in the functioning of the attachment

al
system. Interactions with attachment figures who are available and responsive

ic
in times of need facilitate optimal attachment-system functioning and promote

og
a sense of attachment security, a sense that the world is safe, that attachment
ol
figures are helpful when called on, and that it is possible to explore the
ch
environment curiously and engage effectively and enjoyably with other
y

people. This sense of security is rooted in positive mental representations of

Ps

self and others, which Bowlby called internal working models. When attachment
figures are not reliably available and supportive, however, a sense of security
an

is not attained, negative internal working models are formed, and strategies of
ic

affect regulation other than appropriate proximity seeking (secondary attachment

er

strategies, conceptualized in terms of two major dimensions, avoidance and

Am

anxiety) are adopted.

In studies of adolescents and adults, tests of these theoretical ideas
e

have generally focused on a person’s attachment style, that is, the pattern of
th

relational expectations, emotions, and behavior that results from a particular

by

history of attachment experiences (Fraley & Shaver, 2000). Initially, research

on individual differences in attachment was based on Ainsworth, Blehar,
11

Waters, and Wall’s (1978) three-category typology of attachment patterns in

20

infancy (i.e., secure, anxious, and avoidant) and on Hazan and Shaver’s (1987)
conceptualization of similar adult styles in the romantic relationship domain.
©

Subsequent studies (e.g., Brennan, Clark, & Shaver, 1998) revealed, however,
ht

that attachment styles are more appropriately conceptualized as regions in a

ig

two-dimensional space. The first dimension, attachment anxiety, reflects the

yr

degree to which a person worries that an attachment figure (including adult

op

relationship partners) will not be available in times of need. The second

dimension, attachment-related avoidance, reflects the extent to which a person
C

distrusts relationship partners’ goodwill and strives to maintain behavioral

242 MIKULINCER AND SHAVER

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 243

independence and emotional distance from partners. People who score low
on both dimensions are said to be secure or to have a secure attachment style.
The two dimensions can be measured with reliable and valid self-report scales
and are associated in theoretically predictable ways with various aspects of

A)
personal adjustment and relationship quality (for a review, see Mikulincer &
Shaver, 2007a).

P
(A
Attachment styles are initially formed in interactions with primary care-
givers during early childhood, as a large body of research has shown (Cassidy

n
& Shaver, 2008), but Bowlby (1988) claimed that memorable interactions

io
with others throughout life can alter a person’s working models and move the

at
person from one region of the two-dimensional space to another. Moreover,

ci
although attachment style is often conceptualized as a single global orienta-

so
tion toward close relationships, it is actually rooted in a complex network of

As
cognitive and affective processes and mental representations, which includes
many episodic, context-related, and relationship-specific as well as general

al
attachment representations (Mikulincer & Shaver, 2003). In fact, many studies

ic
indicate that a person’s attachment style can change depending on context

og
and recent experiences (Mikulincer & Shaver, 2007b). This makes it possible
ol
to study the effects of security and insecurity experimentally.
y ch
Ps

ATTACHMENT-SYSTEM FUNCTIONING IN ADULTHOOD

an

On the basis of an extensive review of adult attachment studies,

ic

Mikulincer and Shaver (2003) proposed that when the attachment system
er

is activated, an affirmative answer to the implicit or explicit question “Is

Am

an attachment figure available and likely to be responsive to my needs?”

heightens the sense of attachment security and facilitates the use of constructive
e

emotion-regulation strategies. These strategies are aimed at alleviating distress,

th

maintaining supportive intimate relationships, and bolstering a person’s sense

by

of love-worthiness and self-efficacy. Moreover, they sustain what Shaver and

Mikulincer (2002), following Fredrickson (2001), called a broaden-and-build
11

cycle of attachment security, which expands a person’s resources for main-

20

taining coping flexibility and emotional stability in times of stress, broadens the
person’s perspectives and capacities, and facilitates the incorporation of mental
©

representations of security-enhancing attachment figures into the self. This

ht

broaden-and-build process allows relatively secure individuals to maintain an

ig

authentic sense of personal efficacy, resilience, and optimism even when

yr

social support is temporarily unavailable (Mikulincer & Shaver, 2007a).

op

Perceived unavailability of an attachment figure results in attachment

insecurity, which compounds the distress aroused by the appraisal of a situation
C

as threatening. This state of insecurity forces a decision about the viability of

ATTACHMENT, ANGER, AND AGGRESSION 243

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 244

further (more active) proximity seeking as a protective strategy. The appraisal

of proximity as feasible or essential—because of attachment history, tempera-
mental factors, or contextual cues—results in energetic, insistent attempts to
attain proximity, support, and love. These attempts are called hyperactivating

A)
strategies (Cassidy & Kobak, 1988) because they involve up-regulation of the
attachment system, including constant vigilance and intense concern until an

P
(A
attachment figure is perceived to be available and supportive. Hyperactivating
strategies include attempts to elicit a partner’s involvement, care, and support

n
through clinging and controlling responses (Shaver & Mikulincer, 2002);

io
overdependence on relationship partners as a source of protection (Shaver &

at
Hazan, 1993); and perception of oneself as relatively helpless with respect to

ci
emotion regulation (Mikulincer & Shaver, 2003). These strategies help to

so
explain the psychological correlates and consequences of attachment anxiety.

As
If a person has learned instead that relying on attachment figures is not
a safe or effective way to cope with threats, he or she is likely to downplay

al
such threats, inhibit worries and negative emotions, and defensively engage

ic
in what Bowlby (1982) called compulsive self-reliance. These coping strategies

og
are called deactivating (Cassidy & Kobak, 1988) because their goal is to keep
ol
the attachment system down-regulated rather than experience the frustration
ch
and pain of rejection, punishment for expressing feelings, or abandonment.
y

These strategies help to explain the psychological correlates and consequences

Ps

of avoidant attachment.
In short, each attachment strategy has a major regulatory goal (i.e.,
an

insisting on proximity to an attachment figure or on self-reliance), which goes

ic

along with particular cognitive and affective processes that facilitate goal
er

attainment. These strategies affect the experience, regulation, and expression

Am

of emotions (Mikulincer & Shaver, 2007a), including anger, which is of

special concern in the present chapter. Moreover, the strategies affect the
e

functioning of other behavioral systems, such as exploration and caregiving,

th

as we and others have shown in numerous studies (reviewed by Mikulincer &

by

Shaver, 2007a). As explained by Shaver et al. in Chapter 4 of the present

volume, power and dominance can be viewed as the goals of a separate behav-
11

ioral system. We expect the functioning of that system to be affected by dif-

20

ferences in attachment style (i.e., attachment anxiety and avoidance) just as

the other behavioral systems are.
©
ht
ig

ATTACHMENT AND ANGER

yr
op

In Bowlby’s (1973) analysis of infants’ emotional reactions to separa-

tion from an attachment figure, he viewed anger as a functional reaction
C

to separation because it sometimes motivates an attachment figure to pay

244 MIKULINCER AND SHAVER

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 245

more attention in the future and thereby provide better, more reliable
care. In general, especially for adults, anger is functional to the extent that
it communicates an intense but justifiable reaction to inconsiderate or
undeserved ill treatment, rather than simply being a way to injure or destroy

A)
a relationship partner through acts of revenge (see Chapters 3 and 12, this
volume). Bowlby (1973) called this constructive form of anger the “anger

P
(A
of hope” because it is intended to bring about a better future state of a
relationship. He also mentioned, however, that anger can become so intense

n
that it alienates or injures a partner, in which case it becomes destructive to

io
a relationship and can even lead to violence or death. He called this the

at
“anger of despair.”

ci
Functional anger is typical of people who feel secure in attachment

so
relationships. Mikulincer (1998) found, for example, that when secure adults

As
were hurt or frustrated by relationship partners’ behavior, they were optimistic
about the partners’ willingness to apologize and “reform.” Moreover, secure

al
people’s memories of their reactions to anger-provoking events were charac-

ic
terized by the constructive goal of repairing the relationship, engaging in

og
adaptive problem-solving, and restoring a positive mood following a conflict.
ol
Another study explored the functional nature of secure adolescents’ anger
ch
(Zimmermann, Maier, Winter, & Grossmann, 2001). Adolescent research
y

participants performed a difficult, frustrating problem-solving task with the

Ps

help of a friend, and their disappointment and anger during the task and their
negative behavior toward the friend (e.g., rejecting the friend’s suggestions
an

without discussion) were coded. Disappointment and anger were associated

ic

with more aggressive behavior only in the case of insecure adolescents.

er

Avoidant individuals’ deactivating strategies favor suppression of

Am

anger because anger implies emotional investment in a relationship, which is

incongruent with avoidant people’s emotional distance and extreme self-reliance
e

(Cassidy & Kobak, 1988). Avoidant people’s anger tends to be expressed only
th

in indirect ways and to take the form of nonspecific hostility or generally hate-
by

ful attitudes. Mikulincer (1998) found, for example, that avoidant adults did
not report intense anger in response to provocative experiences, but they
11

were aroused physiologically and attributed hostile intent to a hurtful other,

20

even when the other’s behavior was unintentional. In a laboratory study of

support seeking in dating couples, Rholes, Simpson, and Orina (1999) found
©

in a sample of women that avoidant attachment was associated with hostility

ht

toward their partner (coded from video recordings) when they worried about
ig

an upcoming painful task. Thus, although avoidant individuals are rarely

yr

comfortable describing themselves as needy or angry, they nevertheless react

op

with hostility and hatred.

Attachment-anxious individuals, because of their tendency to intensify
C

distress and ruminate about distressing experiences, are vulnerable to intense

ATTACHMENT, ANGER, AND AGGRESSION 245

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 246

and prolonged bouts of anger. However, their fear of separation, desperate

desire for love, and high dependency may keep them from expressing anger while
causing them to direct it toward themselves. They may react to provocations
and insults with a tangled mixture of resentment, hostility, anger, self-criticism,

A)
fear, sadness, and depression. Indeed, Mikulincer (1998) found that anxious
people’s memories of prior anger-provoking events included a flood of angry

P
(A
feelings, intense rumination, and a variety of negative emotions.
In a study of couple interactions, Simpson, Rholes, and Phillips (1996)

n
found that attachment anxiety was associated with displaying and reporting

io
more anger, hostility, and distress while discussing with a dating partner

at
an unresolved problem in their relationship. In a study of support seeking,

ci
Rholes et al. (1999) found that, although anxious attachment was unrelated

so
to anger toward a dating partner while waiting to undergo an anxiety-arousing

As
experience, after participants were told they would not have to undergo the
stressful task, attachment anxiety was associated with more intense expressions

al
of anger toward one’s partner. This pattern was particularly strong if participants

ic
had been more worried about the upcoming experience and had sought more

og
support from their partner. Thus, it seems that anxious individuals’ need for
ol
reassurance and support caused them to hold back feelings of frustration and
ch
anger while seeking a partner’s support. But once the support was no longer
y

needed (in that particular laboratory setting), the angry feelings surfaced and
Ps

were expressed.
In a study using psychophysiological measures, Diamond and Hicks
an

(2005) presented men with anger-provoking tasks (e.g., performing serial sub-
ic

traction while being criticized by an experimenter), measured anger during

er

and after the tasks, and recorded the men’s vagal tone, a physiological indica-
Am

tor of down-regulation of negative emotions. They found that attachment

anxiety was associated with both greater self-reported anger and lower vagal
e

tone during and after the tasks, suggesting intense anger that was difficult to
th

subdue.
by
11

ATTACHMENT AND AGGRESSION

20

So far, we have shown that attachment insecurities make it difficult for

©

people to confine themselves to the constructive form of anger that Bowlby

ht

called the anger of hope. These insecurities make it more likely that people
ig

will experience anger of despair, which may provoke destructively aggressive

yr

behavior. In this section, we review studies of attachment insecurities and

op

destructive aggression at three levels of analysis: domestic violence in couples;

antisocial behavior that affects communities, such as delinquency and crim-
C

inality; and intergroup (e.g., inter-ethnic or international) violence (for a more

246 MIKULINCER AND SHAVER

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 247

detailed review of the attachment-aggression link at the intergroup level,

see Chapter 14).

Domestic Violence

A)
From an attachment perspective, domestic violence is an exaggerated

P
(A
form of protesting a partner’s hurtful behavior (see Chapters 2, 14, and 20).
It is meant to discourage or prevent a partner from violating or breaking off

n
the relationship (e.g., Bartholomew & Allison, 2006). We expect this kind of

io
behavior to be more common among insecurely attached individuals, especially

at
the anxious ones, who are especially vulnerable to hurt feelings and threats

ci
of abandonment (Mikulincer & Shaver, 2007a). Avoidant individuals, in

so
contrast, might be expected to withdraw from conflict rather than escalate it,

As
because they try to dismiss hurt feelings and avoid expressing vulnerability or
need. Nevertheless, Bartholomew and Allison (2006) found that avoidant

al
people sometimes became violent in the midst of escalating domestic con-

ic
flicts, especially if their partner was anxiously attached and demanded their

og
involvement.
ol
Anxious attachment has been associated with domestic violence in two
ch
kinds of studies (reviewed by Mikulincer & Shaver, 2007a). First, anxious
y

attachment has been found to correlate with measures of domestic violence

Ps

across different samples of couples, whether married or cohabiting, and the

correlation cannot be explained by other relational or personality factors.
an

Second, abusive men who score relatively high on attachment anxiety report
ic

more severe and more frequent acts of coercion and abuse during couple
er

conflicts.
Am

Research also indicates that victims of domestic violence suffer from

attachment insecurities, with most studies finding elevations in attachment
e

anxiety and some finding elevations in avoidance as well (Mikulincer & Shaver,
th

2007a). Because of the cross-sectional nature of these studies, however, the

by

findings may indicate either that attachment insecurity puts people at risk of
being abused or that abuse increases attachment insecurity, or both. Also,
11

because violence in relationships is often reciprocal, many victims are also

20

perpetrators (Bartholomew & Allison, 2006), which creates a strong correlation

between perpetration and victimization (for a more detailed discussion, see
©

Chapter 20, this volume).

ht
ig

Antisocial Behavior
yr
op

Adult attachment researchers have found links between attachment

insecurities and antisocial behavior, such as delinquency and criminality
C

(reviewed by Mikulincer & Shaver, 2007a). Although both anxious and

ATTACHMENT, ANGER, AND AGGRESSION 247

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 248

avoidant individuals are more likely than their secure counterparts to engage
in antisocial behavior, they do so for different reasons. Anxiously attached
people sometimes engage in delinquent or criminal behavior as a way of
crying out for attention and care or of expressing anger and resentment

A)
(Allen, Moore, Kuperminc, & Bell, 1998). Avoidant individuals engage in
antisocial behavior to distance themselves from others (e.g., parents) or to

P
(A
demonstrate, by violating rules and laws, their lack of concern for others
(Allen et al., 1998).

n
Levinson and Fonagy (2004) compared the attachment patterns of

io
22 imprisoned delinquents, 22 patients with personality disorder without a

at
criminal history, and 22 healthy controls. They noted a higher prevalence

ci
of avoidant attachment in the delinquent group than in the other groups.

so
Moreover, delinquents who had committed violent offenses (e.g., murder,

As
malicious wounding) exhibited the same inability or unwillingness to talk
coherently about their emotional experiences that has been noted in other

al
studies of insecure attachment. In a related study, van IJzendoorn et al. (1997)

ic
interviewed 40 male criminals and found that 95% of them had insecure

og
attachment patterns.
ol
If antisocial behavior is associated with attachment insecurity, then
ch
interventions aimed at strengthening a person’s sense of attachment security
y

might help with rehabilitation. In fact, there is evidence from studies of

Ps

residential treatment programs for troubled and delinquent adolescents that

forming and maintaining a secure attachment relationship with a staff member
an

can reduce antisocial behavior (e.g., Born, Chevalier, & Humblet, 1997). In a
ic

yearlong study of adolescents residing in an Israeli treatment center, Gur (2006)

er

found that those who formed secure attachment bonds with staff members had
Am

lower rates of anger, depression, and behavioral problems and more positive
emotional experiences during the year. Adolescents who formed more secure
e

attachment bonds with staff members changed in the direction of security on

th

measures of attachment orientation and exhibited less aggressive behavior

by

toward peers and authorities.

11

Intergroup Aggression
20

The link between attachment insecurity and aggression is also evident

©

in the field of intergroup relations. In a series of five studies, Mikulincer and

ht

Shaver (2001) found that higher levels of self-reported attachment anxiety

ig

were associated with more hostile responses to a variety of out-groups (as defined
yr

by secular Israeli Jewish students): Israeli Arabs, Ultra-Orthodox Jews, Russian

op

immigrants, and homosexuals. In addition, we found that experimental

heightening of the sense of attachment security (by subliminally presenting
C

security-related words or asking study participants to visualize the faces of

248 MIKULINCER AND SHAVER

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 249

security-providing attachment figures) eliminated negative responses to out-

groups. These effects were found even when participants’ sense of personal
value was threatened or their ingroup had been insulted by an outgroup
member. That is, experimentally augmented attachment security reduced the

A)
sense of threat created by encounters with outgroup members and seemed to
eliminate defensive and hostile responses toward outgroup members.

P
(A
Building on these studies, Mikulincer and Shaver (2007b) found that
increasing people’s sense of attachment security reduced actual aggression

n
between contending or warring social groups. Specifically, Israeli Jewish under-

io
graduates participated in a study together with another Israeli Jew or an Israeli

at
Arab (in each case, the same confederate of the experimenter) and were sub-

ci
liminally and repeatedly exposed (for 20 ms on each trial) to the name of their

so
own security-enhancing attachment figure, the name of a familiar person who

As
was not viewed as an attachment figure, or the name of an acquaintance. Fol-
lowing the priming procedure, participants were informed that they would

al
evaluate a food sample and that they had been randomly selected to give the

ic
confederate hot sauce to evaluate. They also learned indirectly that the con-

og
federate strongly disliked spicy foods. (This procedure has been used in other
ol
studies of interpersonal aggression, e.g., McGregor et al., 1998.) The dependent
ch
variable was the amount of hot sauce allocated to the confederate.
y

When participants had been subliminally primed with the name of

Ps

someone who was not an attachment figure, they delivered a larger amount
of hot sauce to the Arab confederate than to the Jewish confederate, a sign
an

of intergroup aggression. But security priming eliminated this difference:

ic

Participants whose sense of security had been enhanced delivered equal

er

(relatively low) amounts of hot sauce to both the Arab and the Jewish con-
Am

federate. In addition, participants scoring higher on attachment anxiety gave

more hot sauce to the outgroup member (Israeli Arab) than to the ingroup
e

member (Israeli Jew). Thus, it seems that people who are either dispositionally
th

secure or induced to feel more secure in a particular setting are better able than
by

their insecure counterparts to tolerate intergroup differences and to refrain

from intergroup aggression.
11
20

ATTACHMENT AND POWER

©
ht

So far, we have considered the association between attachment insecu-

ig

rities and destructive forms of anger and aggression: domestic violence, anti-
yr

social behavior, and intergroup violence. As explained in Chapter 4 of this

op

volume, however, aggression is not exclusively maladaptive or destructive; it

is presumably an adaptive strategy that arose and was maintained in evolution
C

because it contributed to attaining the set goal of a hypothesized power/

ATTACHMENT, ANGER, AND AGGRESSION 249

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 250

dominance behavioral system (for similar analyses, see Chapters 3 and 12).
Following Sroufe and Waters’s (1977) analysis of “felt security” as the proximal
psychological goal of the attachment system, we propose that something we
call “felt power” is the goal of the power/dominance system. This “felt sense”

A)
is associated with feeling that one can control resources without undue
social interference. Only when this goal is blocked by social interference do

P
(A
aggressive behaviors become pervasive, extreme, destructive, or dysfunctional
(see Chapter 4, this volume). In this section, we focus on times when feelings of

n
power and dominance are elevated and aggression serves its adaptive function.

io
In particular, we explore whether and how attachment insecurities affect the

at
experience and exercise of power.

ci
According to the approach/inhibition theory of power (e.g., Galinsky,

so
Gruenfeld, & Magee, 2003; Keltner, Gruenfeld, & Anderson, 2003), experi-

As
encing a sense of power facilitates what Carver and White (1994) called an
approach orientation and Higgins (1998) called a promotion orientation, that is,

al
a motivated state involving reward-related thoughts, heightened attention to

ic
rewards, and positive emotions related to rewards. The sense of power also seems

og
to counteract what Carver and White (1994) called an inhibition orientation
ol
and Higgins (1998) called a prevention orientation, a state aimed at avoiding
ch
threats, heightened attention to threats, activation of threat-related thoughts,
y

and negative emotions. Galinsky et al. (2003) reasoned that a sense of power
Ps

strengthens an approach orientation because powerful people expect to have

greater access to rewards and less interference from others when pursuing
an

rewards. For the same reasons, powerlessness favors an inhibition orientation

ic

because powerless people expect to be subject to more social and material threats
er

and are aware of the constraints that these threats impose on one’s actions.
Am

The approach/inhibition theory of power has received strong empirical

support. For example, Keltner et al. (2003) reviewed evidence that elevated
e

power increases the experience and expression of positive emotions, sensitivity

th

and responsiveness to potential rewards, the automaticity of social cognitions,

by

and the likelihood of approach, disinhibition, and extraverted behavior.

In subsequent studies, experimental augmentation of the sense of power
11

(e.g., asking people to recall an episode in which they had power over others)
20

elevated risk-taking in negotiations; engagement in risky, unprotected sex;

and competitive behavior (e.g., Anderson & Galinsky, 2006; Magee, Galinsky,
©

& Gruenfeld, 2007).

ht

An increased sense of power also protects people from others’ influence

ig

and social pressure (Galinsky, Magee, Gruenfeld, Whitson, & Liljenquist,

yr

2008). At the same time, because powerful people feel they have control over
op

others and do not feel constrained by them, increasing people’s sense of power
may increase their psychological distance from others (Smith & Trope, 2006)
C

and their tendency to objectify others and perceive them as means for attaining

250 MIKULINCER AND SHAVER

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 251

personal goals (Gruenfeld, Inesi, Magee, & Galinsky, 2008). Galinsky, Magee,
Inesi, and Gruenfeld (2006) found that increasing people’s sense of power
reduced their inclination to adopt others’ perspectives and to empathize
with others.

A)
We suspect that these consequences of elevated power are moderated
by people’s attachment orientations because they are likely to depend on how

P
(A
one relates to others and on the goals, wishes, and fears that underlie one’s
social behavior. For example, people who hold negative views of others and

n
dislike closeness and intimacy (i.e., people with an avoidant attachment style)

io
are likely to construe power as a way to gain distance from others and freedom

at
from their influence. In contrast, people who hold positive views of others and

ci
are able to balance dependence and autonomy (i.e., relatively secure people)

so
are probably able to use power and influence to improve their interpersonal

As
relations, respond to others’ needs, and resolve interpersonal conflicts without
deferring too much to a partner’s needs. Moreover, people who fear rejection

al
and abandonment (i.e., anxiously attached people) may feel uneasy when

ic
granted power and be reluctant to act freely, take risks, and step outside the

og
boundaries of conformity. In subsequent sections, we report findings from two
studies that examine these ideas. ol
y ch

Attachment Anxiety Weakens the Link

Ps

Between Power and an Approach Orientation

an

Our first hypothesis is that increasing a person’s sense of power will

ic

strengthen his or her tendency to adopt an approach orientation (e.g., be

er

optimistic and take risks) as long as the person scores low on attachment anxiety
Am

but not if he or she scores high. People with an anxious attachment style are
likely to be ambivalent about power and dominance. On the one hand, they
e

want to have control over relationship partners (Shaver & Hazan, 1993); on
th

the other hand, they may be reluctant to assert themselves (e.g., Bartholomew
by

& Horowitz, 1991), because this could provoke a partner’s resentment, create
conflict, and threaten relationship stability.
11

To test this hypothesis, in Study 1 we examined the effects of priming a

20

sense of power (using Galinsky et al.’s, 2003, technique) on optimism and risk,
taking into account individual differences in attachment anxiety and avoidance.
©

Eighty Israeli undergraduate students (60 women and 20 men) completed in

ht

a class the Experiences in Close Relationships (ECR) measure of attachment

ig

anxiety and avoidance (Brennan et al., 1998), which had alpha reliability
yr

coefficients of .90 and .83 in this sample. Weeks later they came to a labora-
op

tory individually and were randomly assigned to a power priming or a control

condition. In the power priming condition (n = 40), participants were asked
C

to recall a particular incident in which they had power over one or more other

ATTACHMENT, ANGER, AND AGGRESSION 251

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 252

people and to write about what happened, how they felt, and what they did
during and after the episode (for instructions, see Galinsky et al., 2003). In
the control condition (n = 40), participants were asked to recall a particular
TV program they had watched the previous week and to write about it.

A)
All of the participants then completed Weinstein’s (1980) 15-item
optimism scale (with an alpha of .85) and Tversky and Kahneman’s (1981)

P
(A
measure of risk preference. They were told the following:

n
Imagine that you work for a large hi-tech company that is experiencing

io
serious economic troubles and needs to lay off 6000 employees. Plan A

at
will save 2,000 jobs, whereas plan B has a one third probability of saving

ci
all 6,000 jobs but a two thirds probability of saving no jobs.

so
Participants were asked to indicate the extent to which they would favor one

As
option over the other using a scale ranging from 1 (very much prefer program A)
to 7 (very much prefer program B). Higher scores reflected greater preference

al
for the riskier program (program B).

ic
Hierarchical regression analyses examining the unique and interactive

og
effects on optimism and risk preference of being primed with memories of power
ol
(a dummy variable contrasting the power condition with the control condition),
ch
avoidant attachment, and attachment anxiety revealed significant main
y

effects of power priming and attachment anxiety. Replicating Galinsky et al.’s

Ps

(2006) findings, participants in the power condition were more optimistic,

β = .27, p < .05, and more likely to prefer the riskier plan than participants in
an

the control condition, β = .29, p < .01. In addition, more anxiously attached par-
ic

ticipants were less optimistic and less likely to prefer the riskier plan, βs of −.33
er

and −.30, ps < .01. These main effects were qualified by significant interactions
Am

between power priming and anxious attachment (β = −.41, p < .01, for optimism;
β = −.29, p < .01, for risk preference). Simple slope tests revealed that power
e

priming led to greater optimism and risk preference than the control condi-
th

tion mainly when participants scored relatively low on attachment anxiety,

βs of .68 and .58, ps < .01. These effects of power priming were negative or nil
by

and not significant when attachment anxiety was relatively high, βs of −.14
11

and .00. In sum, as expected, attachment anxiety disrupted the previously

20

documented association between power and an approach orientation.

©

Power Increases Objectification of Others

ht

in the Context of Avoidant Attachment

ig
yr

Our second hypothesis is that increasing a person’s sense of power will

op

increase the objectification of others among people scoring high on avoidant

attachment but not among less avoidant people. Avoidant people hold gen-
C

erally negative, uncompassionate views of others and try to remain emotionally

252 MIKULINCER AND SHAVER

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 253

distant and detached from them (see the review by Mikulincer & Shaver, 2007a).
It seems likely that they will use power as an opportunity to act on their pref-
erences for autonomy and distance, their critical view of others, and their
perception of others as objects to be used instrumentally for personal need

A)
satisfaction. In contrast, more secure people, who hold positive, empathic
views of others and are guided by a desire to form mutually satisfying and har-

P
(A
monious relationships, should be less likely to see others as objects even when
they are granted a degree of power over them.

n
To test these ideas we conducted a second study (Study 2) with a new

io
group of 60 Israeli undergraduate students (41 women and 19 men) who

at
completed the ECR scale during a class period (with resulting alphas of .86

ci
for anxiety and .88 for avoidance). Weeks later they came to a laboratory

so
individually and were randomly assigned to a power priming or a control con-

As
dition as in Study 1. All of them were then asked to think about their relation-
ships with three other students in their classes. For each of these students,

al
participants completed Gruenfeld et al.’s (2008) 10-item objectification scale

ic
(e.g., “I tend to contact this person only when I need something from him/her,”

og
“I try to get him/her to do things that will help me succeed”). Item ratings
ol
were made on a 7-point scale ranging from 1 (not at all) to 7 (very much). For
ch
each participant and each target student, we computed an objectification
y

score by averaging the 10 item scores (with alphas ranging from .88 to .91).
Ps

Because the correlations between the three different student scores were high
(rs > .64), we computed a total objectification score by averaging them.
an

Hierarchical regression analyses examining the unique and interactive

ic

effects of power priming, avoidant attachment, and attachment anxiety on

er

objectification revealed a significant main effect of avoidant attachment, β = .34,

Am

p < .01. This effect was qualified, as expected, by a significant interaction

between power priming and avoidant attachment, β = .29, p < .05. Simple
e

slope tests revealed that power priming led to greater objectification of others
th

mainly when participants scored relatively high on avoidance, β = .43, p < .01.
by

The effect was in the opposite direction but not significant when avoidance
was low, β = −.15. Thus, as hypothesized, when people were primed with a sense
11

of power, avoidant participants tended to objectify others more than less

20

avoidant participants, suggesting that secure people might be able to maintain

a sense of power without treating others as objects.
©
ht
ig

CONCLUDING REMARKS
yr
op

As explained in Chapter 4 of this volume, Bowlby (1982) did not say

much about anger, aggression, or power/dominance, partly because he wanted
C

to part company with Freud’s emphasis on sexual and aggressive “instincts”

ATTACHMENT, ANGER, AND AGGRESSION 253

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 254

or “drives” and partly because he was concerned primarily with infant–parent

relationships. He viewed anger in the context of infants’ reactions to separation
or abandonment. Some of his ideas about anger (i.e., anger of hope or anger
of despair) can be extended to the realm of adult close relationships, and we

A)
have summarized some of the literature on that topic here. But infants are not
prepared to occupy powerful or dominant roles in society, so Bowlby said

P
(A
virtually nothing about power and dominance in his books.
Nevertheless, attachment theory has been fruitfully expanded in several

n
directions, including the conceptualization of leadership and group function-

io
ing in terms of attachment dynamics (Mikulincer & Shaver, 2007a). And

at
Bowlby’s behavioral-system construct has proven useful in the study of sexu-

ci
ality, empathy, and altruism from an attachment perspective (Mikulincer &

so
Shaver, 2007a), partly because Bowlby explicitly referred to the existence of

As
sexual and caregiving behavioral systems. Because we now want to extend the
theory into the untouched domain of power and dominance, we have begun

al
to conceptualize the existence of a power/dominance system. If such a system

ic
exists, we expect its operations to be colored by attachment security and the

og
major forms of attachment insecurity. In this chapter, we have unveiled our
ol
first efforts to explore links between attachment and power. We think the
ch
results are interesting and well worth pursuing further.
y
Ps

REFERENCES
an
ic

Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment:
er

Assessed in the strange situation and at home. Hillsdale, NJ: Erlbaum.

Am

Allen, J. P., Moore, C., Kuperminc, G. P., & Bell, K. (1998). Attachment and ado-
lescent psychosocial functioning. Child Development, 69, 1406–1419. doi:10.2307/
e

1132274
th

Anderson, C., & Galinsky, A. D. (2006). Power, optimism, and risk-taking. European
by

Journal of Social Psychology, 36, 511–536. doi:10.1002/ejsp.324

Bartholomew, K., & Allison, C. J. (2006). An attachment perspective on abusive
11

dynamics in intimate relationships. In M. Mikulincer & G. S. Goodman (Eds.),

20

Dynamics of romantic love (pp. 102–127). New York, NY: Guilford Press.
Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults:
©

A test of a four-category model. Journal of Personality and Social Psychology, 61,

ht

226–244. doi:10.1037/0022-3514.61.2.226
ig

Born, M., Chevalier, V., & Humblet, I. (1997). Resilience, desistance, and delinquent
yr

career of adolescent offenders. Journal of Adolescence, 20, 679–694. doi:10.1006/

jado.1997.0119
op

Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. New
C

York, NY: Basic Books.

254 MIKULINCER AND SHAVER

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 255

Bowlby, J. (1980). Attachment and loss: Vol. 3. Sadness and depression. New York, NY:
Basic Books.
Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). New York, NY:
Basic Books.

A)
Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London,

P
England: Routledge.

(A
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult
romantic attachment: An integrative overview. In J. A. Simpson & W. S. Rholes

n
io
(Eds.), Attachment theory and close relationships (pp. 46–76). New York, NY:

at
Guilford Press.

ci
Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation,

so
and affective responses to impending reward and punishment: The BIS/BAS
scales. Journal of Personality and Social Psychology, 67, 319–333. doi:10.1037/

As
0022-3514.67.2.319

al
Cassidy, J., & Kobak, R. R. (1988). Avoidance and its relationship with other defensive

ic
processes. In J. Belsky & T. Nezworski (Eds.), Clinical implications of attachment

og
(pp. 300–323). Hillsdale, NJ: Erlbaum.
ol
Cassidy, J., & Shaver, P. R. (Eds.). (2008). Handbook of attachment: Theory, research, and
ch
clinical applications (2nd ed.). New York, NY: Guilford Press.
Diamond, L. M., & Hicks, A. M. (2005). Attachment style, current relationship secu-
y
Ps

rity, and negative emotions: The mediating role of physiological regulation. Journal
of Social and Personal Relationships, 22, 499–518. doi:10.1177/0265407505054520
an

Fraley, R. C., & Shaver, P. R. (2000). Adult romantic attachment: Theoretical devel-
ic

opments, emerging controversies, and unanswered questions. Review of General

er

Psychology, 4, 132–154. doi:10.1037/1089-2680.4.2.132

Am

Fredrickson, B. L. (2001). The role of positive emotions in positive psychology:

The broaden-and-build theory of positive emotions. American Psychologist, 56,
218–226. doi:10.1037/0003-066X.56.3.218
e
th

Galinsky, A. D., Gruenfeld, D. H., & Magee, J. C. (2003). From power to action.
by

Journal of Personality and Social Psychology, 85, 453–466. doi:10.1037/0022-3514.

85.3.453
11

Galinsky, A. D., Magee, J. C., Gruenfeld, D. H., Whitson, J. A., & Liljenquist, K. A.
20

(2008). Power reduces the press of the situation: Implications for creativity,
conformity, and dissonance. Journal of Personality and Social Psychology, 95,
©

1450–1466. doi:10.1037/a0012633
ht

Galinsky, A. D., Magee, J. C., Inesi, M. E., & Gruenfeld, D. H. (2006). Power and
ig

perspective not taken. Psychological Science, 17, 1068–1074. doi:10.1111/j.1467-

yr

9280.2006.01824.x
op

Gruenfeld, D. H., Inesi, M. E., Magee, J. C., & Galinsky, A. D. (2008). Power and the
objectification of social targets. Journal of Personality and Social Psychology, 95,
C

111–127. doi:10.1037/0022-3514.95.1.111

ATTACHMENT, ANGER, AND AGGRESSION 255

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 256

Gur, O. (2006). Changes in adjustment and attachment-related representations among

high-risk adolescents during residential treatment: The transformational impact of the
functioning of caregiving figures as a secure base (Unpublished doctoral dissertation).
Bar-Ilan University, Ramat Gan, Israel.

A)
Hazan, C., & Shaver, P. R. (1987). Romantic love conceptualized as an attachment
process. Journal of Personality and Social Psychology, 52, 511–524. doi:10.1037/

P
0022-3514.52.3.511

(A
Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motivational

n
principle. In M. P. Zanna (Ed.), Advances in experimental social psychology

io
(Vol. 30, pp. 1–46). New York, NY: Academic Press.

at
Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and inhi-

ci
bition. Psychological Review, 110, 265–284. doi:10.1037/0033-295X.110.2.265

so
Levinson, A., & Fonagy, P. (2004). Offending and attachment: The relationship

As
between interpersonal awareness and offending in a prison population with
psychiatric disorder. Canadian Journal of Psychoanalysis, 12, 225–251.

al
Magee, J. C., Galinsky, A. D., & Gruenfeld, D. H. (2007). Power, propensity to

ic
negotiate, and moving first in competitive interactions. Personality and Social

og
Psychology Bulletin, 33, 200–212. doi:10.1177/0146167206294413
ol
McGregor, H., Lieberman, J., Greenberg, J., Solomon, S., Arndt, J., Simon, L., &
ch
Pyszczynski, T. (1998). Terror management and aggression: Evidence that
y

mortality salience promotes aggression against worldview-threatening individuals.

Ps

Journal of Personality and Social Psychology, 74, 590–605. doi:10.1037/0022-3514.

74.3.590
an

Mikulincer, M. (1998). Adult attachment style and individual differences in functional

ic

versus dysfunctional experiences of anger. Journal of Personality and Social Psy-

er

chology, 74, 513–524. doi:10.1037/0022-3514.74.2.513

Am

Mikulincer, M., & Shaver, P. R. (2001). Attachment theory and intergroup bias:
Evidence that priming the secure base schema attenuates negative reactions to
e

out-groups. Journal of Personality and Social Psychology, 81, 97–115. doi:10.1037/

th

0022-3514.81.1.97
by

Mikulincer, M., & Shaver, P. R. (2003). The attachment behavioral system in adult-
hood: Activation, psychodynamics, and interpersonal processes. In M. P. Zanna
11

(Ed.), Advances in experimental social psychology (Vol. 35, pp. 53–152). New York,
NY: Academic Press.
20

Mikulincer, M., & Shaver, P. R. (2007a). Attachment patterns in adulthood: Structure,

©

dynamics, and change. New York, NY: Guilford Press.

ht

Mikulincer, M., & Shaver, P. R. (2007b). Boosting attachment security to promote

ig

mental health, prosocial values, and intergroup tolerance. Psychological Inquiry,

yr

18, 139–156.
op

Rholes, W. S., Simpson, J. A., & Orina, M. (1999). Attachment and anger in an
anxiety-provoking situation. Journal of Personality and Social Psychology, 76,
C

940–957. doi:10.1037/0022-3514.76.6.940

256 MIKULINCER AND SHAVER

12051-14_CH13-rev2.qxd 5/25/10 11:55 AM Page 257

Shaver, P. R., & Hazan, C. (1993). Adult romantic attachment: Theory and evidence.
In D. Perlman & W. Jones (Eds.), Advances in personal relationships (Vol. 4,
pp. 29–70). London, England: Jessica Kingsley.
Shaver, P. R., & Mikulincer, M. (2002). Attachment-related psychodynamics. Attach-

A)
ment & Human Development, 4, 133–161. doi:10.1080/14616730210154171
Simpson, J. A., Rholes, W. S., & Phillips, D. (1996). Conflict in close relationships:

P
(A
An attachment perspective. Journal of Personality and Social Psychology, 71,
899–914. doi:10.1037/0022-3514.71.5.899

n
Smith, P. K., & Trope, Y. (2006). You focus on the forest when you’re in charge of

io
the trees: Power priming and abstract information processing. Journal of Person-

at
ality and Social Psychology, 90, 578–596. doi:10.1037/0022-3514.90.4.578

ci
Sroufe, L. A., & Waters, E. (1977). Attachment as an organizational construct. Child

so
Development, 48, 1184–1199.

As
Tversky, A., & Kahneman, D. (1981, January). The framing of decisions and the psy-
chology of choice. Science, 211, 453–458. doi:10.1126/science.7455683

al
van IJzendoorn, M. H., Feldbrugge, J., Derks, F. C. H., de Ruiter, C., Verhagen M. F.,

ic
og
Philipse, M. W., . . . Riksen-Walraven, J. M. (1997). Attachment representations
of personality-disordered criminal offenders. American Journal of Orthopsychiatry,
67, 449–459. doi:10.1037/h0080246 ol
ch
Weinstein, N. D. (1980). Unrealistic optimism about future life events. Journal of
y

Personality and Social Psychology, 39, 806–820. doi:10.1037/0022-3514.39.5.806

Ps

Zimmermann, P., Maier, M. A., Winter, M., & Grossmann, K. E. (2001). Attachment
and adolescents’ emotion regulation during a joint problem-solving task with a
an

friend. International Journal of Behavioral Development, 25, 331–343.

ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

ATTACHMENT, ANGER, AND AGGRESSION 257

C
op
yr
ig
ht
©
12051-14_CH13-rev2.qxd

20
11
5/25/10

by
th
e
11:55 AM

Am
er
ic
Page 258

an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 259

14

A)
P
ATTACHMENT AND VIOLENCE:

(A
AN ANGER BORN OF FEAR

n
io
at
ci
DONALD G. DUTTON

so
As
al
ic
Angry coercive behavior, acting in the service of an affectional bond, is

og
not uncommon. It is seen when a mother, whose child has foolishly run
across the road, berates and punishes him with an anger born of fear.
ol — Bowlby, 1973, p. 287
y ch
Ps
an

In this chapter I review the literature on attachment and violence.

ic

I begin by noting that early disruptions of attachment activate the attach-

er

ment behavioral system, a systemic alarm reaction assuaged only by contact

Am

with an attachment figure (Ainsworth, Blehar, Waters, & Wall, 1978;

Bowlby, 1982). Later in life, failures to reduce attachment-related alarm or
e

to find a “safe haven” in times of external threat can produce a spiraling

th

arousal reaction that lowers impulse control and increases the likelihood of
by

violence. I review studies connecting certain types of personality organiza-

tion to intimate partner violence, violence toward the self, and violence in
11

military contexts. These forms of personality organization can be viewed as

20

extreme forms of attachment insecurity persisting into adulthood and mak-

ing an individual susceptible to high aversive arousal and hence to arousal-
©

based violence. Finally, I discuss symbolic attachment as a major incentive

ht

for violence when a culture promises secure attachment in perpetuity if one

ig

martyrs oneself for a “holy” cause.

yr

Bowlby (1982) viewed interpersonal anger as arising from frustrated

op

attachment needs and serving as a form of protest behavior directed at regain-

C

ing or maintaining contact with an attachment figure. In adulthood, such

259
12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 260

protest is often directed at a romantic partner and can take the form of
verbal abuse, coercive control of a partner’s behavior, and even violence
(see Chapters 13 and 20, this volume). Such abusive behavior is most likely
to be precipitated by real or imagined threats of rejection, separation, or

A)
abandonment by the partner. Hence, as Bowlby (1988) stated, “violence . . .
can be understood as the distorted and exaggerated version of behavior that

P
(A
is potentially functional” (p. 12).
This perspective is consistent with a large body of literature indicat-

n
ing that abusive men tend to be insecure and overly dependent on their

io
partners and that jealousy and fears of separation are common triggers of

at
abusive episodes (Dutton, 2006). In adults, tests of the association between

ci
attachment insecurities and violence have indicated that attachment pat-

so
terns that include considerable anxiety (often labeled preoccupied and

As
fearful; e.g., Bartholomew & Horowitz, 1991) incline a person to become
highly emotionally aroused in response to possible separation or rejection

al
(Bartholomew & Allison, 2006; Bartholomew, Henderson, & Dutton, 2001).

ic
og
It follows that individuals with these kinds of insecure attachment may
exhibit higher levels of violence toward their close relationship partners,
ol
either as a way of controlling the threat of separation from a partner or as
ch
an impulsive “acting out” of inner tensions resulting from anticipating and
y

dreading separation.
Ps
an
ic

“PRIMITIVE” ATTACHMENT THREAT: STRESS REACTIONS

er
Am

According to Bowlby (1973, 1982), fear-based activation of the

attachment system is similar to an alarm reaction in animals. It produces
hormonal and neurobiological hyperactivity, including the release of cortisol
e
th

and endogenous opioids (similar to low-dose morphine). Frequent experi-

ences of this kind can affect neural development (Perry, 1997; Schore, 1994,
by

2003a, 2003b; D. J. Siegel, 2001a). There is clear evidence from several

11

independent researchers that early infancy (a period called “primitive” by

psychoanalytic researchers because it is “preoedipal” in Freud’s psychoanalytic
20

theory) is a critical or sensitive period for neural development (Perry, 1997;

©

Pynoos & Eth, 1985; Schore, 1994; D. J. Siegel, 2001b). Neural structures
such as the hypothalamus and the prefrontal and orbitofrontal cortex, which
ht

regulate emotion, develop rapidly at this time (Schore, 1994, 2003a). As

ig

D. J. Siegel (2001b) states: “When certain suboptimal attachment experiences

yr

occur, the mind of the child may not come to function as a well-integrated
op

system” (p. 82).

C

260 DONALD G. DUTTON

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 261

Klein and Mahler’s (e.g., Klein & Riviere, 1937; Mahler, 1971; Mahler,
Pine, & Bergman, 1975) psychoanalytic emphasis on the phase of infant
development in which the child first physically separates from the mother and
initiates a separate sense of identity is supported by current research on

A)
neurological development (Dutton, 2006). Thus, rage is exhibited first in the
context of an intimate relationship, usually with mother (Klein & Riviere, 1937),

P
(A
in a phase of life prior to language development and before autobiographical
memory (Schore, 2003a). Such rage is more common in insecure relationships,

n
and the neural and psychological effects of repeatedly experiencing it in early

io
childhood, incoherently and unmediated by language, may make it more

at
likely to erupt later in life, where it seems especially irrational when it leads

ci
to violence against an intimate adult attachment figure.

so
As
al
THE NEUROBIOLOGY OF ATTACHMENT

ic
og
In an extremely detailed analysis, Schore (1994, 2003a, 2003b) showed
how the capacity for attachment matures into a homeostatic self-regulatory
ol
system near the end of the first year of life. This occurs because dyadic com-
ch
munication with a parental figure generates intense positive affective states
y

in the infant and high levels of dopamine and endogenous opiates. These
Ps

neurotransmitters, in turn, promote growth in the prefrontal cortex, especially

the orbitofrontal cortex, which is critically involved in the attachment process.
an

Attachment therefore has a neuropsychological aspect that is itself based on

ic

dyadic interaction with caregivers during the first year of life (before extensive
er

development of the left hemispheric structures involved in language). Thus,

Am

the problems that abusive individuals have with attachment and emotion
regulation are often grounded in early relationships that are not recalled later
e

in words. This suggests that the most important and long-lasting effects of
th

early family dysfunction may be due not simply to imitation of specific abusive
by

acts but rather to the inability to regulate painful and destructive emotions,
such as panic and rage.
11

D. J. Siegel (2001a) explains that one form of insecure attachment, labeled

20

disorganized/disoriented (Lyons-Ruth & Jacobwitz, 2008; Main & Solomon,

1986, 1990), has been associated with marked impairments in the emotional,
©

social, and cognitive domains, and predisposes a person to a clinical condi-

ht

tion known as dissociation in which the capacity to function in an organized,

ig

coherent manner is impaired (Carlson, 1998). Disorganized attachment in

yr

infancy includes a diverse array of fearful, odd, disorganized, or conflicted

op

behaviors exhibited in the Strange Situation assessment procedure (Ainsworth

et al., 1978). Main (1995) argued that this kind of behavior stems from parents
C

ATTACHMENT AND VIOLENCE 261

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 262

with unresolved losses and traumas who are themselves susceptible to dis-
sociative states. These parents generate in their young children “fright without
solution.”
Recent studies have shown that youths with a history of disorganized

A)
attachment are at great risk of expressing hostility with peers and have the
potential for extreme interpersonal violence (Lyons-Ruth, Bronfman, &

P
(A
Atwood, 1999; Lyons-Ruth & Jacobwitz, 2008). This implies that lack of
resolution of trauma in a parent, leading to what Lyons-Ruth and her colleagues

n
call hostile or helpless behavior, can cause parents to create frightening,

io
paradoxical, and unsolvable problematic situations for their children. The

at
adult attachment figure, who, according to Bowlby (1982) is equipped with a

ci
caregiving behavioral system that evolved because it provided a safe haven

so
and secure base for children who were then more likely to live and reproduce,

As
paradoxically becomes also a source of alarm, leaving a dependent child with
no clear strategy for getting attachment needs met.

al
ic
og
EMPIRICAL STUDIES OF ATTACHMENT STYLE, ANGER,
AND INTIMATE PARTNER VIOLENCE ol
y ch

Adult attachment patterns or styles can be conceptualized in terms of

Ps

two continuous dimensions, attachment anxiety and avoidance, as explained

elsewhere in this volume (Chapter 13). Mikulincer (1998) examined a college
an

sample and found that experiences and expressions of anger were related to
ic

attachment insecurities. Participants who scored relatively high on avoidance

er

attempted to deny their anger but nevertheless revealed it in the form of

Am

smoldering hostility and physiological arousal, of which the avoidant indi-

viduals seemed to be unaware. The more anxious participants, in contrast,
e

exhibited intense anger and a lack of anger control.

th

Mikulincer’s findings motivated me to reexamine anger data that my

by

associates and I had collected from abusive men using the subscales of the
Siegel Multidimensional Anger Inventory (MAI; J. M. Siegel, 1986): anger-in,
11

anger-out, and frequency, duration, and magnitude of anger and hostility. In our
20

study, fearful attachment (a combination of attachment anxiety and avoidance,

conceptually similar to infant disorganized attachment) correlated consistently
©

higher than other adult attachment patterns with every subscale on the MAI.
ht

In my work (e.g., Dutton, 2006), I have described what I call “the abusive per-
ig

sonality in males,” which includes several characteristics that are significantly

yr

correlated with partners’ reports of abusiveness. This personality pattern includes

op

what I call a “fearful–angry” attachment style, and I hypothesized that some

aspect of male socialization helped shape the fear to anger. Intimate abuse,
C

I argue, is literally “an anger born of fear” (Bowlby 1973, p. 287).

262 DONALD G. DUTTON

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 263

The relationship of attachment to anger and violence has been studied

in a sample selected for problems in intimate relationships, a court mandated
group of men convicted of wife assault (N = 160). In this group, my colleagues
and I (Dutton, Starzomski, Saunders, & Bartholomew, 1994) examined

A)
associations between attachment style (measured with the Relationship
Styles Questionnaire [RSQ]; Griffin & Bartholomew, 1994) and emotional,

P
(A
cognitive, and behavioral reactions to intimacy. Specifically, the men completed
a battery of questionnaires including the RSQ and measures of borderline per-

n
sonality traits (Oldham et al., 1985), trauma symptoms (Briere & Runtz, 1989),

io
trait anger (J. M. Siegel, 1986), recalled parental treatment (Perris, Jacobsson,

at
Lindstrom, von Knorring, & Perris, 1980), verbal and physical abusiveness

ci
(Straus, 1992; Tolman, 1989), and jealousy (Mathes & Severa, 1981). The

so
men’s spouses also completed the latter two scales to describe their husbands.

As
We found that the more anxious attachment styles, fearful and preoccupied,
were significantly correlated with most features of intimate abusiveness and

al
with spouses’ reports of abusiveness. The secure attachment style, even in this

ic
self-selected, court-mandated population, was negatively correlated with

og
these features of abusiveness.
ol
A recent study of dating violence in a college sample examined associations
ch
between attachment insecurities and these same features of abusiveness in
y

women (Clift, 2008). A comparison of their findings with those obtained by

Ps

Dutton et al. (1994) indicated that there is remarkable similarity across gen-
ders in the associations between forms of attachment insecurity and certain
an

features of abusiveness.
ic

Henderson and her colleagues (Henderson, Bartholomew, Trinke,

er

& Kwong, 2005) also found attachment anxiety (in this case, preoccupied
Am

attachment) to be associated with women’s violence toward their male partners.

In fact, preoccupied attachment was associated with both perpetration and
e

receipt of violence, regardless of gender. Bartholomew, Oram, and Landolt

th

(2008) also found attachment anxiety to be associated with intimate partner

by

violence in gay male relationships. In addition, Follingstad, Bradley, Helff,

and Laughlin (2002) found that attachment anxiety and angry temperament
11

were related to controlling behavior and physical violence in both partners

20

in heterosexual dating relationships.

Mauricio, Tein, and Lopez (2007) examined 192 men who had been
©

court mandated to participate in a batterer intervention program. They

ht

completed measures of anxious and avoidant attachment (the Experiences in

ig

Close Relationships questionnaire; Brennan, Clark, & Shaver, 1998), border-

yr

line and antisocial personality disorders, psychological and physical violence,

op

and social desirability. The authors used structural equation modeling to test
hypotheses that associations between anxious attachment and physical and
C

psychological violence were mediated by borderline personality disorder and

ATTACHMENT AND VIOLENCE 263

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 264

that associations between avoidant attachment and physical and psychological

violence were mediated by antisocial personality disorder. Social desirability was
included in the models as a covariate. The results indicated that personality
disorders fully mediated the associations between avoidant attachment and

A)
physical and psychological abuse, and personality disorders fully mediated the
association between anxious attachment and psychological abuse but only par-

P
(A
tially mediated the link between anxious attachment and psychological abuse.
These studies suggest that attachment insecurity, especially anxious

n
attachment, may be a risk factor for violence, but for it to result in actual

io
violence it has to “crystallize” into something more: a chronically angry

at
temperament or a disturbance of the self that includes angry acting out.

ci
so
As
THE ABUSIVE PERSONALITY:
ATTACHMENT AND BORDERLINE PERSONALITY

al
ic
My early work on the “abusive personality” was completed before the

og
studies reviewed in the previous section were conducted, but it foreshadowed
ol
the importance of both attachment insecurities and personality dysfunctions
ch
(see also Westen & Shedler, 1999). At the time, I focused on borderline per-
y

sonality symptoms because of clinical descriptions of borderlines as exhibit-

Ps

ing intermittent rage in intimate relationships (Gunderson, 1984). One of

the main correlates of borderline functioning is attachment anxiety. The
an

Diagnostic and Statistical Manual of Mental Disorders (4th ed., American Psy-
ic

chiatric Association, 1994) describes “frantic efforts to avoid real or imagined

er

abandonment” (p. 710). However, at the time of our early research no one
Am

had connected borderline features or attachment insecurities with each other

or with abusiveness.
e

Using a research strategy already mentioned (i.e., studying male per-

th

petrators in court mandated groups and gathering data from them and their
by

female partners), we (Dutton et al., 1994) found that fearful attachment

(a combination of anxiety, anger, and discomfort with reliance on one’s
11

partner) was correlated significantly with wives’ reports of abusiveness and

20

perpetrators’ self-reports of jealousy, anger, and trauma symptoms. Men who

were abusive in intimate relationships exhibited both attachment insecurities
©

and features of borderline personality. Subsequently, samples from several

ht

populations (i.e., court-mandated men in treatment programs, colleges

ig

students, working-class control men, clinical outpatients, gay men from urban
yr

communities) have also turned up strong and significant correlations (in the
op

.45 to .65 range) between fearful attachment and borderline personality

organization (as measured with the Borderline Personality Organization Scale;
C

Oldham et al., 1985).

264 DONALD G. DUTTON

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 265

A central clinical feature of borderline individuals is their impulsivity.

In examining an impulsive group of spouse abusers (compared with those who
used violence instrumentally), we found that scores on fearful attachment
were significantly higher in the impulsive group (Tweed & Dutton, 1998), a

A)
finding replicated by Edwards, Scott, Yarvis, Paizis, and Panizzon (2003) using
an impulsive aggression questionnaire that assessed frequency of losing control,

P
(A
being suicidal, and displaying an extreme temper. Subsequent research with large
samples of men (e.g., Henderson, Bartholomew, & Dutton, 1997; Mauricio

n
et al., 2007) has largely confirmed the associations between attachment

io
insecurity (especially of anxious varieties) and borderline traits, impulsivity,

at
and abusiveness. The same pattern appears in abusive women as well (e.g., Clift

ci
2008; Henning, Jones, & Holford, 2003).

so
We consider severely insecure early attachment to be a likely key to

As
this personality constellation. The “modal family” for our court-mandated
samples was notable for its abusiveness and lack of a safe haven and secure base

al
for the children. Future research should explore the development of abusive-

ic
ness in more detail, and further integrative theorizing should be attempted to

og
bring into a single picture what we have learned and are still learning about early
ol
attachment experiences, neural development, impulsivity and poor emotion
ch
regulation, and clinically significant features of borderline personality and
y

other personality disorders. Some of the pieces of this important story can be
Ps

gleaned from other chapters in this volume (e.g., Chapters 1, 2, 7, and 9)

and from chapters in the most recent edition of the Handbook of Attachment
an

(Cassidy & Shaver, 2008), which includes many new findings regarding
ic

the neural underpinnings of attachment phenomena, the relation between

er

genes and experiences in attachment relationships, and connections between

Am

attachment patterns and clinical disorders.

e
th

DYADIC PROCESSES CONNECTING ATTACHMENT

by

AND VIOLENCE
11

Work by Bartholomew and her colleagues suggests that a dyadic or rela-

20

tionship perspective is important for understanding attachment dynamics.

When pursuing this research strategy, it is important to assess attachment
©

styles of both partners in a relationship. For example, Allison, Bartholomew,

ht

Mayseless, and Dutton (2008) used thematic interviews to assess attachment

ig

dynamics in heterosexual couples identified as involving male partner violence.

yr

The interviews revealed two attachment-related strategies, pursuit and distanc-

op

ing, which sometimes led to abuse. As a pursuit strategy, violence forced one
partner to focus on the other. As a distancing strategy, violence served to push
C

a partner away. Bartholomew et al. (2008) suggested that optimal social distance

ATTACHMENT AND VIOLENCE 265

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 266

rather than physical proximity and contact (the goals of the attachment system
highlighted by Ainsworth et al., 1978) might be considered the state in adult
relationships that lowers or terminates attachment-system activation.
Referring to the frequent reciprocity of negative affect in relationships,

A)
Bartholomew and Allison (2006) analyzed the dyadic nature of interpersonal
violence. Surveys in the United States indicate that bilateral violence is the

P
(A
most common form in intimate relationships (e.g., Stets & Straus, 1989;
Whitaker, Haileyesus, Swahn, & Saltzman, 2007), and in laboratory studies

n
of marital interactions, reciprocal negative affect is a consistent predictor of

io
relationship dysfunction and deterioration (e.g., Cordova, Jacobson, Gottman,

at
Rushe, & Cox, 1993; Leonard & Roberts, 1998; Margolin, John, & Gleberman,

ci
1988). Future research should therefore consider both partners’ attachment

so
insecurities and their connections with both personality disorder symptoms

As
and destructive reciprocal hostility and impulsive violence.

al
ic
OTHER MANIFESTATIONS OF ATTACHMENT THREAT

og
ol
Beyond the issues raised by relationship dynamics, several matters having
ch
to do with attachment insecurity and violence deserve attention. These include
y

suicide, destructive behavior in military contexts, religion-based aggressive

Ps

behavior, and violence against an outgroup as a means of defending oneself

from death awareness.
an
ic

Attachment Disruption and Aggression Toward the Self

er
Am

One rarely reported aspect of attachment-related violence is suicide,

which is especially common among men who have recently divorced. Using
e

data from the U.S. National Morbidity Study, Kposowa (2000) found that
th

divorced or separated persons were over twice as likely as married persons

by

to commit suicide. Men were four times more likely than women to commit
suicide under these conditions, and White males had the highest rate of all.
11

If suicide is considered to be a form of aggression directed toward the self, we

20

should consider why men are more prone than women to enact this form of
aggression following relationship dissolution.
©
ht

Attachment and Military Violence

ig
yr

The attachment behavioral system can be triggered by exposure to any

op

kind of danger. Anecdotal reports from battlefields indicate that wounded

and dying soldiers often call out for their mothers (Ferguson, 2006), recalling
C

the well-documented reactions of young children. Under these extremely

266 DONALD G. DUTTON

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 267

frightening conditions, soldiers sometimes engage in what Mawson (1987)

called “transient criminality.” He reviewed evidence that under conditions of
extreme stress, such as found in natural disasters and combat, there are notable
psychological changes of the following kind: partial identity diffusion, a sense

A)
of depersonalization, impaired sense of self-worth, impairments in memory
and perceptual functioning, a partial loss of abstract standards including moral

P
(A
and legal rules, and a general decline in intellectual functioning (e.g., loss of
concentration, decline in problem-solving ability).

n
According to Mawson (1987), combat stress produces chronic increases

io
in sympathetic nervous system arousal, which in turn produces “seeking

at
behavior” of the kind Bowlby (1973, 1982) attributed to the attachment

ci
behavioral system. In combat, because no security or familiar source of security

so
is to be found, further increases in sympathetic arousal are generated. Under

As
these conditions, a person’s cognitive map begins to disintegrate; more patterned,
abstract, differentiated mental processes situating the person in a complex of

al
familiar people and places and normative obligations dissolve. According to

ic
Baumeister (1990), the situated identity or individuated self also breaks down

og
and consciousness is altered so that it is completely focused on the present
ol
action, with autonomic arousal remaining very high. Instead of seeking the
ch
familiar (to reduce distress), the soldier is likely to engage in impulsive, enor-
y

mously destructive behavior.

Ps

This is one way to think about the infamous My Lai massacre, perpetrated
by a U.S. Army platoon during the Vietnam War. It followed weeks of losing
an

comrades to booby-trapped mines while being unable to find the “enemy” on

ic

a search and destroy mission (Hersh, 1970). The testimony given at the war
er

crimes trial of the perpetrators revealed a thought process that resembled

Am

paranoid schizophrenia (Dutton, 2007). The defendants described “booby-

trapped babies” lobbed at them by mothers and referred to “the Battle of My
e

Lai” although none of the villagers were armed or fired a weapon. This kind
th

of mental breakdown in a situation in which violence is natural is often viewed

by

by social psychologists as something that could happen to anyone under

extreme stress. An attachment analysis suggests, however, that there may be
11

important individual differences in behavior even under these conditions, as

20

in fact was the case at My Lai (Hersh, 1970).

©

Attachment in Perpetuity and Holy War

ht
ig

Religious ideology has one great advantage over political ideology in

yr

generating violence. It can offer everlasting salvation, as Pope Urban II did to

op

launch the First Crusade. The same strategy was used with Japanese kamikaze
pilots in World War II, and it is used now to inspire Islamic suicide bombers
C

(Dutton, 2007). The belief that is central to this kind of activity may be the

ATTACHMENT AND VIOLENCE 267

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 268

most powerful motivating belief for human beings: that they will live with their
loved ones (including an all-powerful symbolic parental figure) in everlasting
bliss (i.e., they will enjoy attachment security in perpetuity). Cheung-Blunden
and Blunden (2008), for example, found that Christian students exposed to

A)
images of the World Trade Center towers in flames on 9/11 were more likely
than atheists to define the enemy in religious terms and to condone bombing

P
(A
and killing as a form of reprisal. Both Christianity and Islam have done this
repeatedly throughout their histories, despite both the Bible and the Koran

n
advocating peace and forgiveness.

io
A further motivating image resides in religious ideology: the notion of

at
Armageddon. Asbridge (2004) describes how the recapturing of Jerusalem

ci
was consistent with the Christian belief that the “last days” before the Second

so
Coming of Christ, foretold in the Bible, could come to pass only after Jerusalem

As
was again in Christian hands. Phillips’ 2006 volume, American Theocracy,
points to the same belief among a sizeable number of fundamentalist Ameri-

al
cans (called “end timers”) today. He sees this belief as supportive of the 2003

ic
invasion of Iraq, which end timers viewed as a way to reduce a threat to Israel

og
and Jerusalem.
ol
Although weaponry has improved immensely in the millennium since the
ch
crusades, the ideological belief structure and psychological need to belong to a
y

like-minded “tribe” with a shared “worldview” (often articulated in religious

Ps

terms) still persists. This raises a question of whether religion constitutes a

symbolic form of attachment (Granqvist & Kirkpatrick, 2008; Kirkpatrick,
an

2005). Asbridge (2004) describes crusaders, still covered in blood and carrying
ic

booty, kneeling at the Holy Sepulchre to pray. This was not, he assures the
er

reader, seen as contradictory in medieval times. Although the extreme violence

Am

displayed in response to a perceived threat to a central belief is sometimes

described as primitive or “inhuman,” it has been demonstrated across human
e

history and diverse cultures with alarming consistency (Dutton, 2007).

th
by

Attachment and Terror Management Studies:

Violence Against a Worldview-Threatening Outgroup
11
20

Is secure attachment the antidote for death anxiety? Although Bowlby

(1973) described an anger born of fear and Becker (1973) described evil as the
©

“disguise of panic,” for some strange reason attachment research and terror
ht

management theory research (TMT; Greenberg, Solomon, & Pyszcynski, 1997)

ig

rarely overlapped until recently. If death terror is a primary human motive, as

yr

TMT suggests, and secure attachment is a relief from terror, then attachment
op

and TMT appear to be two sides of the same coin. Developmentally, attachment
has to precede death terror, because attachment, as we described earlier, is a
C

primitive process predating oedipal issues and coinciding with separation–

268 DONALD G. DUTTON

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 269

individuation around age 1.5 to 2 years (Schore, 1994, 2003a, 2003b), and
children in the preschool years have little understanding of the universality,
irreversibility, and inevitability of death (Nagy, 1948). However, studies show-
ing that some of a person’s defenses against death awareness are unrelated to

A)
death in any logical or semantic way (Pyszczynski, Greenberg, & Solomon, 1999)
raise the question of whether unconscious awareness of death is prewired or

P
(A
precedes conscious development of notions of death.
Following this reasoning, Mikulincer and Florian (2000) performed five

n
studies examining the contribution of attachment style to the psychological

io
effects of experimentally enhanced mortality salience. Although defending

at
one’s worldview by punishing criminals or dissenters has been portrayed as the

ci
normative or natural defense against mortality salience (Greenberg et al., 1997),

so
Mikulincer and Florian (2000) showed that this reaction is more characteristic

As
of people who score high on measures of attachment insecurity than of those
who score low. Specifically, experimentally induced death reminders produced

al
more severe judgments and punishments of moral transgressors only among

ic
insecurely attached participants, whether they were primarily anxious or

og
avoidant. Secure participants reacted to mortality salience with an increased
ol
sense of symbolic immortality—a transformational, constructive strategy that
ch
leads a person to invest in his or her children’s care and to engage in creative
y

activities whose products will live on after one’s death—and a more intense
Ps

desire for intimacy in close relationships (Mikulincer & Florian, 2000).

Taubman–Ben-Ari, Findler, and Mikulincer (2002) also found that less
an

insecure individuals reacted to mortality salience with greater willingness to

ic

initiate social interactions and more positive appraisals of their interpersonal

er

competence.
Am

The essential study, however—one examining attachment security as a

moderator of aggression and violence motivated by fear of death—seems
e

not to have been conducted. The theoretical question of whether insecure

th

individuals would be more prone to outgroup violence during times of

by

heightened stress and attachment-system activation remains to be conducted.

However, the manipulation of attachment bonds has historically been used
11

to shape child killers. In Rwanda, for example, child soldiers were forced to
20

kill a resident of their own village to break attachment to the tribe and foster
attachment to the killer group (Dutton, 2007; Human Rights Watch, 1999).
©
ht
ig

CONCLUSIONS
yr
op

What, then, is the connection between attachment and violence?

I have argued here that secure attachment diminishes both fear and anger
C

born of fear and the converse, that insecure attachment increases both. When

ATTACHMENT AND VIOLENCE 269

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 270

panic-driven anger is aroused, violence becomes much more likely. This process
is most salient when anger arises in intimate relationships, especially ones
perceived by the perpetrator to be the main or only potential source of security.
It manifests as rage reactions to separation or threats of separation. If we take

A)
one more theoretical step and focus on symbolic attachment in the form of a
tribe or family, then the evolutionary roots of Bowlby’s (1982) formulation of

P
(A
attachment become salient. Threats to the tribe or, by extension, to the tribe’s
ideology or future ability to thrive can generate extreme violence that may

n
have roots in attachment. Attachment and its ability to make us feel safe from

io
death is perhaps the strongest human motive. The lure of secure attachment in

at
perpetuity, unfortunately and tragically, can motivate people to kill themselves

ci
and strangers in a “holy war.”

so
As
REFERENCES

al
ic
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment:

og
Assessed in the Strange Situation and at home. Hillsdale, NJ: Erlbaum.
ol
Allison, C. J., Bartholomew, K., Mayseless, O., & Dutton, D. G. (2008). Love as
ch
a battlefield: Attachment and relationship dynamics in couples identified
y

for male partner violence. Journal of Family Issues, 29, 125–150. doi:10.1177/
Ps

0192513X07306980
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental
an

disorders (4th ed.). Washington, DC: Author.

ic

Asbridge, T. (2004). The first crusade: A new history. Oxford, England: Oxford
er

University Press.
Am

Bartholomew, K., & Allison, C. J. (2006). An attachment perspective on abusive

dynamics in intimate relationships. In M. Mikulincer & G. S. Goodman (Eds.),
e

Dynamics of romantic love: Attachment, caregiving, and sex (pp. 102–127). New York,
th

NY: Guilford Press.

by

Bartholomew, K., Henderson, A. J. Z., & Dutton, D. G. (2001). Insecure attachment

and abuse in intimate relationships. In C. Clulow (Ed.), Adult attachment and
11

couple psychotherapy: The “secure base” concept in practice and research (pp. 43–61).
London, England: Routledge.
20

Bartholomew, K., & Horowitz, L. W. (1991). Attachment styles among young adults:
©

A test of a four-category model. Journal of Personality and Social Psychology, 61,

226–244. doi:10.1037/0022-3514.61.2.226
ht
ig

Bartholomew, K., Oram, D., & Landolt, M. A. (2008). Correlates of partner abuse in
male same sex relationships. Violence and Victims, 23, 348–364.
yr
op

Baumeister, R. F. (1990). Suicide as an escape from self. Psychological Review, 97,

90–113. doi:10.1037/0033-295X.97.1.90
C

Becker, E. (1973). The denial of death. New York, NY: Free Press.

270 DONALD G. DUTTON

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 271

Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. New
York, NY: Basic Books.
Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). New York, NY:
Basic Books.

A)
Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London,

P
England: Routledge.

(A
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult
romantic attachment: An integrative overview. In J. A. Simpson & W. S. Rholes

n
(Eds.), Attachment theory and close relationships (pp. 46–76). New York, NY:

io
Guilford Press.

at
ci
Briere, J., & Runtz, M. (1989). The Trauma Symptom Checklist (TSC-33): Early

so
data on a new scale. Journal of Interpersonal Violence, 4, 151–163. doi:10.1177/
088626089004002002

As
Carlson, E. A. (1998). A prospective longitudinal study of attachment disorganization/

al
disorientation. Child Development, 69, 1107–1128.

ic
Cassidy, J., & Shaver, P. R. (Eds.). (2008). Handbook of attachment: Theory, research,

og
and clinical applications (2nd ed.). New York, NY: Guilford Press.
ol
Cheung-Blunden, V., & Blunden, B. (2008). Paving the road to war with group
ch
membership, appraisal antecedents, and anger. Aggressive Behavior, 34, 175–189.
doi:10.1002/ab.20234
y
Ps

Clift, R. J. (2008). The abusive personality in women in dating relationships. (Unpublished

doctoral dissertation). University of British Columbia, Vancouver, Canada.
an

Cordova, J. V., Jacobson, N. S., Gottman, J. M., Rushe, R., & Cox, G. (1993). Negative
ic

reciprocity and communication in couples with a violent husband. Journal of

er

Abnormal Psychology, 102, 559–564. doi:10.1037/0021-843X.102.4.559

Am

Dutton, D. G. (2006). The abusive personality (2nd ed.). New York, NY: Guilford
Press.
e

Dutton, D. G. (2007). The psychology of genocide, massacres, and extreme violence.

th

New York, NY: Praeger.

by

Dutton, D. G., Starzomski, A., Saunders, K., & Bartholomew, K. (1994). Intimacy-
anger and insecure attachment as precursors of abuse in intimate relationships.
11

Journal of Applied Social Psychology, 24, 1367–1386. doi:10.1111/j.1559-1816.1994.

tb01554.x
20

Edwards, D. W., Scott, C. L., Yarvis, R. M., Paizis, C. L., & Panizzon, M. S. (2003).
©

Impulsiveness, impulsive aggression, personality disorder and spousal violence.

ht

Violence and Victims, 18, 3–14. doi:10.1891/vivi.2003.18.1.3

ig

Ferguson, N. (2006). The war of the world. New York, NY: Penguin.
yr

Follingstad, D. R., Bradley, R. G., Helff, C. M., & Laughlin, J. E. (2002). A model for
op

predicting dating violence: Anxious attachment, angry temperament and need

for relationship control. Violence and Victims, 17, 35–47. doi:10.1891/vivi.17.1.
C

35.33639

ATTACHMENT AND VIOLENCE 271

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 272

Granqvist, P., & Kirkpatrick, L. A. (2008). Attachment and religious representations

and behavior. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment:
Theory, research, and clinical applications (2nd ed., pp. 906–933). New York, NY:
Guilford Press.

A)
Greenberg, J., Solomon, S., & Pyszcynski, T. (1997). Terror management theory of
self-esteem and cultural worldviews: Empirical assessments and conceptual

P
refinements. In M. P. Zanna (Ed.), Advances in experimental social psychology

(A
(Vol. 29, pp. 61–141). San Diego, CA: Academic Press.

n
Griffin, D. W., & Bartholomew, K. (1994). The metaphysics of measurement: The

io
case of adult attachment. In K. Bartholomew & D. Perlman (Eds.), Advances

at
in personal relationships: Attachment processes in adulthood (Vol. 5, pp. 17–52).

ci
London, England: Jessica Kinsley.

so
Gunderson, J. G. (1984). Borderline personality disorder. Washington, DC: American

As
Psychiatric Press.
Henderson, A. J. Z., Bartholomew, K., & Dutton, D. G. (1997). He loves me, he

al
loves me not: Attachment and separation resolution of abused women. Journal

ic
of Family Violence, 12, 169–192. doi:10.1023/A:1022836711637

og
Henderson, A. J. Z., Bartholomew, K., Trinke, S., & Kwong, M. J. (2005). When loving
ol
means hurting: An exploration of attachment and intimate abuse in a community
ch
sample. Journal of Family Violence, 20, 219–230. doi:10.1007/s10896-005-5985-y
y

Henning, K., Jones, A., & Holford, R. (2003). Treatment needs of women arrested
Ps

for domestic violence: A comparison with male offenders. Journal of Interpersonal

Violence, 18, 839–856. doi:10.1177/0886260503253876
an

Hersh, S. (1970). My Lai: A report on the massacre and its aftermath. New York, NY:
ic

Vintage Books.
er

Human Rights Watch. (1999). Leave none to tell the tale: Genocide in Rwanda. New
Am

York, NY: Author.

Kirkpatrick, L. A. (2005). Attachment, evolution, and the psychology of religion. New York,
e

NY: Guilford Press.

th

Klein, M., & Riviere, J. (1937). Love, hate, and reparation. New York, NY: Norton.
by

Kposowa, A. J. (2000). Marital status and suicide in the National Longitudinal

Mortality Study. Journal of Epidemiology and Community Health, 54, 254–261.
11

doi:10.1136/jech.54.4.254
20

Leonard, K. E., & Roberts, L. J. (1998). The effects of alcohol on the marital inter-
actions of aggressive and nonaggressive husbands and their wives. Journal of
©

Abnormal Psychology, 107, 602–615. doi:10.1037/0021-843X.107.4.602

ht

Lyons-Ruth, K., Bronfman, E., & Atwood, G. (1999). A relational diathesis model
ig

of hostile-helpless states of mind: Expressions in mother–infant interaction.

yr

In J. Solomon & C. George (Eds.), Attachment disorganization (pp. 33–70). New

op

York, NY: Guilford Press.

Lyons-Ruth, K., & Jacobwitz, D. (2008). Attachment disorganization: Genetic
C

factors, parenting contexts, and developmental transformation from infancy to

272 DONALD G. DUTTON

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 273

adulthood. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory,

research, and clinical applications (2nd ed., pp. 666–697). New York, NY: Guilford
Press.
Mahler, M. (1971). A study of the separation–individuation process and its possible

A)
application to borderline phenomena in the psychoanalytic situation. The
Psychoanalytic Study of the Child, 26, 403–424.

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

n
Main, M. (1995). Attachment: Overview with implications for clinical work. In

io
S. Goldberg, R. Muir, & J. Kerr (Eds.), Attachment theory: Social, developmental,

at
and clinical perspectives (pp. 407–474). Hillsdale, NJ: Erlbaum.

ci
Main, M., & Solomon, J. (1986). Discovery of a new, insecure disorganized/disoriented

so
attachment pattern: Procedures, findings and implications for the classification

As
of behavior. In T. B. Brazelton & M. W. Yogman (Eds.), Affective development
in infancy (pp. 95–124). Norwood, NJ: Ablex.

al
Main, M., & Solomon, J. (1990). Procedures for identifying infants as disorganized/

ic
disoriented during the Ainsworth strange situation. In M. T. Greenberg,

og
D. Cicchetti, & M. Cummings (Eds.), Attachment in the preschool years: Theory,
ol
research, and intervention (pp. 121–160). Chicago, IL: University of Chicago Press.
ch
Margolin, G., John, R. S., & Gleberman, L. (1988). Affective responses to conflictual
y

discussions in violent and nonviolent couples. Journal of Consulting and Clinical

Ps

Psychology, 56, 24–33. doi:10.1037/0022-006X.56.1.24

Mathes, E. W., & Severa, N. (1981). Jealousy, romantic love, and liking: Theoretical
an

considerations and preliminary scale development. Psychological Reports, 49, 23–31.

ic

Mauricio, A. M., Tein, J. Y., & Lopez, F. G. (2007). Borderline and antisocial person-
er

ality scores as mediators between attachment and intimate partner violence.

Am

Violence and Victims, 22, 139–157. doi:10.1891/088667007780477339

Mawson, A. W. (1987). Transient criminality: A model of stress induced crime. New York,
e

NY: Praeger.
th

Mikulincer, M. (1998). Adult attachment style and individual differences in func-

by

tional versus dysfunctional experiences of anger. Journal of Personality and Social

Psychology, 74, 513–524. doi:10.1037/0022-3514.74.2.513
11

Mikulincer, M., & Florian, V. (2000). Exploring individual differences in reactions

20

to mortality salience: Does attachment style regulate terror management

mechanisms? Journal of Personality and Social Psychology, 79, 260–273. doi:10.1037/
©

0022-3514.79.2.260
ht

Nagy, M. (1948). The child’s theories concerning death. The Journal of Genetic
ig

Psychology, 73, 3–27.

yr

Oldham, J., Clarkin, J., Appelbaum, A., Carr, A., Kernberg, P., Lotterman, A., &
op

Haas, G. (1985). A self-report instrument for Borderline Personality Organization.

In T. H. McGlashan (Ed.), The borderline: Current empirical research (pp. 1–18).
C

Washington, DC: American Psychiatric Press.

ATTACHMENT AND VIOLENCE 273

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 274

Perris, C., Jacobsson, L., Lindstrom, H., von Knorring, L., & Perris, H. (1980). Devel-
opment of a new inventory for assessing memories of parental rearing behaviour.
Acta Psychiatrica Scandinavica, 61, 265–274. doi:10.1111/j.1600-0447.1980.
tb00581.x

A)
Perry, B. (1997). Incubated in terror: Neurodevelopmental factors in the cycle of
violence. In J. D. Osofsky (Ed.), Children, youth and violence: Searching for solutions

P
(pp. 124–149). New York, NY: Guilford Press.

(A
Phillips, K. (2006). American theocracy. New York, NY: Viking.

n
io
Pynoos, R. S., & Eth, S. (1985). Children traumatized by witnessing acts of personal
violence: Homicide, rape, or suicide behavior. In S. Eth & R. S. Pynoos (Eds.),

at
Posttraumatic stress disorder in children (pp. 17–44). Washington, DC: American

ci
Psychiatric Press.

so
Pyszczynski, T., Greenberg, J., & Solomon, S. (1999). A dual-process model of defense

As
against conscious and unconscious death-related thoughts: An extension of
terror management theory. Psychological Review, 106, 835–845. doi:10.1037/

al
0033-295X.106.4.835

ic
og
Schore, A. N. (1994). Affect regulation and the origin of the self: The neurobiology of
emotional development. Hillsdale, NJ: Erlbaum.
ol
Schore, A. N. (2003a). Affect dysregulation and the disorders of the self. New York, NY:
ch
Norton.
y
Ps

Schore, A. N. (2003b). Affect regulation and the repair of the self. New York, NY: Norton.
Siegel, D. J. (2001a). The developing mind: How relationships and the brain interact to
an

shape who we are. New York, NY: Guilford Press.

ic

Siegel, D. J. (2001b). Toward an interpersonal neurobiology of the developing mind:

er

Attachment relationships, “mindsight,” and neural integration. Infant Mental

Am

Health Journal, 22, 67–94. doi:10.1002/1097-0355(200101/04)22:1<67::AID-

IMHJ3>3.0.CO;2-G
e

Siegel, J. M. (1986). The Multidimensional Anger Inventory. Journal of Personality

th

and Social Psychology, 51, 191–200. doi:10.1037/0022-3514.51.1.191

by

Stets, J., & Straus, M. A. (1989). The marriage license as a hitting license: A com-
parison of dating, cohabiting and married couples. Journal of Family Violence, 4,
11

161–180. doi:10.1007/BF01006627
20

Straus, M. A. (1992). Measuring intrafamily conflict and violence: The Conflict

Tactics Scale. In M. A. Straus & R. J. Gelles (Eds.), Physical violence in American
©

families (pp. 29–46). New Brunswick, NJ: Transaction Publishers.

ht

Taubman-Ben-Ari, O., Findler, L., & Mikulincer, M. (2002). The effects of mortality
ig

salience on relationship strivings and beliefs: The moderating role of attach-

yr

ment style. The British Journal of Social Psychology, 41, 419–441. doi:10.1348/
op

014466602760344296
Tolman, R. M. (1989). The development of a measure of physiological maltreatment
C

of women by their male partners. Violence and Victims, 4, 159–177.

274 DONALD G. DUTTON

12051-15_CH14-rev2.qxd 5/25/10 11:56 AM Page 275

Tweed, R. G., & Dutton, D. G. (1998). A comparison of instrumental and impulsive

subgroups of batterers. Violence and Victims, 13, 217–230.
Westen, D., & Shedler, J. (1999). Revising and assessing Axis II, Part 1: Developing
a clinically and empirically valid assessment instrument. The American Journal

A)
of Psychiatry, 156, 258–272.
Whitaker, D. J., Haileyesus, T., Swahn, M., & Saltzman, L. (2007). Differences in

P
(A
frequency of violence and reported injury between relationships with reciprocal
and non-reciprocal intimate partner violence. American Journal of Public Health,

n
97, 941–947. doi:10.2105/AJPH.2005.079020

io
at
ci
so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

ATTACHMENT AND VIOLENCE 275

C
op
yr
ig
ht
©
12051-15_CH14-rev2.qxd

20
11
5/25/10

by
th
e
11:56 AM

Am
er
ic
Page 276

an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 277

15

PA)
RESPECTING OTHERS AND BEING

(A
RESPECTED CAN REDUCE

n
io
AGGRESSION IN PARENT–CHILD

at
ci
RELATIONS AND IN SCHOOLS

so
As
OFRA MAYSELESS AND MIRI SCHARF

al
ic
og
ol
y ch
Ps
an

This chapter examines the role that respect plays in mitigating aggression
ic

in two developmental contexts: parent–child relations and schools. The role of

er

respect in reducing aggression and violence has been discussed and examined
Am

primarily by social psychologists who study adult relationships. In marriages,

legal proceedings, politics, and intergroup conflicts, respect has been recognized
e

as an important form of positive regard that helps to diffuse aggressive impulses

th

(e.g., de Cremer & Tyler, 2005; Gottman, 1994; Janoff-Bulman & Werther,
by

2008). It is interesting that the term respect has rarely been used within
the literature of developmental psychology, which examines the contexts in
11

which children grow and develop. Hence we know little about the expression,
20

experience, and effects of respect in developmental contexts; the possible

inverse association between respect and aggression; and the ways in which
©

respect can be encouraged and developed.

ht

Building on conceptualizations in the literatures of philosophy and social

ig

psychology, we propose a conceptual framework for the study of respect in

yr

developmental contexts. In the following sections, we discuss two kinds of

op

respect, unconditional and contingent, and distinguish between four related

but distinct ways in which respect and disrespect are involved in preventing
C

277
12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 278

or fostering aggression: (a) respecting others, (b) being respected, (c) being
disrespected or humiliated, and (d) respecting oneself. We then examine the
role of respect in parent–child relationships and at school. Although respect
per se, under that particular name, has rarely been examined in these contexts,

A)
several core characteristics of respect have been considered and discussed,
and their association with aggression and the misuse of power has been inves-

P
(A
tigated. In a final section we advocate an increased emphasis on respect in
parent–child relationships, schools, and other contexts, because of its potential

n
for reducing aggression and violence.

io
at
ci
UNCONDITIONAL AND CONTINGENT FORMS OF RESPECT

so
As
The concept of respect is complex and multifaceted. It refers to several
distinct yet interconnected processes and seems to have somewhat different

al
implications in different contexts and relationships. Following previous

ic
writings (e.g., Frei & Shaver, 2002), we suggest that respect is an attitude rather

og
than an emotion and includes cognitive evaluations, feelings, and behaviors
ol
with possible disparities among these different components. People may, for
ch
example, behave respectfully toward others but internally despise them and view
y

them as immoral and unworthy. Conversely, a person may feel respect toward
Ps

someone but not show it behaviorally. This may be the case, for instance, when
adolescents internally respect their parents but at times behave in a defiant and
an

disrespectful manner toward them.

ic

Various researchers and thinkers have discussed different types or

er

components of respect (Darwall, 1977; Frei & Shaver, 2002; Hendrick &
Am

Hendrick, 2006; Janoff-Bulman & Werther, 2008; Langdon, 2007; Lawrence-

Lightfoot, 2000; Roland & Foxx, 2003). Here, we will refer mainly to two broad
e

types of respect that have been discussed by most researchers, although under
th

different names.
by

Unconditional Respect
11
20

The most general meaning of respect refers to a broad humanistic tendency

to value each person as a worthy human being. We term this kind of respect
©

unconditional. It is based on a moral contention that every human being has

ht

basic rights to freedom, dignity, and autonomy (e.g., Rawls, 1971; see also
ig

Chapter 10, this volume, for a related conceptualization of respect in what

yr

the authors call “cultures of dignity”). Some researchers rely on the writings
op

of philosophers such as Kant (1959, p. 46), who viewed humans as rational

beings and “ends in themselves” who deserve unconditional respect. Others
C

(Lightfoot, 2000) rely on Piaget (1932), who considered the ability to recognize

278 MAYSELESS AND SCHARF

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 279

others as equals, and to appreciate them as having different and valuable

points of view to be based on a major developmental accomplishment that he
labeled decentration. The developmental emergence of this ability marks a move
from cognitive egocentrism to a capacity for understanding and considering

A)
other points of view.
At a deeper level, such a stance often reflects a spiritual sense or belief

P
(A
that all humans have a common divine origin and share a divine quality. This
belief implies a moral imperative to value, appreciate, and respect others, just

n
as one respects the divine. This moral imperative is shared by many religions

io
and religious movements, old and new, Eastern and Western, monotheistic

at
or not (Smith, 1991).

ci
so
Contingent Respect

As
The second broad meaning of respect, which, following Janoff-Bulman

al
and Werther (2008), we label contingent respect, refers to an attitude toward

ic
an individual who possesses or embodies certain qualities or attributes or

og
who has attained a certain admirable status. This general category of respect
ol
includes several types. For example, respect for social power (Langdon, 2007)
ch
includes respect for people with authority, such as parents or teachers, whose
y

power is bestowed by their role or status regardless of their own specific qualities
Ps

(e.g., warmth, competence, moral behavior; see Chapter 10, this volume, for
a similar analysis of respect in what the author calls “face cultures”). A different
an

kind of respect can be accorded individuals based on their moral character

ic

and integrity, for example, being hard working, trustworthy, sincere, altruis-
er

tic, or honest (Darwall, 1977). A third kind of respect is a response to a per-

Am

son’s special competence or a particular attribute that is valued by society

(Chapter 10, this volume, offers a related analysis of respect in cultures of
e

honor). Janoff-Bulman and Werther (2008) say that such respect is accorded
th

to individuals or groups perceived as having the greatest potential to contribute

by

to one’s own group or organization. Whereas unconditional respect is not

earned by any specific deed or accomplishment, the recipients of the other
11

kinds of respect—the conditional forms—must earn their respect-worthiness

20

by actions, efforts, or attainment of a certain social role or status.

©

Respect for Others and Its Relation to Aggression

ht
ig

Unconditional respect is closely related to foregoing aggression and

yr

violence. If a person respects others unconditionally, he or she values them

op

in their own right and acknowledges that they are entitled to autonomy,
privacy, and dignity and should not be injured, insulted, or humiliated.
C

Such a person is polite and allows others to have a voice; above all, because

RESPECTING OTHERS AND BEING RESPECTED 279

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 280

such a person assumes others’ divine or humanistic core, he or she refrains

from any act that might damage or hurt the other person physically or
psychologically.
Contingent respect for others is also likely to reduce aggression. First,

A)
people are naturally inclined to nurture and protect an entity (whether a person,
a group, or “nature”) that is valued and not to hurt, damage, or destroy it.

P
(A
Second, there is a moral aspect to contingent respect. A respected individual
or a respected group is considered to be morally worthy and deserving of kind

n
treatment (McCullough, 2008; see also Chapters 10 and 12, this volume).

io
Hence, there is a moral prohibition against aggression toward a respected

at
individual or group. Furthermore, in line with previous conceptual analyses

ci
(Frei & Shaver, 2002; Langdon, 2007), we contend that respecting others

so
typically involves not just a general moral valuing of them but also a commit-

As
ment to nurture, cherish, and support them. Thus, respect calls for a prosocial,
caring orientation rather than an antagonistic or antisocial orientation.

al
Finally, respecting others, particularly close relationship partners, encourages

ic
reciprocation of respect (Kumashiro, Finkel, & Rusbult, 2002). In relation-

og
ships where one feels respected, there are fewer reasons to be frustrated, angry,
or aggressive. ol
y ch

Being Respected and Its Relation to Aggression

Ps

Not just respecting others but being respected by others is likely to reduce
an

aggression. First, when an individual is unconditionally respected, he or she

ic

feels valued, protected, and secure. In this situation there is little reason to
er

behave aggressively (see Chapters 13 and 14). Further, as discussed more fully
Am

in the next section, being contingently respected fulfills people’s basic needs,
in close relationships and groups, for honor, dignity, and voice. Feeling respected
e

therefore eliminates a common source of anger and aggression and contributes

th

to an atmosphere of mutual understanding and appreciation (Miller, 2001; see

by

also Chapter 10, this volume). Second, being and feeling respected by others
often implies an expectation on the part of others that one will act morally.
11

Such expectations tend to be fulfilled (e.g., Jussim, 1986), and they can atten-
20

uate, or even prevent, aggression, even in situations where frustration and

anger might otherwise lead to aggressive acts. Finally, being respected often
©

provides a person with greater social influence, so there is less need to resort
ht

to aggression to achieve desired ends.

ig
yr

Feeling Disrespected and Aggression

op

On the flip side of respecting others and receiving their respect is the
C

possibility of losing respect or being disrespected. Not being especially singled

280 MAYSELESS AND SCHARF

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 281

out for respect is not necessarily negative; it may not lead to any particular
feeling. But being disrespected is an extremely negative experience; it often
amounts to being devalued, demeaned, or insulted, and it may include having
one’s rights to autonomy and dignity violated. Researchers have addressed

A)
this kind of disrespect in the context of expectations regarding fair and just
treatment and reactions to perceived injustice (e.g., Heuer, Blumenthal,

P
(A
Douglas, & Weinblatt, 1999; Miller, 2001). According to Miller (2001), peo-
ple generally believe they are entitled to fairness in the allocation of resources

n
(distributive justice) and fairness in procedures used to determine the allocation

io
of these resources (procedural justice). When such fairness—especially with

at
regard to procedural justice—is enacted, individuals tend to feel respected

ci
even if their requests have not been granted. When individuals feel that their

so
right to polite, fair, and respectful treatment has been violated, or that their

As
honor has been violated, they are likely to feel hurt, frustrated, and disrespected,
and therefore to react aggressively to restore self-esteem, save face, or educate

al
the offender (see Chapters 1 and 2, this volume). In line with these contentions,

ic
disrespectful treatment, which denies people what they believe is rightfully

og
theirs, is likely to cause both anger and aggression (Bettencourt & Miller, 1996;
ol
see also Chapter 10, this volume). Moreover, being subjected to disrespectful
ch
treatment is considered to justify aggression. People are less critical of aggressive
y

acts when they are viewed as retaliation for disrespectful treatment (Harvey
Ps

& Enzle, 1978; see also Chapter 12, this volume).

One interesting point in these discussions is the importance of behavioral
an

signs of respect that serve as social signals of appreciation and recognition of

ic

a person’s status (Janoff-Bulman & Werther, 2008; see also Chapter 12, this
er

volume). For example, when discussing the negative effects of disrespect, it is

Am

often observed that public disrespect in front of an audience is more threat-

ening, and hence has higher chances of evoking aggression, than a private
e

display (Bies & Moag, 1986; Pitt-Rivers, 1965).

th
by

Self-Respect and Aggression

11

Self-respect, a sense of having personal moral standards and a sense of

20

autonomy, freedom, and dignity (Roland & Foxx, 2003), is less discussed in
the social psychological literature than respect for other people. Some of the
©

negative effects of being disrespected or humiliated, and much of the resulting

ht

urge to retaliate, are said to derive from the damage incurred to one’s self-respect
ig

(e.g., Roland & Foxx, 2003). Maintaining self-respect requires living and
yr

behaving in ways that accord with one’s moral standards and expectations.
op

Respecting oneself makes it more likely that one will not tolerate and accept
disrespectful treatment by others (Roland & Foxx, 2003, p. 250). The relevance
C

of self-respect to aggression and violence is thus clear. A self-respecting

RESPECTING OTHERS AND BEING RESPECTED 281

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 282

individual is likely to be committed to behaving morally and respectfully toward

others and be motivated to do good. Hence he or she should not normally be
aggressive or violent. Furthermore, self-respect is normally associated with a
capacity for rationality and self-control. These qualities, in turn, are associated

A)
with lower levels of aggression and violence (see Chapters 1 and 2, this volume).
Finally, individuals are likely to be less aggressive toward others who exhibit

P
(A
self-respect and dignity (Dillon, 2007).
How and where do people learn self-respect and respect of others? Two

n
developmental contexts in which socialization of these dispositions occurs are

io
(a) parent–child relationships and (b) schools. Both are contexts in which

at
socialization agents are imbued with a moral and legal duty to educate and

ci
raise children to become competent adults and good citizens. How then does

so
respect, with its different forms and facets, develop and function in these

As
contexts?

al
ic
RESPECT IN PARENT–CHILD RELATIONS

og
AND ITS RELATION TO AGGRESSION
ol
ch
Respect, particularly of children toward their parents, has often been
y

discussed. The fifth of the Bible’s Ten Commandments is to “Honor your father
Ps

and your mother, so that your days may be long in the land that the Lord your
God is giving you.” A similar moral and religious mandate is embedded in
an

another ancient tradition, Chinese Confucian thought, which makes filial piety,
ic

love and respect for one’s parents and ancestors, one of the most important
er

virtues. This is respect of the contingent type, which is common in unilateral

Am

power relationships.
Children’s respect for their parents involves demonstrations of honor
e

and esteem toward parents and polite compliance with their instructions and
th

requests (Langdon, 2007; Lightfoot, 2000). A child who disobeys his or her
by

parents or shows defiance is considered disrespectful, whereas such behavior

in a more egalitarian peer relationship might not be considered disrespectful.
11

The underlying assumption seems to be that parents have a moral right to be

20

obeyed, esteemed, and honored by their children because of their parental role,
almost without consideration of the parents’ actual expertise, competence,
©

knowledge, or virtue. This is quite different from the conceptualization of

ht

respect within an egalitarian relationship, such as a romantic relationship,

ig

in which respect implies appreciation based on actual merit (Hendrick &

yr

Hendrick, 2006).
op

Parents, too, are expected to respect their children. In his book on

10 principles of good parenting, Steinberg (2004), a prominent developmental
C

researcher, includes respecting your child as one of the principles. What is

282 MAYSELESS AND SCHARF

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 283

meant by respect in that case? As authority figures, parents are expected to

consider their children’s point of view and honor their autonomous decision
processes and autonomous pursuits, while still providing rules and regulations
and monitoring their implementation. Parents’ respect for their children can

A)
be seen in the fair and just procedures they implement and in their acknowl-
edgement of their children’s right to have a “voice.” This respect seems to be

P
(A
of the unconditional type, because it relates to the children’s value as human
beings and not to their specific moral character or accomplishments. Conse-

n
quently, although respect in these relationships is expected to be mutual and

io
is often described as such, the ways in which it is expected to be expressed

at
differ somewhat in their qualities.

ci
Both types of respect are related to aggression. When children respect,

so
obey, and honor their parents, and comply with their parents’ requests, they

As
evince capacities for self-control and delay of gratification as well as appreci-
ation of authority, rules, and regulations. Each of these qualities can reduce

al
anger and aggression that children display toward their parents and that parents

ic
display toward their child (see Chapters 1, 2, and 9). Furthermore, children’s

og
respect for their parents usually puts the parents in the position of authority
ol
figures whom their children wish to emulate, thereby minimizing disagreements.
ch
Such a clear and predictable context enhances the child’s sense of confidence
y

and security. Securely attached children, who have a sense of direction and
Ps

meaning, are usually less prone to aggression as either an instrumental means to

attain a desired outcome or an uncontrolled reaction to an offence or frustration
an

(see Chapters 13 & 14). Furthermore, parents’ modeling of respect for their
ic

children tends to be emulated and internalized; hence, respected children tend

er

to respect themselves and the people with whom they interact (see Chapter 7
Am

for evidence of these intergenerational transmission processes). In addition,

children who have been treated with respect tend to develop senses of security,
e

autonomy, and competence, which render them happier, more self-controlled,

th

and less often frustrated (see Chapter 7). Each of these qualities should reduce
by

the prevalence of angry, aggressive behavior.

In a study of Singapore adolescents (Sim, 2000), regard for parents was
11

associated with lower levels of antisocial behavior. Regard for parents further
20

moderated the association between parental support and children’s self-esteem,

and mediated the association between parental monitoring and children’s
©

susceptibility to antisocial peer pressure. Similarly, in a large representative

ht

sample of Israeli youth (Scharf & Mayseless, 2005), respect for parents
ig

(e.g., “I respect them”) predicted lower levels of aggression at school, over and
yr

above the contribution of perceived parental support and acceptance. In this

op

study, we distinguished between respect as compliance (e.g., “It’s important

for me to do what my parent would like me to do”) and respect as a positive
C

valuation of parents (e.g., “They are wise and have had rich life experiences”).

RESPECTING OTHERS AND BEING RESPECTED 283

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 284

It is interesting that both forms of respect uniquely predicted less aggression

at school, thus underscoring the importance of two different types of contingent
respect, one based on the parental role as an authority figure and the other
based on merit because of specific positive attributes (e.g., wisdom).

A)
The effects of parental respect for children have been extensively studied
by researchers examining the topic of disrespect for children’s rights to autonomy

P
(A
and individuation. Parenting attitudes and practices that result in rejection
of the child or invalidation of the child’s autonomy—practices such as guilt

n
induction and aversive psychological control—have clear negative conse-

io
quences for the child’s affect regulation and externalizing behavior problems

at
(Barber, 2002; see also Chapter 7, this volume). Psychological control involves

ci
intrusive parenting practices such as instilling anxiety, inducing guilt, and

so
withdrawing love in order to deny or suppress a child’s thoughts, feelings, or

As
desired actions (Barber, 2002; Mayseless & Scharf, 2009). In many studies,
psychological control has been strongly associated with adverse outcomes,

al
including delinquency and antisocial behavior (Barber, 2002).

ic
Parents too, when they feel disrespected by their children, tend to be

og
aggressive and abusive. Though not explicitly using the term respect, devel-
ol
opmental researchers have examined the ramifications of parents’ sense of
ch
humiliation and lack of power because of disrespect and defiance on the part of
y

their children. In particular, Bugental and colleagues (Bugental & Lewis, 1999;
Ps

Martorell & Bugental, 2006) have proposed a model of the misuse of power
by people who perceive themselves as powerless yet are placed in a position
an

of authority. Bugental and colleagues proposed and demonstrated that parents

ic

who feel powerless are more stressed by their children’s misbehavior and tend
er

more frequently to resort to coercive and abusive practices than parents who
Am

perceive themselves to have greater power in the parent–child relationship.

Furthermore, the tendency to use abusive practices (e.g., spanking) is heightened
e

when children are more challenging (Bugental & Happaney, 2004). We suspect,
th

given these researchers’ descriptions of what they observed, that parents’ sense
by

of low power in the parent–child relationship was related to being disobeyed

and feeling disrespected by their children.
11

Researchers have also noted variability among cultures and ethnic

20

groups with regard to the importance of respect in parent–child relationships

(Rubin & Chung, 2006). For example, respect as a demonstration of filial
©

piety is highly emphasized in some Asian cultures, especially Chinese culture

ht

(Ikels, 2004). In Western cultures, too, there are variations with regard to
ig

the importance of respect for parents. For example, Dixon, Graber, and
yr

Brooks-Gunn (2008) found higher respect for parental authority among

op

African American and Latina girls compared with European American girls.
Furthermore, low levels of respect were associated with more intense arguments
C

in ethnic groups that placed higher value on respect (e.g., African American

284 MAYSELESS AND SCHARF

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 285

mother–daughter dyads and Latina mother–daughter dyads) than in other

groups (e.g., European American).
In a binational study that included samples of middle-class families with
late adolescent girls in the United States and Israel, we found another inter-

A)
esting example of cultural differences. The study included a videotaped inter-
action in which parents and their adolescent child were asked to discuss a

P
(A
conflictual issue. Based on the 10-minute interaction with each parent, several
aspects of respect were coded. Autonomy-inhibiting behaviors included over-

n
personalizing the disagreement and pressuring the other person to agree, which

io
were ways of expressing disrespect for the other’s space, boundaries, and views;

at
whereas relatedness-inhibiting behaviors included expressing hostility and rudely

ci
interrupting the other, showing disrespect for the other’s right to be heard.

so
Israeli girls were more autonomy-inhibiting toward both parents than U.S. girls.

As
In addition, Israeli girls were more relatedness-inhibiting toward their mothers
compared with the U.S. girls. These observations accord well with claims for

al
low levels of respect toward authority in general, and for parents in particular,

ic
in Israeli society (Scharf & Mayseless, 2005; Golden & Mayseless, 2008).

og
ol
ch
RESPECT IN SCHOOLS AND ITS RELATION TO AGGRESSION
y
Ps

School is another developmental context in which respect is important.

Respect in schools has been studied in terms of various constructs and in
an

various kinds of relationships, for example, the general school climate,

ic

teacher–student relations, and peer relations. Many researchers have discussed

er

the importance of respect in the school context, suggesting that respecting each
Am

student and his or her potential for growth and development should be a core
aspect of education and a central component of school climate (Noddings,
e

1996). This contention has been emphasized in discussions of minority students,

th

students with special needs, and students from high-risk backgrounds or envi-
by

ronments (e.g., Battistich, Solomon, Watson, & Schaps, 1997).

This concern is part of a humanistic perspective on education and a
11

moral perspective that advocates caring for students as valued human beings
20

and ensuring that they flourish and actualize their potential (Noddings, 2005).
Respecting individual differences in this context often means that authority
©

figures are expected to create a school climate in which each student feels
ht

valued and competent and is helped to realize his or her potential. Such a
ig

climate is expected to prevent aggression and violence and to promote cooper-

yr

ation and benevolence (e.g., Mulcahy & Casella, 2005). Although past research
op

has not assessed respect directly, it has demonstrated associations between a

caring school climate and lower levels of violence, and between interventions
C

aimed at creating such a climate and decreases in violence and aggression

RESPECTING OTHERS AND BEING RESPECTED 285

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 286

(e.g., Benbenishty & Astor, 2005; Kasen, Berenson, Cohen, & Johnson, 2004)
In one of the few studies that directly focused on the expression of respect,
LaRusso, Romer, and Selman (2008) found, in a representative sample of U.S.
high schools, that greater perceived teacher support and regard for students’

A)
perspectives was associated with students’ perception of their schools as
having respectful climates and with positive outcomes, such as lower levels

P
(A
of drug use.
Students are generally expected to have and show respect for peers and

n
teachers. Showing respect for peers often means not being aggressive toward

io
them either physically or relationally, being polite to them, and refraining

at
from damaging their belongings and blemishing their reputations (DioGuardi

ci
& Theodore, 2006) Within such an egalitarian context, mutual respect proved

so
to be important in promoting friendships (Zongkui, Chumei, & Hsueh, 2006).

As
On the flip side, disrespect for peers who have low social power has been
considered one of the most serious causes of aggression at school: peer victim-

al
ization and bullying. Peer victimization has been defined as repeated exposure

ic
to physical and verbal aggressive actions by peers (Olweus, 1997). Several

og
studies have identified peer bullying as one of the major problems in U.S. schools
ol
(Espelage & Swearer, 2004). It is interesting that when discussing the factors
ch
in a school that affect peer bullying, Batsche and Knoff (1994) suggested that
y

lack of respect for peers and the presence of teachers who seem to overlook the
Ps

problem and not intervene are important. It is not surprising that some of
the best-known intervention programs mention respect in their titles, although
an

the focus on respect in the actual interventions is not strong (e.g., the Expect
ic

Respect Program described by Meraviglia, Becker, Rosenbluth, Sanchez, &

er

Robertson, 2003; the Respect Program described by Ertesvåg & Vaaland, 2007).
Am

Students are clearly expected to respect their teachers, and this type
of respect closely resembles that expected from children in parent–child
e

relationships. Respect for teachers is based on the teacher’s role as an author-

th

ity figure who is responsible for the children’s acquisition of culturally valued
by

knowledge and competencies, proper behavior, and a sense of value and

well-being. In this context the student, especially in elementary school, is
11

expected to obey the teachers and show respect by abiding by the teachers’
20

rules and meeting their expectations (e.g., Hsueh, Zhou, Cohen, Hundley, &
Deptula, 2005). Thus, in teacher–student relations, students are expected to
©

show respect that is contingent on each teacher’s role as an authority figure in

ht

charge of their education, whereas teachers are expected to show unconditional

ig

respect for their students as valuable human beings.

yr

In line with the similarity between parents’ and teachers’ roles, several
op

researchers have likened the teacher to a parent (e.g., Wentzel, 2002). For
example, teachers’ respect for students’ autonomy, as in parent–child relation-
C

ships, was associated with fewer behavioral problems (Wentzel, 2002).

286 MAYSELESS AND SCHARF

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 287

Bugental and her colleagues (Bugental, Lyon, Lin, McGrath, & Bimbela, 1999)
have similarly likened teachers and parents in their authority roles and found
that a teacher’s low perception of power is associated with intrusive and
coercive educational practices.

A)
As children grow older, respect is expected to be more reciprocal; that
is, both teachers and students are expected to show respect of the unconditional

P
(A
type although, in addition, students have contingent respect for their teachers
based on meritorious qualities, such as expertise, knowledge, or moral integrity

n
(e.g., Chunmei, Zongkui, & Hse, 2005). In this context, students often respect

io
their teachers in response to the respect they receive from them (Noddings,

at
1996). All types of respect—of teachers for students and of students for teachers

ci
and peers—are expected to be negatively associated with aggression. For

so
example, victims of violence (i.e., both pure victims and bullies who are also

As
victims) report feeling less respected at school than other students feel
(Morrison, 2006). Langdon and Preble (2008) examined both adult respect

al
(e.g., “Most of my teachers treat students with fairness and respect”) and peer

ic
respect (e.g., “Students treat each other with fairness and respect at this school”)

og
in school using a large sample of 5th- through 12th-grade students. Each type of
ol
respect uniquely predicted lower levels of bullying at the school after controlling
ch
for background variables such as gender and ethnic origin.
y

In general, we would predict that respect from various sources—parents,

Ps

peers, and teachers—are all relevant to lowering levels of aggression and

violence. In our own study of a large representative sample of Israeli youth
an

(N = ∼ 3500), we examined the association between adolescents’ respect for

ic

parents and respect for teachers on the one hand and various outcomes on
er

the other. We found that respect for teachers was negatively associated with
Am

aggression after statistically controlling respect for parents and perceived

acceptance by parents, which were also associated with lower aggression
e

(Scharf & Mayseless, 2005).

th

Another indication of the role of various sources of respect in attenuating

by

aggression comes from a study by Knafo, Daniel, and Khoury-Kassabri (2008).

They examined the importance of students’ hierarchy of values in predicting
11

violence and aggression at school. Though they did not specifically use the
20

term respect, they assessed values that are closely related to facets of respect.
For example, the value called universalism, defined as “understanding, appre-
©

ciation, tolerance, and protection of the welfare of all people and of nature”
ht

(p. 654), reflects the unconditional type of respect, whereas the value called
ig

conformity, defined as “limiting actions and urges that might violate social
yr

expectations and norms” (p. 654), reflects the contingent type of respect for
op

authority. Using a large sample of Jewish and Arab students in Israel, Knafo
et al. found that youth in both ethnic groups who endorsed these values were
C

less inclined to be aggressive and violent at school. Furthermore, in schools

RESPECTING OTHERS AND BEING RESPECTED 287

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 288

in which violent behavior was more frequent, these values, and in particular
universalism, were strongly negatively associated with self-reported violence
and provided a stronger protective effect against violence than in schools
where violent behavior was less frequent. It thus appears that values that

A)
reflect respect, either unconditional or contingent, are associated with lower
levels of aggression and violence, and such values seem to be able to counteract

P
(A
contextual factors that favor aggression. In sum, although there are currently
only a handful of studies examining respect in schools directly, there are

n
promising empirical indications that respect of various kinds—teachers for

io
students, students for teachers, and peers for peers—in addition to internalized

at
values that foster respect are associated with lower levels of aggression and

ci
violence at school.

so
As
WHY IS RESPECT SO IMPORTANT AND HOW IS IT DIFFERENT

al
FROM OTHER ATTITUDES OR EMOTIONS?

ic
og
Respect is often conceptually and empirically associated with other atti-
ol
tudes and emotions. For example, examining respect in close relationships,
ch
Frei and Shaver (2002) found that when participants were asked to identify
y

central features of respect in romantic relationships, loving and caring seemed

Ps

to be central, in addition to more predictable qualities such as honesty, trust-

worthiness, reliability, and being considerate. Similarly, Hendrick and Hendrick
an

(2006) devised a scale to assess respect for a romantic partner, and it focused
ic

on approval, communication, interest, and care. In addition, high status

er

and power within a group was discussed as contributing to respect.

Am

These characteristics might play an important role in mitigating aggres-

sion, but they are not equivalent to respect. Respect should not be confused
e

with love and caring or with general power and value within a group. First,
th

unlike empathy or caring, which are “warm” emotions that incline a person
by

to take responsibility for the well-being of a needy other, respect is a moral

attitude of appreciation of the other and a moral imperative to refrain from
11

harming the other. It is a relatively “cooler” disposition. In general, love and

20

caring are strong buffers against aggression, but they do not always have this
beneficial effect. In fact, a large number of clinical case studies of marital vio-
©

lence and child abuse demonstrate that love and caring can sometimes lead
ht

to violence and aggression when they are not associated with respect for the
ig

other’s autonomy and his or her value as a separate and unique human being
yr

(see Chapters 19 and 20). Similarly, viewing another person as having high
op

status or great power is not synonymous with respect and may not be associated
with lower levels of aggression toward the person. In fact, history is full of
C

incidents in which people were aggressive toward others precisely because

288 MAYSELESS AND SCHARF

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 289

they valued them and wanted to possess them or tried to destroy them because
of their unjustifiably high status or undeserved power (see Chapter 4). Respect
for others is quite different from this love and status envy. Respect includes a
moral imperative not to harm the other. The value accorded to the other

A)
entails consideration and acceptance of the other’s right to dignity and
autonomy.

P
(A
n
HOW CAN RESPECT BE FOSTERED?

io
at
In every context in which children are reared and educated, a general

ci
attitude of respect can be demonstrated and taught. One of the most influential

so
ways in which respect can be taught and transmitted is by modeling it in daily

As
life. Respect breeds respect, and values and attitudes in particular are inter-
nalized by emulating authority figures (Grusec & Kuczynski, 1997). To foster

al
and teach respect, parents, teachers, and other authority figures should behave

ic
respectfully toward others. Behaving respectfully means honoring and enabling

og
the other’s autonomy, showing interest and care, allowing the other to have
ol
a voice even when not approving its message, and appreciating the other’s
ch
natural goodness. Contingent respect is learned when the unique qualities of
y

each individual are acknowledged, valued, and appreciated. In the school

Ps

context, this requires becoming better acquainted with each student’s unique-
ness and allowing each student to be successful in his or her own way and in
an

his or her own preferred domains.

ic

Another behavior that is important to model is not allowing disrespectful

er

behavior toward oneself or others; that is, to foster respect, authority figures
Am

need to intervene when they observe disrespectful behavior. Such an inter-

vention should in itself be done respectfully, but it should be clear and decisive.
e

Children and adults learn respect or disrespect partly from the general climate
th

in their environment. Observing disrespect on the part of parents toward each

by

other, or disrespect shown by a school principal toward teachers, might be

almost as harmful to a child’s development of respect as being personally dis-
11

respected. In particular, to foster respect, authority figures should not allow

20

disrespectful behavior toward themselves or among peers.

Another way to foster respect is to promote children’s self-respect.
©

Self-respect is linked to a sense of autonomy and self-control that includes

ht

personal and moral standards (Roland & Foxx, 2003). Authority figures can
ig

promote these aspects of self-respect by targeting them as socialization goals and

yr

promoting moral values, self-control, and autonomy. Thus, besides modeling

op

respect and intervening when disrespect is shown, authority figures should

teach children about respect more directly and verbally. This includes clear
C

discussions of values and the importance of good character and moral conduct.

RESPECTING OTHERS AND BEING RESPECTED 289

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 290

Articulating clear expectations for behaviors that are consistent with respect,
as discussed in this chapter, provides children with concrete and much needed
direction. Following through with helpful scaffolding (e.g., reminders, obser-
vations, examples) helps them internalize these standards and make them

A)
their own. This can be done by helping children articulate and discuss their
feelings, hesitations, and decision processes with regard to their values.

P
(A
Contemporary parents and teachers sometimes shy away from a directive
approach because they see themselves as living in a child-centered world. But

n
good directions, as can be seen in athletic coaching situations, can be beneficial

io
to novices.

at
ci
so
CONCLUSION

As
In sum, the explicit study of respect within contexts in which children

al
are raised, socialized, and educated is still fairly rare, and its results are not

ic
conceptually well integrated. Although constructs related to respect have been

og
examined (e.g., psychological control by parents, acknowledgement of ethnic
ol
diversity in schools), research efforts to conceptualize and measure different
ch
types of respect in home and school contexts, their associations with various
y

outcomes, and their unique contributions to these outcomes are scarce. We

Ps

hope that the conceptual framework elucidated in this chapter will open
new possibilities to study respect and its vicissitudes, including reductions
an

in aggression and violence, in developmental contexts.

ic
er
Am

REFERENCES
e

Barber, B. K. (Ed.). (2002). Intrusive parenting: How psychological control affects children
th

and adolescents. Washington, DC: American Psychological Association.

by

Batsche, G. M., & Knoff, H. M. (1994). Bullies and their victims: Understanding a
pervasive problem in the schools. School Psychology Review, 23, 165–174.
11

Battistich, V., Solomon, D., Watson, M., & Schaps, E. (1997). Caring school commu-
nities. Educational Psychologist, 32, 137–151. doi:10.1207/s15326985ep3203_1
20

Benbenishty, R., & Astor, R. A. (2005). School violence in context: Culture, neighborhood,
©

family, school, and gender. New York, NY: Oxford University Press.
ht

Bettencourt, B. A., & Miller, N. (1996). Sex differences in aggression as a function of

ig

provocation: A meta-analysis. Psychological Bulletin, 119, 422–447. doi:10.1037/

yr

0033-2909.119.3.422
op

Bies, R. J., & Moag, J. S. (1986). Interactional justice: Communication criteria for
fairness. In B. Sheppard (Ed.), Research on negotiation in organizations (pp. 43–55).
C

Greenwich, CT: JAI Press.

290 MAYSELESS AND SCHARF

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 291

Bugental, D. B., & Happaney, K. (2004). Predicting infant maltreatment in low-

income families: The interactive effects of maternal attributions and child
status at birth. Developmental Psychology, 40, 234–243. doi:10.1037/0012-1649.
40.2.234

A)
Bugental, D. B., & Lewis, J. C. (1999). The paradoxical misuse of power by those who
see themselves as powerless: How does it happen? Journal of Social Issues, 55,

P
(A
51–64. doi:10.1111/0022-4537.00104/
Bugental, D. B., Lyon, J. E., Lin, E. K., McGrath, E. P., & Bimbela, A. (1999). Children

n
“tune out” in response to the ambiguous communication style of powerless adults.

io
Child Development, 70, 214–230. doi:10.1111/1467-8624.00016

at
ci
Chunmei, Z., Zongkui, Z., & Hse, Y. (2005). The conception of respect and its devel-

so
opment in childhood. Psychological Science (China), 28, 337–341.

As
Darwall, S. L. (1977). Two kinds of respect. Ethics, 88, 36–48. doi:10.1086/292054
de Cremer, D., & Tyler, T. R. (2005). Am I respected or not? Inclusion and reputation

al
as issues in group membership. Social Justice Research, 18, 121–153. doi:10.1007/

ic
s11211-005-7366-3

og
Dillon, R. S. (2007). Arrogance, self respect, and personhood. Journal of Consciousness
Studies, 14, 101–126. ol
ch
DioGuardi, R. J., & Theodore, L. A. (2006). Understanding and addressing peer
y

victimization among students. In S. R. Jimerson & M. Furlong (Eds.), Handbook

Ps

of school violence and school safety: From research to practice (pp. 339–352). Mahwah,
NJ: Erlbaum.
an

Dixon, S. V., Graber, J., & Brooks-Gunn, J. (2008). The roles of respect for parental
ic

authority and parenting practices in parent–child conflict among African Amer-

er

ican, Latino, and European American families. Journal of Family Psychology, 22,
Am

1–10. doi:10.1037/0893-3200.22.1.1
Ertesvåg, S. K., & Vaaland, G. S. (2007). Prevention and reduction of behavioral
e

problems in school: An evaluation of the Respect program. Educational Psychology,

th

27, 713–736. doi:10.1080/01443410701309258

by

Espelage, D. L., & Swearer, S. M. (2004). Bullying in American schools: A social–ecological

perspective on prevention and intervention. Hillsdale, NJ: Erlbaum.
11

Frei, J., & Shaver, P. R. (2002). Respect in close relationships: Prototype definition,
20

self-report assessment, and initial correlates. Personal Relationships, 9, 121–139.

doi:10.1111/1475-6811.00008
©

Golden, D., & Mayseless, O. (2008). On the alert in an unpredictable environment.

ht

Culture and Psychology, 14, 155–179. doi:10.1177/1354067X08088553

ig

Gottman, J. M. (1994). What predicts divorce? The relationship between marital processes
yr

and marital outcomes. Hillsdale, NJ: Erlbaum.

op

Grusec, J. E., & Kuczynski, L. (1997). Parenting and children’s internalization of values:
C

A handbook of contemporary theory. Hoboken, NJ: Wiley.

RESPECTING OTHERS AND BEING RESPECTED 291

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 292

Harvey, M. D., & Enzle, M. E. (1978). Effects of retaliation latency and provocation
level on judged blameworthiness for retaliatory aggression. Personality and Social
Psychology Bulletin, 4, 579–582. doi:10.1177/014616727800400417
Hendrick, S. S., & Hendrick, C. (2006). Measuring respect in close relationships.

A)
Journal of Social and Personal Relationships, 23, 881–899. doi:10.1177/0265407
506070471

P
(A
Heuer, L., Blumenthal, E., Douglas, A., & Weinblatt, T. (1999). A deservingness
approach to respect as a relationally based fairness judgment. Personality and

n
Social Psychology Bulletin, 25, 1279–1292. doi:10.1177/0146167299258009

io
Hsueh, Y., Zhou, Z., Cohen, R., Hundley, R. J., & Deptula, D. P. (2005). Knowing

at
and showing respect: Chinese and U.S. children’s understanding of respect and

ci
its association to their friendships. Journal of Psychology in Chinese Societies, 6,

so
229–260.

As
Ikels, C. (Ed.). (2004). Filial piety: Practice and discourse in contemporary East Asia.
Stanford, CA: Stanford University Press.

al
Janoff-Bulman, R., & Werther, A. (2008). The social psychology of respect: Implications

ic
for delegitimization and reconciliation. In A. Nadler, T. Malloy, & J. D. Fisher

og
(Eds.), The social psychology of inter-group reconciliation (pp. 145–170). New York,
ol
NY: Oxford University Press. doi:10.1093/acprof:oso/9780195300314.003.0008
ch
Jussim, L. (1986). Self-fulfilling prophecies: A theoretical and integrative review.
y

Psychological Review, 93, 429–445. doi:10.1037/0033-295X.93.4.429

Ps

Kant, I. (1959). Foundations of metaphysics of morals (L. W. Beck, Trans.). Indianapolis,

IN: Bobbs-Merrill.
an

Kasen, S., Berenson, K., Cohen, P., & Johnson, J. G. (2004). The effects of school
ic

climate on changes in aggressive and other behaviors related to bullying. In

er

R. L. Espelage & S. M. Swearer (Eds.), Bullying in American schools: A social–

Am

ecological perspective on prevention and intervention (pp. 187–210). Hillsdale, NJ:

Erlbaum.
e

Knafo, A., Daniel, E., & Khoury-Kassabri, M. (2008). Values as protective factors against
th

violent behavior in Jewish and Arab high schools in Israel. Child Development,
79, 652–667. doi:10.1111/j.1467-8624.2008.01149.x
by

Kumashiro, M., Finkel, E. J., & Rusbult, C. E. (2002). Self-respect and prorelationship
11

behavior in marital relationships. Journal of Personality, 70, 1009–1049.

20

Langdon, S. W. (2007). Conceptualizations of respect: Qualitative and quantitative

evidence of four (five) themes. Journal of Psychology, 141, 469–484. doi:10.3200/
©

JRLP.141.5.469-484
ht

Langdon, S. W., & Preble, W. (2008). The relationship between levels of perceived
ig

respect and bullying in 5th through 12th graders. Adolescence, 43, 485–503.
yr

LaRusso, M. D., Romer, D., & Selman, R. L. (2008). Teachers as builders of respectful
op

school climates: Implications for adolescent drug use norms and depressive symp-
toms in high school. Journal of Youth and Adolescence, 37, 386–398. doi:10.1007/
C

s10964-007-9212-4

292 MAYSELESS AND SCHARF

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 293

Lawrence-Lightfoot, S. (2000). Respect: An exploration. New York, NY: Perseus.

Lightfoot, C. (2000). On respect. New Ideas in Psychology, 18, 177–185. doi:10.1016/
S0732-118X(00)00007-6
Martorell, G. A., & Bugental, D. (2006). Maternal variations in stress reactivity:

A)
Implications for harsh parenting practices with very young children. Journal of
Family Psychology, 20, 641–647. doi:10.1037/0893-3200.20.4.641

P
(A
Mayseless, O., & Scharf, M. (2009). Too close for comfort: Inadequate boundaries
with parents and individuation in late adolescent girls. American Journal of

n
Orthopsychiatry, 79, 191–202. doi:10.1037/a0015623

io
at
McCullough, M. (2008). Beyond revenge: The evolution of the forgiveness instinct. San
Francisco, CA: Jossey-Bass.

ci
so
Meraviglia, M. G., Becker, H., Rosenbluth, B., Sanchez, E., & Robertson, T. (2003).
The Expect Respect Project: Creating a positive elementary school climate. Jour-

As
nal of Interpersonal Violence, 18, 1347–1360. doi:10.1177/0886260503257457

al
Miller, D. T. (2001). Disrespect and the experience of injustice. Annual Review of

ic
Psychology, 52, 527–553. doi:10.1146/annurev.psych.52.1.527

og
Morrison, B. (2006). School bullying and restorative justice: Toward a theoretical
understanding of the role of respect, pride, and shame. Journal of Social Issues,
ol
62, 371–392. doi:10.1111/j.1540-4560.2006.00455.x
ch
Mulcahy, D. G., & Casella, R. (2005). Violence and caring in school and society:
y

Educational studies. Educational Studies, 37, 244–255. doi:10.1207/s15326993

Ps

es3703_3
an

Noddings, N. (1996). Learning to care and to be cared for. In A. M. Hoffman (Ed.),

Schools, violence, and society (pp. 185–198). Westport, CT: Praeger.
ic
er

Noddings, N. (2005). What does it mean to educate the whole child? Educational
Leadership, 63, 8–13.
Am

Olweus, D. (1997). Bully/victim problems in school: Facts and intervention. European

Journal of Psychology of Education, 12, 495–510. doi:10.1007/BF03172807
e
th

Piaget, J. (1932). The moral judgment of the child. Oxford, England: Harcourt, Brace.
Pitt-Rivers, J. (1965). Honour and social status. In J. G. Peristiany (Ed.), The nature
by

of human society: Honor and shame (pp. 19–78). London, England: Weidenfeld
11

& Nicholson.
Rawls, J. (1971). A theory of justice. Cambridge, MA: Harvard University Press.
20

Roland, C. E., & Foxx, R. M. (2003). Self-respect: A neglected concept. Philosophical

©

Psychology, 16, 247–288. doi:10.1080/09515080307764

ht

Rubin, K. H., & Chung, O. B. (2006). Parenting beliefs, behaviors, and parent–child
ig

relations: A cross-cultural perspective. New York, NY: Psychology Press.

yr

Scharf, M., & Mayseless, O. (2005). Parental authority and its associations with social and
op

emotional functioning in school: Final research report. Unpublished report submitted

to the Ministry of Education, Israel (in Hebrew). Faculty of Education, University
C

of Haifa, Haifa, Israel.

RESPECTING OTHERS AND BEING RESPECTED 293

12051-16_CH15-rev2.qxd 5/25/10 11:57 AM Page 294

Sim, T. N. (2000). Adolescent psychosocial competence: The importance and role

of regard for parents. Journal of Research on Adolescence, 10, 49–64. doi:10.1207/
SJRA1001_3
Smith, H. (1991). The world’s religions: Our great wisdom traditions. New York, NY:

A)
HarperCollins.
Steinberg, L. (2004). The 10 basic principles of good parenting. New York, NY: Simon

P
(A
& Schuster.
Wentzel, K. R. (2002). Are effective teachers like good parents? Teaching styles

n
and student adjustment in early adolescence. Child Development, 73, 287–301.

io
doi:10.1111/1467-8624.00406

at
Zongkui, Z., Chumei, Z., & Hsueh, Y. (2006). Relations between conception of

ci
respect and peer relationships in childhood. Acta Psychologica Sinica, 38, 232–239.

so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

294 MAYSELESS AND SCHARF

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 295

IV

A)
P
(A
AGGRESSION AT

n
io
THE SOCIETAL LEVEL

at
ci
so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C
C
op
yr
ig
ht
©
20
12051-17_PT4_CH16-rev2.qxd

11
by
5/25/10

th
e
Am
11:58 AM

er
ic
an
Page 296

Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 297

16

PA)
AN EXISTENTIAL PERSPECTIVE

(A
ON VIOLENT SOLUTIONS TO

n
io
ETHNO–POLITICAL CONFLICT

at
ci
so
GILAD HIRSCHBERGER AND TOM PYSZCZYNSKI

As
al
ic
og
War would end if the dead could return.
ol
—Stanley Baldwin, British Prime Minister
y ch
Ps
an

The outbreak of violence between Israel and Hamas on the eve of the
ic

year 2009 surprised hardly anyone. The fragile ceasefire had come to an end,
er

and the renewed missile attacks on towns and villages in the south of Israel
Am

were the appetizer preceding the inevitable main course of massive violent
retribution. But, although the ebbing and flowing of violence in this region
e

has become habitual, for many it is accompanied by an increased sense of

th

frustration caused by the inability to move beyond violence to find a rational

by

solution that will bring peace to both Israelis and Palestinians.

And what would be the rational solution to this bitter and seemingly
11

intractable conflict? The dictum that “war itself is the enemy,” attributed to
20

the Prussian philosopher von Clausewitz, has never resonated more strongly.
The violent clashes between Israel and the Palestinians have resulted in
©

significant losses to both sides with little gain, if any, to justify the price. Every
ht

round of violence ends with a new shaky agreement that differs ever so slightly
ig

from the shaky agreement that preceded the most recent violent outbreak.
yr

Thus, time and again when the dust settles from the futile attempt to subdue
op

the other, Israelis and Palestinians find themselves with no viable option
but to find a way to live peacefully with one another. But no one seems capable
C

of finding a way to achieve this peace (see Chapter 17).

297
12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 298

Even more perplexing is the fact that since the 1993 peace accord
between Israel and the Palestine Liberation Organization (PLO), relations
between the groups have significantly deteriorated and violence has reached
an unprecedented level. The Israeli disengagement from Gaza in the summer

A)
of 2005 was intended to reduce friction between Israelis and Palestinians, but
it failed to live up to its promise of breaking the cycle of violence and may

P
(A
have inadvertently contributed to the recent escalation of violent conflict.
How can we explain that despite mounting evidence of the futility of violence,

n
and the obvious need for reconciliation among the various factions, there

io
seems to be no end in sight to perpetual warfare? And why has every step

at
taken toward peace paradoxically resulted in a spiral of bitter violence?

ci
One possible answer that many Israelis and Palestinians seem to adhere

so
to is that the other side has proven to be duplicitous, inhumane, and ruthless,

As
using peace as a cover for malevolent intentions (see Chapter 17). In the
present chapter we provide an alternative answer to these questions and

al
argue that powerful psychological forces underlying ethno–political conflict

ic
hamper the ability to achieve peace, even when peace seems to be a rational

og
solution that would benefit all. We base our analysis on terror management
ol
theory (e.g., Greenberg, Pyszczynski, & Solomon, 1997) and demonstrate, with
ch
a body of research conducted recently in Iran, Israel, Europe, and the United
y

States, how existential concerns underlie the proclivity to choose violent

Ps

solutions to ethno–political conflicts. We also show that the link between

mortality concerns and ethno–political violence is not inevitable, and that
an

at times mortality concerns can even reduce violent inclinations. We suggest

ic

that a better understanding of the role played by mortality concerns in polit-

er

ical reasoning provides insight into ways to move beyond violence and to
Am

promote peace.
e
th

TERROR MANAGEMENT THEORY

by

Terror management theory (TMT; e.g., Greenberg et al., 1997) contends

11

that the instinctive animal desire for continued life juxtaposed with the
20

uniquely human awareness of the inevitability of death creates a potential for

paralyzing terror; effective regulation of ongoing human behavior requires that
©

this potential for terror be effectively managed. According to TMT, existential

ht

terror is managed and security is provided by (a) a cultural worldview that

ig

provides an explanation for existence, standards through which one can

yr

attain a sense of personal value, and the promise of literal or symbolic immor-
op

tality to those who live up to these standards; (b) self-esteem, which is acquired
by believing in the cultural worldview and living up to its standards; and
C

(c) close interpersonal relationships.

298 HIRSCHBERGER AND PYSZCZYNSKI

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 299

Terror management defenses are fragile social constructions that require

ongoing validation from others if they are to promote effective functioning.
Faith in one’s worldview, the sense of personal value derived from living up
to social and cultural standards, and the anxiety-buffering effectiveness of

A)
these structures is bolstered by others sharing one’s beliefs and diminished
when others adhere to a different system of values and beliefs. According to

P
(A
TMT, because these terror-management processes protect people from deeply
rooted existential fears, much social thought and behavior is oriented toward

n
maintaining them and defending them against threats.

io
Because of the fragile nature of culturally derived forms of defense,

at
threats to these symbolic constructions undermine the emotional security that

ci
they provide and motivate people to protect their death-denying mechanisms

so
to ward off the threats. When others hold beliefs that conflict with one’s own

As
basic conceptions and proclaim their culture’s superiority or moral righteousness,
they imply that one’s worldview is incorrect and one’s culture is inferior, which

al
undermines one’s sense of value and meaning. Such threats to worldviews and

ic
self-esteem strip away an individual’s symbolic defensive shield and then people

og
may express anger and derogate the source of the threats or choose to demon-
ol
strate their group’s superiority by subduing, defeating, or even annihilating
ch
groups that challenge their worldview.
y

Terror management studies have tested these theoretical propositions

Ps

by priming thoughts of death (i.e., heightening mortality salience [MS]) and

examining cultural worldview defenses. The results of over 400 empirical
an

studies conducted in 21 countries have provided support for the theoretical

ic

propositions of TMT. For example, studies have found that MS increases the
er

motivation to invest in one’s worldview and also leads to avoidance, derogation,

Am

and aggression against worldview-threatening others (e.g., Greenberg et al.,

1990; Hirschberger & Ein-Dor, 2006; McGregor et al., 1998; Pyszczynski
e

et al., 2006). These effects appear to be unique to thoughts about death.

th

Other anxiety-producing activities, such as thinking about giving a speech,

by

imagining physical pain, or worrying about life after college do not produce the
same reactions as MS (e.g., Greenberg, Pyszczynski, Solomon, Simon, & Breus,
11

1994; Greenberg Simon, Porteus, Pyszczynski, & Solomon, 1995). Moreover,

20

TMT effects are not accompanied by an increase in negative feelings or

mediated by such feelings (Pyszczynski, Greenberg, & Solomon, 1999).
©
ht
ig

TERROR MANAGEMENT AND INTERGROUP CONFLICT

yr
op

Much of the focus of terror management research has been on inter-

group relations, showing that the need to defend symbolic death-denying
C

mental structures often results in extreme reactions toward people who uphold

AN EXISTENTIAL PERSPECTIVE ON VIOLENT SOLUTIONS 299

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 300

different cultural, religious, or national worldviews (for a review, see Pyszczynski,

Solomon, & Greenberg, 2003). For example, studies have shown that MS leads
Christian participants to derogate a Jewish person (Greenberg et al., 1990),
American college students to behave more aggressively toward those with dif-

A)
ferent political orientations than their own (McGregor et al., 1998), White
Americans to express sympathy for a White racist (Greenberg, Schimel,

P
(A
Martens, Pyszczynski, & Solomon, 2001), Italians to view their own nation
as superior to other European countries (Castano, 2004), and even Israeli chil-

n
dren as young as 11 to react more negatively to an immigrant child from Russia

io
and more positively toward a child from Israel (Florian & Mikulincer, 1998).

at
These studies have consistently shown that brief, unobtrusive reminders

ci
of mortality lead people to view their group in a more positive light and view

so
other groups in a more negative light to the extent of derogating and even

As
aggressing against other groups. Thus, it may seem that reminders of mortality
instantly turn people into ethnocentric, prejudiced, and potentially violent

al
automatons. Fortunately, TMT views the effects of MS on intergroup attitudes

ic
as more complex and maintains that different people may react to MS in dif-

og
ferent ways, depending on individual differences and the situational context.
ol
For example, research has shown that whereas people with a conservative
ch
political orientation respond to MS with greater intolerance, people with a
y

liberal political orientation respond to MS with greater tolerance (Greenberg,

Ps

Simon, Pyszczynski, Solomon, & Chatel, 1992). Other research has shown that
MS leads to heightened ingroup affiliation only when the ingroup is portrayed
an

as strong and successful but leads to disaffiliation from elements of the worldview
ic

that reflect weakness or inferiority (Arndt, Greenberg, Schimel, Pyszczynski,

er

& Solomon, 2002; Dechesne, Greenberg, Arndt, & Schimel, 2000).

Am

In the realm of political conflict, additional variables may enter into the
equation and further complicate the effect of existential concerns on violent
e

inclinations so that mortality reminders may lead to support of violent solu-

th

tions to conflict in some cases but to a rejection of violent means in others.

by

In this chapter we demonstrate, first, that MS increases support for violent

solutions to conflict but that this link between death concerns and violent
11

outcomes depends on three major conditions: (a) social and national consensus
20

on the use of violence, (b) a sense that violence is justified and necessary, and
(c) a sense that violence is imminent and unavoidable. In the next step, we
©

outline conditions that may disrupt the link between death concerns and
ht

support for violent solutions to conflict, and we demonstrate that thinking

ig

rationally about violence and considering the consequences of violence may

yr

moderate and even reverse the effects of MS on violent inclinations. In the final
op

section of our analysis we focus on moving beyond violence and demonstrate

that emphasizing basic human values and human similarities promotes more
C

peaceful motives even when death is salient.

300 HIRSCHBERGER AND PYSZCZYNSKI

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 301

EXISTENTIAL CONCERNS PROMOTE POLITICAL VIOLENCE

It is in the sphere of terrorism and counterterrorism that fear’s most

harmful manifestations flourish.
—Irene Kahn, Amnesty International

PA)
The hypothesis that death awareness motivates violent solutions to

(A
political conflict is the most straightforward application of TMT to the realm
of political psychology. War and terrorism simplify conflict and dichotomize

n
groups into “us” and “them,” “good” and “evil.” Violence also offers hope for a

io
clear-cut and long-lasting resolution of the conflict, the potential (or illusion)

at
of pronouncing winners and losers, and a better future following victory. These

ci
so
attributes of political violence make it particularly attractive when death is
salient, because under these conditions people are motivated to promote the

As
triumph of their group and the thorough defeat of the opposition.
Indeed, research has revealed that among conservative Americans MS

al
leads to greater support of extreme violence against countries or organizations

ic
that pose a threat to the United States (Pyszczynski et al., 2006, Study 2). It

og
has led to Israeli settlers in the Gaza Strip, and their supporters who refused
ol
to accept the 2005 disengagement plan, to support more violent resistance
ch
(Hirschberger & Ein-Dor, 2006). Research has also shown that MS led par-
y

ticipants in Iran (Pyszczynski et al., 2006, Study 1) and in Britain (Routledge

Ps

& Arndt, 2008) to express greater willingness to sacrifice their life for their
country. However, despite the seemingly clear link between mortality concerns
an

and support for intergroup violence, we contend that this link is neither auto-
ic

matic nor inevitable. Instead, it depends on three major conditions: perceived

er

consensus, justice, and inevitability of conflict.

Am

Consensus
e
th

Organized forms of violence, such as war and terrorism, depend to a large

by

extent on the broad support of the populations for whom these actions are pur-
portedly undertaken. When leaders receive the support of their people, they feel
11

less restrained in sending young men and women to the battlefield, and they are
20

under less pressure from their publics to end the war. However, consensus is
fragile, and often the price of war and the inability to achieve the expected
©

outcomes instill doubt as to the legitimacy or efficacy of violent policies, con-

ht

cerns about the costs of such policies (in terms of loss of life, resources, and
ig

international respect), and the legitimacy and wisdom of the leader and his
yr

or her decisions. In such cases, public support at the beginning of a conflict

op

can transform into disillusionment, frustration, and anger. Such was the
case in the American war in Vietnam and more recently in the war in Iraq,
C

where initial consensus and hope dissipated when the war failed to deliver

AN EXISTENTIAL PERSPECTIVE ON VIOLENT SOLUTIONS 301

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 302

its promise. In Israel, the Lebanon War of 1982 and, to some extent, the
Lebanon War of 2006 were launched following missile attacks on Northern
Israel and enjoyed public support until the death toll rose, and it became evident
that the declared goals of the war were not attainable by violent means.

A)
From a terror management perspective, consensus is a necessary ingredient
for the functioning of the cultural worldview as an effective anxiety buffer.

P
(A
Because worldviews are symbolic social constructions that are fragile and
susceptible to disconfirming information, they require constant consensual

n
validation. For consensus to be established and maintained it is necessary to

io
believe that the threat is of such magnitude that massive use of force is the

at
only effective response.

ci
Consensus is clearly a requirement for military action in democratic

so
societies, where leaders receive their legitimacy from the support of their

As
constituents. There is good reason to believe that consensus is important
for violent behavior in nondemocratic societies as well. For example, in the

al
Palestinian territories, suicide bombers are exalted as martyrs. Their pictures

ic
are posted on walls and buildings, and they are admired by children who view

og
them as role models. Such idolization of people who are ready to kill themselves
ol
and others for what is considered a holy cause is necessary for the propagation
ch
of suicidal terrorism because in return for a shortened life, terrorists gain
y

fame, adoration, respect, and honor (not to mention several afterlife virgins).
Ps

According to TMT, long-lasting fame and admiration may be more appealing

than a longer life because they provide a sense of symbolic immortality, the
an

feeling that certain aspects of the self will survive physical death.
ic

Empirical evidence supports the contention that MS elicits more consen-

er

sus regarding violent responses to conflict as well as the contention that con-
Am

sensus is a necessary precondition for MS to elicit support for ethno–political

violence. In a series of studies, Landau and his colleagues (Landau, Johns, et al.,
e

2004) demonstrated that shortly after the beginning of the Iraq War, reminders
th

of death increased support for American president George W. Bush and his
by

counterterrorist policies. Other studies show that without a sense of consensus,

MS may not lead to support of violence against other groups.
11

In a unique study conducted in Iran (Pyszczynski et al., 2006), partici-

20

pants were assigned to either an MS or control condition and then read a

description of a student portrayed as holding commonly expressed views on
©

political issues. Half of the participants read a description in which the student
ht

expressed support for martyrdom attacks against Western targets, and the
ig

other half read a description in which the student expressed disapproval of

yr

suicidal terrorism. The results indicated that MS led to greater support of

op

martyrdom attacks only when the student expressed pro-martyrdom attitudes.

However, when the student voiced opposition to suicidal terrorism, MS did
C

not have a significant effect on support of violence. A more recent study

302 HIRSCHBERGER AND PYSZCZYNSKI

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 303

demonstrated that MS increased support for terrorist violence among Iranians

but that this effect was eliminated when participants were led to believe that
the majority of their countrymen disapproved of such tactics (Abdollahi,
Henthorn, & Pyszczynski, 2009). These results indicate that for MS to elicit

A)
support for suicidal terrorism, participants had to feel that there was consensus
among their peers supporting such violence.

P
(A
Justice

n
io
Consensus for violent solutions to conflict may be established when

at
people feel that they are fighting for a just and noble cause. However, just

ci
because a cause is noble does not mean it will be achieved using violent means

so
or that violence is the most effective route to attain the desired goal. Early

As
scholars of war and conflict viewed the use of violence as a rational option that
could advance a country’s interests, as Clausewitz (1832/1976) contended,

al
“War is not a mere act of policy but a true political instrument, a continuation

ic
of political activity by other means,” implying that the decision to engage in

og
war is the product of a rational cost–benefit analysis. From this perspective,
ol
war is waged when leaders conclude that war is a more efficient way to achieve
ch
political goals than other available means.
y

Expected utility theory (e.g., Bueno de Mesquita, 1988) formalizes this

Ps

assumption and posits that by delineating the costs and benefits of conflict,
political scientists may better understand the motives for going to war.
an

However, over the years scholars of conflict resolution have concluded that
ic

rational factors alone do little to explain the outburst of violent conflict and
er

that emotional and motivational factors such as anger, resentment, and

Am

revenge play a pivotal role in the decision to engage in war (e.g., Baumeister
& Butz, 2005).
e

The decision to use violence in interethnic conflict may be driven

th

so powerfully by emotional factors such as anger, revenge, and the need to

by

restore a sense of justice, that utilitarian considerations, such as whether war

is a good instrumental means of achieving a desired outcome, are ignored
11

(see Chapter 17). In an analysis of five major international conflicts (including

20

the two world wars), Welch (1993) concluded that the motivation to achieve
justice or restore justice has been a major factor in most global conflicts. Welch
©

further contended that the sense of injustice involves powerful emotions that
ht

often hinder a rational analysis of costs and benefits, and may lead to decisions
ig

that in retrospect seem hasty and impulsive.

yr

TMT provides an opportunity to move beyond a rational analysis of

op

violent conflict to better understand the underlying motives that instigate

violence, even when violence makes little rational sense. In this section we
C

argue that when death is salient, justice motives gain prominence and may

AN EXISTENTIAL PERSPECTIVE ON VIOLENT SOLUTIONS 303

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 304

override utilitarian considerations to the extent that violence may be exercised

even when it is clearly counter to rational self-interest. Previous terror man-
agement research has already demonstrated that MS increases the propensity
to strive for a just world (Hirschberger, 2006; Landau, Solomon, et al., 2004),

A)
and Pyszczynski et al. (2003) argued that part of the appeal of President Bush
and his counter-terrorist polices immediately following 9/11 was that existential

P
(A
fear increased the desire to vanquish evil and restore justice.
In a series of four studies conducted in 2008 (Hirschberger, Pyszczynski,

n
& Ein-Dor, 2009b), we examined whether MS would increase justice motives

io
for violence and whether the motivation to achieve or restore justice would

at
be greater than rational utilitarian considerations. In Study 1, Israeli partici-

ci
pants were randomly assigned to MS and control conditions and then read a

so
description of a missile attack from the Gaza Strip on an Israeli town. Some

As
participants were told that security experts believed that a military incursion
into Gaza was likely to significantly diminish Hamas’s ability to fire more

al
missiles (utility condition). Other participants were told that security experts

ic
believed that an incursion into Gaza would not reduce Hamas’s ability to fire

og
more missiles, nor would it effectively deter Hamas, but it would restore a
ol
sense of justice to the Israeli public (justice condition). All participants were
ch
asked to indicate their support for a military incursion into Gaza. The results
y

revealed that in both the justice and utility conditions, MS led to greater
Ps

support of a military strike. These findings suggest that reminders of death

increase support of violence not only when violence has a clear purpose and
an

is executed to obtain a concrete result but also when violence is considered

ic

to be ineffective but will contribute to feelings of greater justice. These findings

er

seem to imply that people still desire violence even when they understand
Am

that it will not have the desired effect.

These findings indicated that MS leads to greater support of violence
e

for both justice and utility reasons, but they did not indicate whether justice
th

motives are more prominent than utility motives when death is salient. In the
by

next study we developed a scale to measure justice and utility motives for
violence, which would enable us to measure these motives within each subject
11

rather than manipulate them between subjects as in Study 1. In Study 2, we

20

developed the Justice, Utility, and Peace Inventory (JUPI), which consists of
questions favoring violence to restore justice (justice factor; e.g., “A military
©

strike on Gaza will make the Palestinians pay for their crimes”), favoring
ht

violence for utilitarian reasons (utility factor; e.g., “A military strike on Gaza
ig

will reduce missile attacks against Sderot”), or opposing violence altogether

yr

(peace factor; e.g., “A military strike on Gaza hurts the chances for peace”).
op

A factor analysis confirmed the factorial structure of the JUPI, and correlations
between the JUPI and other relevant instruments, such as the Right-Wing
C

Authoritarianism Scale (Altemeyer, 1981) and the Need for Cognitive Closure

304 HIRSCHBERGER AND PYSZCZYNSKI

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 305

Scale (Webster & Kruglanski, 1994), established the construct, convergent,

and discriminant validity of the JUPI.
Based on this second study, we examined the impact of MS on JUPI
factors. Following the MS procedure, participants were instructed to read a

A)
description of a missile attack from Gaza on an Israeli town, as in the first
study. However, half of the participants read that the attack resulted only in

P
(A
some minor damage and no casualties (mild outcome condition), and the
other half read the description used in the first study wherein several people

n
were killed or wounded during the attack (severe outcome condition). Then

io
all participants completed the JUPI. Results revealed that in the mild outcome

at
condition MS had no significant impact on the JUPI. However, in the severe

ci
outcome condition MS led to greater endorsement of the justice factor

so
compared with the control condition. There was no significant effect of MS

As
on the utility and peace items.
The results of this study indicate that when participants are given the

al
opportunity to endorse both justice and utility items, MS has a significant

ic
effect only on the justice items and not on the utility items. Furthermore, the

og
impact of MS on justice motives for violence was evident only when the
ol
outcome of the attack was severe and not when it was mild. Perhaps for MS
ch
to increase justice motives one needs to be in an enraged state of mind that
y

overrides rational considerations.

Ps

To test this possibility, participants in a fourth study first read an essay

that either recommended making decisions based on rational considerations
an

or an essay that argued that decisions are best made on an emotional, intuitive
ic

basis. Then, participants completed the MS procedure and read a description

er

of a panel of experts unanimously concluding that the appropriate response

Am

to a recent terrorist attack in Tel Aviv would be a limited attack targeting

only the person responsible for the terrorist attack, rather than engaging in a
e

full-scale attack against Gaza. Following this description, participants were asked
th

to indicate whether, given the choice, they would favor a limited attack, as
by

recommended, or a full scale attack, and they were asked to indicate to what
extent they felt confident about their choice. Thus, MS, cognitive mode, and
11

decision regarding the type of attack served as the independent variables.

20

The level of confidence they expressed in the decision they made served as
the dependent variable. Results revealed that participants who favored a
©

limited attack were not significantly affected by experimental conditions.

ht

However, among those favoring a full-scale attack, MS led to greater support

ig

of an attack when participants were induced to make decisions based on

yr

emotions and intuitions. The results of this study suggest that confidence in
op

the decision to endorse a full-scale attack that is considered by experts to be

counterproductive is greater under MS conditions but only after participants
C

are induced to think with their gut rather than their mind.

AN EXISTENTIAL PERSPECTIVE ON VIOLENT SOLUTIONS 305

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 306

The prominence of justice and revenge motives in times of war is

demonstrated in the following story. In the Gaza War of 2009 the three
daughters of a Palestinian physician, Dr. Az-a-Din Abu El-Aish, who works
at a large Israeli hospital, were killed while in their home. In his anguish and

A)
grief Dr. Abu El-Aish pleaded in a press conference that the violence be
stopped and that Israelis and Palestinians find a way to live in peace. He also

P
(A
insisted that there was no reason to target his house, as there were no terrorists
shooting from it. A mother of an Israeli soldier angrily interrupted his speech

n
and accused him of harboring weapons or terrorists; otherwise why would

io
anyone bomb his house?

at
At first sight, this rude and insensitive interruption could be seen as the

ci
epitome of coldheartedness. However, from a terror management perspective

so
this behavior is different only in style, but not in essence, from other desper-

As
ate attempts to defend the cultural worldview at all costs. From this point of
view, Dr. Abu-Aish represented for two major reasons a severe threat to

al
the predominant Israeli worldview that supported the war. First, he insisted

ic
that there were no hostile activities taking place in his house, undermining the

og
position that all the casualties of the war were justified. For the accusing
ol
woman (a mother of a soldier), the possibility that some of the killing in Gaza
ch
could not be explained or justified posed an unbearable threat to her belief
y

system. Second, in spite of his devastating loss, the doctor remained steadfast
Ps

in his belief in peace and coexistence, threatening the need to believe that the
other side of the conflict consists only of inhumane, cruel, and savage terrorists.
an

In her almost instinctive attack on the doctor (which she later regretted), the
ic

Israeli woman defended her worldview from the possibility that the war was
er

anything but a just, moral, and necessary clash between the forces of good
Am

and evil. This small episode in a larger war demonstrates the resistance of a
population at war to any information that might undermine consensus that
e

the war is righteous and justified. It is also a powerful demonstration of the

th

workings of the justice motive and the mental acrobatics people will perform to
by

convince themselves that wrongdoings committed on their behalf are justified.

11

The Inevitability of Violence

20

Rallying public support for a war not only requires that people perceive the
©

cause as just and view their group as representing forces of good fighting against
ht

evil, but people also need to believe that there is no alternative and that war
ig

is imminent and unavoidable (see Chapter 17). For example, the American
yr

public perceived the September 11th attacks as the beginning of a violent and
op

inevitable clash with radical Islam, and following the terrorist attacks support
for President Bush and his war on terrorism was high. In contrast, the 2004
C

terrorist attacks on Madrid had the opposite effect: Support for the ruling Partido

306 HIRSCHBERGER AND PYSZCZYNSKI

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 307

Popular party diminished, probably in part because the Spanish population

viewed the Madrid bombings as a reaction to Spanish support of the American
war in Iraq and not as a direct conflict between Spain and radical Islam. In their
view, the terrorist attacks were not inevitable and could be stopped if Spain

A)
changed its foreign policy and withdrew its support from the war in Iraq.
The difference between an imminent and inevitable war from an

P
(A
American perspective, and an undesirable and unnecessary war from a Spanish
perspective, led to diametrically opposed reactions to the attacks in the two

n
countries. For Americans, the seemingly inevitable route to violent conflict

io
left the impression of no alternative options and perhaps elevated terror

at
management mechanisms in the form of a symbolic war against evil. For the

ci
Spanish, who perceived the terrorist attacks as the price they were paying for

so
the ill-considered policy of their government, the elevated death awareness

As
following the attacks made their own vulnerability to harm salient, and
they reacted by replacing their leadership with a more peaceful one. The fact

al
that the Madrid attacks occurred several days before the election probably

ic
contributed to the Spanish sentiment that power was in their hands to avoid

og
an unnecessary violent confrontation.
ol
Recent research lends support to this analysis and indicates that when
ch
violence seems inevitable, MS leads to greater support for violent solutions
y

to conflict. In one study (Hirschberger et al., 2009c, Study 1), MS led to greater
Ps

support of a preemptive strike against Iran after participants read a speech

that was purportedly delivered by an Iranian leader calling for the destruction
an

of the State of Israel and for the continued development of Iran’s nuclear
ic

program. In a similar study (Hirschberger et al., 2009c, Study 3), Israelis who
er

lived in a region attacked by missiles during the 2006 Lebanon War expressed
Am

greater support of a preemptive strike against Hezbollah following MS, but

only if they first read a passage describing Hezbollah as preparing for an
e

imminent war with Israel.

th
by

SELF-PROTECTION MODERATES POLITICAL VIOLENCE

11
20

Previous terror management research has focused primarily on an

abstract and nonimminent threat of death and on symbolic defenses used to quell
©

existential anxieties. However, research conducted on current geopolitical

ht

conflicts and on the populations immersed in these conflicts must consider

ig

the fact that real-life conflict has not only symbolic but also real implications.
yr

That is, as much as people engaged in conflict are concerned about protect-
op

ing their symbolic death-denying structures, so too are they concerned about
their own physical safety. Reminders of personal mortality have been shown
C

to engage motivation to defend the worldview, but it is also likely that they bring

AN EXISTENTIAL PERSPECTIVE ON VIOLENT SOLUTIONS 307

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 308

to mind the possibility of dying in a war. What determines people’s choice of war
or peace when mortality is salient? So far, we have demonstrated that MS leads
to greater support for political violence when social consensus is high, when the
cause is perceived as just, and when conflict seems inevitable. In this section we

A)
argue that MS will lead to less violent intentions when (a) perceived personal
vulnerability to conflict-related injury or death is high, (b) adversary rhetoric

P
(A
raises the possibility of a nonviolent solution, (c) experts advise that violence is
counterproductive, and (d) people are induced to think rationally.

n
To examine the role of perceived vulnerability and adversary intent

io
in moderating the link between MS and political violence, we (Hirschberger

at
et al., 2009c) conducted a series of studies to examine the dynamic interplay

ci
between symbolic terror management defenses, concrete self-protection from

so
physical danger, and their effects on support of violent solutions to political

As
conflict. In Study 1, we focused on the growing tensions between Israel and
Iran over the development of Iran’s nuclear program. Participants completed

al
the typical MS procedure and were then randomly assigned to read either a

ic
conflict-escalating speech by an Iranian leader against Israel and the West or

og
a conflict-de-escalating speech that implies that violence is not inevitable.
ol
Participants then rated their support of extremely violent reactions against
ch
Iran, including a preemptive nuclear strike. Results revealed that in the esca-
y

lating rhetoric scenario condition, MS increased support of extreme violence

Ps

against Iran, but in the de-escalating scenario the opposite pattern was observed,
and MS decreased support of extreme preemptive violence relative to the
an

control condition.
ic

Our interpretation of the results of this study is that when violence

er

seemed imminent and unavoidable, as in the escalation scenario, MS increased

Am

violent motives. However, because MS makes salient not only the need for a
symbolic worldview but also the fact of personal vulnerability to harm and to
e

death, when there were reasons to believe that violence might be averted, the
th

need for personal safety overrode the defense of the symbolic worldview, and
by

the motivation for violence was reduced.

To further test this explanation, we manipulated personal vulnerability
11

in Study 2. Following the MS procedure and a description of the current state

20

of tensions with Iran, participants were randomly assigned to two groups. The
first group was asked to reflect on the possibility that they or their loved ones
©

might be hurt in a future conflict between Iran and Israel. The second group
ht

was asked to reflect on the content of the passage. All participants then
ig

completed the same measure as in Study 1. Results revealed that participants

yr

who reflected only on the content of the passage responded to MS with increased
op

support for preemptive violence. However, participants who reflected on their

personal vulnerability to conflict-related harm responded to MS with decreased
C

support for preemptive violence.

308 HIRSCHBERGER AND PYSZCZYNSKI

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 309

Based on the results of Studies 1 and 2, we attempted to better under-

stand the impact on support of violence of the interaction between perceived
adversary intent and perceived personal vulnerability. In Study 3, rather than
manipulating personal vulnerability as in Study 2, we chose to focus on two

A)
groups of participants that differed in their level of exposure to war-related
violence (matched on other potentially confounding variables). The first group

P
(A
consisted of participants who lived in Northern Israel during the Second
Lebanon War against Hezbollah (summer, 2006) and had directly experienced

n
missile attacks. The second group consisted of participants who lived in other

io
parts of Israel and had never been directly exposed to conflict-related violence.

at
All participants completed the MS procedure and then read either a conflict-

ci
escalating speech or a conflict-de-escalating speech by a leader of Hezbollah,

so
and answered questions on support of a preemptive attack against Hezbollah.

As
Results revealed that for participants in the no-war-exposure group,
MS led to greater support of violence regardless of Hezbollah rhetoric. However,

al
among participants in the war-exposure group, MS led to greater support of

ic
violence in the escalation scenario but led to a reduced support of violence

og
among participants in the de-escalation scenario. These findings suggest that
ol
among persons who feel less vulnerable to conflict-related violence, MS leads
ch
to increased support of violence regardless of whether the adversary’s rhetoric
y

is peaceful or belligerent. Ideological threat trumps practical concerns for one’s

Ps

safety. However, among participants who had experienced the war firsthand, the
influence of MS on support of violence was contingent on adversary rhetoric:
an

When it seemed that war was imminent, it increased support of violence, but
ic

when war seemed avoidable it had the opposite effect and reduced support of
er

violence. When one’s own life and that of one’s family is on the line, people are
Am

more likely to consider nonviolent options when they perceive that violence
can be averted.
e

Similar findings were obtained in the aforementioned research on justice

th

and utility motives for violence (Hirschberger et al., 2009b). In Study 1,

by

which measured whether utilitarian and justice motivations for violence

would increase support of violence under MS conditions, a third group of
11

participants were told that security experts believed that an incursion into
20

Gaza would be counterproductive and was likely only to increase attacks against
Israel (futility of violence condition). Counter to the other conditions, in this
©

group MS significantly reduced support for political violence. In Study 3 of

ht

this research, which examined whether inducing rational or intuitive thinking

ig

influences the impact of MS on support of massive violent retribution, we not

yr

only found that when participants were induced to decide intuitively and
op

emotionally, as previously mentioned, MS led to greater support of a full-fledged

attack against Gaza but we also found that when participants were induced to
C

think rationally, MS reduced support of violent retributions.

AN EXISTENTIAL PERSPECTIVE ON VIOLENT SOLUTIONS 309

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 310

MOVING BEYOND VIOLENCE

The findings described in the previous section, demonstrating that

self-protective concerns may override the need to defend the cultural worldview

A)
and may lead to a reduction in violent motivations under MS conditions,
are encouraging. But they still suggest that humans are defensive violence-prone

P
(A
creatures in one way or another. In this section of the chapter we demonstrate
that focusing on common humanity with others, on compassionate religious

n
values, and on feelings of psychological security enables people to transcend

io
their fear and respond to MS with greater tolerance of others.

at
In a study examining support of political violence in a sample of Israelis

ci
(Hirschberger et al., 2009a), participants were assigned to either an MS con-

so
dition, a pain salience condition, a “Holocaust as a crime against the Jewish

As
people” condition, or a “Holocaust as a crime against humanity” condition
(based on Wohl & Branscombe, 2005). They then answered questions tapping

al
support of violent solutions to conflict. In the MS condition and the “Holocaust

ic
as a crime against the Jewish people” condition, support of violence was signifi-

og
cantly higher compared with the pain salience condition and the “Holocaust
ol
as a crime against humanity” condition. These results suggest that describing
ch
the Holocaust as a crime against the Jewish people led to effects similar to MS.
y

However, framing the Holocaust as a crime against humanity reduced support

Ps

for violent solutions to ethno–political conflict, probably because such portrayal

of the Holocaust induced a sense of common humanity.
an

Another series of studies demonstrated that compassionate religious

ic

teachings moderate the impact of MS on support for violence against an

er

adversary among Christian fundamentalists in the United States and Shiite

Am

Muslims in Iran (Rothschild, Abdollahi, & Pyszczynski, 2008). In both cases,

MS led to greater support of violence against the other after reading a neutral
e

text and even after reading about nonbiblical compassionate values. However,
th

after reading about the compassionate teachings of Jesus or the compassionate

by

teachings from the Koran, MS led participants to significantly reduce their

support of violence against the other.
11

Research has also shown that feelings of security in close relationships

20

reduce support for violence (see Chapters 13 and 14). On this basis, Weise and
his colleagues (2008) primed American participants with thoughts of death
©

and then asked them to visualize a warm and accepting personal relationship.
ht

Results revealed that the effect of MS on support for harsh counterterrorism

ig

measures was reduced after participants thought about close personal relation-
yr

ships. Similarly, in a study conducted on an Israeli sample (Hirschberger,

op

Arias Ben-Tal, Pyszczynski, & Ein-Dor, 2009), participants had to write a brief
comment following MS about a sketch of a woman breastfeeding her child
C

310 HIRSCHBERGER AND PYSZCZYNSKI

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 311

(attachment security condition) or about a sketch of a group of women working

in a field (neutral condition). They were then asked to rate their approval of
military force against the Hamas in Gaza. The effect of MS on support for
military force was significant in the neutral condition but not in the attachment

A)
security condition.

P
(A
CONCLUSIONS

n
io
The conflict in the Middle East that currently involves Israel, Arab

at
nations, Iran, Europe, and the United States is concerned with disputes over

ci
land, water, oil, terrorism, occupation, historical rights, and religious promises.

so
These important reasons notwithstanding, we have suggested in this chapter

As
that human existential concerns also contribute to the perpetuation of this
conflict and the inability to reach a peaceful solution. We have reviewed

al
research that sheds light on the underlying mechanisms that transform

ic
unobtrusive, brief reminders of personal mortality into powerful motives to

og
subdue an enemy. We demonstrated that perceived consensus, belief in a just
ol
cause, and belief that war is inevitable are powerful catalysts that amplify
ch
the impact of mortality concerns on support for violent solutions to conflict.
y

However, we also demonstrated that when people focus on the personal price
Ps

they might pay in a war, when they perceive the adversary as harboring less
malevolent intentions, and when they are induced to think rationally MS has
an

a strikingly different effect and leads to reduced motivation for violence.

ic

Moreover, when people think of their common humanity with others, focus
er

on compassionate religious teachings, or feel secure in their relationships the

Am

effect of MS on support for violence is diminished.

To achieve peace in the Middle East, all involved parties will have to
e

make painful compromises on the concrete issues at stake, but they will also
th

have to change the way they think about their adversaries and transform the
by

symbolic belief structures that contribute to the perpetuation of violence. We

have shown here that deep-seated existential concerns lie at the heart of
11

violent conflicts and render them resistant to change. However, our research
20

also suggests that the path to real change and to peace requires that, as people
and as nations, we confront our deepest fears with courage rather than denying
©

them. Although many political plans for peace have been proposed over the
ht

years, none has been successful. Perhaps in addition to diplomacy and com-
ig

promises aimed at producing a just resolution of the conflict, policies aimed at

yr

changing the psychological forces that promote hatred to those that promote
op

peace are needed to involve in the process of peacemaking people on all sides
of the conflict.
C

AN EXISTENTIAL PERSPECTIVE ON VIOLENT SOLUTIONS 311

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 312

REFERENCES

Abdollahi, A., Henthorn, C., & Pyszczynski, T. (2009). Experimental peace psychol-
ogy: Priming consensus mitigates aggression against out-groups under mortality

A)
salience. Behavioral Sciences of Terrorism and Political Aggression, 2, 30–37.
Altemeyer, B. (1981). Right-wing authoritarianism. Winnipeg, Canada: University of

P
(A
Manitoba Press.
Arndt, J., Greenberg, J., Schimel, J., Pyszczynski, T., & Solomon, S. (2002). To belong

n
or not to belong, that is the question: Terror management and identification

io
with gender and ethnicity. Journal of Personality and Social Psychology, 83, 26–43.

at
doi:10.1037/0022-3514.83.1.26

ci
so
Baumeister, R. F., & Butz, J. (2005). Roots of hate, violence, and evil. In R. J. Sternberg
(Ed.), The psychology of hate (pp. 87–102). Washington, DC: American Psy-

As
chological Association. doi:10.1037/10930-005
Bueno de Mesquita, B. (1988). The contribution of expected utility theory to the study

al
of international conflict. The Journal of Interdisciplinary History, 18, 629–652.

ic
doi:10.2307/204818

og
Castano, E. (2004). In case of death, cling to the ingroup. European Journal of Social
Psychology, 34, 375–384. doi:10.1002/ejsp.211
ol
ch
Clausewitz, C. V. (1976). On war (M. Howard & P. Paret, Eds. &Trans.). Princeton, NJ:
y

Princeton University Press.

Ps

Dechesne, M., Greenberg, J., Arndt, J., & Schimel, J. (2000). Terror management
an

and the vicissitudes of sports fan affiliation: The effects of mortality salience on
optimism and fan identification. European Journal of Social Psychology, 30, 813–835.
ic

doi:10.1002/1099-0992(200011/12)30:6<813::AID-EJSP17>3.0.CO;2-M
er

Florian, V., & Mikulincer, M. (1998). Terror management in childhood: Does death
Am

conceptualization moderate the effects of mortality salience on acceptance

of similar and different others? Personality and Social Psychology Bulletin, 24,
e

1104–1112. doi:10.1177/01461672982410007
th

Greenberg, J., Pyszczynski, T., & Solomon, S. (1997). Terror management theory
by

of self-esteem and cultural worldviews: Empirical assessments and conceptual

refinements. In M. P. Zanna (Ed.), Advances in experimental social psychology
11

(Vol. 29, pp. 61–139). San Diego, CA: Academic Press.

20

Greenberg, J., Pyszczynski, T., Solomon, S., Rosenblatt, A., Veeder, M., Kirkland, S.,
& Lyon, D. (1990). Evidence for terror management theory II: The effects of
©

mortality salience on reactions to those who threaten or bolster the cultural

ht

worldview. Journal of Personality and Social Psychology, 58, 308–318. doi:10.1037/

ig

0022-3514.58.2.308
yr

Greenberg, J., Pyszczynski, T., Solomon, S., Simon, L., & Breus, M. (1994). The role
op

of consciousness and accessibility of death related thoughts in mortality salience

effects. Journal of Personality and Social Psychology, 67, 627–637. doi:10.1037/
C

0022-3514.67.4.627

312 HIRSCHBERGER AND PYSZCZYNSKI

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 313

Greenberg, J., Schimel, J., Martens, A., Pyszczynski, T., & Solomon, S. (2001). Sym-
pathy for the devil: Evidence that reminding whites of their mortality promotes
more favorable reactions to white racists. Motivation and Emotion, 25, 113–133.
doi:10.1023/A:1010613909207

A)
Greenberg, J., Simon, L., Porteus, J., Pyszczynski, T., & Solomon, S. (1995). Evidence
of a terror management function of cultural icons: The effects of mortality salience

P
on the inappropriate use of cherished cultural symbols. Personality and Social

(A
Psychology Bulletin, 21, 1221–1228. doi:10.1177/01461672952111010

n
Greenberg, J., Simon, L., Pyszczynski, T., Solomon, S., & Chatel, D. (1992). Terror

io
management and tolerance: Does mortality salience always intensify negative

at
reactions to others who threaten one’s worldview? Journal of Personality and

ci
Social Psychology, 63, 212–220. doi:10.1037/0022-3514.63.2.212

so
Hirschberger, G. (2006). Terror management and attributions of blame to innocent

As
victims: Reconciling compassionate and defensive responses. Journal of Personality
and Social Psychology, 91, 832–844. doi:10.1037/0022-3514.91.5.832

al
Hirschberger, G., Arias Ben-Tal, K., Pyszczynski, T., & Ein-Dor, T. (2009). Peace and

ic
psychological security: Evidence that security primes moderate the impact of mortality

og
salience on violent inclinations. Unpublished manuscript, Interdisciplinary Center
(IDC) Herzliya, Israel. ol
ch
Hirschberger, G., & Ein-Dor, T. (2006). Defenders of a lost cause: Terror management
y

and violent resistance to the disengagement plan. Personality and Social Psychology
Ps

Bulletin, 32, 761–769. doi:10.1177/0146167206286628

an

Hirschberger, G., Pyszczynski, T., Canetti, D., & Ein-Dor, T. (2009). Never again:
Evidence that reminders of the holocaust elicit support for violent solutions to contem-
ic

porary conflicts. Unpublished manuscript Interdisciplinary Center (IDC) Herzliya,

er

Israel.
Am

Hirschberger, G., Pyszczynski, T., & Ein-Dor, T. (2009a). The commonalities and differ-
ences between group mortality and individual mortality: A terror management perspec-
e

tive on past victimization effects. Unpublished manuscript, Bar-Ilan University,

th

Ramat-Gan, Israel.
by

Hirschberger, G., Pyszczynski, T., & Ein-Dor, T. (2009b). Death and the deliverance
of justice: Mortality salience increases support for vindictive yet inefficient violent
11

retributions. Unpublished manuscript, Interdisciplinary Center (IDC) Herzliya,

20

Israel.
Hirschberger, G., Pyszczynski, T., & Ein-Dor, T. (2009c). Vulnerability and vigilance:
©

Threat awareness and perceived adversary intent moderate the impact of mortality
ht

salience on intergroup violence. Personality and Social Psychology Bulletin, 35,

ig

597–607. doi:10.1177/0146167208331093
yr

Landau, M. J., Johns, M., Greenberg, J., Pyszczynski, T., Martens, A., Goldenberg, J. L.,
op

& Solomon, S. (2004). A function of form: Terror management and structuring

the social world. Journal of Personality and Social Psychology, 87, 190–210.
C

doi:10.1037/0022-3514.87.2.190

AN EXISTENTIAL PERSPECTIVE ON VIOLENT SOLUTIONS 313

12051-17_PT4_CH16-rev2.qxd 5/25/10 11:58 AM Page 314

Landau, M. J., Solomon, S., Greenberg, J., Cohen, F., Pyszczynski, T., Arndt, J., . . .
Cook, A. (2004). Deliver us from evil: The effects of mortality salience and
reminders of 9/11 on support for president George W. Bush. Personality and Social
Psychology Bulletin, 30, 1136–1150. doi:10.1177/0146167204267988

A)
McGregor, H., Lieberman, J. D., Solomon, S., Greenberg, J., Arndt, J., Simon, L., &
Pyszczynski, T. (1998). Terror management and aggression: Evidence that

P
mortality salience motivates aggression against worldview threatening others.

(A
Journal of Personality and Social Psychology, 74, 590–605. doi:10.1037/0022-3514.

n
74.3.590

io
Pyszczynski, T., Abdollahi, A., Solomon, S., Greenberg, J., Cohen, F., & Weise, D.

at
(2006). Mortality salience, martyrdom, and military might: The great Satan

ci
versus the axis of evil. Personality and Social Psychology Bulletin, 32, 525–537.

so
doi:10.1177/0146167205282157

As
Pyszczynski, T., Greenberg, J., & Solomon, S. (1999). A dual process model of defense
against conscious and unconscious death-related thoughts: An extension of

al
terror management theory. Psychological Review, 106, 835–845. doi:10.1037/

ic
0033-295X.106.4.835

og
Pyszczynski, T., Solomon, S., & Greenberg, J. (2003). In the wake of 9/11: The psychology
ol
of terror. Washington, DC: American Psychological Association. doi:10.1037/
10478-000
ch
Rothschild, Z. K., Abdollahi, A., & Pyszczynski, T. (2008). Does peace have a prayer?
y
Ps

The effect of mortality salience, compassionate values, and religious fundamentalism

on hostility toward ourgroups. Manuscript submitted for publication, University
an

of Colorado at Colorado Springs.

Routledge, C., & Arndt, J. (2008). Self-sacrifice as self-defense: Mortality salience
ic

increases efforts to affirm a symbolic immortal self at the expense of the physical
er

self. European Journal of Social Psychology, 38, 531–541. doi:10.1002/ejsp.442

Am

Webster, D., & Kruglanski, A. (1994). Individual differences in need for cognitive
closure. Journal of Personality and Social Psychology, 67, 1049–1062. doi:10.1037
e

/0022-3514.67.6.1049
th

Weise, D. R., Pyszczynski, T., Cox, C., Arndt, J., Greenberg, J., Solomon, S., &
by

Kosloff, S. (2008). Interpersonal politics: The role of terror management and

attachment processes in shaping political processes. Psychological Science, 19,
11

448–455. doi:10.1111/j.1467-9280.2008.02108.x
20

Welch, D. (1993). Justice and the genesis of war. Cambridge, England: Cambridge
University Press. doi:10.1017/CBO9780511521805
©

Wohl, M. J. A., & Branscombe, N. R. (2005). Forgiveness and collective guilt

ht

assignment to historical perpetrator groups depend on level of social category

ig

inclusiveness. Journal of Personality and Social Psychology, 88, 288–303. doi:10.1037/

yr

0022-3514.88.2.288
op
C

314 HIRSCHBERGER AND PYSZCZYNSKI

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 315

17

PA)
THE EMOTIONAL ROOTS

(A
OF INTERGROUP AGGRESSION:

n
io
THE DISTINCT ROLES OF ANGER

at
ci
AND HATRED

so
As
ERAN HALPERIN

al
ic
og
ol
y ch
Ps
an

Prolonged intergroup conflicts are often marked by belligerent actions,

ic

provocative statements, and mutual insults. Notable examples are destruc-

er

tive terrorist attacks, the killing of innocent civilians, the kidnapping of

Am

citizens or soldiers, offensive maneuvers of military forces, and extremely

threatening speeches by political and religious leaders. In most such cases,
e

members of conflicting societies or groups view the conflict through a uni-

th

dimensional, biased lens and therefore perceive the other group’s actions
by

as unjust, unfair, and incompatible with acceptable norms (White, 1970).

Often these perceptions are amplified by blindness to the previous wrong-
11

doings of one’s own group, which are considered by the opponents to

20

be the reasons for their current actions or statements (Bar-Tal & Halperin,
2010).
©

Given these biased appraisals, it is not surprising that intergroup anger

ht

is a pivotal emotion in every conflict (see Chapters 1 and 9 for cognitive analy-
ig

ses of anger arousal). Think for example of the emotional experiences of U.S.
yr

citizens who watched the televised 9/11 terrorist attacks on the World Trade
op

Center almost as they were happening. Surely rage (among other emotions)
was central to their experience. Similarly, it is not hard to imagine the rapid
C

heartbeats, the sweaty palms and faces, and the extreme anger felt by Jewish

315
12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 316

Israelis who sat down to a Passover dinner on March 27, 2002, and suddenly
heard about the destructive suicide bombing at the Park Hotel in Netanya,
which resulted in the deaths of 30 citizens dressed in festive holiday clothing.
The main issue in this chapter is the role played by public anger in the U.S.

A)
decision to invade Iraq or in Israel’s decision to initiate a large-scale military
action against the Palestinians in March of 2002.

P
We know that emotions in general, and anger in particular, play a

(A
pivotal role in driving people to aggressive behavior (see Chapters 6 and 12).

n
It has long been assumed in social psychology that anger is the most pow-

io
erful emotional determinant of aggressive behavior (e.g., Berkowitz, 1993).

at
Accordingly, there is evidence that flare-ups of public anger automatically

ci
lead to widespread support for vengeful aggression (Huddy, Feldman, &

so
Cassese, 2007; Skitka, Bauman, Aramovich, & Morgan, 2006). In reality,

As
however, despite the prevalence of ingroup anger following an outgroup
provocation, public opinion is often divided about the best response (Maoz

al
& McCauley, 2008). Some people clearly advocate an aggressive response,

ic
but others counsel self-restraint. These differences in opinion can be explained,

og
at least in part, by differences in rational calculations concerning the costs
ol
and benefits of aggressive reactions. But I argue in this chapter that rational
ch
analyses are not sufficient to explain public opinion; in addition, a deeper
y

understanding of the emotions involved in prolonged intergroup conflicts

Ps

is needed.
Specifically, I contend that in the context of prolonged intergroup
an

conflict, short-term anger generates public support for an aggressive response

ic

mainly if it is backed by long-term, extremely negative sentiments toward

er

the opposing group. In other words, in the absence of long-term hatred for
Am

the outgroup (and especially when a degree of empathy exists), anger will not
necessarily lead to support for aggression, and in some cases it may even lead
e

to peaceful outcomes (Fischer & Roseman, 2007; Halperin, 2008b). Similarly,

th

long-term hatred for an opponent will not necessarily cause aggression in the
by

absence of an event that triggers explosive anger. That is, only the combined
influences of short-term anger and long-term hatred are likely to produce
11

strong support for large-scale aggression.

The next section provides basic concepts concerning emotions and
20

emotional sentiments within the context of intergroup conflicts. This con-

©

ceptual foundation is then followed by presentation of an appraisal-based model

of emotions in intergroup conflicts. The model highlights relations among
ht

long-term emotional sentiments, short-term emotions, and attitudes about

ig

intergroup violence. I then review recent findings concerning the roles of anger
yr

and hatred in the context of the Middle-East conflict. Finally, I present new
op

data gathered in that context to illustrate the joint effects of anger and hatred
C

on support for intergroup violence.

316 ERAN HALPERIN

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 317

EMOTIONS, SENTIMENTS, AND AGGRESSION:

BASIC CONCEPTS

If we wish to understand the role of emotion in generating support for

A)
aggressive action in intergroup conflicts, we need a working definition of
emotion. For this purpose, I accept William James’s (1884) perspective

P
(A
on emotions as response tendencies. According to this view (as elaborated, for
example, by Frijda, 1986, and Scherer, 1984), emotions are flexible response

n
sequences evoked when a person evaluates a situation as offering important

io
challenges or opportunities (Tooby & Cosmides, 1990). In other words,

at
emotions transform a substantive event into a motive to respond to it in a

ci
particular manner (Zajonc, 1998).

so
Core components of emotion include subjective feelings, bodily changes,

As
facial expressions, and other physiological reactions (Shaver, Schwartz,
Kirson, & O’Connor, 1987). These components help to distinguish emo-

al
tions from other psychological phenomena, such as attitudes or beliefs

ic
(Cacioppo & Gardner, 1999). Following Averill (1994), I view emotions as

og
also comprising stories that help people interpret events and guide their
ol
behavioral reactions. This perspective highlights two additional compo-
ch
nents, cognitive appraisals and emotional goals that guide response tenden-
y

cies, which I view as central to the links between emotions and support for
Ps

intergroup aggression.
Extensive research has illuminated the role of cognitive appraisal in the
an

generation of discrete emotions (e.g., Lazarus, 1991; Roseman, 1984; Scherer,

ic

2004; Shaver et al., 1987; Smith & Ellsworth 1985). It is now well established
er

that in most situations, emotions include an evaluation (either conscious or

Am

unconscious) of the emotion-eliciting stimulus or event that shapes a person’s

behavior during and after the event. In other words, a discrete emotion is likely
e

to motivate support for aggressive behavior only if its associated cognitive

th

appraisals correspond with the psychological preconditions for that kind of

by

behavior.
The emotional behavior that one can observe from outside is thought
11

to be the expression of what Frijda (1986; Frijda, Kuipers, & ter Schure, 1989)
20

called “action tendencies.” Roseman (1984) more specifically distinguished

among actions, action tendencies, and emotional goals. Although actual
©

behavior depends on numerous external factors and cannot usually be

ht

predicted by the arousal of specific emotions alone, the action tendencies

ig

and emotional goals associated with particular emotions are usually present
yr

(Frijda et al., 1989; Roseman, 2002). In this view, a discrete emotion such as
op

anger will cause a person to support aggression against an outgroup only if

he or she believes that a violent response will serve the goals activated by
C

the emotion.

THE EMOTIONAL ROOTS OF INTERGROUP AGGRESSION 317

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 318

Emotions are just one of several kinds of affective responses (Gross, 2007).
For the purposes of this chapter, it is important to distinguish between emo-
tions and sentiments. As I have just explained, emotions are multicomponent
responses to particular kinds of events. Sentiments, by contrast, are enduring

A)
configurations of emotions or emotional predispositions (Arnold, 1960;
Frijda, 1986; Lazarus, 1994). According to this view, an emotional sentiment

P
(A
is a relatively stable emotional disposition toward a person, group, or symbol
(Halperin, 2008a). Emotional sentiments also differ from another kind of

n
affective state, moods, in that moods typically do not have well-defined objects

io
(Oatley, Keltner, & Jenkins, 2006), whereas emotional sentiments do. In my

at
view, as I explain in subsequent sections of this chapter, only a combination

ci
of a specific long-term sentiment and a momentarily triggered discrete emotion

so
toward an outgroup will lead to support for intergroup aggression.

As
The emotional sentiments and emotions relevant to the present discussion
reflect intergroup and not interpersonal phenomena. In recent years, there

al
has been growing interest in the concept of intergroup emotions, emotions that

ic
are felt by individuals as a result of their membership in or identification with

og
a certain group or society and are targeted toward another group (Iyer & Leach,
ol
2008; Mackie, Devos, & Smith, 2000; Smith & Mackie, 2008). Research has
ch
shown that people may experience emotions not only in response to personal
y

experiences and activities but also in response to events that affect other
Ps

members of a group with which they identify or in response to activities in

which other group members have taken part (Mackie et al., 2000; Smith,
an

Seger, & Mackie, 2007; Wohl, Branscombe, & Klar, 2006; Yzerbyt, Dumont,
ic

Wigboldus, & Gordin, 2003). I want to focus here on ways in which events
er

experienced only indirectly (although directly by other group members) elicit

Am

intergroup emotions toward an outgroup and thereby arouse support for

aggressive political action toward that outgroup.
e

I adopt a bottom-up perspective, according to which emotions, atti-

th

tudes, and actions of individuals and groups influence the course of a conflict.
by

At least in democratic societies, public support for aggressive retaliation to

outgroup provocations can encourage political leaders to initiate military
11

actions. This “psychology of the people” is especially important in the context

20

of intractable conflicts, because such conflicts demand extensive societal invest-

ment and are often perceived by both sides as total and zero sum (Bar-Tal, 1998;
©

Kriesberg, 1993).
ht

The nature of these conflicts makes it likely that sentiments will be

ig

established that extend way beyond any immediate provocative event. The
yr

combination of repeated events over a long period of time and repeated

op

rehashing of these events in the media is likely to result in stable group-related

emotional sentiments, such as hatred or despair, toward the opponent and the
C

continuing conflict (Kelman & Fisher, 2003). The interaction between these

318 ERAN HALPERIN

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 319

emotional sentiments and the emotions that arise in response to particular

events, as well as their joint influence on support for aggression, can be under-
stood in terms of emotion-related appraisals.

A)
EMOTIONS AND AGGRESSION IN INTERGROUP CONFLICT:

P
(A
AN APPRAISAL-BASED MODEL

n
The model in Figure 17.1 shows how exposure to a particular conflict-

io
related event is processed and transformed into support for certain political

at
and military actions. The model has been tested among Israeli Jews and found

ci
to be valid and analytically useful (Halperin, 2008a). According to the model,

so
the link between exposure to a conflict-related event and a person’s attitude

As
concerning the required response to the outgroup is mediated by a discrete
emotional response elicited by the event. In turn, this emotional response is

al
influenced by the individual’s long-term emotional sentiments toward the

ic
outgroup (Halperin, Sharvit, & Gross, 2010). The model is sufficiently

og
general to capture processes initiated by various conflict-related events, includ-
ol
ing positive events (e.g., a new peace proposal) or negative ones (e.g., a terror
ch
attack). In the current chapter, however, I use the model to explain how a
y
Ps
an

long-term
ic

emotional
sentiment
er

about the
adversary
Am
e
th

exposure to reactive support

by

cognitive
provocative/ emotions— for
appraisals of
aggressive personal or aggressive
the events
11

event group-based action

20
©
ht

non-affective
factors (e.g.,
ig

ideology,
yr

personality)
op
C

Figure 17.1. An appraisal-based framework for understanding the influence of emo-

tions on support for aggressive action.

THE EMOTIONAL ROOTS OF INTERGROUP AGGRESSION 319

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 320

discrete emotional reaction to outgroup provocation determines whether a

person will support an aggressive response.
The process begins with a particular event or with the arrival of new
information related to the conflict. Although events can be experienced

A)
personally, in most cases they are experienced directly by only a few group
members and transmitted to other group members through leaders, the mass

P
(A
media, or one or more social networks. If a person identifies with the same
group as the directly affected individuals, he or she will experience group-based

n
emotions (Mackie et al., 2000; Smith, 1993; Yzerbyt et al., 2003).

io
Such events will elicit individual and group-based emotions and the

at
ensuing political response tendencies, depending on how the events are

ci
appraised. For example, if a violent act committed by outgroup members against

so
one or more members of the ingroup is appraised as unjust and is accompanied

As
by appraisal of the ingroup as strong, it will induce anger (Halperin, 2008b;
Huddy et al., 2007; Shaver et al., 1987). Hence, the cognitive appraisal of an

al
event is a crucial factor in determining the kind of emotion that will be evoked

ic
by the event.

og
When an event occurs in the context of a long-term conflict, the appraisal
ol
of the event is influenced by three sets of factors: (a) the way in which the event
ch
is presented to the public (i.e., the framing of the event by leaders and media
y

reporters); (b) a broad array of nonaffective factors, such as ideology, personal-

Ps

ity, moral values, and implicit schemas or theories (e.g., Halperin, 2008a;
Sharvit, Halperin, & Rosler, 2008); and (c) most important, a set of long-term
an

emotional sentiments that bias the cognitive appraisal of specific events.

ic

The importance of this last set of factors is explained by the appraisal

er

tendency framework of Lerner and Keltner (2000), according to which each

Am

emotion activates a cognitive predisposition to interpret future events in line

with the central appraisal dimensions that triggered the emotion. For example,
e

long-term external threats to one’s group are likely to cause bouts of fear and
th

insecurity, and these emotions encourage longer-term sentiments that cause

by

group members to be more sensitive to potentially threatening cues.

This analysis suggests that the same event can be appraised differently by
11

individuals who harbor different long-term emotional sentiments about the

20

opponent. Furthermore, although it is reasonable to assume that a hostile, vio-

lent event such as a terror attack or a kidnapping of innocent civilians will elicit
©

strong anger in all members of the victimized society, the nature of that anger
ht

as well as its associated political response tendencies will differ depending

ig

on the level of long-term hatred felt by an individual toward the opponent

yr

(Halperin, Sharvit, et al., 2010). To fully appreciate this analysis, it is necessary

op

to understand the psychological preconditions that support aggressive action

and to understand the correspondence between these preconditions and the
C

unique nature of each kind of emotional phenomenon (hatred or anger).

320 ERAN HALPERIN

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 321

INTERGROUP AGGRESSION: NECESSARY PRECONDITIONS

The most important preconditions for intergroup conflict and intergroup

aggression are a perceived conflict of interest between the groups and percep-

A)
tion of a long-term threat posed by the outgroup (Maoz & McCauley, 2008;
Struch & Schwartz, 1989). Yet in the majority of cases, conflict of interest or

P
(A
even perceived threat alone will not bring about violence in the absence of
an immediate provocation.

n
In terms of the more immediate antecedents of intergroup aggression, it

io
seems that public support for aggression and violence results, in most cases,

at
from a shared belief that the outgroup committed a grievous, unjust action

ci
(or made a provocative, unfair statement) aimed at the ingroup (see Chapter 16).

so
To be an adequate trigger of violence, the action or statement must usually be

As
perceived as an extreme deviation from moral norms and a notable deviation
from the routine (hence acceptable) reciprocal intergroup hostility.

al
A second precondition for supporting aggressive action is perceiving the

ic
ingroup as strong enough to overpower the outgroup in a future military

og
battle. (See Chapter 3 for a similar analysis at the interpersonal level, and
ol
Chapter 16 for a similar analysis at the intergroup level.) This perception of
ch
strength and controllability provides the confidence necessary to undertake
y

aggressive action. Similar to the appraised sense of relative power is a sense

Ps

of willingness or eagerness to accept the risks involved in entering a violent

battle. People who have a dispositional tendency to avoid risks are less likely
an

to support a risky military campaign.

ic

Finally, given the high probable costs (in resources, human lives, and
er

international condemnation) of a belligerent campaign, many people will

Am

support them only as a last resort (see Chapter 16). In other words, most
members who believe that their group is capable of correcting a perceived wrong
e

without resorting to military action (e.g., by relying on diplomacy) will oppose

th

the use of military power, at least for the time being. Moreover, in most cases,
by

public support for an aggressive response to an outgroup’s provocation is highly

dependent on believing that the outgroup action provocation stemmed from
11

a stable, irreversible, and evil disposition central to the outgroup members’

20

character. If this is not the case, other paths of correction may be perceived
as equally beneficial and much less costly than the aggressive one.
©
ht
ig

INTERGROUP ANGER AND HATRED: SIMILAR BUT DIFFERENT

yr
op

I turn now to differences in the cognitive appraisals and behavioral

tendencies associated with anger and hatred. In a series of studies of the
C

Israeli–Palestinian conflict, I attempted to uncover the unique cognitive

THE EMOTIONAL ROOTS OF INTERGROUP AGGRESSION 321

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 322

and motivational components of each of these two affective phenomena

(Halperin, 2008b). In one of the studies (Study 2), 241 Israeli students read
about a particular conflict-related event (e.g., a terror attack), followed by
a short description of the way the protagonist in each account (an ingroup

A)
member) appraised the event. They were then asked to name the emotions
experienced by the protagonist in relation to the event.

P
(A
Both anger and hatred were associated with blaming the outgroup
for the conflict-related event and with appraising the outgroup’s behavior

n
as unpleasant, hurtful, and contrary to the ingroup’s interests and goals.

io
Nevertheless, in line with previous theoretical writings about anger and hatred

at
(e.g., Ben-Zeev, 1992; Royzman, McCauley, & Rozin, 2005; Sternberg, 2003),

ci
there was also a clear distinction between the appraisals associated with each

so
of these phenomena. The negative appraisals associated with anger focused

As
solely on unfairness of the outgroup’s specific action, whereas the appraisals
associated with hatred focused on the nature of the outgroup itself, suggesting

al
that its actions were not aimed at achieving instrumental goals but stemmed

ic
from a malevolent disposition to hurt the ingroup.

og
Important as they may be, cognitive appraisals are just one part of the
ol
emotion process. It is well established that the appraisals associated with each
ch
emotion underpin the unique content of its corresponding motivational
y

tendencies; these action tendencies are a logical response to the “diagnosis”

Ps

of the problematic situation inherent in the appraisals (e.g., Frijda et al., 1989).
For example, fear aroused by seeing a tiger about to attack leads to running
an

away or hiding behind a barrier of some kind; but fear that one is about to miss
ic

the last train home leads to running toward the door of the train. Therefore, in
er

another study (Halperin, 2008b, Study 3) I examined the emotional goals and
Am

response tendencies that are associated with anger and hatred in the context
of intergroup conflict.
e

Specifically, I created a simple conflict-related scenario to be presented

th

to 313 Israeli interviewees in a national-representative sample, asking these

by

people to rank the level of each of several emotional reactions (i.e., hatred,
anger), emotional goals, and response tendencies as reactions to the scenario.
11

In a subsequent analysis of the data, I regressed each of two reactions (i.e., anger,
20

hatred) on the emotional goals and response tendencies while controlling for
sociopolitical variables (e.g., political position, educational level).
©

As expected, there were interesting dissimilarities between anger and

ht

hatred. Hatred was the only emotion associated with the goals of exclusion
ig

(i.e., removing outgroup members from one’s life) and attack (i.e., hurting
yr

the outgroup members). However, anger was associated with the goal of cor-
op

rection (i.e., improving the behavior of outgroup members). Thus, whereas

anger causes people to take action to right a specific wrong (see Chapters 4,
C

13, and 14, for similar analyses of anger), hatred reflects avoidance of any deal-

322 ERAN HALPERIN

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 323

ings with the outgroup based on having given up on the outgroup’s capacity
for change.
The findings regarding the response tendencies associated with hatred
and anger were complex. Naturally, participants who reported relatively high

A)
levels of hatred were more likely to support “denying Palestinians’ basic
political and social rights” as well as “physical and violent actions toward the

P
(A
Palestinians.” It is interesting that although anger was also associated with
support for “physical and violent actions toward the Palestinians,” it was

n
associated as well with “support for educational channels to create perceptual

io
change among Palestinians.” This latter response was obviously a more con-

at
structive approach to changing the Palestinians’ (perceived) unjust behavior.

ci
It seems, based on my research and previous studies by other researchers,

so
that people who feel angry have one key emotional goal: They wish to correct

As
and redirect behavior that they perceive to have been unfair and unjustified.
Researchers have documented that when people feel angry, they believe they

al
have (or at least deserve to have) high control over the situation, are more

ic
willing to take risks, and believe they have the ability to create beneficial

og
changes (Lerner, Gonzalez, Small, & Fischhoff, 2003; Lerner & Keltner, 2001;
ol
Mackie et al., 2000; see also Chapters 3, 4, and 12, this volume). These com-
ch
ponents of anger correspond with the first three psychological preconditions
y

required to support aggression, discussed earlier in this chapter. And indeed,

Ps

previous studies have shown that anger motivates confrontation (Berkowitz,

1993), attack, and even killing the elicitor of anger (Roseman, Wiest, &
an

Swartz, 1994).
ic

Yet observations of the response tendencies associated with anger, as

er

well as findings from a study by Fischer and Roseman (2007), indicate that
Am

although some angry people choose aggression and destruction as responses,

others choose more constructive responses such as education and efforts to
e

achieve reconciliation. An interesting question, therefore, is what leads

th

some angry individuals to choose aggression and others to choose a construc-

by

tive response. As I explain in the next section, the most important factor
moderating the effect of anger on aggressive behavior is the level of long-term
11

hatred. Viewing an outgroup as inherently evil and unlikely to reform under-

20

mines any consideration of seeking constructive change.

©
ht

THE INTERACTION OF ANGER AND HATRED

ig

AND SUPPORT FOR INTERGROUP AGGRESSION

yr
op

As already mentioned, much of the research literature points to the

conclusion that anger is the most important and proximal emotional antecedent
C

of aggressive political action (Cheung-Blunden & Blunden, 2008; Huddy et al.,

THE EMOTIONAL ROOTS OF INTERGROUP AGGRESSION 323

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 324

2007; Lerner et al., 2003; Skitka et al., 2006). However, the findings I have
summarized here suggest that anger interacts with hatred to fuel aggression.
This conclusion is supported by recent studies of associations between inter-
group emotions and political intolerance. Political intolerance is usually defined

A)
as the willingness to deny the political rights of out-group members (Sullivan,
Piereson, & Marcus, 1979). This willingness is commonly viewed as a political

P
(A
version of aggressive intentions toward the outgroup, so it is important for
understanding the emotional underpinnings of intergroup aggression.

n
Two studies, one conducted in the post-9/11 period in the United States

io
(Skitka, Bauman, & Mullen, 2004) and the other in the midst of the second

at
Palestinian uprising in the Middle East (Halperin, Canetti-Nisim, & Hirsch-

ci
Hoefler, 2009), reveal similar patterns of political intolerance related to anger

so
and hatred. In a two-wave national field study (N = 550), Skitka et al. (2004)

As
examined the effects of anger and fear on political intolerance toward Arab
Americans, Muslims, and first-generation immigrants 4 months after the terror

al
attacks of 9/11. They found that anger had no direct influence on political

ic
intolerance and that most of the indirect effect of anger on intolerance was

og
mediated by moral outrage and outgroup derogation. Unfortunately, hatred
ol
of the outgroup was not measured as a separate variable in this study, but
ch
arguably, moral outrage and outgroup derogation might be proxies of hatred.
y

In a more direct examination of the roles played by anger and hatred in

Ps

political intolerance, my colleagues and I (Halperin, Canetti-Nisim, et al., 2009)

conducted four surveys of large representative samples of Israeli Jews. The
an

surveys varied in their design (two panel surveys and two cross sectional surveys),
ic

their context (some of the surveys were conducted during war and others
er

at times of relative calm in the conflict between Israelis and Arabs), and the
Am

political intolerance measures used (“most disliked group” vs. “Palestinians”

as the defined outgroup). The results of the different surveys were essentially
e

the same: Anger toward the outgroup led to political intolerance only if it was
th

supported by hatred. Furthermore, in some cases anger in the absence of hatred

by

was associated with reduced political intolerance.

In search of even more direct support for my growing understanding of
11

the interplay of anger and hatred, I examined survey-based data sets in which
20

explicit measures of support for military action were among the dependent
variables. The first analysis was based on a nationwide representative telephone
©

survey conducted among Israeli Jews in March of 2008. The sample consisted of
ht

781 Jewish-Israeli respondents (403 women, 378 men). The reported political
ig

affiliation of participants in the survey mirrored the distribution in Israeli

yr

society at the time: 37.3% were identified as rightists, 45.2% as centrists, and
op

17.6% as leftists.
Among other items, the questionnaire included one assessing long-term
C

hatred of Palestinians, two items (i.e., anger and rage) capturing anger about

324 ERAN HALPERIN

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 325

Palestinian actions, and two items concerning support for aggressive actions
against Palestinians (i.e., “In a time of significant Palestinian threat, Israel should
use unconventional warfare”; “Only by using force can you achieve anything
in the Middle East”). For all of these items, participants used a 6-point response

A)
scale ranging from 1 (not at all) to 6 (very much).
Support for aggressive action toward Palestinians was regressed on hatred,

P
(A
anger, and the interaction of the two while controlling for sociopolitical variables
(e.g., political position, educational level). No main effect of anger (β = −.10)

n
or hatred (β = −.17) was found, but their interaction was significant (β = .40,

io
p < .05). As shown in Figure 17.2, anger was associated with heightened

at
support for aggressive actions only in the presence of long-term hatred.

ci
Given the provocative nature of the findings, replication of the results

so
was called for. Therefore, I examined data from another nationwide survey

As
conducted in Israel during August and September of 2008. This sample
included 500 respondents, of whom 48.4% were men and 51.6% women.

al
The mean age was 45.5 years (SD = 16.49), and 41% of the respondents

ic
defined themselves as rightists, 28.9% as centrists, and 18.3% as leftists. Socio-

og
demographic measures and scales assessing anger and hatred were similar to
ol
the ones used in the previous survey. Support for military action was measured
ch
in a slightly different way, with two items: support for initiating a large military
y

operation of the Israeli army and support for using severe military action, even
Ps

if it meant harming innocent civilians (α = .74).

an

5
ic
er

4.5
Am

4.17
4
e
Aggressive Action

th

3.5 3.27
3.24
by

Low Hatred
3 3.09 High Hatred
11

2.5
20

2
©
ht

1.5
ig
yr

1
op

Low Anger High Anger

Figure 17.2. Regression lines depicting the anger × hatred interaction affecting sup-
C

port for aggressive action toward Palestinians in the first nationwide survey (N = 781).

THE EMOTIONAL ROOTS OF INTERGROUP AGGRESSION 325

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 326

Again, support for aggressive action against Palestinians was regressed

on hatred, anger, and their interaction, while controlling for sociopolitical
variables. Except for minor dissimilarities in the effects of sociopolitical variables
on support for aggressive action, the findings replicated those of the first survey.
There was no main effect of anger (β = −.04) or hatred (β = −.11) on support

A)
for aggression, but their interaction was significant (β = .45, p < .01). Again, only

P
(A
the combination of anger and hatred was significantly associated with support
for aggressive action.

n
io
at
CONCLUSION

ci
so
Protracted intergroup conflicts inherently include mutual provocations,

As
violent responses, and diverse transgressions by both parties to the conflict.
As a result, members of societies involved in such conflicts experience frequent

al
episodes of anger toward the outgroup, its leaders, and its representatives.

ic
This anger, in various degrees, is an inevitable element of every intergroup

og
conflict. The research literature on emotions occurring in the context of conflict
ol
shows that anger evoked by an antagonist’s behavior is highly associated with
ch
support for retaliatory aggression (Huddy et al., 2007; Skitka et al., 2006).
y

In turn, widespread public support for aggression may legitimize and even
Ps

encourage political decision makers to initiate large-scale military action

(Zaller, 1994). This dynamic can easily escalate into a vicious cycle of reciprocal
an

violence that causes a tragic loss of life on both sides of the conflict. However,
ic

the theoretical model and the bulk of empirical data presented in this chapter
er

suggest that the seemingly direct link between anger and the cycle of violence
Am

is not unavoidable.
The link between anger and support for aggression is dependent on the
e

level of hatred one feels for the outgroup. Anger in the absence of hatred, even
th

though it may have been triggered by appraisals of the outgroup’s behavior as

by

painful and unjust, can arouse a desire to correct the wrongdoing without
necessarily harming or destroying the outgroup (Fischer & Roseman, 2007).
11

Anger that emerges on top of prolonged hatred, often combined with an

20

appraisal of the outgroup as inherently evil and unlikely to change, is likely to

arouse support for aggressive retaliation unchecked by empathy with members
©

of the outgroup.
ht

Another implication of this analysis is that despite prolonged hatred of

ig

the outgroup, members of the ingroup may not be motivated to engage in

yr

large-scale aggression if their goals are not shifted in that direction by a spike
op

in anger. This is because hatred by itself lacks the anger-related sense of

power and self-righteousness that makes aggression seem possible as well as
C

justified.

326 ERAN HALPERIN

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 327

The analysis of anger’s role that I have sketched here suggests a need for
studies that identify moderating factors that can reduce the destructive effects
of hatred and anger. Possibilities include other emotional sentiments such as
despair, empathy, or compassion, but also relevant are personality factors

A)
(e.g., authoritarianism, need for structure, attachment security), implicit theo-
ries about individuals and groups, adherence to moral values, socioeconomic

P
status, and political ideology.

(A
I hope I have made readers more aware of the role of intergroup emotions

n
as influences on public opinion, political decision making, and intergroup vio-

io
lence. More research is needed on ways to down-regulate “destructive emo-

at
tions” (Goleman, 2003; Halperin, Sharvit, & Gross, in press), up-regulate

ci
more prosocial emotions (Mikulincer & Shaver, 2009), and identify other

so
emotional sentiments and dispositions that might be used to foster more

As
favorable interactions with various emotions. In the long run, basic and
applied research along these lines may contribute to a reduction in retalia-

al
tory aggression.

ic
og
REFERENCES ol
ch
Arnold, M. B. (1960). Emotion and personality (Vols. 1 & 2). New York, NY: Columbia
y
Ps

University Press.
Averill, J. R. (1994). In the eyes of the beholder. In P. Ekman & R. J. Davidson (Eds.),
an

The nature of emotion: Fundamental questions (pp. 7–15). New York, NY: Oxford
University Press.
ic
er

Bar-Tal, D. (1998). Societal beliefs in times of intractable conflict: The Israeli case.
The International Journal of Conflict Management, 9, 22–50. doi:10.1108/eb022803
Am

Bar-Tal, D., & Halperin, E. (2009). Overcoming psychological barriers to peace mak-
ing: The influence of mediating beliefs about losses. In M. Mikulincer & P. R.
e

Shaver (Eds.), Prosocial motives, emotions, and behavior The better angels of our
th

nature (pp. 431–448). Washington, DC: American Psychological Association.

by

Bar-Tal, D., & Halperin, E. (2010). Socio-psychological barriers to conflict

resolution. In D. Bar-Tal (Ed.), Intergroup conflicts and their resolution: Social
11

psychological perspective. New York, NY: Psychology Press.

20

Ben-Zeev, A. (1992). Anger and hate. Journal of Social Philosophy, 2, 85–110. doi:10.1111/
j.1467-9833.1992.tb00295.x
©

Berkowitz, L. (1993). Aggression: Its causes, consequences, and control. Philadelphia, PA:
ht

Temple University Press.

ig

Cacioppo, J. T., & Gardner, W. L. (1999). Emotion. Annual Review of Psychology, 50,
yr

191–214. doi:10.1146/annurev.psych.50.1.191
op

Cheung-Blunden, V., & Blunden, B. (2008). The emotional construal of war: Anger,
fear, and other negative emotions. Peace and Conflict, 14, 123–150. doi:10.1080/
C

10781910802017289

THE EMOTIONAL ROOTS OF INTERGROUP AGGRESSION 327

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 328

Fischer, A. H., & Roseman, I. J. (2007). Beat them or ban them: The characteristics
and social functions of anger and contempt. Journal of Personality and Social
Psychology, 93, 103–115. doi:10.1037/0022-3514.93.1.103
Frijda, N. H. (1986). The emotions. Cambridge, England: Cambridge University Press.

A)
Frijda, N. H., Kuipers, P., & ter Schure, E. (1989). Relations among emotion, appraisal,

P
and emotional action readiness. Journal of Personality and Social Psychology, 57,

(A
212–228. doi:10.1037/0022-3514.57.2.212
Goleman, D. (2003). Destructive emotions: How can we overcome them? New York, NY:

n
Bantam Dell.

io
at
Gross, J. J. (Ed.). (2007). Handbook of emotion regulation. New York, NY: Guilford

ci
Press.

so
Halperin, E. (2008a). Emotional barriers to peace: Negative emotions and public opinion
about the peace process in the Middle East. Paper presented at the 31st annual meet-

As
ing of the International Society of Political Psychology, Paris, France.

al
Halperin, E. (2008b). Group-based hatred in intractable conflict in Israel. The Journal

ic
of Conflict Resolution, 52, 713–736. doi:10.1177/0022002708314665

og
Halperin, E., Canetti-Nisim, D., & Hirsch-Hoefler, S. (2009). Emotional antecedents
ol
of political intolerance: The central role of group-based hatred. Political Psychology,
30, 93–123. doi:10.1111/j.1467-9221.2008.00682.x
ch
Halperin, E., Sharvit, K., Gross, J. J. (2010). Emotions and emotion regulation in
y
Ps

conflicts. In D. Bar-Tal (Ed.), Intergroup conflicts and their resolution: Social

psychological perspective. New York, NY: Psychology Press.
an

Huddy, L., Feldman, S., & Cassese, E. (2007). On the distinct political effects of
ic

anxiety and anger. In A. Crigler, M. MacKuen, G. Marcus, & W. R. Neuman

er

(Eds.), The dynamics of emotion in political thinking and behavior (pp. 202–230).
Chicago, IL: Chicago University Press.
Am

Iyer, A., & Leach, C. W. (2008). Emotion in inter-group relations. European Review
of Social Psychology, 19, 86–125. doi:10.1080/10463280802079738
e
th

James, W. (1884). What is an emotion? Mind, os-IX, 188–205. doi:10.1093/mind/

os-IX.34.188
by

Kelman, H. C., & Fisher, R. J. (2003). Conflict analysis and resolution. In D. O. Sears,
11

L. Huddy, & R. Jervis (Eds.), Oxford handbook of political psychology (pp. 315–353).
New York, NY: Oxford University Press.
20

Kriesberg, L. (1993). Intractable conflict. Peace Review, 5, 417–421.

©

Lazarus, R. S. (1991). Emotion and adaptation. New York, NY: Oxford University Press.
ht

Lazarus, R. S. (1994). Universal antecedents of the emotions. In P. Ekman & R. J.

ig

Davidson (Eds.), The nature of emotion: Fundamental questions (pp. 163–171).

yr

New York, NY: Oxford University Press.

op

Lerner, J. S., Gonzalez, R. M., Small, D. A., & Fischhoff, B. (2003). Effects of fear and
anger on perceived risk of terrorism: A national field experiment. Psychological
C

Science, 14, 144–150.

328 ERAN HALPERIN

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 329

Lerner, J. S., & Keltner, D. (2000). Beyond valence: Toward a model of emotion-
specific influences on judgment and choice. Cognition and Emotion, 14, 473–493.
doi:10.1080/026999300402763
Lerner, J. S., & Keltner, D. (2001). Fear, anger, and risk. Journal of Personality and Social

A)
Psychology, 81, 1 146-1159.
Mackie, D. M., Devos, T., & Smith, E. R. (2000). Intergroup emotions: Explaining

P
(A
offensive actions in an intergroup context. Journal of Personality and Social
Psychology, 79, 602–616. doi:10.1037/0022-3514.79.4.602

n
Maoz, I., & McCauley, C. (2008). Threat, dehumanization and support for retaliatory-

io
aggressive policies in asymmetric conflict. The Journal of Conflict Resolution, 52,

at
93–116. doi:10.1177/0022002707308597

ci
so
Mikulincer, M., & Shaver, P. R. (Eds.). (2009). Prosocial motives, emotions, and
behavior: The better angels of our nature. Washington, DC: American Psychological

As
Association.
Oatley, K., Keltner, D., & Jenkins, J. M. (2006). Understanding emotions (2nd ed.).

al
Maldin, MA: Blackwell.

ic
og
Roseman, I. J. (1984). Cognitive determinants of emotions: A structural theory. In
P. R. Shaver (Ed.), Review of personality and social psychology (Vol. 5, pp. 11–36).
Beverly Hills, CA: Sage. ol
ch
Roseman, I. J. (2002). Dislike, anger, and contempt: Interpersonal distancing, attack,
y

and exclusion emotions. Emotion Researcher, 16, 5–6.

Ps

Roseman, I. J., Wiest, C., & Swartz, T. S. (1994). Phenomenology, behaviors, and
an

goals differentiate discrete emotions. Journal of Personality and Social Psychology,

67, 206–221.
ic

Royzman, E. B., McCauley, C., & Rozin, P. (2005). From Plato to Putnam: Four
er

ways to think about hate. In R. J. Sternberg (Ed.), The psychology of hate

Am

(pp. 3–35). Washington, DC: American Psychological Association. doi:10.1037/

10930-001
e

Scherer, K. R. (1984). Emotion as a multicomponent process: A model and some

th

cross-cultural data. In P. R. Shaver (Ed.), Review of personality and social psychology

by

(Vol. 5, pp. 37–63). Beverly Hills, CA: Sage.

Scherer, K. R. (2004). Feelings integrate the central representation of appraisal-driven
11

response organization in emotion. In A. S. R. Manstead, N. Frijda, & A. Fischer

20

(Eds.), Feeling and emotions: The Amsterdam Symposium (pp. 136–157). Cambridge,
England: Cambridge University Press.
©

Sharvit, K., Halperin, E., & Rosler, N. (2008). Forces of stability and change in prolonged
ht

occupation: Image threat, emotions, and justifying beliefs. Paper presented at the
ig

31st annual meeting of the International Society of Political Psychology, Paris,

yr

France.
op

Shaver, P. R., Schwartz, J., Kirson, D., & O’Connor, C. (1987). Emotion knowledge:
Further exploration of a prototype approach. Journal of Personality and Social
C

Psychology, 52, 1061–1086. doi:10.1037/0022-3514.52.6.1061

THE EMOTIONAL ROOTS OF INTERGROUP AGGRESSION 329

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 330

Skitka, L. J., Bauman, C. W., Aramovich, N. P., & Morgan, G. C. (2006). Confronta-
tional and preventative policy responses to terrorism: Anger wants a fight and
fear wants “them” to go away. Basic and Applied Social Psychology, 28, 375–384.
doi:10.1207/s15324834basp2804_11

A)
Skitka, L. J., Bauman, C. W., & Mullen, E. (2004). Political tolerance and coming
to psychological closure following the September 11, 2001, terrorist attacks: An

P
integrative approach. Personality and Social Psychology Bulletin, 30, 743–756.

(A
doi:10.1177/0146167204263968

n
Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in emotions.

io
Journal of Personality and Social Psychology, 48, 813–838. doi:10.1037/0022-3514.

at
48.4.813

ci
Smith, E. R. (1993). Social identity and social emotions: Toward a new conceptual-

so
ization of prejudice. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition

As
and stereotyping: Interactive processes in group perception (pp. 297–315). San Diego,
CA: Academic Press.

al
Smith, E. R., & Mackie, D. M. (2008). Intergroup emotions. In M. Lewis, J. M.

ic
Haviland-Jones, & L. F. Barrett (Eds.), Handbook of emotions (3rd ed., pp. 428–439).

og
New York, NY: Guilford Press.
ol
Smith, E. R., Seger, C. R., & Mackie, D. M. (2007). Can emotions be truly group level?
ch
Evidence for four conceptual criteria. Journal of Personality and Social Psychology,
y

93, 431–446. doi:10.1037/0022-3514.93.3.431

Ps

Sternberg, R. J. (2003). A duplex theory of hate: Development and application to

terrorism, massacres, and genocide. Review of General Psychology, 7, 299–328.
an

doi:10.1037/1089-2680.7.3.299
ic

Struch, N., & Schwartz, S. H. (1989). Intergroup aggression: Its predictors and
er

distinctness from in-group bias. Journal of Personality and Social Psychology, 56,
Am

364–373. doi:10.1037/0022-3514.56.3.364
Sullivan, J. L., Piereson, J., & Marcus, G. E. (1979). An alternative conceptualization
e

of political tolerance: Illusory increases 1950s–1970s. The American Political Science

th

Review, 73, 781–794. doi:10.2307/1955404

by

Tooby, J., & Cosmides, L. (1990). The past explains the present: Emotional adaptations
and the structure of ancestral environments. Ethology and Sociobiology, 11, 375–424.
11

doi:10.1016/0162-3095(90)90017-Z
20

White, R. K. (1970). Nobody wanted war: Misperception in Vietnam and other wars.
Garden City, NY: Anchor Books.
©

Wohl, M. J. A., Branscombe, N. R., & Klar, Y. (2006). Collective guilt: Emotional
ht

reactions when one’s group has done wrong or been wronged. European Review
ig

of Social Psychology, 17, 1–37. doi:10.1080/10463280600574815

yr

Yzerbyt, V., Dumont, M., Wigboldus, D., & Gordin, E. (2003). I feel for us: The
op

impact of categorization and identification on emotions and action ten-

dencies. The British Journal of Social Psychology, 42, 533–549. doi:10.1348/
C

014466603322595266

330 ERAN HALPERIN

12051-18_CH17-rev2.qxd 5/25/10 11:58 AM Page 331

Zajonc, R. B. (1998). Emotions. In D. Gilbert, S. T. Fiske, & G. Lindzey (Eds.),

The handbook of social psychology (4th ed., Vol. 1, pp. 591–632). Boston, MA:
McGraw-Hill.
Zaller, J. (1994). Strategic politicians, public opinion, and the Gulf crisis. In W. L.

A)
Bennett & D. L. Paletz (Eds.), Taken by storm: The media, public opinion, and U.S.
foreign policy in the Gulf War (pp. 250–276). Chicago, IL: University of Chicago

P
Press.

(A
n
io
at
ci
so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

THE EMOTIONAL ROOTS OF INTERGROUP AGGRESSION 331

C
op
yr
ig
ht
©
12051-18_CH17-rev2.qxd

20
11
5/25/10

by
th
e
11:58 AM

Am
er
ic
Page 332

an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 333

18

A)
P
TENSION AND HARMONY

(A
IN INTERGROUP RELATIONS

n
io
at
ci
TAMAR SAGUY, NICOLE TAUSCH, JOHN F. DOVIDIO,

so
FELICIA PRATTO, AND PURNIMA SINGH

As
al
ic
og
ol
y ch
Ps
an

Aggression between groups is an extreme manifestation of inter-

ic

group tension (see Chapters 16 and 17). This form of tension can result in
er

the killing and displacement of innocent people, the destruction of infra-

Am

structure and property, and setback in the process of reconciliation. There-

fore, tension between groups has traditionally been viewed as categorically
e

harmful. In sharp contrast, harmony between groups, which encompasses

th

positive perceptions and orientations toward outgroup members, is often

by

viewed as ideal. These categorical views of harmony as “good” and tension

as “bad” have guided much thinking in the social sciences, as reflected in
11

research on conflict resolution (Hewstone & Cairns, 2001), prejudice

20

reduction (see Paluck & Green, 2009, for a review), and intergroup contact
(Pettigrew, 1998). The primary goal in these areas of research has typically
©

been to eliminate tension between groups and create harmony.

ht

Nevertheless, not all forms of harmony are necessarily beneficial to

ig

intergroup relations (Jackman, 1994), and not all forms of intergroup tension
yr
op
C

Preparation of this manuscript was supported by NSF Grant # BCS-0613218.

333
12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 334

are inevitably destructive (Varshney, 2002). Particularly when the broader

social context is taken into account, viewing harmony as good and tension
as bad can be an oversimplification of the dynamics of intergroup relations.
The context that often characterizes relations between groups is one of

A)
systematic inequality (e.g., Jews and Arabs in Israel, Blacks and Whites in the
United States or South Africa, Muslims and Hindus in India, Muslims and

P
(A
non-Muslims in the United Kingdom). As exemplified by historical changes,
societal tension can be quite constructive in creating pressure for changing

n
unequal structures in the direction of equality. Social struggles such as those

io
initiated by the U.S. civil rights movement or by the African National Congress

at
in South Africa were aimed at disturbing apparent amity in order to “open

ci
the eyes of blind prejudice” (King, 1964, p. 35) and incite action for social

so
justice (see Sharp, 2005).

As
Because the disruption of harmony is often necessary for stimulating
change toward social equality, efforts to create harmony between groups may in

al
fact work to undermine these potential changes (see Wright & Lubensky, 2009).

ic
Thus, interventions to create societal harmony can potentially have the ironic

og
effect of sustaining negative patterns of intergroup relations, particularly
ol
those pertaining to group-based inequality. In the present chapter, we consider
ch
research on the effects of harmony-inducing strategies on psychological factors
y

related to changes in group-based inequality.

Ps

Our emphasis is on the consequences of intergroup contact, which is

psychology’s most influential and frequently used strategy for creating
an

harmony, or improving relations, between groups (Dovidio, Gaertner, &

ic

Kawakami, 2003; Pettigrew, 1998). In the majority of cases, intergroup

er

contact is implemented to improve relations between members of groups

Am

that are unequal in the social system. Although it is well established that
intergroup contact relates to improved attitudes toward outgroup members
e

(Pettigrew & Tropp, 2006), little is known about the effect of contact and
th

improved attitudes on factors that relate to social change toward equality.

by

Our central argument is that because harmony undermines tension, and

because tension is crucial for social change, a by-product of contact inter-
11

ventions may be the paradoxical stability of unequal social systems that

20

interveners hoped to change.

This chapter is organized into three sections. In the first section, we
©

focus on intergroup tension and discuss factors that can bring about change
ht

toward equality. In the second section, we consider how intergroup contact, as

ig

a strategy that aims to increase harmony, can work to undermine these factors.
yr

Finally, we present empirical evidence that converges on the conclusion that

op

contact may reduce the potential for change. We discuss the implications of
these results for intergroup relations in general and for intergroup hostility
C

and aggression more specifically.

334 SAGUY ET AL.

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 335

PREDICTORS OF SOCIAL CHANGE TOWARD EQUALITY

Our starting point is the assumption that social inequality is likely to

remain stable insofar as group members do not oppose it. Therefore, when

A)
considering antecedents of social change, we focus on psychological processes
that predict group members’ motivations to resist the status quo. We begin by

P
(A
explicating who the likely agents of change are in a hierarchical system and
then turn to consider the more specific psychological processes involved in

n
the development of social actions that can increase equality.

io
at
Who Are the Likely Agents of Change? The Psychologies

ci
of Advantaged and Disadvantaged Group Members

so
As
Without exception, societies are hierarchically organized such that at least
one group controls a greater share of valued resources (e.g., political power,

al
land, economic wealth, educational opportunities, access to health care) than

ic
do other groups (Jackman, 2001; Sidanius & Pratto, 1999). Group-based

og
hierarchy is reflected in almost every aspect of social life, from poverty rates
ol
and school attrition rates to prison sentences and mortality rates—favoring
ch
members of advantaged groups over those in disadvantaged groups (Feagin,
y

2006; Jackman, 2001; Smooha, 2005; Ulmer & Johnson, 2004). Moreover,
Ps

members of disadvantaged groups, compared with members of advantaged

groups, are subjected to discrimination and social injustice in a wide range of
an

domains, such as interviewing for jobs and being quoted a price for a house or
ic

a car (e.g., Ayres, 1991; Bertrand & Mullainathan, 2004).

er

Group-based disparities accompanied by differential social treatment

Am

produce divergent daily realities for members of advantaged and disadvantaged

groups. Whereas members of disadvantaged groups find many doors to economic
e

opportunities closed, suffer higher rates of unemployment, have a difficult

th

time climbing the social ladder, and experience legal authorities as a source
by

of violence and intimidation, advantaged group members experience far

more economic security, opportunities to advance, and social acceptance.
11

Thus, although the world may seem fair, hospitable, and inviting to members of
20

advantaged groups, it often appears to be unjust, dangerous, and exclusionary to

members of disadvantaged groups (Jones, Engelman, Turner, & Campbell,
©

2009). These different realities and divergent perspectives form the basis for
ht

different motives and goals that advantaged and disadvantaged group members
ig

often have regarding the status quo, as explained by prominent theories in

yr

sociology and social psychology.

op

In his pioneering work on race relations, for example, Herbert Blumer

(1958) proposed that membership in an advantaged group is associated with a
C

need to protect the dominance of one’s group, which in turn can account for

TENSION AND HARMONY IN INTERGROUP RELATIONS 335

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 336

phenomena such as prejudice, discrimination, and opposition to egalitarian poli-

cies (Bobo & Hutchings, 1996). Similarly, according to realistic group-conflict
theory (LeVine & Campbell, 1972), motives to advance or protect the interests
of one’s group are at the root of intergroup processes. Whereas members of

A)
advantaged groups are motivated to defend the existing social arrangements that
benefit them, members of disadvantaged groups are motivated to gain more

P
(A
resources, and thus to change the status quo so that their groups’ position in the
social hierarchy can improve. In line with these theories, research on social dom-

n
inance theory (Pratto, Sidanius, Stallworth, & Malle, 1994; Sidanius & Pratto,

io
1999) demonstrates that individuals who hold more power in society tend to

at
view the social hierarchy as natural and even necessary, whereas members of dis-

ci
advantaged groups are more likely to see the hierarchy as in need of change.

so
Taken together, these views suggest that because members of advantaged

As
groups benefit both practically (Bobo & Hutchings, 1996) and psychologically
(Tajfel & Turner, 1979) from hierarchical social arrangements, they are not

al
likely to oppose the status quo. Undoubtedly, because they hold positions of

ic
power there is much value in the attempts to reduce potential discriminatory

og
behavior, which can occur in a variety of interpersonal contexts (e.g., in hiring
ol
decisions, when shopping in retail stores). Nevertheless, expecting advantaged
ch
group members to mobilize for structural-level change fails to consider the
y

psychology associated with their dominant group position. In contrast, because

Ps

they can benefit from changes toward greater equality, members of disadvantaged
groups are generally more likely to be motivated to challenge the status quo
an

and to raise related societal tension so that their group position can improve
ic

(Saguy, Dovidio, & Pratto, 2008; Scheepers, Spears, Doosje, & Manstead,
er

2006; Tajfel & Turner, 1979). Thus, for both practical and psychological rea-
Am

sons, collective action aimed at advancing more equality typically arises from
disadvantaged groups.
e

Broadly speaking, collective action can be either normative or non-

th

normative (Martin, 1986; Wright, Taylor, & Moghaddam, 1990). Normative

by

collective action refers to behaviors such as protesting at legal rallies and

organizing or signing a petition to government officials. Nonnormative action
11

refers to more violent types of protest such as destroying property and, in extreme
20

cases, harming innocent individuals. In the current program of research, we

built on previous work on collective action, which has largely centered on
©

understanding predictors of normative action (see van Zomeren, Postmes, &

ht

Spears, 2008, for a meta-analysis). Whereas there are parallels between predic-
ig

tors of normative and nonnormative action, a fuller account of nonnormative

yr

forms of action can be found in research on political violence and terrorism

op

(Lemieux, 2006; Moghaddam & Marsella, 2004). We next consider the con-
ditions that can increase or attenuate tendencies to partake in normative
C

forms of collective action.

336 SAGUY ET AL.

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 337

Predictors of Collective Action

Although members of disadvantaged groups are likely to desire social

change in the direction of greater equality, there are cases in which this

A)
motivation is relatively subdued. A large body of research suggests that a key
(but not sole) determinant of collective action on the part of disadvantaged

P
(A
group members is their recognition that intergroup inequality exists and that
they are unjustly disadvantaged within the hierarchical system (see van Zomeren

n
et al., 2008, for a meta-analysis). This awareness of unjust inequality depends

io
on group members engaging in intergroup comparisons (Tajfel & Turner, 1979;

at
Wright & Lubensky, 2009); that is, they must perceive themselves as members

ci
of a group and compare their group’s standing with that of other groups. In

so
the context of social inequality, intergroup comparisons are likely to result in

As
a sense of relative disadvantage or unjust deprivation within the social system
(Walker & Smith, 2002).

al
What makes a sense of deprivation a powerful motivator for action is

ic
the emotions associated with it, such as anger and resentment (van Zomeren,

og
Spears, Fischer, & Leach, 2004). Beyond the recognition of deprivation that
ol
can incite these emotions, members of disadvantaged groups need to perceive
ch
an external source as responsible for their situation, typically the privileged
y

outgroup and/or the social system (Simon & Klandermans, 2001; Smith &
Ps

Walker, 2009). It is unlikely that collective resistance would occur if indi-

viduals attributed their own disadvantage to themselves or to their ingroup.
an

Alternatively, when an external, outgroup-related source is identified, negative

ic

emotions such as anger can be directed toward it and can also arouse and guide
er

relevant action tendencies.

Am

Because intergroup comparisons are central to the process of recognizing

inequality, factors that reduce the tendency to engage in such comparisons
e

can also affect the way group members view their social standing. Research on
th

perceived discrimination is consistent with this idea in showing that factors

by

that reduce the emphasis on group boundaries and on unjust social structures
reduce disadvantaged group members’ perceptions of discrimination. For
11

example, when primed with beliefs about individual merit, women who were
20

discriminated against were more likely to attribute their disadvantage to their

own internal characteristics than to unfair treatment (McCoy & Major, 2007).
©

In another study, participants who were rejected from a desirable position

ht

despite their good performance were significantly less likely to act on behalf of
ig

their group when the illegitimacy of the decision was not emphasized (compared
yr

with when it was emphasized; Wright, 1997). Thus, although relative depriva-
op

tion is often an objective marker of any social hierarchy (Jackman, 1994, 2001),
factors that blur group boundaries can reduce the extent to which disadvantaged
C

group members are aware of the inequality and of their position in it.

TENSION AND HARMONY IN INTERGROUP RELATIONS 337

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 338

Because recognizing ingroup disadvantage and attributing it to external

factors (typically related to the outgroup) are crucial for mobilizing members
of an oppressed group to act for change, reduced awareness of inequality can
have consequences for social hierarchy. Consistent with this idea, group-

A)
based hierarchy has been theorized to remain stable insofar as disadvantaged
group members deny their disadvantage or make internal attributions for their

P
(A
subordination (Jackman, 1994; Sidanius & Pratto, 1999). As we argue next,
the emphasis on commonalities that is part of positive contact can reduce

n
group members’ awareness of group-based inequality. Because members of dis-

io
advantaged groups are likely agents of social change, their reduced awareness

at
can ultimately stabilize the existing social hierarchy.

ci
so
As
INTERGROUP CONTACT AND SOCIAL CHANGE

al
The fundamental premise of contact theory is that intergroup relations can

ic
improve by bringing members of opposing groups together under conditions

og
that involve institutional support, cooperation, equal status, and potential for
ol
personal acquaintance (Allport, 1954; Pettigrew & Tropp, 2006). The theory
ch
has stimulated a large body of research and has been applied in a wide variety
y

of settings, notably those marked by intergroup aggression (see Brown &

Ps

Hewstone, 2005, for a review). For example, interventions informed by

contact theory were implemented for reducing tension between Catholics
an

and Protestants in Northern Ireland (Hughes, 2001) and between Jews and
ic

Palestinians in the Middle East (Maoz, 2004).

er

Although intergroup contact is the most widely used framework for

Am

improving intergroup relations, the association between contact and change

toward equality has received surprisingly little research attention. In some cases,
e

researchers have examined the effects of contact on support for egalitarian

th

policies, yet the majority of this work has focused on stated support for such
by

policies on the part of members of advantaged groups, not on actual egalitarian

behavior (e.g., Pettigrew, Wagner, & Christ, 2007). Moreover, the way contact
11

affects disadvantaged group members’ perceptions of inequality, which is a

20

chief instigator of social change toward equality (Simon & Klandermans, 2001),
has rarely been examined.
©

A fundamental component of positive intergroup contact is the focus on

ht

cooperative, commonality-focused aspects. Indeed, contact is typically opera-

ig

tionalized as structured intergroup encounters that emphasize commonalities

yr

between the groups (e.g., Gaertner, Mann, Murrell, & Dovidio, 1989) or as
op

cross-group friendships (Pettigrew, 1998). Psychologically, then, one of the

main ways in which positive intergroup contact operates is by reducing the
C

salience of the psychological distinction between the ingroup and the outgroup

338 SAGUY ET AL.

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 339

(Gaertner & Dovidio, 2000). In the case of cross-group friendships, a highly

personalized form of contact, the goal is to reduce the salience of group bound-
aries so that people are perceived as individuals and not as representatives of
their group (Miller, 2002). In the case of commonality-based encounters, the

A)
goal is to redefine original group boundaries so that members of both groups are
perceived as part of a single more inclusive category (Gaertner & Dovidio, 2000).

P
(A
Either way, weakening the salience of original group boundaries is likely
to reduce the extent to which group members are focusing on group differences,

n
including those pertaining to differences in resources and power. Thus, the

io
focus on commonalities is likely to reduce awareness of structural inequality,

at
which is a crucial component in motivating disadvantaged group members to

ci
advance change toward equality. The fact that this reduced awareness is likely

so
to be coupled with positive attitudes toward the outgroup, makes the reduced

As
motivation for change even more likely, because disadvantaged group members
are less likely to attribute unfair or unjust acts to members of the outgroup

al
(see Smith & Walker, 2009). In the next section we present empirical evidence

ic
suggesting that intergroup contact can undermine the conditions necessary

og
for social change toward equality (see Wright & Lubensky, 2009, for a similar
argument). ol
y ch
Ps

RESEARCH SUPPORT
an

We propose that beyond improving attitudes, the commonality-focused

ic

nature of intergroup contact can also affect the way group members view
er

social inequality. Past research has established that forms of contact that are
Am

considered “optimal,” which are often operationalized as cross-group friend-

ships or as commonality-focused encounters, improve attitudes toward the
e

outgroup (Pettigrew & Tropp, 2006). Because of the blurring of intergroup

th

boundaries, we propose that these forms of contact can also reduce awareness
by

of intergroup inequality. Reduced awareness of inequality along with positive

attitudes toward the advantaged group can inflate disadvantaged group mem-
11

bers’ perceptions of the fairness of the advantaged group, thus encouraging

20

optimism about the prospects of equality between the groups. Because recog-
nition of inequality and external attribution of disadvantage are necessary
©

motivators of collective action, we would expect this optimism ultimately to

ht

reduce motivation to advance social change.

ig

To test these ideas we conducted a laboratory experiment and two field

yr

studies concerning the effect of intergroup contact on perceptions and motives

op

related to social inequality. In the first study we experimentally manipulated

group membership and related power (advantaged vs. disadvantaged) and type
C

of contact (commonality-focused vs. differences-focused). According to

TENSION AND HARMONY IN INTERGROUP RELATIONS 339

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 340

our theorizing, a focus on commonalities, compared with differences, should

reduce awareness of group-based inequality and can thereby affect the way
disadvantaged group members perceive their social standing. Therefore, we
were particularly interested in whether, for disadvantaged group members,

A)
commonality-focused contact would reduce awareness of inequality in gen-
eral and of relative disadvantage more specifically. In the subsequent two field

P
(A
studies we examined whether a similar effect would occur among members of
naturally occurring disadvantaged groups: Arabs in Israel and Muslims in

n
India. In these field studies, we further tested whether perceptions of group-based

io
inequality predict reduced motivation for social change, our main outcome

at
of interest.

ci
In the laboratory study (Saguy, Tausch, Dovidio, & Pratto, 2009, Study 1),

so
power between two groups to which members were randomly assigned was

As
manipulated by giving the advantaged group the opportunity to assign extra
course credits to the two groups. Before the advantaged group members

al
allocated the credits, members of both groups interacted under the influence

ic
of instructions to focus on either intergroup commonalities or differences.

og
Consistent with prior research, commonality-focused interactions, compared
ol
with differences-focused interactions, produced more positive intergroup
ch
attitudes in both advantaged and disadvantaged group members. In addition,
y

however, commonality-focused contact was related to reduced awareness of

Ps

the inequality between the groups. Moreover, members of the disadvantaged

group expected the advantaged group to distribute the credits in a more
an

equitable fashion following commonality-focused rather than differences-

ic

focused interactions. Mediation analysis revealed that this expectation was

er

explained by improved attitudes and reduced awareness of inequality. Thus,

Am

commonality-focused contact strengthened perceptions of outgroup fairness,

reflected in expectations for equality between the groups.
e

However, when the disadvantaged group members’ expectations were

th

compared with the advantaged group’s actual allocations, there was a sig-
by

nificant discrepancy. As the disadvantaged group members anticipated,

advantaged group members were substantially biased against them in the
11

allocation of credits after differences-focused contact. However, although

20

disadvantaged group members expected a more equal distribution of credits

after commonality-focused contact, advantaged group members were just
©

as biased in this condition as in the difference-focused interaction. This effect

ht

is consistent with extensive research showing that advantaged groups are

ig

motivated to maintain their power (Blumer, 1958; Sidanius & Pratto, 1999).
yr

Furthermore, as argued by Dixon and colleagues (Durrheim & Dixon, 2004),

op

whereas members of advantaged groups may support equality in principle,

which is a likely attitudinal outcome of favorable contact, they may still not
C

act to create equality in practice. Thus, the results from the laboratory experi-

340 SAGUY ET AL.

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 341

ment support our predictions by showing that for members of disadvantaged

groups, commonality-focused contact resulted in inaccurate perception of
their social standing.
Although the experimental nature of this study permitted causal analysis,

A)
the intergroup relations were situation-based and short-lived. They may
therefore not reflect processes that occur in more naturalistic intergroup

P
(A
contexts. For instance, members of disadvantaged groups might initially be
overly optimistic regarding outgroup fairness but not show the same effect

n
following repeated intergroup experiences. Our goal in the next set of studies

io
was to examine whether among members of real groups, intergroup contact

at
relates to reduced awareness of inequality and increased perceptions of out-

ci
group fairness. The naturalistic contexts also enabled us to explore whether these

so
outcomes were related to reduced support for social change among members

As
of disadvantaged groups.
Participants in the second study (Saguy et al., 2009, Study 2) were Arabs

al
in Israel, a national minority that suffers notable disadvantage compared with

ic
Jews (e.g., in academic achievement, income, political power; Smooha, 2005).

og
We examined the statistical associations among friendships with Jews (a type
ol
of positive contact that is particularly likely to involve a focus on common-
ch
alities; Aron et al., 2004), attitudes toward Jews, awareness of inequality, and
y

perceptions of Jews as fair. We further measured Arabs’ support for social

Ps

change toward equality (e.g., by asking them the degree to which they support
legislation guaranteeing equal work opportunities for Arabs). Drawing on col-
an

lective action research (Simon & Klandermans, 2001; van Zomeren et al., 2008)
ic

we expected that both reduced awareness of inequality and positive outgroup

er

orientations (i.e., improved attitudes toward Jews and viewing Jews as fair)
Am

would undermine disadvantaged group members’ support for change toward

equality.
e

Consistent with the results of our laboratory experiment, more positive

th

contact with Jews was associated with more positive attitudes toward Jews and
by

with reduced awareness of inequality between Jews and Arabs. In addition,

improved attitudes were associated with increased perceptions of Jews as fair.
11

Moreover, and consistent with our theorizing, both perceptions of Jews as fair
20

and reduced awareness of inequality were associated with reduced support

for social change. Thus, through its effects on the way disadvantaged group
©

members viewed the social inequality and members of the other group, contact
ht

was associated with a decrease in support for social change. The overall model
ig

testing the proposed links between variables fit the data well and better than
yr

alternative models (see Figure 18.1).

op

Results of a third study (Tausch, Saguy, & Singh, 2009) replicated these
findings, this time among Muslims in India. Muslims in India are disadvantaged
C

compared with other minorities and, most notably, compared with Hindus, the

TENSION AND HARMONY IN INTERGROUP RELATIONS 341

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 342

Outgroup
β = .31** attitude β = .13
β = .38**
b=.67** Support
β = .14† Outgroup β = –.24**
Positive contact for

A)
fairness
social change
b=.67** β = .11

P
β = 7**
β = –.18*
Inequality

(A
as just

n
io
Figure 18.1. Path model examining the mediated relationships among contact and
support for social change. The beta value between “positive contact” and “outgroup

at
fairness” = .14†; the beta value between “positive contact” and “inequality as

ci
just” = .22**.

so
*p < .05. **p < .01. †p < .09

As
dominant majority group. Having more Hindu friends was related to improved

al
attitudes toward Hindus but also to reduced awareness of inequality between

ic
Muslims and Hindus. In addition, these outcomes predicted stronger percep-

og
tions of Hindus as fair, which in turn were related to weaker collective action
ol
tendencies (measured as intentions to participate in various actions that could
ch
improve the position of Muslims in India).
y

Taken together, results from the field studies and the laboratory exper-
Ps

iment converge to suggest that experiences of positive, commonality-focused

contact can affect the way disadvantaged group members view the social
an

inequality and their own disadvantage. These perceptions, which are parts of
ic

an overly optimistic view of intergroup relations, can reduce motivation to

er

challenge existing social inequality. These effects are consistent with our
Am

theorizing that the harmony created by contact, reflected in improved attitudes

between groups, can paradoxically undermine the potential for social action
e

that might bring about an increase in equality.

th

More evidence for the connection between intergroup contact and

by

reduced motivation for social change is currently emerging. Wright and

Lubensky (2009), who examined data from a survey of African American
11

and Latino/Latina students at a predominantly White university, obtained

20

compatible findings. Positive intergroup contact was associated with more

favorable attitudes toward Whites and with less support for collective
©

action. In addition, mediation analyses revealed that the negative effect of

ht

contact on collective action was partly a result of reduced ethnic identifi-

ig

cation. Similarly, Dixon, Durrheim, and Tredoux (2007), in a survey study

yr

of Black and White respondents in South Africa, found that more positive
op

intergroup contact was associated with Black South Africans’ decreased

support for social policies that might enhance racial equality. Black South
C

Africans who reported more positive contact with Whites were less support-

342 SAGUY ET AL.

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 343

ive of compensatory policies promoting the interests of Blacks in education

and employment. Thus, consistent with our findings from the Arab sample,
positive contact predicted less support to legal remedies for inequality offered
by institutional authorities.

A)
These recent findings, coupled with our own studies reported here, suggest
that the harmony created by intergroup contact can undermine tension that

P
(A
might bring about change toward equality. This suggests that commonality-
focused contact can be used as a strategy, or a tool, for maintaining the hierarchy

n
that favors advantaged groups (see Jackman, 1994). If indeed harmony reduces

io
the potential for change, advantaged group members might use it to appease

at
members of disadvantaged groups. An additional implication of the current

ci
findings, which is highly relevant to the issue of intergroup aggression, is

so
that in the long term, the effect of commonality-focused contact might not

As
necessarily lead to a reduction in intergroup conflict, particularly if (or when)
disadvantaged group members are disillusioned and realize that their expec-

al
tations of fairness are not met. In the final two sections, we discuss these issues,

ic
among others, and offer potential solutions to some of the problems raised by

og
the reported findings.
ol
y ch

DISCUSSION AND FUTURE DIRECTIONS

Ps

At the beginning of this chapter, we advocated a more refined and less

an

categorical view of intergroup harmony as good and intergroup tension as bad.

ic

We proposed that tension can be adaptive for intergroup relations, because it

er

can foster social change that brings about greater equality and justice between
Am

groups. For this reason, promoting harmony could be maladaptive because, by

reducing potentially constructive tension, it might indirectly reinforce existing
e

inequality.
th

In general, our research supports this view by showing that the harmony
by

encouraged by contact may not translate into greater equality between groups.
The orientations of both disadvantaged and advantaged groups can contribute
11

to this “irony of harmony.” For disadvantaged group members, harmony-

20

inducing strategies can turn attention away from social inequities and cause
members of the advantaged group to be perceived as fair. Both of these outcomes
©

can lead disadvantaged group members to relax their efforts to promote change
ht

toward equality. For members of advantaged groups, improved attitudes as a

ig

result of contact do not necessarily translate to more egalitarian intergroup

yr

behavior. This finding is consistent with research showing that changes in

op

attitudes produced by contact may reflect mainly a commitment to equality

as a principle rather than equality as a reality (Durrheim & Dixon, 2004;
C

but see, Pettigrew et al., 2007).

TENSION AND HARMONY IN INTERGROUP RELATIONS 343

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 344

Future research might fruitfully examine processes related to intergroup

harmony that might impede egalitarian actions on the part of advantaged
groups members. One possibility is that when intergroup relations are positive,
members of advantaged groups feel less urgency to make changes in the social

A)
system. Their knowledge that the disadvantaged group members are content
can reduce concerns about social inequality and make actual egalitarian

P
(A
behavior seem less necessary. A related possibility is that members of advantaged
groups may strategically try to create forms of positive social relations to sustain

n
the status quo (Jackman, 1994). Indeed, the threat of change in the system is

io
reduced if the likely agents of change (i.e., members of the disadvantaged

at
group) see less need for action. Therefore, from the point of view of those who

ci
wish to sustain the hierarchy, commonality-focused contact can be functional,

so
if somewhat deceptive.

As
Future research might also examine the effects of commonality-focused
contact over time, especially the effects on members of the disadvantaged group.

al
As we have demonstrated, disadvantaged group members may at first become

ic
optimistic, through positive contact, about intergroup relations and the

og
advantaged group’s benevolence. In the longer term, however, disadvantaged
ol
group members may become “disillusioned,” particularly if they are repeatedly
ch
appeased without achieving any real change. In addition, their situation might
y

actually get worse rather than improve if members of the advantaged group
Ps

feel no pressure to be truly egalitarian. This turn of events is likely to produce

distrust, disillusionment, and strong resentment, which might fuel more extreme
an

forms of conflict. This possibility points to the potential fragility of harmony

ic

created by commonality-focused contact and to the need to develop inter-

er

ventions that would help achieve a more stable and genuine improvement in
Am

intergroup relations.
e
th

CONCLUSION AND IMPLICATIONS

by

Harmony-inducing strategies can have obvious and relatively immedi-

11

ate positive consequences for intergroup attitudes. Nevertheless, our work

20

suggests that these benefits can turn out to be superficial and may even impede,
in the long run, constructive societal changes. Practical efforts to improve
©

intergroup relations might therefore profit from moving beyond questions

ht

about how to create harmony and eliminate tension to studying ways to cre-
ig

ate harmony without eliminating constructive tension. This requires an

yr

approach that views harmony and tension between groups as processes that
op

can coexist. Rather than overlooking the fact that groups operate in a system
of asymmetrical power, group-based differences, particularly those pertain-
C

ing to power, can be acknowledged alongside a focus on commonalities. This

344 SAGUY ET AL.

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 345

can allow the pressure for social change to continue while providing an
avenue for communication and exchange (see Halabi, 2004). In addition, to
the extent that recognizing both commonality and group-based differences
and inequalities helps people to extend moral principles across group lines,

A)
advantaged group members may more readily recognize the illegitimacy of
group-based disparities and become motivated to respond fairly in a way that

P
(A
supersedes separate group interests.

n
io
REFERENCES

at
ci
so
Allport, G. W. (1954). The nature of prejudice. Reading, MA: Addison-Wesley.
Aron, A., McLaughlin-Volpe, T., Mashek, D., Lewandowski, G., Wright, S. C., &

As
Aron, E. N. (2004). Including others in the self. In W. Stroebe & M. Hewstone

al
(Eds.), European review of social psychology (Vol. 15, pp. 101–132). Hove, England:

ic
Psychology Press.

og
Ayres, I. (1991). Fair driving: Gender and race discrimination in retail car negotiations.
Harvard Law Review, 104, 817–872. doi:10.2307/1341506
ol
ch
Bertrand, M., & Mullainathan, S. (2004). Are Emily and Greg more employable than
Lakisha and Jamal? The American Economic Review, 94, 991–1013. doi:10.1257/
y

0002828042002561
Ps

Blumer, H. (1958). Race prejudice as a sense of group position. Pacific Sociological

an

Review, 1, 3–7.
ic

Bobo, L., & Hutchings, V. L. (1996). Perceptions of racial group competition: Extend-
er

ing Blumer’s theory of group position to a multiracial social context. American

Sociological Review, 61, 951–972. doi:10.2307/2096302
Am

Brown, R., & Hewstone, M. (2005). An integrative theory of intergroup contact. In

M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 37, pp. 255–343).
e
th

San Diego, CA: Academic Press.

Dixon, J., Durrheim, K., & Tredoux, C. (2007). Intergroup contact and attitudes
by

toward the principle and practice of racial equality. Psychological Science, 18,
11

867–872. doi:10.1111/j.1467-9280.2007.01993.x
Dovidio, J. F., Gaertner, S. L., & Kawakami, K. (2003). The contact hypothesis: The
20

past, present, and the future. Group Processes & Intergroup Relations, 6, 5–21.
©

doi:10.1177/1368430203006001009
ht

Durrheim, K., & Dixon, J. A. (2004). Attitudes in the fiber of everyday life:
Desegregation and the discourse of racial evaluation. American Psychologist, 59,
ig

626–636. doi:10.1037/0003-066X.59.7.626
yr
op

Feagin, J. R. (2006). Systematic racism: A theory of oppression. New York, NY: Routledge.
Gaertner, S. L., & Dovidio, J. F. (2000). Reducing intergroup bias: The common ingroup
C

identity model. Philadelphia, PA: Psychology Press.

TENSION AND HARMONY IN INTERGROUP RELATIONS 345

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 346

Gaertner, S. L., Mann, J., Murrell, A., & Dovidio, J. F. (1989). Reducing intergroup bias:
The benefits of recategorization. Journal of Personality and Social Psychology, 57,
239–249. doi:10.1037/0022-3514.57.2.239
Halabi, R. (Ed). (2004). Israeli and Palestinian identities in dialogue: The school for peace

A)
approach. New Brunswick, NJ: Rutgers University Press.
Hewstone, M., & Cairns, E. (2001). Social psychology and intergroup conflict. In

P
(A
D. Chirot & M. E. P. Seligman (Eds.), Ethnopolitical warfare: Causes, consequences,
and possible solutions (pp. 319–342). Washington, DC: American Psychological

n
Association. doi:10.1037/10396-020

io
Hughes, J. (2001). Constitutional reform in Northern Ireland: Implications for

at
community relations policy and practice. The International Journal of Conflict

ci
Management, 12, 154–173. doi:10.1108/eb022854

so
Jackman, M. R. (1994). The velvet glove. Berkeley, CA: University of California

As
Press.
Jackman, M. R. (2001). License to kill: Violence and legitimacy in expropriative social

al
relations. In J. T. Jost & B. Major (Eds.), The psychology of legitimacy: Emerging

ic
perspectives on ideology, justice, and intergroup relations (pp. 437–467). New York,

og
NY: Cambridge University Press.
ol
Jones, J. M., Engelman, S., Turner, C., & Campbell, S. (2009). Worlds apart: The
ch
universality of racism leads to divergent social realities. In S. Demoulin, J. P. Leyens,
y

& J. F. Dovidio (Eds.), Intergroup misunderstandings: Impact of divergent social

Ps

realities (pp. 117–134). Philadelphia, PA: Psychology Press.

King, M. L. (1964). Why we can’t wait. New York, NY: Penguin Books.
an

Lemieux, A. F. (2006). Social psychological approaches to understanding and pre-

ic

venting terrorism: Toward an interdisciplinary perspective. Journal of Security

er

Education, 1, 75–84. doi:10.1300/J460v01n04_07

Am

LeVine, R. A., & Campbell, D. T. (1972). Ethnocentrism: Theories of conflict, ethnic

attitudes, and group behavior. New York, NY: Wiley.
e

Maoz, I. (2004). Coexistence is in the eye of the beholder: Evaluating intergroup

th

encounter interventions between Jews and Arabs in Israel. Journal of Social Issues,
by

60, 437–452. doi:10.1111/j.0022-4537.2004.00119.x

Martin, J. (1986). The tolerance of injustice. In J. M. Olson, C P. Herman, & M. P
11

Zanna (Eds), Relative deprivation and social comparison: The Ontario Symposium
20

(Vol. 4, pp. 217–242). Hillsdale, NJ: Erlbaum.

McCoy, S. K., & Major, B. (2007). Priming meritocracy and the psychological justi-
©

fication of inequality. Journal of Experimental Social Psychology, 43, 341–351.

ht

doi:10.1016/j.jesp.2006.04.009
ig

Miller, N. (2002). Personalization and the promise of contact theory. Journal of Social
yr

Issues, 58, 387–410. doi:10.1111/1540-4560.00267

op

Moghaddam, F. M., & Marsella, A. J. (2004). Understanding terrorism: Psychosocial

roots, causes and consequences. Washington, DC: American Psychological Asso-
C

ciation. doi:10.1037/10621-000

346 SAGUY ET AL.

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 347

Paluck, E. L., & Green, D. P. (2009). Prejudice reduction: What works? A critical look
at evidence from the field and the laboratory. Annual Review of Psychology, 60,
339–367. doi:10.1146/annurev.psych.60.110707.163607
Pettigrew, T. F. (1998). Intergroup contact theory. Annual Review of Psychology, 49,

A)
65–85. doi:10.1146/annurev.psych.49.1.65
Pettigrew, T. F., & Tropp, L. (2006). A meta-analytic test of intergroup contact

P
(A
theory. Journal of Personality and Social Psychology, 90, 751–783. doi:10.1037/
0022-3514.90.5.751

n
Pettigrew, T. F., Wagner, U., & Christ, O. (2007). Who opposes immigration?

io
Comparing German with North American findings. Du Bois Review, 4, 19–39.

at
Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance

ci
orientation: A personality variable predicting social and political attitudes.

so
Journal of Personality and Social Psychology, 67, 741–763. doi:10.1037/0022-

As
3514.67.4.741
Saguy, T., Dovidio, J. F., & Pratto, F. (2008). Beyond contact: Intergroup contact in the

al
context of power relations. Personality and Social Psychology Bulletin, 34, 432–445.

ic
doi:10.1177/0146167207311200

og
Saguy, T., Tausch, N., Dovidio, J. F., & Pratto, F. (2009). The irony of harmony:
ol
Intergroup contact can produce false expectations for equality. Psychological
ch
Science, 20, 114–121. doi:10.1111/j.1467-9280.2008.02261.x
y

Scheepers, D., Spears, R., Doosje, B., & Manstead, A. S. R. (2006). Diversity in
Ps

in-group bias: Structural factors, situational features, and social functions. Journal
of Personality and Social Psychology, 90, 944–960. doi:10.1037/0022-3514.90.6.944
an

Sharp, G. (2005). Waging nonviolent struggle: 20th century practice and 21st century
ic

potential. Boston, MA: Porter Sargent.

er

Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social hierarchy
Am

and oppression. New York, NY: Cambridge University Press.

Simon, B., & Klandermans, B. (2001). Politicized collective identity: A social–
e

psychological analysis. American Psychologist, 56, 319–331. doi:10.1037/0003-066X.

th

56.4.319
by

Smith, H. J., & Walker, I. (2009). Feeling relative deprivation: The rocky road
from comparisons to actions. In U. Wagner, L. R., Tropp, G. Finchilescu, &
11

C. Tredoux (Eds.), Improving intergroup relations: Building on the legacy of

20

Thomas F. Pettigrew (pp.227–243). Malden, MA: Wiley-Blackwell. doi:10.1002/

9781444303117.ch15
©

Smooha, S. (2005). Index of Arab–Jewish relations in Israel 2004. Haifa, Israel: The
ht

Jewish–Arab Center, University of Haifa.

ig

Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In

yr

W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations

op

(pp. 33–48). Monterey, CA: Brooks/Cole.

Tausch, N., Saguy, T., & Singh, P. (2009). Contact between Muslims and Hindus:
C

Benefits and limitations. Unpublished Manuscript.

TENSION AND HARMONY IN INTERGROUP RELATIONS 347

12051-19_CH18-rev2.qxd 5/25/10 11:59 AM Page 348

Ulmer, J., & Johnson, B. D. (2004). Sentencing in context: A multilevel analysis.

Criminology, 42, 137–178. doi:10.1111/j.1745-9125.2004.tb00516.x
van Zomeren, M., Postmes, T., & Spears, R. (2008). Toward an integrative social
identity model of collective action: A quantitative research synthesis of three

A)
sociopsychological perspectives. Psychological Bulletin, 134, 504–535. doi:10.1037/
0033-2909.134.4.504

P
(A
van Zomeren, M., Spears, R., Fischer, A. H., & Leach, C. W. (2004). Put your money
where your mouth is! Explaining collective action tendencies through group-based

n
anger and group efficacy. Journal of Personality and Social Psychology, 87, 649–664.

io
doi:10.1037/0022-3514.87.5.649

at
Varshney, A. (2002). Ethnic conflict and civic life: Hindus and Muslims in India. New

ci
Haven, CT: Yale University Press.

so
Walker, I., & Smith, H. J. (2002). Relative deprivation: Specification, development, and

As
integration. Cambridge, England: Cambridge University Press.
Wright, S. C. (1997). Ambiguity, social influence, and collective action: Generating

al
collective protest in response to tokenism. Personality and Social Psychology

ic
Bulletin, 23, 1277–1290. doi:10.1177/01461672972312005

og
Wright, S. C., & Lubensky, M. (2009). The struggle for social equality: Collective
ol
action versus prejudice reduction. In S. Demoulin, J. P. Leyens, & J. F. Dovidio
ch
(Eds.), Intergroup misunderstandings: Impact of divergent social realities (pp. 291–310).
y

New York, NY: Psychology Press.

Ps

Wright, S. C., Taylor, D. M., & Moghaddam, F. M. (1990). Responding to member-

ship in a disadvantaged group: From acceptance to collective protest. Journal of
an

Personality and Social Psychology, 58, 994–1003. doi:10.1037/0022-3514.58.6.994

ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

348 SAGUY ET AL.

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 349

A)
P
(A
CONSEQUENCES

n
io
OF AGGRESSION:

at
ci
THE VICTIM PERSPECTIVE

so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C
C
op
yr
ig
ht
©
20
12051-20_PT5-CH19-rev2.qxd

11
by
5/25/10

th
e
Am
12:00 PM

er
ic
an
Page 350

Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 351

19

A)
P
INFLUENCE OF VIOLENCE AND

(A
AGGRESSION ON CHILDREN’S

n
io
PSYCHOLOGICAL DEVELOPMENT:

at
ci
TRAUMA, ATTACHMENT,

so
AND MEMORY

As
al
SHEREE L. TOTH, LATONYA S. HARRIS, GAIL S. GOODMAN,

ic
AND DANTE CICCHETTI

og
ol
y ch
Ps
an

The focus of this section of the volume—the consequences of violence

ic

and aggression—is particularly relevant to child maltreatment, because

er

aggression and violence directed at children have important implications for

Am

their psychosocial development and long-term life outcomes. One issue of

particular interest is the effect of aggression and violence on children’s mem-
e

ory, a topic that has recently received attention from researchers. Scores of
th

studies spanning several decades of research have shown unequivocally that

by

children traumatized by aggression and violence within their families,

including by sexual abuse, physical abuse, or emotional abuse, are at height-
11

ened risk of poor mental health outcomes. Only more recently have the
20

implications for children’s memory been the focus of study. Understanding

the effects of child maltreatment on central psychological processes such as
©

memory is important for clinical, legal, and societal interventions and for
ht

basic psychological theory. As Zigler (1998) insisted over a decade ago, it is

ig
yr
op

This material is based in part on work supported by the National Science Foundation under Grants
No. 0545413 and 0851420 to Dr. Gail S. Goodman.
C

Any opinions, findings, and conclusions or recommendations expressed in this material are those of
the author(s) and do not necessarily reflect the views of the National Science Foundation.

351
12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 352

important for researchers to recognize their obligations to help formulate

policies and practices that benefit the populations they study. Given chil-
dren’s vulnerability and the enduring effects of maltreatment experiences,
Zigler’s claim is particularly relevant.

A)
In this chapter, we consider how violence can affect children’s socio-
emotional and cognitive development, including their memories. Rather than

P
(A
providing a comprehensive review, we direct our attention to investigations
guided by a developmental psychopathology perspective. In so doing, we pro-

n
vide an overview of the sequelae of child maltreatment in relation to emotion

io
regulation, attachment, and psychopathology. We describe how these psycho-

at
social phenomena provide important perspectives from which to understand

ci
so
maltreated individuals’ memories for emotional or traumatic events, an issue
of special importance to both the field of developmental psychology and the

As
legal system.

al
ic
og
PATHOGENIC RELATIONAL ENVIRONMENTS
AND CHILD MALTREATMENT ol
ch
Child maltreatment, in the form of aggression and violence against chil-
y
Ps

dren, is one of the most profound failures of a social environment to support

normal development (Chapter 9). Maltreating families do not provide many
an

of the experiences that theories of normal development view as important for

healthy development. Given the cumulative nature of psychological develop-
ic
er

ment, early traumatic experiences have implications for psychological

functioning and psychopathology across the life span. It is important to note,
Am

however, that maltreatment does not automatically doom all children to neg-
ative outcomes. In fact, some children are surprisingly resilient even under
e
th

seemingly overwhelming odds against favorable development. Longitudinal

research suggests that resilience is promoted by a child having at least one sup-
by

portive and security-enhancing attachment relationship; that is, a relationship

11

with a person who believes in the child and offers support when needed
(Werner, 1993). This places attachment theory at the heart of theorizing about
20

child maltreatment and its effects.

©
ht

THE SEQUELAE OF CHILD MALTREATMENT

ig
yr

Research indicates that there are numerous sequelae of child maltreat-

op

ment. Here we address several of them, specifically emotion regulation, attach-

C

ment, psychopathology, and memory.

352 TOTH ET AL.

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 353

Emotion Regulation and Child Maltreatment

An early issue in infant development is the ability to regulate affective

experience. Emotion regulation includes the intra- and extraorganismic fac-

A)
tors by which emotional arousal is redirected, controlled, modulated, and mod-
ified so that an individual can function adaptively in emotionally challenging

P
(A
situations (Cicchetti, Ganiban, & Barnett, 1991). The ability to regulate emo-
tions develops within the context of early parent–child interactions and rela-

n
tionships (Maughan & Cicchetti, 2002; Thompson, 1999, 1994).

io
Child maltreatment poses a significant threat to the optimal develop-

at
ment of affect regulation. Although there are important individual differences,

ci
maltreated children as a group exhibit numerous deficits in the recognition,

so
expression, and understanding of emotions. Even in the early months of life,

As
distortions in affect differentiation can occur in children with a history of mal-
treatment (Gaensbauer & Hiatt, 1984; Gaensbauer, Mrazek, & Harmon,

al
1981). Specifically, either excessive amounts of negative affect or blunted

ic
affect may be observed in such cases.

og
Early problems in emotion regulation can lay a foundation for future dif-
ol
ficulties in modulating affect. For example, in one study, physically abused
ch
preschool boys who witnessed an angry simulated interaction involving their
y

mothers were more behaviorally reactive (e.g., more aggressive) than

Ps

nonabused boys (Cummings, Hennessy, Rabideau, & Cicchetti, 1994). In a

related investigation, physically abused boys (compared with nonabused
an

boys) who viewed videotaped vignettes of angry and friendly interactions

ic

reported experiencing more distress in response to angry adult interactions.

er

Physically abused boys also reported greater fear in response to angry interac-
Am

tions between adults, especially when the conflicts were not resolved (Hen-
nessy, Rabideau, Cicchetti, & Cummings, 1994). Finally, in an investigation
e

of 4- to 6-year-old children, approximately 80% of maltreated preschoolers

th

exhibited patterns of emotion dysregulation in response to witnessing an

by

angry interaction between adults (Maughan & Cicchetti, 2002). Findings

such as these support a sensitization hypothesis, wherein repeated exposure
11

to anger and familial violence results in greater emotional reactivity.

20

Difficulties in mentally representing and processing affective information

are also observed in children with histories of maltreatment. For example, phys-
©

ically abused children show biased responses to angry emotional expressions

ht

(Pollak, Cicchetti, Hornung, & Reed, 2000). Pollak and Sinha (2002) reported
ig

that physically abused children required less sensory input than comparison
yr

children to accurately identify facial displays of anger (see also Pollak & Kistler,
op

2002). In an investigation of the attentional mechanisms underlying physically

abused children’s over-attention to angry expressions, Pollak and Tolley-Schell
C

(2003) discovered that such children notice such expressions readily and find

INFLUENCE OF VIOLENCE AND AGGRESSION ON CHILDREN 353

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 354

it difficult to disengage attention from them. These results suggest that early
adverse experiences influence maltreated children’s selective attention to
threat-related emotional signals.
Deviations in understanding negative affect and in processing negative

A)
affective signals are also evident in children who have experienced forms of
abuse other than physical, including sexual abuse, emotional maltreatment,

P
(A
and neglect. Such difficulties are associated with undercontrolled and aggres-
sive behavior in school settings (Rogosch, Cicchetti, & Aber, 1995). In an

n
investigation of attributional processes, Toth, Cicchetti, and Kim (2002)

io
found that children’s attributions for positive and negative events moderated

at
externalizing behavior problems (e.g., aggression) and that children’s percep-

ci
tions of their mothers mediated both internalizing and externalizing symp-

so
toms. Maltreated children had higher levels of internalizing and externalizing

As
behavior problems than did nonmaltreated children. However, maltreated
children with higher positive attribution scores had lower levels of external-

al
izing symptomatology, and those with lower levels of negative attribution

ic
scores had lower levels of externalizing behavior problems. These findings

og
suggest that children’s positive attributional styles exert a protective force
ol
against the damaging effects of child maltreatment. Negative perceptions of
ch
mothers served as an independent risk factor for children’s internalizing and
y

externalizing behavior problems, regardless of maltreatment status. Finally,

Ps

maltreated children had less positive perceptions of their mothers. In attach-

ment theory, these perceptions are part of what Bowlby (1982) called internal
an

working models of relationship partners.

ic

Drawing on models of social information processing, Dodge and col-

er

leagues (Dodge, Pettit, Bates, & Valente, 1995; see also Chapter 9, this vol-
Am

ume) reported that children who were physically abused during the first 5 years
of life were later defensively hypervigilant to hostile cues and failed to attend
e

adequately to relevant nonhostile cues. The abused children attributed hostil-

th

ity to others in situations where most people would not make such attributions.
by

In addition, physically abused children acquired large repertoires of highly

accessible aggressive responses to interpersonal problems, so that when they
11

were provoked in everyday peer relations, aggressive retaliatory responses were

20

likely.
A number of cross-sectional investigations have examined associations
©

between affect-regulatory problems and behavioral dysregulation among mal-

ht

treated children. Maltreated preschool and school-age children exhibit a range

ig

of dysregulated behaviors that are often characterized by disruptive and aggres-

yr

sive actions. Maltreated toddlers tend to react to peer distress with poorly reg-
op

ulated and situationally inappropriate affect and behavior, including anger, fear,
and aggression, as opposed to the more normatively expected responses of
C

empathy and concern (Klimes-Dougan & Kistner, 1990). Shields and Cicchetti

354 TOTH ET AL.

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 355

(1998) found that maltreated children were more likely than nonmaltreated
children to be aggressive, with physically abused children at heightened risk of
reactive aggression (i.e., reacting to perceived provocations). Maltreated chil-
dren also evidenced attention deficits, and subclinical or nonpathological dis-

A)
sociation was more likely among children with histories of physical or sexual
abuse. A history of abuse also predicted emotion dysregulation, affective labil-

P
(A
ity and negativity, and socially inappropriate emotional expressions. In that
study, emotion dysregulation was the mechanism whereby maltreatment

n
resulted in reactive aggression (see Chapters 2, 6, and 7, this volume).

io
Overall, the effects of maltreatment on emotion regulation are profound,

at
and they have important effects on socioemotional development including

ci
aggressive behavior. As we describe later, the resulting emotion-regulation dif-

so
ficulties may also have important effects on memories of trauma, which are

As
important clinically and forensically.

al
Attachment and Child Maltreatment

ic
og
Establishing a secure attachment relationship between an infant and his
ol
or her primary caregiver is an important developmental task during the first
ch
year of life (Sroufe, 1979; Thompson, 1999). Attachment theorists have
y

posited that, as development proceeds, a secure attachment relationship pro-

Ps

vides a secure base from which to explore the environment and, ultimately,
contributes to the integration of cognitive, affective, and behavioral capacities
an

that influence ongoing and future relationships as well as self-understanding

ic

(Bowlby, 1973, 1980, 1982; Chapters 13, 14, and 18, this volume). Not
er

surprisingly, maltreated children are at heightened risk of developing insecure

Am

attachment relationships with their caregivers. Researchers have found that

maltreated children are likely to be insecurely attached, with rates of insecu-
e

rity reaching as high as 95% (Cicchetti, Rogosch, & Toth, 2006; Crittenden,
th

1988; Lyons-Ruth, Connell, Zoll, & Stahl, 1987). Furthermore, in both cross-
by

sectional and longitudinal studies, maltreated infants and toddlers have ele-
vated rates of disorganized/disoriented (Type D) attachment, an atypical form of
11

attachment organization characterized in infancy by behavioral freezing, daz-

20

ing, stilling, and apprehension of the caregiver; this is characterized later in

life by increased risk of serious mental health problems (Barnett, Ganiban, &
©

Cicchetti, 1999; Carlson, Cicchetti, Barnett, & Braunwald, 1989; Cicchetti

ht

et al., 2006; Lyons-Ruth et al., 1987).

ig

Although particularly salient during infancy and toddlerhood, attach-

yr

ment continues to be important across the life course, because internal work-
op

ing models (i.e., mental representations) of self, others, and close relationships
are carried forward into later stages of development (Chapters 4, 13, and 14).
C

For example, children maltreated in infancy and toddlerhood are more likely

INFLUENCE OF VIOLENCE AND AGGRESSION ON CHILDREN 355

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 356

than nonmaltreated children to exhibit insecure attachment during all the

preschool years (Cicchetti & Barnett, 1991a; Crittenden, 1988), and insecure
attachment is common among such children during the school years, when it
can be assessed with self-report measures (Lynch & Cicchetti, 1991).

A)
There is evidence (summarized, e.g., in the comprehensive volume edited
by Grossmann, Grossmann, & Waters, 2005) that child–parent attachment lays

P
(A
a foundation for later “coherence of mind” with respect to attachment (Hesse,
2008) as well as emotion-regulation strategies relevant to attachment issues

n
(Chapters 13 and 14, this volume). As such, children’s attachment security is

io
likely to affect memory and other cognitive processes in ways that are impor-

at
tant to close relationships, clinical assessment and intervention, and forensic

ci
interviews.

so
As
Psychopathology and Child Maltreatment

al
Although not all maltreated children will have serious mental health

ic
problems, disruptions in psychological development during the early years of

og
life can certainly contribute to the emergence of psychopathology later on. In
ol
general, the literature on maltreatment reveals a greater preponderance of psy-
ch
chiatric symptoms and diagnoses in maltreated than in nonmaltreated chil-
y

dren, and the range of related disorders is broad. Physical abuse and neglect are
Ps

associated with higher levels of child depressive symptomatology (Cicchetti

& Rogosch, 2001; Kaufman, 1991; Manly, Kim, Rogosch, & Cicchetti, 2001;
an

Toth & Cicchetti, 1996; Toth, Manly, & Cicchetti, 1992). In addition, mal-
ic

treated children have an increased risk of developing depression in adulthood

er

(Widom, DuMont, & Czaja, 2007). Moreover, the risk of adult depression
Am

increases even further for adults who were maltreated in the first 5 years of life
(Kaplow & Widom, 2007). The same is true for conduct disorder and delin-
e

quency (Smith & Thornberry, 1995). Higher rates of attention deficit hyper-
th

activity disorder, oppositional disorder, and posttraumatic stress disorder

by

(PTSD) have also been reported in maltreated children (Famularo, Kinscherff,

& Fenton, 1992). Childhood maltreatment is associated with personality dis-
11

orders, substance abuse, suicidal and self-injurious behavior, somatization, anx-

20

iety, and dissociation (Johnson, Cohen, Brown, Smailes, & Bernstein, 1999;
Luntz & Widom, 1994; Macfie, Cicchetti, & Toth, 2001a, 2001b; Malinosky-
©

Rummell & Hansen, 1993; McLeer, Callaghan, Henry, & Wallen, 1994; Yates,
ht

Carlson, & Egeland, 2008).

ig

Of the various kinds of psychopathology that can develop subsequent to

yr

child maltreatment, PTSD is of special interest to memory researchers, because

op

PTSD has sometimes been conceptualized as a memory disorder (involving

flashbacks and other troubling and unwanted memories). However, researchers
C

have recently begun to reconceptualize PTSD in terms of normal memory for

356 TOTH ET AL.

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 357

trauma (Rubin, Berntsen, & Bohni, 2008). These memory issues are addressed
in the following section.

Memory and Child Maltreatment

A)
Increasing evidence suggests that being maltreated in childhood not

P
(A
only affects emotion regulation, attachment, and mental health but also
affects cognitive functioning (Harris et al., 2008), in particular, memory. In

n
fact, emotion regulation, attachment, psychopathology, and maltreatment all

io
come together to influence memory processes.

at
Research reviewed earlier in this chapter indicates that victims of mal-

ci
treatment often overattend to trauma-related information, as in the case,

so
mentioned earlier, of children with histories of physical abuse having diffi-

As
culty disengaging attention from anger cues while experiencing heightened
arousal and emotional dysregulation. One consequence is that many maltreat-

al
ment victims, especially if they suffer from the trauma-related psychopathol-

ic
ogy of PTSD, may have unusually accurate memories of childhood traumas

og
(Alexander et al., 2005), because they focused great attention on the events
ol
in question and have remembered them in detail ever since. In other words,
ch
the oversensitization to trauma cues, mentioned earlier, may make trauma-
y

related memories particularly robust and enduring. However, a subset of mal-

Ps

treatment victims tries to avoid memories and reminders of their traumatic

experiences, which may interfere with accurate memory. Such avoidance is
an

due to a need to down-regulate strong emotion as a means of coping.

ic

Attachment theory provides a framework for conceptualizing individual

er

differences in the ways in which aggression and violence toward children may
Am

influence memory, especially memory for traumatic childhood experiences

(e.g., Alexander, Quas, & Goodman, 2002; Goodman, Quas, Batterman-
e

Faunce, Riddlesberger, & Kuhn, 1996). According to Bowlby (1980; see also
th

Main, 1990), insecure–avoidant children, whose bids for care have been repeat-
by

edly rejected or belittled, are theorized to develop a defensive, largely uncon-

scious strategy called “defensive exclusion” (Bowlby, 1980). This strategy limits
11

the processing of stressful information, with the goal of preventing the intense
20

activation of attachment behavior that might remind a person of prior

attachment-related rejections, separations, or losses (Chae et al., 2010). Such
©

limits on cognitive processing have important implications for memory.

ht

Relations between patterns of insecure attachment and memory are evi-

ig

dent in studies of maltreated samples, especially when participants are asked

yr

to recall their abusive experiences. Goodman et al. (1992) first examined the
op

psychosocial effects of testifying against assailants in criminal court. Their

study was based on a sample of child victims of sexual abuse (Goodman et al.,
C

1992). About 13 years after their court cases, the victims were interviewed

INFLUENCE OF VIOLENCE AND AGGRESSION ON CHILDREN 357

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 358

about their abusive experiences (Edelstein et al., 2005; Goodman et al., 2003).
The results indicated that victims who scored higher on a measure of avoidant
attachment had poorer (i.e., less detailed and less accurate) memories the more
severe their assaults had been, a pattern apparently reflecting defensive exclu-

A)
sion. In sharp contrast, victims with lower scores on avoidant attachment were
actually more accurate about their more severe abuse experiences.

P
(A
Coping strategies, which are likely to be affected by child maltreatment,
may explain, at least in part, the connection between avoidant attachment,

n
avoidant coping (sometimes called distancing coping), and memory. In a recent

io
study of autobiographical memory, adults and adolescents with or without

at
childhood sexual abuse histories were asked to recount childhood events

ci
(Harris et al., 2008). Their accounts were scored for the degree of specificity of

so
the memories (e.g., specificity as to time and place of the incidents described;

As
see Williams, 1996). In line with Bowlby’s (1980) conception of defensive
avoidance, Harris et al. (2008) predicted that survivors of childhood sexual

al
abuse would adopt a nonspecific memory retrieval style as a way of avoiding

ic
unpleasant and intrusive specific memories (e.g., Hermans, Defranc, Raes,

og
Williams, & Eelen, 2005). Harris et al. found that memory nonspecificity was
ol
associated with distancing coping processes (i.e., cognitive efforts to detach
ch
oneself from and minimize the significance of the situation, as measured by
the Ways of Coping Questionnaire; Folkman & Lazarus, 1988), which was
y
Ps

expected based on attachment theory and research (Hazan & Shaver, 1987;
Mikulincer & Shaver, 2007). The relation between distancing coping and
an

autobiographical-memory nonspecificity was robust; the results held even when

ic

measures of PTSD symptomatology and general psychopathology were statisti-

er

cally controlled in regression analyses. The findings support Bowlby’s (1980)

theory of defensive avoidance.
Am

Overall, the evidence suggests that attachment orientation, emotion-

regulation and coping strategies, and psychopathology need to be examined
e
th

when assessing memories of individuals with maltreatment histories. In some

cases, such as with children and adults who suffer from PTSD and overattend
by

to trauma cues, memory for trauma may be robust and accurate. In other cases,
such as with children and adults with an avoidant attachment style and who
11

use avoidant emotion-regulation and coping strategies, memory for trauma

20

may be suppressed or weakened, and for that reason may also be less accurate.
©
ht

METHODOLOGICAL CONSIDERATIONS
ig
yr

Having briefly reviewed research on child maltreatment and some of its

effects and moderators, we consider next a few methodological issues in the
op

design and conduct of this research. We are especially interested here in how
C

child maltreatment is defined and in issues of internal validity.

358 TOTH ET AL.

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 359

Defining Child Maltreatment

How is child maltreatment defined and operationalized? Historically,

research on child maltreatment has been hampered by the lack of clearly artic-

A)
ulated approaches to defining maltreatment (Barnett, Manly, & Cicchetti,
1993; National Research Council, 1993) and by a long history of disagreement

P
among researchers, lawmakers, and clinicians (Aber & Zigler, 1981; Barnett

(A
et al., 1993). Many researchers have maintained that the definition of mal-

n
treatment should focus on specific acts that endanger a child (Barnett et al.,

io
1993; Cicchetti & Barnett, 1991b; Zuravin, 1991). However, because mal-

at
treatment is largely determined by legal considerations, it is often identified by

ci
social service workers. Yet there is still not a broad consensus among them

so
regarding acceptable disciplinary practices (e.g., spanking) and what counts as

As
child maltreatment (Barnett, Manly, & Cicchetti, 1993; Black & Dubowitz,
1999). There is also a lack of agreement on whether child maltreatment should

al
be defined based on the actions of the perpetrator, the effects on the child, or

ic
a combination of the two (Barnett et al., 1993).

og
Operationalizations of the maltreatment variable have ranged from inves-
ol
tigator observations of “poor parenting” through reliance on parent and/or child
ch
reports, to utilization of Child Protective Services records. These issues are not
easily resolved, but they need to be considered when evaluating studies of the
y
Ps

effects of parental aggression on children.

an

Internal Validity
ic
er

When conducting research on child maltreatment, clearly, one cannot

Am

randomly assign children to abused and nonabused groups. As a result, con-

siderable caution is needed when attempting to draw causal inferences about
child maltreatment. There are many co-occurring and intertwined variables
e
th

to ferret out and statistically control. Nevertheless, carefully designed studies

can provide substantial insight into the possible effects of child maltreatment
by

on children’s development, including their emotion regulation, attachment,

11

mental health, and memory.

20

CONCLUSION
©
ht

Research has demonstrated that multiple psychological processes and

ig

outcomes are affected when children experience aggression and violence, par-
yr

ticularly in their homes. We have highlighted the importance of examining

op

the influence of maltreatment on several domains of psychological function-

ing. Many studies, conducted over many years, have implicated maltreatment
C

in disrupting emotion regulation, child–parent attachment, and children’s

INFLUENCE OF VIOLENCE AND AGGRESSION ON CHILDREN 359

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 360

current and later mental health. It is now well documented that violence
affects children’s emotional memories as well and their feelings and behavior.
Further research should examine the extent to which violence adversely
affects other cognitive domains such as executive functions, problem solving,

A)
planning, and learning. Although research in the area of child maltreatment
is not without major challenges and complexities, the potential theoretical,

P
clinical, and social policy implications of such research provide compelling

(A
reasons to increase its sophistication and applicability in the future.

n
io
at
REFERENCES

ci
so
Aber, J. L., & Zigler, E. (1981). Developmental considerations in the definition of

As
child maltreatment. New Directions for Child Development, 1981(11), 1–29.
doi:10.1002/cd.23219811103

al
Alexander, K. W., Quas, J. A., & Goodman, G. S. (2002). Theoretical advances in

ic
understanding children’s memory for distressing events: The role of attachment.

og
Developmental Review, 22, 490–519. doi:10.1016/S0273-2297(02)00004-7
Alexander, K. W., Quas, J. A., Goodman, G. S., Ghetti, S., Edelstein, R. S., Redlich,
ol
A. D., . . . Jones, D. P. H. (2005). Traumatic impact predicts long-term memory
ch
for documented child sexual abuse. Psychological Science, 16, 33–40. doi:10.1111/
y

j.0956-7976.2005.00777.x
Ps

Barnett, D., Ganiban, J., & Cicchetti, D. (1999). Maltreatment, negative expressiv-
an

ity, and the development of Type D attachments from 12- to 24-months of

age. Monographs of the Society for Research in Child Development, 64, 97–118.
ic

doi:10.1111/1540-5834.00035
er

Barnett, D., Manly, J. T., & Cicchetti, D. (1993). Defining child maltreatment: The
Am

interface between policy and research. In D. Cicchetti & S. L. Toth (Eds.), Child
abuse, child development, and social policy (pp. 7–74). Norwood, NJ: Ablex.
e

Black, M. M., & Dubowitz, H. (1999). Child neglect: Research recommendations

th

and future directions. In H. Dubowitz (Ed.), Neglected children: Research, prac-

by

tice, and policy (pp. 261–277). Thousand Oaks, CA: Sage.

Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. New
11

York, NY: Basic Books.

20

Bowlby, J. (1980). Attachment and loss: Vol. 3. Sadness and depression. New York, NY:
Basic Books.
©

Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). New York, NY:
ht

Basic Books.
ig

Carlson, V., Cicchetti, D., Barnett, D., & Braunwald, K. (1989). Disorganized/
yr

disoriented attachment relationships in maltreated infants. Developmental

op

Psychology, 25, 525–531. doi:10.1037/0012-1649.25.4.525

Chae, Y. (2010). The influence of child attachment style and parental attachment on mem-
C

ory accuracy and suggestibility on young children. Manuscript in preparation.

360 TOTH ET AL.

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 361

Cicchetti, D., & Barnett, D. (1991a). Attachment organization in preschool-aged

maltreated children. Development and Psychopathology, 3, 397–411 doi:10.1017/
S0954579400007598
Cicchetti, D., & Barnett, D. (1991b). Toward the development of a scientific nosology

A)
of child maltreatment. In W. Grove & D. Cicchetti (Eds.), Thinking clearly about
psychology: Essays in honor of Paul E. Meehl: Vol. 2. Personality and psychopathology

P
(pp. 346–377). Minneapolis, MN: University of Minnesota Press.

(A
Cicchetti, D., Ganiban, J., & Barnett, D. (1991). Contributions from the study of

n
high-risk populations to understanding the development of emotion regulation.

io
In J. Garber & K. A. Dodge (Eds.), The development of emotion regulation and

at
dysregulation (pp. 15–48). New York, NY: Cambridge University Press.

ci
Cicchetti, D., & Rogosch, F. A. (2001). The impact of child maltreatment and psycho-

so
pathology upon neuroendocrine functioning. Development and Psychopathology,

As
13, 783–804.
Cicchetti, D., Rogosch, F. A., & Toth, S. L. (2006). Fostering secure attachment in

al
infants in maltreating families through preventive interventions. Development

ic
and Psychopathology, 18, 623–649. doi:10.1017/S0954579406060329

og
Crittenden, P. M. (1988). Relationships at risk. In J. Belsky & T. Nezworski (Eds.),
ol
The clinical implications of attachment (pp. 136–174). Hillsdale, NJ: Erlbaum.
ch
Cummings, E. M., Hennessy, K. D., Rabideau, G. J., & Cicchetti, D. (1994).
y

Responses of physically abused boys to interadult anger involving their mothers.

Ps

Development and Psychopathology, 6, 31–41.

Dodge, K. A., Pettit, G. S., Bates, J. E., & Valente, E. (1995). Social information-
an

processing patterns partially mediate the effect of early physical abuse on later
ic

conduct problems. Journal of Abnormal Psychology, 104, 632–643. doi:10.1037/

er

0021-843X.104.4.632
Am

Edelstein, R. S., Ghetti, S., Quas, J. A., Goodman, G. S., Alexander, K. W., Redlich,
A. D., & Cordón, I. M. (2005). Individual differences in emotional memory:
Adult attachment and long-term memory for child sexual abuse. Personality and
e
th

Social Psychology Bulletin, 31, 1537–1548. doi:10.1177/0146167205277095

Famularo, R., Kinscherff, R. T., & Fenton, T. (1992). Psychiatric diagnoses of mal-
by

treated children. Journal of the American Academy of Child and Adolescent Psychi-
11

atry, 31, 863–867. doi:10.1097/00004583-199209000-00013

Folkman, S., & Lazarus, R. S. (1988). Manual for the Ways of Coping Questionnaire.
20

Mountain View, CA: Consulting Psychologists Press.

©

Gaensbauer, T. J., & Hiatt, S. W. (1984). Facial communication of emotion in early

infancy. Journal of Pediatric Psychology, 9, 205–217.
ht
ig

Gaensbauer, T. J., Mrazek, D., & Harmon, R. (1981). Emotional expression in abused
yr

and/or neglected infants. In N. Frude (Ed.), Psychological approaches to child abuse

(pp. 120–135). Totowa, NJ: Rowman & Littlefield.
op

Goodman, G. S., Ghetti, S., Quas, J. A., Edelstein, R. S., Alexander, K. W., Redlich,
C

A. D., . . . Jones, D. P. H. (2003). A prospective study of memory for child sex-

INFLUENCE OF VIOLENCE AND AGGRESSION ON CHILDREN 361

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 362

ual abuse: New findings relevant to the repressed-memory controversy. Psycho-

logical Science, 14, 113–118. doi:10.1111/1467-9280.01428
Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesberger, M. M., &
Kuhn, J. (1997). Children’s reactions to and memory for a stressful event: Influ-

A)
ences of age, anatomical dolls, knowledge, and parental attachment. Applied
Developmental Science, 1, 54–75. doi:10.1207/s1532480xads0102_1

P
(A
Goodman, G. S., Taub, E. P., Jones, D., England, P., Port, L., Rudy, L., & Prado, L.
(1992). Testifying in court. Monographs of the Society for Research in Child Devel-

n
opment, 57 (Serial no. 229).

io
Grossmann, K. E., Grossmann, K., & Waters, E. (Eds.). (2005). Attachment from

at
infancy to adulthood: The major longitudinal studies. New York, NY: Guilford Press.

ci
so
Harris, L. S., Block, S. D., Augusti, E. M., Larson, R. P., Culver, M. D., Pineda,
R. R., . . . Goodman, G. S. (2008, May). The relation between coping style and

As
autobiographical memory specificity in maltreated and nonmaltreated adolescents and
adults. Poster session presented at the meeting of the annual convention of the

al
Association for Psychological Science, Chicago, IL.

ic
og
Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment
process. Journal of Personality and Social Psychology, 52, 511–524. doi:10.1037/
0022-3514.52.3.511 ol
ch
Hennessy, K. D., Rabideau, G. J., Cicchetti, D., & Cummings, E. M. (1994).
y

Responses of physically abused and nonabused children to different forms of

Ps

interadult anger. Child Development, 65, 815–828. doi:10.2307/1131420

an

Hermans, D., Defranc, A., Raes, F., Williams, J. M. G., & Eelen, P. (2005). Reduced
autobiographical memory specificity as an avoidant coping style. The British Jour-
ic

nal of Clinical Psychology, 44, 583–589. doi:10.1348/014466505X53461

er

Hesse, E. (2008). The Adult Attachment Interview: Protocol, method of analysis,

Am

and empirical studies. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attach-

ment: Theory, research, and clinical applications (pp. 552–598). New York, NY:
e

Guilford Press.
th

Johnson, J. G., Cohen, P., Brown, J., Smailes, E. M., & Bernstein, D. P. (1999). Child-
by

hood maltreatment increases risk for personality disorders during early adulthood.
Archives of General Psychiatry, 56, 600–606. doi:10.1001/archpsyc.56.7.600
11

Kaplow, J. B., & Widom, C. S. (2007). Age of onset of child maltreatment predicts
20

long-term mental health outcomes. Journal of Abnormal Psychology, 116, 176–187.

doi:10.1037/0021-843X.116.1.176
©

Kaufman, J. (1991). Depressive disorders in maltreated children. Journal of the Amer-

ht

ican Academy of Child and Adolescent Psychiatry, 30, 257–265. doi:10.1097/

ig

00004583-199103000-00014
yr

Klimes-Dougan, B., & Kistner, J. (1990). Physically abused preschoolers’ responses

op

to peers’ distress. Developmental Psychology, 26, 599–602. doi:10.1037/0012-

C

1649.26.4.599

362 TOTH ET AL.

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 363

Luntz, B. K., & Widom, C. S. (1994). Antisocial personality disorder in abused and
neglected children grown up. The American Journal of Psychiatry, 151, 670–674.
Lynch, M., & Cicchetti, D. (1991). Patterns of relatedness in maltreated and non-
maltreated children: Connections among multiple representational models.

A)
Development and Psychopathology, 3, 207–226. doi:10.1017/S0954579400000080

P
Lyons-Ruth, K., Connell, D., Zoll, D., & Stahl, J. (1987). Infants at social risk:

(A
Relationships among infant maltreatment, maternal behavior, and infant
attachment behavior. Developmental Psychology, 23, 223–232. doi:10.1037/

n
0012-1649.23.2.223

io
Macfie, J., Cicchetti, D., & Toth, S. L. (2001a). The development of dissociation

at
in maltreated preschool-aged children. Development and Psychopathology, 13,

ci
233–254. doi:10.1017/S0954579401002036

so
Macfie, J., Cicchetti, D., & Toth, S. L. (2001b). Dissociation in maltreated versus

As
nonmaltreated preschool-aged children. Child Abuse & Neglect, 25, 1253–1267.
doi:10.1016/S0145-2134(01)00266-6

al
Main, M. (1990). Cross-cultural studies of attachment organization: Recent studies,

ic
og
changing methodologies, and the concept of conditional strategies. Human
Development, 33, 48–61. doi:10.1159/000276502
ol
Malinosky-Rummell, R., & Hansen, D. (1993). Long-term consequences of child-
ch
hood physical abuse. Psychological Bulletin, 114, 68–79. doi:10.1037/0033-
y

2909.114.1.68
Ps

Manly, J. T., Kim, J. E., Rogosch, F. A., & Cicchetti, D. (2001). Dimensions of child
an

maltreatment and children’s adjustment: Contributions of developmental timing

and subtype. Development and Psychopathology, 13, 759–782.
ic

Maughan, A., & Cicchetti, D. (2002). The impact of child maltreatment and
er

interadult violence on children’s emotion regulation abilities. Child Development,

Am

73, 1525–1542. doi:10.1111/1467-8624.00488

McLeer, S. V., Callaghan, M., Henry, D., & Wallen, J. (1994). Psychiatric disorders in
e

sexually abused children. Journal of the American Academy of Child and Adolescent
th

Psychiatry, 33, 313–319. doi:10.1097/00004583-199403000-00003

by

Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynamics,

and change. New York, NY: Guilford Press.
11

National Research Council. (1993). Understanding child abuse and neglect. Washington,
20

DC: National Academy of Sciences.

Pollak, S. D., Cicchetti, D., Hornung, K., & Reed, A. (2000). Recognizing emotion in
©

faces: Developmental effects of child abuse and neglect. Developmental Psychology,

ht

36, 679–688. doi:10.1037/0012-1649.36.5.679

ig

Pollak, S. D., & Kistler, D. (2002). Early experience alters the development of categor-
yr

ical representations for facial expressions of emotion. Proceedings of the National

op

Academy of Sciences of the United States of America, 99, 9072–9076. doi:10.1073/

C

pnas.142165999

INFLUENCE OF VIOLENCE AND AGGRESSION ON CHILDREN 363

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 364

Pollak, S. D., & Sinha, P. (2002). Enhanced perceptual sensitivity for anger among
physically abused children. Developmental Psychology, 38, 784–791. doi:10.1037/
0012-1649.38.5.784
Pollak, S. D., & Tolley-Schell, S. A. (2003). Selective attention to facial emotion
in physically abused children. Journal of Abnormal Psychology, 112, 323–338.

A)
doi:10.1037/0021-843X.112.3.323

P
Rogosch, F. A., Cicchetti, D., & Aber, J. L. (1995). The role of child maltreatment in

(A
early deviations in cognitive and affective processing abilities and later peer rela-

n
tionship problems. Development and Psychopathology, 7, 591–609. doi:10.1017/

io
S0954579400006738

at
Rubin, D., Berntsen, D., & Bohni, M. (2008). A memory based model of posttraumatic

ci
stress disorder. Psychological Review, 115, 985–1011. doi:10.1037/a0013730

so
Shields, A., & Cicchetti, D. (1998). Reactive aggression among maltreated children:

As
The contributions of attention and emotion dysregulation. Journal of Clinical
Child Psychology, 27, 381–395. doi:10.1207/s15374424jccp2704_2

al
Smith, C. A., & Thornberry, T. P. (1995). The relationship between childhood

ic
maltreatment and adolescent involvement in delinquency. Criminology, 33,

og
451–481. doi:10.1111/j.1745-9125.1995.tb01186.x
ol
Sroufe, L. A. (1979). The coherence of individual development: Early care, attach-
ch
ment, and subsequent developmental issues. American Psychologist, 34, 834–841.
doi:10.1037/0003-066X.34.10.834
y
Ps

Thompson, R. A. (1994). Emotion regulation: A theme in search of a definition. Mono-

graphs of the Society for Research in Child Development, 59, 25–52. doi:10.2307/
an

1166137
ic

Thompson, R. A. (1999). Early attachment and later development. In J. Cassidy &

er

P. R. Shaver (Eds.), Handbook of attachment (pp. 265–286). New York, NY:

Guilford Press.
Am

Toth, S. L., & Cicchetti, D. (1996). Patterns of relatedness and depressive symp-
tomatology in maltreated children. Journal of Consulting and Clinical Psychology,
e

64, 32–41. doi:10.1037/0022-006X.64.1.32

th

Toth, S. L., Cicchetti, D., & Kim, J. E. (2002). Relations among children’s percep-
by

tions of maternal behavior, attributional styles, and behavioral symptomatology

in maltreated children. Journal of Abnormal Child Psychology, 30, 487–501.
11

doi:10.1023/A:1019868914685
20

Toth, S. L., Manly, J. T., & Cicchetti, D. (1992). Child maltreatment and vulnera-
bility to depression. Development and Psychopathology, 4, 97–112. doi:10.1017/
©

S0954579400005587
ht

Werner, E. E. (1993). Risk, resilience, and recovery: Perspectives from the Kauai
ig

Longitudinal Study. Development and Psychopathology, 5, 503–515. doi:10.1017/

yr

S095457940000612X
op

Widom, C. S., DuMont, K., & Czaja, S. J. (2007). A prospective investigation of major
depressive disorder and comorbidity in abused and neglected children grown up.
C

Archives of General Psychiatry, 64, 49–56. doi:10.1001/archpsyc.64.1.49

364 TOTH ET AL.

12051-20_PT5-CH19-rev2.qxd 5/25/10 12:00 PM Page 365

Williams, J. M. G. (1996). Depression and the specificity of autobiographical mem-

ory. In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical mem-
ory (pp. 244–267). New York, NY: Cambridge University Press. doi:10.1017/
CBO9780511527913.010

A)
Yates, T. M., Carlson, E. A., & Egeland, B. R. (2008). A prospective study of child
maltreatment and self-injurious behavior in a community sample. Development

P
and Psychopathology, 20, 651–671. doi:10.1017/S0954579408000321

(A
Zigler, E. (1998). A place of value for applied and policy studies. Child Development,

n
69, 532–542.

io
Zuravin, S. J. (1991). Research definitions of child physical abuse and neglect: Cur-

at
rent problems. In R. H. Starr & D. A. Wolfe (Eds.), The effects of child abuse and

ci
neglect (pp. 100–128). New York, NY: Guilford Press.

so
As
al
ic
og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

INFLUENCE OF VIOLENCE AND AGGRESSION ON CHILDREN 365

C
op
yr
ig
ht
©
20
12051-20_PT5-CH19-rev2.qxd

11
by
5/25/10

th
e
Am
12:00 PM

er
ic
an
Page 366

Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 367

20

PA)
THE PARADOX OF PARTNER

(A
AGGRESSION: BEING COMMITTED

n
io
TO AN AGGRESSIVE PARTNER

at
ci
so
XIMENA B. ARRIAGA AND NICOLE M. CAPEZZA

As
al
ic
og
ol
y ch
Ps
an

Aggression in relationships is a serious problem. In the United States,

ic

for example, approximately, 1.3 million women and 835,000 men are physi-
er

cally assaulted by a romantic partner annually, with the majority of reported

Am

acts (about 75%) being mild forms of aggression, such as pushing or slapping
(Tjaden & Theonnes, 2000). Each year, about 5% of physically abused
e

women (approximately 44,000) suffer serious injuries from being beaten up,
th

choked, or assaulted with a weapon. (Such severe acts are significantly more
by

likely to be perpetrated by men against women than the reverse; Rennison &
Welchans, 2000; see also Chapters 2, 13, and 14, this volume.)
11

Whereas other forms of aggression often occur in contexts where aggres-

20

sion might be expected (e.g., between antagonistic groups), aggression commit-

ted by a relationship partner violates most people’s fundamental hopes and
©

expectations for a close relationship. Targets of partner aggression who are com-
ht

mitted to their relationship are left to make sense of a paradox: The presumed
ig

source of love and intimacy is also the source of pain. This paradox often causes
yr

outsiders to wonder why the victims of partner aggression remain with their
op

aggressive partner.
In this chapter, we suggest that perceptions or interpretations of the part-
C

ner’s acts play a central role in continuing the relationship, and we consider

367
12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 368

the possibility that individuals perceive aggressive partner acts as less severe to
the extent that they are committed. Victims who downplay their partner’s
aggression fail to protect themselves and thus run the risk of being hurt further.
Moreover, when members of a society fail to recognize that partner aggression

A)
has negative consequences and should not be tolerated, partner aggression
becomes more likely to continue as perpetrators avoid punishment and targets

P
(A
fail to get needed protection and support.
We begin by defining partner aggression and delineating the different

n
forms it takes. We then review our own and other research on what is known

io
about perceptions of partner aggression in general, and about how victims’ per-

at
ceptions compare with others’ perceptions. We then shift the focus from per-

ci
ceptions of partner aggression in general to how victims perceive their own

so
partner’s acts. We use concepts from interdependence and consistency theories

As
to suggest that the closer and more committed the victim feels to the perpetra-
tor, the more likely she is to downplay his aggressive acts. We end by discussing

al
how a victim’s well-being is affected by continuing a relationship in which she

ic
downplays her partner’s aggressive acts.

og
ol
ch
WHAT DO WE MEAN BY PARTNER AGGRESSION?
y
Ps

Our focus is on acts that occur between partners in an intimate relation-

ship and that are intended to inflict harm. These include physically aggressive
an

acts, such as hitting, punching, and kicking. They also include psychologically
ic

aggressive acts, such as yelling, derogating, threatening, and otherwise attempt-

er

ing to control and dominate another person (O’Leary, 1999).

Am

For several reasons, we are particularly interested in perceptions of psy-

chological aggression. Severe forms of psychological aggression, such as humil-
e

iating, degrading, and threatening a partner, are highly correlated with

th

occurrences of physical aggression (Murphy & Hoover, 1999). Moreover,

by

recent research has revealed, somewhat surprisingly, that a pattern of sustained

psychological aggression is as or more damaging to a victim than physical
11

aggression. For instance, over 70% of physically abused women report that
20

emotional abuse had a more damaging effect on their self-esteem and health
than physical aggression (Follingstad, Rutledge, Berg, Hause, & Polek, 1990).
©

Psychological aggression not only has a strong, deleterious effect on a victim’s

ht

mental health (e.g., Arias & Pape, 1999), but it has also been linked to a num-
ig

ber of adverse physical health outcomes. Victims of psychological aggression

yr

are just as likely as victims of physical aggression to suffer from chronic neck
op

or back pain, migraines, stomach ulcers, spastic colon, and gastrointestinal

symptoms, among other health problems (Coker et al., 2002; Coker, Smith,
C

368 ARRIAGA AND CAPEZZA

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 369

Bethea, King, & McKeown, 2000). Thus, psychological aggression has clear
mental and physical health outcomes for victims.
Yet, as we describe later in this chapter, psychological aggression is not
perceived to be as serious as physical aggression, despite the growing literature

A)
on the severe consequences of psychological aggression suggesting otherwise.
This may make victims themselves and people in general less likely to counter

P
(A
partner psychological aggression than physical aggression.
Much of our analysis focuses on commitment, a strong subjective force

n
that keeps people in relationships (Rusbult, 1983). A victim’s feeling tied to

io
a partner, imagining a long-term future with the partner, and intending to

at
remain in the relationship—all of which are aspects of commitment (Arriaga

ci
& Agnew, 2001)—create a subjective state that motivates more benign inter-

so
pretations of negative partner acts (Rusbult, Olsen, Davis, & Hannon, 2001).

As
To date, we have examined commitment and type of aggression (i.e., physi-
cal vs. psychological) as important factors predicting perceptions of partner

al
aggression, but there are likely many other personality, relational, and social

ic
circumstances that affect perceptions. We begin our analysis by establishing

og
how partner aggression is generally perceived.
ol
y ch

HOW IS PARTNER AGGRESSION GENERALLY PERCEIVED?

Ps

Perceptions of partner aggression vary. One source of variation is time,

an

in that perceptions are more negative today than they used to be. A second
ic

source of variation in perceptions is the specific type of aggression being

er

considered.
Am

General U.S. Norms

e
th

In the United States, acts of physical aggression in intimate relationships

by

may have been tolerated 40 years ago, but they are now generally considered
unacceptable (Gelles, 1993). Representative U.S. samples have shown that
11

acts of physical aggression (e.g., punch, slap) are considered to be cases of

20

domestic violence and are thus unlawful. For instance, 98.8% of respondents
in one study (Carlson & Worden, 2005) reported that a husband punching his
©

wife constitutes domestic violence, and 91.3% considered slapping to be an

ht

act of domestic violence.

ig

Although these U.S. norms against physical abuse have become stable
yr

parts of American culture, there is an absence of comparable norms against

op

psychological aggression. For instance, only 53.8% of respondents consider

a husband insulting his wife by calling her “a stupid slob” to be domestic
C

THE PARADOX OF PARTNER AGGRESSION 369

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 370

violence (Carlson & Worden, 2005). Other studies of representative American

samples (e.g., Sorenson & Taylor, 2005) have shown that sanctions against a
perpetrator (e.g., arrests, restraining orders) are supported more for physically
aggressive behaviors (e.g., slapping, punching, beating) than for psycho-

A)
logically aggressive behaviors (e.g., belittling, humiliating, threatening).
On one hand, this may not seem surprising given that physical aggression

P
(A
can leave visible injuries whereas psychological aggression leaves no physical
marks. But on the other hand, U.S. society’s failure to condemn psycho-

n
logical aggression is surprising given that its effects are as or more damag-

io
ing to the victim’s long-term well-being than physical aggression. It is as if

at
American adults adhere to the familiar childhood retort that “Sticks and

ci
stones may break my bones, but words will never hurt me.” Unfortunately,

so
that long-lived maxim is untrue.

As
Research that attempts to identify individual characteristics that make a
person prone to tolerate domestic violence has yielded mixed findings. For

al
example, several studies examining the sex of the perceiver suggest that

ic
women find a perpetrator’s aggressive behavior more blameworthy and less

og
acceptable than do men (e.g., Cauffman, Feldman, Jensen, & Arnett, 2000;
ol
Pierce & Harris, 1993), but not all studies have found similar sex differences
ch
(e.g., Capezza & Arriaga, 2008a). Although there may be some variation
y

among ethnic groups and subcultures, the belief that acts of partner aggression
Ps

are best avoided seems to be widely held in the United States.

an

Physical Versus Psychological Aggression

ic
er

Our own research (Capezza & Arriaga, 2008a, 2008b) suggests that per-
Am

ceptions are strongly influenced by the nature of the partner acts and that not
all forms of partner aggression are perceived equally. In one study (Capezza &
e

Arriaga, 2008b), college students who had previously been or were currently
th

in a relationship read a hypothetical scenario in which a marital couple has an

by

argument and the husband becomes aggressive. The levels of physical aggres-
sion (i.e., none, low, high) and psychological aggression (i.e., low, high) were
11

crossed in a between-subjects factorial design. After reading the scenario, par-

20

ticipants completed measures of their perceptions of (a) the perpetrator’s

behavior and (b) the conflict.
©

The results revealed that participants generally held negative perceptions

ht

regarding the event, with mean ratings beyond the midpoint and close to the
ig

scale anchor indicating negative perceptions. The main effect for physical
yr

aggression was strong: Across all dependent variables, participants perceived

op

the perpetrator and conflict in more negative ways with increasing levels of
physical aggression, confirming that norms exist among the college students we
C

sampled against using physical force in a relationship. The effect for psycholog-

370 ARRIAGA AND CAPEZZA

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 371

ical aggression was not as robust, however: One dependent variable failed to
show any effect, one showed a more negative perception for high (vs. low) psy-
chological aggression, and one showed the effect only when physical aggression
was absent (i.e., a simple effect within the no physical aggression condition). A

A)
similar study (Capezza & Arriaga, 2008a) revealed that participants’ percep-
tions did not vary between a vignette depicting verbal aggression (e.g., yelling

P
(A
and swearing) versus one depicting severe emotional aggression (e.g., ridicul-
ing, degrading, highly threatening behaviors), despite the documented serious

n
consequences of emotional aggression.

io
People generally see physical partner aggression as negative. Views of

at
psychological aggression, although not positive or neutral, are not uniformly

ci
negative. It stands to reason that victims of partner aggression should share

so
these perceptions; that is, they should view their partner’s behavior negatively,

As
particularly when the acts qualify as physical aggression. Social psychological
theories, however, suggest that this commonsense prediction may be incorrect,

al
as we explain in the next section outlining our theoretical analysis of victim

ic
perceptions.

og
ol
ch
WHY MIGHT VICTIM PERCEPTIONS DIFFER
y

FROM OTHERS’ PERCEPTIONS?

Ps

At first glance, it seems counterintuitive to suggest that victims of part-

an

ner aggression might downplay their partner’s aggressive acts. One might
ic

expect victims to seek help and sympathy and to complain that their partners’
er

acts are severe and damaging. Why and how, then, might victims downplay
Am

their partner’s aggressive acts? Two social psychological theoretical frame-

works are particularly relevant when answering this question.
e
th

Consistency Theories
by

In social psychology, consistency theories suggest that victims who feel

11

particularly tied or committed to their partners would be likely to perceive

20

their partner’s aggressive acts less negatively than victims who are less com-
mitted to their partners. Heider’s (1958) balance theory, for example, suggests
©

what he called a positive unit relation between the two partners and a positive
ht

unit relation between the perpetrator and his aggressive act. At the outset, the
ig

sentiment relation between the victim and the perpetrator’s aggressive act
yr

would be negative, comprising a state of imbalance among the three relations

op

(i.e., two positive and one negative relation). If the unit relation linking the
two partners is strong, that is, if the ties keeping them together are strong and
C

the victim is committed, then this would create pressure to shift the sentiment

THE PARADOX OF PARTNER AGGRESSION 371

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 372

relation between the victim and the perpetrator’s aggressive act from negative
to neutral or positive. In short, the victim would adopt a less negative view of
the partner’s aggressive act.
In addition to commitment, the nature of the aggressive behavior may

A)
also make a difference. Overtly aggressive acts (e.g., hitting, punching) are more
difficult to reinterpret as being benign than are less physical forms of aggression

P
(A
(e.g., belittling, degrading). Regardless of commitment level, an individual may
be more likely to adopt a neutral sentiment relationship toward an act of ver-

n
bal aggression than toward an act of physical aggression.

io
Festinger’s (1957) cognitive dissonance theory would make a similar pre-

at
diction. The theory suggests that accepting two opposing beliefs results in

ci
mental discomfort that motivates a person to change one of the beliefs to be

so
consistent with the other. Less committed individuals faced with an aggressive

As
(and thus negative) partner act might come to feel more negative about the
relationship. More committed individuals, however, are motivated to feel pos-

al
itive about the partner and therefore would adopt a less negative perception

ic
of the partner act. Similarly, less physical forms of aggression would not create

og
as much dissonance as would overtly physical forms of aggression.
ol
ch
Interdependence Theory
y
Ps

A second theoretical framework that is useful for understanding why vic-

tims might downplay partner aggression is interdependence theory (Kelley, 1979).
an

This theory provides an analysis of thoughts and actions based on a person’s

ic

interaction situation (e.g., a partner interaction) and the person’s broad goals
er

toward the relationship. Highly committed individuals are strongly affected by

Am

what the partner does; they are “dependent” on the partner to the extent that
their interaction experience is strongly affected by what the partner does or
e

what the two partners do in unison.

th

Contentious interpersonal situations, including ones in which a partner

by

is aggressive, trigger in most people an inclination to retaliate with comparable

behavior (Kelley, 1979). However, what often redirects the expected reaction
11

and determines the actual response is a person’s broader goals for the relation-
20

ship. Actual behavior reflects any of several broad goals, such as wanting to help
the partner, promote the relationship, or be slightly ahead of the partner
©

(McClintock, 1972). When faced with a highly negative interaction—for

ht

example, when the partner is aggressive—highly committed individuals will

ig

either be motivated to respond in ways that salvage the relationship (e.g.,

yr

downplay or justify the aggressive act) or reduce their motivation to save the
op

relationship by becoming less committed and dependent. Research has shown

that, compared with less committed individuals, more committed individu-
C

als respond to a negative partner behavior by redirecting their reaction from

372 ARRIAGA AND CAPEZZA

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 373

retaliation to an attempt to diffuse the contentious situation (Yovetich &

Rusbult, 1994).
To the extent that psychological aggression does not have the social taboo
attached to physical aggression, situations in which a partner is psychologically

A)
aggressive may be perceived as less contentious than physical aggression situa-
tions. Victims may find it easier to respond in benign ways toward less rather

P
(A
than more overt forms of aggression, irrespective of their commitment level.
In addition to predicting responses to contentious partner situations,

n
interdependence theory suggests that people become acclimated to events in

io
their relationship by forming expectations of typical behavior, namely a stan-

at
dard of comparison or “comparison level” (Thibaut & Kelley, 1959). Individ-

ci
uals gauge their satisfaction level based on whether events in their relationship

so
exceed their expectations (causing high satisfaction) or fall short of their

As
expectations (causing low satisfaction). As such, victims who repeatedly expe-
rience partner aggression may come to expect the occurrence of aggression in

al
their current and future intimate relationships. The more aggression individ-

ic
uals experience, the more likely they are to perceive the acts as normal rather

og
than severely negative.
ol
Taken together, these theoretical frameworks call into question the idea
ch
that a target of partner aggression would immediately take a negative stance
y

toward the partner’s aggressive actions. It depends on factors such as the indi-
Ps

vidual’s level of commitment, the nature of the aggressive acts, and the current
comparison level or level of expectations.
an
ic
er

REINTERPRETING THE PAST: DO VICTIMS DOWNPLAY

Am

THEIR PARTNER’S PAST AGGRESSION?

e

Earlier we reviewed research examining how people in general view

th

aggression in relationships. Given the theoretical analysis we described, a

by

likely source of systematic variation in perceptions of partner aggression is

whether one is currently being victimized. Because current victims are linked
11

to an aggressive partner, they may have a vested interest in holding more tol-
20

erant perceptions of partner aggression. Indeed, a recent worldwide study by

the World Heath Organization (Garcia-Moreno, Heise, Jansen, Ellsberg, &
©

Watts, 2005) revealed that perceiving a male partner’s aggression as normal

ht

or justified—specifically, accepting a man’s beating of his wife—was more

ig

pervasive among women who had experienced such aggression than among
yr

those who had not.

op

Much (but not all) of our own research on this issue has primarily
involved female college students. We focus on women mainly to narrow our
C

analysis at the outset, with the intention of expanding it to male victims once

THE PARADOX OF PARTNER AGGRESSION 373

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 374

the basic processes of entrapment are well understood. By no means do we

suggest that men are not victims of partner aggression; in fact, research has
shown that they can be at risk with respect to less severe aggression (Johnson
& Ferraro, 2000).

A)
We focus on college students for several reasons. From a practical stand-
point, it has been convenient to collect data from college samples. More

P
(A
important, however, on theoretical grounds, we have been interested in the
forces that keep victims connected to an aggressive partner when there are no

n
legally binding reasons for remaining with the partner (which is the case in

io
marriage), and when victims reside in an environment where there are many

at
other potential partners (which interdependence theory conceptualizes in

ci
terms of the comparison level for alternatives; Kelley & Thibaut, 1978).

so
Because the objective circumstances are such that leaving one’s abusive part-

As
ner should not be difficult, it stands to reason that the subjective circumstances
keeping victims in the relationship must be strong.

al
ic
Research Comparing General Perceptions of

og
Current Victims, Past Victims, and Nonvictims
ol
ch
We sought to examine further the link between being victimized and
y

holding tolerant beliefs regarding partner aggression (Arriaga, 2007). We pre-

Ps

dicted that current or past victims would be more tolerant than nonvictims,
based on the idea in interdependence theory that individuals shift their point
an

of reference for expected behavior based on their own experiences. The more
ic

aggression one has experienced, the more aggression might come to be expected
er

in relationships, and thus the less likely it would be viewed as grounds for end-
Am

ing a relationship.
Our critical prediction, however, was with respect to differences between
e

current and past victims. If these two groups do not differ in their perceptions
th

of partner aggression, it suggests that these perceptions are not motivated by

by

one’s current relationship goals. They are influenced only by the victim’s expe-
riences and expectations. Consistency and interdependence theories, how-
11

ever, suggest that current victims should be more tolerant of partner aggression
20

than past victims. Current victims remain involved with their partner and thus
have an interest in holding perceptions that are consistent with this involve-
©

ment. We anticipated support for a motivated cognition process, whereby cur-

ht

rent victims would be more tolerant in general of partner aggression than past
ig

victims, who in turn would be more tolerant than nonvictims.

yr

Across two studies, we recruited female college students who had pre-
viously been or currently were in a relationship (Study 1, n = 186; Study 2,
op

n = 156). Participants completed several measures, including their own rela-

C

tionship status (i.e., currently dating or not), and their victimization status

374 ARRIAGA AND CAPEZZA

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 375

(i.e., current victim, victim in a past relationship, never been victimized) as

indicated by a scale modeled after the Conflict Tactics Scale (CTS; Straus
& Gelles, 1990). The CTS asks participants to indicate the number of times
their current partner has engaged in specific verbally, emotionally, or phys-

A)
ically aggressive acts (e.g., partner insulted or swore at you, partner grabbed
and shook you).

P
(A
Participants also completed a scale that asked about the same and addi-
tional specific partner aggressive acts; on this scale, they were instructed to indi-

n
cate the degree to which having a partner commit that act would be grounds

io
for ending a relationship. We also included one nonaggressive conflict act as a

at
comparison point. The scale anchors were definitely would not end a relationship

ci
versus definitely would end a relationship. Responses were recoded and averaged

so
to create three dependent variables, with higher numbers indicating more tol-

As
erance for a particular type of act: (a) a nonaggressive act (e.g., refused to talk
about an issue with you), (b) psychologically aggressive acts (e.g., insulted or

al
swore at you, intentionally destroyed your belongings), and (c) physically

ic
aggressive acts (e.g., grabbed and shook you). Study 2 also assessed the partici-

og
pant’s perpetration status and eliminated participants who had been perpetra-
ol
tors more than victims. All analyses controlled for whether the participant was
ch
currently in a relationship. Study 1 also controlled for the format of the survey
y

(paper and pencil vs. online). Study 2 exclusively relied on online surveys, but
Ps

controlled for participant’s amount of perpetration.

Generally, both samples perceived physical aggression to be grounds for
an

ending a relationship. Both samples were more tolerant of psychological aggres-

ic

sion than physical aggression and were less uniform (i.e., more variable) in their
er

ratings of psychological aggression. Both samples were relatively tolerant of a

Am

nonaggressive act. In short, not all aggressive acts are perceived similarly, and
there is more tolerance of psychological aggression than of physical aggression.
e

Despite these general trends, in both studies the participant’s victimiza-

th

tion status was significantly and positively associated with perceptions of what
by

would be grounds for ending a relationship. This provides evidence that stan-
dards shift based on one’s experience, consistent with interdependence theory’s
11

comparison level construct. Moreover, perceptions of partner aggression were

20

motivated by the amount of connection with a partner. Current victims were

more tolerant of psychological aggression than past victims, who were more tol-
©

erant than nonvictims. In short, the more connected a victim is, the more moti-
ht

vated the victim may be to tolerate partner aggression.

ig

We now turn to a more direct indicator of connection to a partner: a

yr

person’s level of commitment. We also focus more specifically on perceptions

op

of acts that have already occurred, rather than perceptions of what might be
grounds for ending a relationship if they were to occur. That is, we shift our
C

focus from hypothetical, possible acts to ones that have actually occurred.

THE PARADOX OF PARTNER AGGRESSION 375

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 376

Research on Commitment and Partner Aggression

Consistency theories provide a framework for anticipating differences

between individuals who are currently involved with an abusive partner and

A)
those who are not. However, these theories have not provided precise ways to
conceptualize and measure variations in closeness or connection among rela-

P
(A
tionship partners. In that respect, interdependence theory has been useful.
Rusbult’s concept of relationship commitment (Rusbult, 1983; Rusbult et al.,

n
2001)—thoughts and feelings that reflect wanting to continue the relation-

io
ship and being attached to the partner, thoughts and feelings that stem from

at
being strongly affected by a partner—is useful for predicting variations in per-

ci
ceptions among current daters. In several lines of work, our basic prediction

so
has been the same: The more committed a victim is to a partner, the more

As
likely the victim will be to avoid maintaining strongly negative views of the
partner’s aggressive acts.

al
One study (Arriaga, 2002) testing this basic prediction examined whether

ic
victims of partner aggression “spin” (reinterpret) their partner’s aggressive acts

og
as “just joking around.” Students at a community college and a large university
ol
were recruited for a survey study. Of the 82 who reported currently being in a
ch
romantic relationship, 54 (18 males and 36 females) reported that their current
y

partner had engaged in at least one aggressive act. Participants completed var-
Ps

ious relationship measures, including a widely used measure of commitment

(Rusbult, Martz, & Agnew, 1998) and the CTS (Straus & Gelles, 1990) to indi-
an

cate the number of times their current partner had been aggressive, that is, to
ic

indicate the amount of conflict-related violence.

er

Participants also completed a scale modeled after the CTS, listing the
Am

same aggressive acts but varying the instructions to elicit instances when the
partner was joking around. Some of the acts listed—the same as those used
e

earlier in the survey—might well have occurred when the partner was joking
th

around (e.g., pushing) but other acts were unequivocally aggressive (e.g.,
by

kicking, beating up, striking with a weapon). Participants who reported these
severely aggressive acts in a joking context were deemed to be reinterpreting
11

unambiguously violent behaviors as being less serious or nonviolent than they

20

really were. The number of times these severe acts occurred provided an indi-
cator of joking violence.
©

In general, participants who reported more conflict violence also reported

more joking violence, r(54) = .36, p < .01. The association of conflict violence
ht
ig

with joking violence was moderated, however, by commitment level, suggest-

yr

ing that highly committed individuals did not share the same pattern of percep-
op

tions that characterized less committed individuals. Less committed individuals

who reported conflict violence were no more or less likely to report joking vio-
C

lence, r(22) = −.06, p = .79. In contrast, highly committed individuals were

376 ARRIAGA AND CAPEZZA

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 377

much more likely to have their reports of joking violence associated with their
reports of conflict violence, r(32) = .59, p < .01, suggesting that they reinter-
preted their partner’s aggression in ways that would make it easier to accept
while continuing the relationship.

A)
The highest levels of conflict violence were reported by individuals who
were highly committed and who reported joking violence. For highly commit-

P
(A
ted individuals, accepting partner aggression and reinterpreting it go hand in
hand. As would be predicted from cognitive dissonance theory, individuals

n
who are strongly tied (i.e., committed) to their partner feel a sense of entrap-

io
ment (Rusbult & Martz, 1995) and find it difficult to report even particularly

at
difficult instances of aggression (Arriaga, 2002).

ci
so
As
PERCEPTIONS OF THE FUTURE: HOW EMOTIONALLY AFFECTED
DO VICTIMS ANTICIPATE BEING FROM A BREAKUP?

al
ic
So far we have focused on victims’ motivated perceptions of their part-

og
ner’s past aggressive behaviors. Interdependence theory suggests that victims
ol
might suffer from motivated, distorted perceptions of a range of interpersonal
ch
situations with their partner, not just those involving aggression. A person who
y

has no overarching goals for interactions with the partner would perceive things
Ps

differently than a person who is committed to a relationship with the partner

(Kelley, 1979). We therefore anticipated that commitment would predict moti-
an

vated perceptions of how affected one would be if the relationship ended.

ic

Research in the affective forecasting literature (e.g., Wilson & Gilbert,

er

2003) shows that people overestimate how emotionally affected they are likely
Am

to be by positive or negative events. For example, people predict that they

would be happier on winning the lottery than lottery winners actually are. Aca-
e

demics predict they would be much less happy if denied tenure than is actually
th

the case among those who actually were denied tenured. In one of several stud-
by

ies discussed by Wilson and Gilbert (2003), Gilbert, Pinel, Wilson, Blumberg,
and Wheatley (1998) examined the bias in forecasting future happiness among
11

relationship partners. They found that partners thought they would feel much
20

less happy following a breakup than was actually the case among those who had
recently experienced a breakup.
©

Would victims of partner aggression show the same bias, whereby they
ht

overestimate how unhappy they might feel if their relationship with the
ig

aggressive partner ended? The commonsense view is that victims should not
yr

be as devastated as nonvictims. Victims often are more depressed within their

op

relationships than are nonvictims, and ending the relationship would provide
an opportunity to undo the negative consequences of being a victim of part-
C

ner aggression.

THE PARADOX OF PARTNER AGGRESSION 377

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 378

The social psychological prediction, however, goes against this common

sense. The same factors that keep a person in a relationship may be the ones
that influence biased perceptions of future happiness or unhappiness. Our
recent research (including data still being collected at the time this chapter was

A)
written) suggests that victims exhibit the same bias as nonvictims in predicting
their (un)happiness following a breakup, even though victims report less hap-

P
(A
piness overall and less satisfaction in their relationships as compared with non-
victims. In an initial cross-sectional survey study (n = 165), dating participants

n
were asked to forecast their happiness immediately following, 6 months after,

io
and 1 year after the dissolution of their relationship. Their responses were com-

at
pared with current happiness ratings of individuals who had been in a past

ci
aggressive relationship and who experienced the aggression either within the

so
last month, 6 months, or 1 year.

As
Several interesting findings emerged for victims of partner aggression.
First, current victims forecasted they would feel much less happy if their rela-

al
tionship were to end than was actually the case among past victims (i.e., those

ic
whose aggressive relationship had already ended). This was the case for every

og
time frame (i.e., immediately after, 6 months after, and 1 year after the dissolu-
ol
tion). Second, current victims reported significantly lower current happiness
ch
and relationship satisfaction than individuals whose current partners were not
y

aggressive (i.e., nonvictims). In short, current victims were less happy and less
Ps

satisfied than their nonvictim counterparts but just as likely to overestimate the
negative impact on their well-being of relationship dissolution. This finding
an

contradicts the belief that current unhappiness in an unsatisfying relationship

ic

will motivate victims to forecast more happiness if their relationship ends.

er

Instead, victims forecast levels of unhappiness comparable with those predicted

Am

by nonvictims.
Third, regardless of current victimization status, level of commitment
e

was strongly related to forecasting future unhappiness were the relationship

th

to end, r(76) = −.58, p < .001. We are currently collecting longitudinal data
by

to compare with the findings from the cross-sectional study. In short, regard-
less of whether perceptions concern past partner aggression or the future of a
11

relationship, victims of partner aggression exhibit the same motivated cogni-

20

tions as nonvictims, cognitions strongly associated with level of commitment.

©
ht

IMPLICATIONS FOR INTERVENTIONS

ig

AND VICTIM WELL-BEING

yr
op

We have shown that victims of partner aggression find themselves in a

very challenging situation: They seek a partner who will love them and yet
C

their partner is capable of harming them. Downplaying partner aggression may

378 ARRIAGA AND CAPEZZA

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 379

help victims manage the paradox that arises from choosing to remain with an
aggressive partner (Dunham & Senn, 2000) and is one way of coping with a
severely troubling situation. As Dunham and Senn (2000) noted, this coping
strategy may have benefits and costs. In the short term, victims can find a

A)
“mental space” within which to sustain their relationships. In the long term,
however, denying or justifying a partner’s aggression may keep a victim from

P
(A
recognizing the acts as aggressive, detecting their negative consequences, and
seeking help.

n
An obvious way to avoid the devastating effects of partner aggression is

io
to prevent it from occurring. Major advances have been made over the last

at
decades in understanding perpetrators of partner aggression (Arriaga &

ci
Capezza, 2005; see also Chapters 2 and 14, this volume). Preventing the nega-

so
tive consequences of partner aggression, however, is not limited to ending the

As
violence itself. It also involves helping victims of violence who are at risk of
negative mental and/or physical health outcomes (Arriaga & Capezza, 2005).

al
A major challenge in helping victims is simply finding them. Victims who

ic
downplay aggression often do not perceive a problem and thus fail to seek help.

og
Unfortunately, aggressive partners often convince victims that the situation is
ol
not so bad, that other relationships would be worse, or that things will get bet-
ch
ter. At the start of their relationship, the perpetrator may have had qualities to
y

attract the victim, and she may hope that those qualities will prevail and replace
Ps

the aggressive tendencies. It should come as no surprise, then, that a victim may
decline over time—feel more depressed, anxious, and uncertain—and yet not
an

pinpoint a “serious” problem.

ic

Other victims recognize a problem but cannot leave their partner. They
er

may be financially dependent on the partner (Rusbult & Martz, 1995). Alter-
Am

natively, they may feel increasingly dependent as a result of feeling unworthy

of anyone else. Psychologically aggressive partners often convince the victim
e

that she is to blame for the aggression. As she comes to feel increasingly respon-
th

sible and intent on repairing the relationship, she becomes further exposed to
by

dangerous conditions—isolation from supportive others, coming to feel that she

will “never do better”—which may make her less able or motivated to seek help
11

(Ferraro & Johnson, 1983). More psychological aggression triggers a downward

20

spiral of more victim self-blame, self-loathing, and loss of self, which increases
dependence and susceptibility to future aggression (Kirkwood, 1993). Many
©

victims accurately fear more violence if they try to leave.

ht

Interventions need to be tailored depending on the specific circum-

ig

stances of the victim. Financial dependence might best be addressed with

yr

immediate and sustained financial assistance (e.g., providing shelter, job train-
op

ing, childcare, and assistance in securing a home and employment). Emotional

dependence might best be addressed by (a) redefining what constitutes a
C

healthy, loving, and committed relationship; (b) recognizing highly aggressive

THE PARADOX OF PARTNER AGGRESSION 379

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 380

partner acts as destructive and understanding how they exert their nefarious
effects on the victim; and (c) restoring self-esteem, self-confidence, and a sense
that one is capable of happiness alone or in another relationship. Denying
aggression might best be addressed by a similar approach, especially by identi-

A)
fying destructive partner acts.
Would reducing commitment and dependence be the basis of an effec-

P
(A
tive intervention? It is violent partners, not commitment or dependence, that
causes harm to victims. Commitment and dependence are states that charac-

n
terize a multitude of close relationships, are crucial to relationship well-being,

io
and predict a vast array of relationship-maintenance behaviors. Indeed,

at
healthy, committed relationships may be a source of physical and emotional

ci
well-being (Kiecolt-Glaser & Newton, 2001). It is not commitment per se

so
that needs to change but personal and societal norms that are conducive

As
to the acceptance of violence. As such, interventions might address find-
ing nurturing partners and sustaining a strong commitment to them rather

al
than mistakenly suggesting that commitment or dependence per se is the

ic
source of distress.

og
Reducing the toll of partner aggression on mental and physical health
ol
requires changing perceptions of partner aggression. As victims adopt less
ch
embellished, more negative perceptions of the partner’s behavior, they
y

become more emotionally ready to pursue their own goals with or without the
Ps

partner. As societies strengthen and articulate negative views of all forms of

partner aggression—physical and psychological—they may be more likely to
an

(a) help and, importantly, support victims; (b) ostracize those who engage in
ic

partner aggression; and (c) support policies that will eradicate all forms of
er

partner aggression. Eradicating partner aggression will eliminate the paradox-

Am

ical situations in which victims must choose between sustaining unwarranted

partner-enhancing perceptions and sustaining their own well-being.
e
th
by

REFERENCES
11

Arias, I., & Pape, K. T. (1999). Psychological abuse: Implications for adjustment and
commitment to leave violent partners. Violence and Victims, 14, 55–67.
20

Arriaga, X. B. (2002). Joking violence among highly committed individuals. Journal

©

of Interpersonal Violence, 17, 591–610. doi:10.1177/0886260502017006001

ht

Arriaga, X. B. (2007, January). The cognitive entrapment of relationship violence victims.

ig

Invited address at the Relationships Preconference of the annual meeting of the

yr

Society for Personality and Social Psychology, Memphis, TN.

op

Arriaga, X. B., & Agnew, C. R. (2001). Being committed: Affective, cognitive, and
conative components of relationship commitment. Personality and Social Psy-
C

chology Bulletin, 27, 1190–1203. doi:10.1177/0146167201279011

380 ARRIAGA AND CAPEZZA

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 381

Arriaga, X. B., & Capezza, N. M. (2005). Targets of partner violence: The impor-
tance of understanding coping trajectories. Journal of Interpersonal Violence, 20,
89–99. doi:10.1177/0886260504268600
Capezza, N. M., & Arriaga, X. B. (2008a). Factors associated with acceptance of psy-
chological aggression against women. Violence Against Women, 14, 612–633.

A)
doi:10.1177/1077801208319004

P
Capezza, N. M., & Arriaga, X. B. (2008b). You can degrade but you can’t hit: Differ-

(A
ences in perceptions of psychological versus physical aggression. Journal of Social

n
and Personal Relationships, 25, 225–245. doi:10.1177/0265407507087957

io
Carlson, B. E., & Worden, A. P. (2005). Attitude and beliefs about domestic vio-

at
lence: Results of a public opinion survey: 1. Definitions of domestic violence,

ci
criminal domestic violence, and prevalence. Journal of Interpersonal Violence, 20,

so
1197–1218. doi:10.1177/0886260505278530

As
Cauffman, E., Feldman, S. S., Jensen, L. A., & Arnett, J. J. (2000). The (un)accept-
ability of violence against peers and dates. Journal of Adolescent Research, 15,

al
652–673. doi:10.1177/0743558400156003

ic
Coker, A. L., Davis, K. E., Arias, I., Desai, S., Sanderson, M., Brandt, H. M., &

og
Smith, P. H. (2002). Physical and mental health effects of intimate partner vio-
lence for men and women. American Journal of Preventive Medicine, 23, 260–268.
ol
doi:10.1016/S0749-3797(02)00514-7
ch
Coker, A. L., Smith, P. H., Bethea, L., King, M. R., & McKeown, R. E. (2000). Phys-
y

ical health consequences of physical and psychological intimate partner violence.

Ps

Archives of Family Medicine, 9, 451–457. doi:10.1001/archfami.9.5.451

an

Dunham, K., & Senn, C. Y. (2000). Minimizing negative experiences: Women’s

disclosure of partner abuse. Journal of Interpersonal Violence, 15, 251–261.
ic

doi:10.1177/088626000015003002
er

Ferraro, K. J., & Johnson, J. M. (1983). How women experience battering: The
Am

process of victimization. Social Problems, 30, 325–339. doi:10.1525/sp.1983.30.

3.03a00080
e

Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford Univer-

th

sity Press.
by

Follingstad, D. R., Rutledge, L. L., Berg, B. J., Hause, E. S., & Polek, D. S. (1990).
The role of emotional abuse in physically abusive relationships. Journal of Fam-
11

ily Violence, 5, 107–120.

20

Garcia-Moreno, C., Heise, L., Jansen, H. A. F. M., Ellsberg, M., & Watts, C. (2005,
November). Violence against women. Science, 310, 1282–1283. doi:10.1126/
©

science.1121400
ht

Gelles, R. J. (1993). Through a sociological lens: Social structure and family vio-
ig

lence. In R. J. Gelles & D. R. Loseke (Eds.), Current controversies on family vio-

yr

lence (pp. 31–46). Newbury Park, CA: Sage.

op

Gilbert, D. T., Pinel, E. C., Wilson, T. D., Blumberg, S. J., & Wheatley, T. (1998).
Immune neglect: A source of durability bias in affective forecasting. Journal of
C

Personality and Social Psychology, 75, 617–638. doi:10.1037/0022-3514.75.3.617

THE PARADOX OF PARTNER AGGRESSION 381

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 382

Heider, F. (1958). The psychology of interpersonal perception. New York, NY: Wiley.
doi:10.1037/10628-000
Johnson, M. P., & Ferraro, K. P. (2000). Research on domestic violence in the 1990s:
Making distinctions. Journal of Marriage and the Family, 62, 948–963. doi:10.1111/

A)
j.1741-3737.2000.00948.x

P
Kelley, H. H. (1979). Personal relationships: Their structures and properties. Hillsdale,

(A
NJ: Erlbaum.
Kelley, H. H., & Thibaut, J. W. (1978). Interpersonal relations: A theory of interdepen-

n
dence. New York, NY: Wiley

io
at
Kiecolt-Glaser, J. K., & Newton, T. L. (2001). Marriage and health: His and hers.

ci
Psychological Bulletin, 127, 472–503. doi:10.1037/0033-2909.127.4.472

so
Kirkwood, C. (1993). Leaving abusive partners: From the scars of survival to the wisdom
for change. Newbury Park, CA: Sage.

As
McClintock, C. G. (1972). Social motives: A set of propositions. Behavioral Science,

al
17, 438–454. doi:10.1002/bs.3830170505

ic
Murphy, C. M., & Hoover, S. A. (1999). Measuring emotional abuse in dating rela-

og
tionships as a multifactorial construct. Violence and Victims, 14, 39–53.
ol
O’Leary, K. D. (1999). Psychological abuse: A variable deserving critical attention
ch
in domestic violence. Violence and Victims, 14, 3–23.
y

Pierce, M. C., & Harris, R. J. (1993). The effect of provocation, race, and injury
Ps

description on men’s and women’s perceptions of a wife-battering incident.

Journal of Applied Social Psychology, 23, 767–790. doi:10.1111/j.1559-1816.
an

1993.tb01006.x
ic

Rennison, C. M., & Welchans, S. (2000). Intimate partner violence (Report No. NCJ
er

178247). Washington, DC: Bureau of Justice Statistics, U.S. Department of

Am

Justice.
Rusbult, C. E. (1983). A longitudinal test of the investment model: The development
e

(and deterioration) of satisfaction and commitment in heterosexual involve-

th

ments. Journal of Personality and Social Psychology, 45, 101–117. doi:10.1037/

0022-3514.45.1.101
by

Rusbult, C. E., & Martz, J. M. (1995). Remaining in an abusive relationship: An

11

investment model analysis of nonvoluntary commitment. Personality and Social

Psychology Bulletin, 21, 558–571. doi:10.1177/0146167295216002
20

Rusbult, C. E., Martz, J. M., & Agnew, C. R. (1998). The Investment Model Scale:
©

Measuring commitment level, satisfaction level, quality of alternatives, and

ht

investment size. Personal Relationships, 5, 357–387. doi:10.1111/j.1475-6811.

1998.tb00177.x
ig
yr

Rusbult, C. E., Olsen, N., Davis, J. L., & Hannon, P. A. (2001). Commitment and
op

relationship maintenance mechanisms. In J. Harvey & A. Wenzel (Eds.), Close

romantic relationships: Maintenance and enhancement (pp. 87–113). Mahwah, NJ:
C

Erlbaum.

382 ARRIAGA AND CAPEZZA

12051-21_CH20-rev2.qxd 5/25/10 12:00 PM Page 383

Sorenson, S. B., & Taylor, C. A. (2005). Female aggression toward male intimate
partners: An examination of social norms in a community-based sample. Psychol-
ogy of Women Quarterly, 29, 78–96. doi:10.1111/j.1471-6402.2005.00170.x
Straus, M. A., & Gelles R. J. (1990). Physical violence in American families: Risk factors

A)
and adaptations to violence in 8,145 families. New Brunswick, NJ: Transaction.
Thibaut, J. W., & Kelley, H. H. (1959). The social psychology of groups. New York,

P
(A
NY: Wiley.
Tjaden, P., & Theonnes, N. (2000). Prevalence, incidence, and consequences of violence

n
against women: Findings from the National Violence Against Women Survey (Report

io
No. NCJ 183781). Washington, DC: National Institute of Justice, U.S. Depart-

at
ment of Justice.

ci
Wilson, T. D., & Gilbert, D. T. (2003). Affective forecasting. In M. Zanna (Ed.),

so
Advances in experimental social psychology (Vol. 35, pp. 345–411). San Diego,

As
CA: Academic Press.
Yovetich, N. A., & Rusbult, C. E. (1994). Accommodative behavior in close rela-

al
tionships: Exploring transformation of motivation. Journal of Experimental Social

ic
Psychology, 30, 138–164. doi:10.1006/jesp.1994.1007

og
ol
y ch
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

THE PARADOX OF PARTNER AGGRESSION 383

C
op
yr
ig
ht
©
12051-21_CH20-rev2.qxd

20
11
5/25/10

by
th
e
12:00 PM

Am
er
ic
Page 384

an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 385

21

PA)
THE PSYCHOLOGICAL TOLL OF

(A
EXPOSURE TO POLITICAL VIOLENCE:

n
io
THE ISRAELI EXPERIENCE

at
ci
so
ZAHAVA SOLOMON AND KARNI GINZBURG

As
al
ic
og
ol
y ch
Ps
an

The proclivity of human beings for aggression, which too often culmi-
ic

nates in war, entails a high price in human suffering. Combatants, mostly men,
er

are naturally the ones who suffer most severely. In constant danger themselves,
Am

they witness the injury and death of friends and enemies and are repeatedly
exposed to the gruesome sights and sounds of slaughter. They struggle with fear
e

and loneliness as well as the more tangible deprivations of food, drink, and
th

sleep. The enormous destructive power of modern weapons and the uncertain-
by

ties of modern guerrilla warfare add to the already massive stress of war.
People have recognized the inevitability of psychological wounds of war
11

since biblical times at least. However, although war-induced psychological

20

disorders receive public attention during and immediately after a war, little
attention is devoted to the long-term psychological tolls of war. Hence, until
©

recently, knowledge about this topic has been quite fragmented, and lessons
ht

learned in one war have easily been forgotten before the next one begins
ig

(Mangelsdorff, 1985).
yr

Life in Israel is characterized by recurrent and prolonged exposure to mil-

op

itary and political violence. After completing 3 years of compulsory military

service, all able-bodied men are required to serve in the Israel Defense Forces
C

385
12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 386

(IDF) reserves, which means that they are more or less continually exposed to
military violence or training for war. Thus, the postwar experiences of Israeli
veterans differ from those of veterans in other countries, where soldiers usu-
ally return to a stable civilian society after a war ends. Given that the context

A)
of postwar recovery has a substantial effect on traumatized veterans’ mental
health, one cannot necessarily extrapolate to Israeli veterans from findings of

P
(A
studies conducted in other parts of the world.
In this chapter, we bring together findings from 3 decades of Israeli

n
studies of male combatants designed to shed light on the immediate and

io
long-term effects of recurrent and prolonged exposure to military violence.

at
In particular, the chapter is based on our own prospective longitudinal

ci
assessment of two cohorts: (a) veterans of the First Lebanon War (1982) and

so
(b) ex-prisoners of war (POWs) from the Yom Kippur War (1973).

As
al
COMBAT-INDUCED STRESS DISORDERS

ic
og
The most common and conspicuous psychological sequelae of combat are
ol
combat stress reaction, an acute response, and posttraumatic stress disorder, a
ch
chronic disorder.
y
Ps

On the Battlefield: Combat Stress Reactions

an

Exposure to battlefield violence is bound to arouse anxiety, which is a

ic

perfectly normal response to an imminent threat. The arousal of moderate

er

anxiety seems to be adaptive because it increases concentration and vigilance,

Am

which soldiers need on the battlefield. It is perhaps surprising that most com-
batants manage their anxiety fairly well and remain psychologically immune
e

to the terrible violence of modern warfare. They continue to function as sol-

th

diers, do not endanger themselves or their fellow comrades, and do not ask to
by

be evacuated (Solomon, 1993).

Not all combatants are as fortunate, however. A small but important pro-
11

portion of combatants are overwhelmed by their battlefield anxiety (Solomon,

20

1993). They perceive the threat as so intense, prolonged, and uncontrollable

that they feel totally vulnerable, helpless, and powerless. These are character-
©

istic signs of battlefield-related psychological breakdown, known in various

ht

places and eras as combat stress reaction (CSR), shell shock, combat fatigue, or war
ig

neurosis, among other terms (Solomon, 1993). CSR occurs when a soldier is
yr

stripped of his psychological defenses, feels so overwhelmed by the threat that

op

he becomes powerless to fight, and is inundated by feelings of utter helplessness

and anxiety. In such a state, the soldier is a danger to himself and others and is
C

no longer able to perform military duties.

386 SOLOMON AND GINZBURG

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 387

CSR is characterized by polymorphic and labile symptoms (Solomon,

1993). Psychosomatic symptoms, for example, range from loss of bladder and
bowel control, trembling, stuttering, and vomiting, to conversion reactions,
such as blindness and paralysis without identified organic cause. Cognitive

A)
symptoms include confusion, temporal and spatial disorientation, and impaired
attention, memory, judgment, and decision making. The main emotional

P
(A
symptoms are alternating states of paralyzing anxiety and deep depression.
Behavioral symptoms include manifestations of these emotions: great agitation,

n
on the one hand, and apathy and withdrawal, on the other. Some of the symp-

io
toms are quite bizarre, with soldiers tearing off their uniforms and running amok

at
toward the enemy or becoming frozen in their tracks, or clinging to a piece of

ci
clothing. These manifestations change as rapidly as the emotional state that

so
underlies them and can be quite perplexing to an observer (Solomon, 1993).

As
The reported prevalence rates of CSR vary considerably, both within and
among wars (Solomon, 1993). The reported rates among World War II’s

al
Allied soldiers ranged from 10% to 48% of injured soldiers. In the Vietnam

ic
War, rates were significantly lower, with the official figure being 1.2% of the

og
total number of American soldiers wounded on the battlefield. In the Yom
ol
Kippur War, the official count was 10% of Israeli soldiers wounded in action,
ch
although in some units it was as high as 70%. In the First Lebanon War, the
y

official figure was 23% of Israelis wounded in action, meaning that one of every
Ps

four war casualties was a psychiatric casualty. More recently, rates of only 3%
of CSR casualties were reported among British veterans evacuated from the
an

battle zone in Iraq (Turner et al., 2005). This huge variation seems to reflect
ic

differences in identification and counting of CSR casualties as well as differ-

er

ences in the amount and severity of destruction and atrocities to which sol-
Am

diers are exposed on the battlefield. However, despite this variability, the
reported rates clearly indicate that CSR is an inevitable and common conse-
e

quence of war.
th
by

The Aftermath of War: Posttraumatic Stress Disorder and Comorbidities

11

CSR can be a transient state for some veterans, but for others it marks
20

the beginning of a process of posttraumatic decline. At the end of a war, the

debilitating effects of CSR may abate in some cases. In other veterans, how-
©

ever, profound and prolonged psychological sequelae may remain and consol-
ht

idate into chronic and pervasive psychiatric disorders, such as posttraumatic

ig

stress disorder (PTSD). PTSD is a constellation of repeated reexperiencing of a

yr

traumatic event (e.g., unwanted mental intrusions, memories, images, dreams),

op

reduced involvement with the external world (i.e., trauma-related avoidance),

cognitive–affective hyperarousal, and impaired functioning (American Psychi-
C

atric Association, 1994).

THE PSYCHOLOGICAL TOLL OF EXPOSURE TO POLITICAL VIOLENCE 387

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 388

To understand these individual differences in long-term effects of expo-

sure to war violence, we initiated in 1982 a prospective longitudinal study of
all identified Israeli CSR casualties of the First Lebanon War and a matched
control group (for details, see Solomon, 1993). Data collection in this unique

A)
study commenced during the war itself, and the follow-up study is still con-
tinuing. Hundreds of veterans from both study groups have been interviewed

P
at four time points: 1, 2, 3, and 20 years after their active participation in the

(A
1982 war.

n
io
Chronic Posttraumatic Stress Disorder

at
ci
According to currently prevailing psychiatric formulations, PTSD is the

so
most common psychological injury of exposure to war violence (American
Psychiatric Association, 1994). Still, few empirical investigations have

As
assessed PTSD in identified CSR casualties. The results of our 20-year follow-

al
up study of all treated Israeli CSR casualties of the First Lebanon War

ic
(Solomon & Mikulincer, 2006) revealed that 64% of identified CSR casual-

og
ties suffered from PTSD symptoms 1 year after the war. At 2 years after the
war, the rate of PTSD casualties decreased to 59% and then further to 40%
ol
at 3 years. The most interesting and perhaps surprising finding, however, was
ch
that the rate of PTSD casualties increased again to 53% 20 years after the war.
y

In other words, the war does not end for many traumatized veterans; instead,
Ps

CSR marks the beginning of a lifelong psychological vulnerability.

Combat-induced PTSD was also identified among some veterans in our
an

control group who were not identified as CSR casualties on the battlefield.
ic

Fourteen percent of them reported suffering from PTSD 1 year after the war,
er

22% after 2 years, 11% after 3 years, and 27% 20 years later. These figures
Am

point to a long-term detrimental effect of war violence even for soldiers who
survive the immediate stress of battlefield without any identified psychologi-
e

cal breakdown.
th

It is important to note that the afflicted soldiers in our control group did
by

not seek help. Many of them were probably unaware that they suffered from
CSR symptoms or believed that these symptoms were a natural and inevitable
11

outcome of their horrific battlefield experiences. Others probably did realize

20

their plight but were reluctant to seek help. It is likely that these veterans
with silent PTSD are a mere fraction of a much larger number of psychiatric
©

casualties of war whose distress is similarly unidentified and untreated. This

ht

observation is consistent with studies demonstrating that many traumatized

ig

American veterans suffered from PTSD for years without seeking help.
yr

Although we realized that PTSD often goes untreated, it was still sur-
prising and troubling in the Israeli context. Had the PTSD casualties sought
op

help for their war-related disturbances at any IDF mental health clinic, they
C

388 SOLOMON AND GINZBURG

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 389

would have averted the risk of being sent back to the front. In our view, this
disinclination to seek help can be explained in part by the finding that PTSD
among our control veterans was less severe and less distressing than that of
CSR veterans (Solomon, 1993). Earlier studies have shown that the propen-

A)
sity to seek treatment for psychiatric disorders and the propensity to adopt the
sick role are closely related to symptom severity (e.g., Nadler, 1983).

P
(A
Alternatively, it is possible that PTSD casualties in the control group
were highly motivated to continue serving in the army and were not interested

n
in obtaining any potential secondary gain. At the time of the study, masculine

io
identity was strongly associated with military service in Israel. Identifying one-

at
self as ill may be costly in terms of both self-esteem and social approval. Vet-

ci
erans who sustained a CSR had to contend with the implications of their

so
breakdown, with the shame and guilt of having let down their buddies and

As
having betrayed the trust placed in them by their family and nation. Consid-
ering the great importance attributed to the army in Israel, the silent PTSD

al
sufferers’ sense of failure and injured manhood must weigh heavily on them

ic
and perhaps contributes to their continuing PTSD.

og
The high rates of PTSD casualties in both of our study groups are
ol
intriguing; they may reflect the recurrent exposure to military and political
ch
violence in Israel. All of our study participants, like other Israeli men of their
y

age, have served in the IDF reserves and could have been recalled for active
Ps

military duty during the first and second intifadas (Palestinian rebellions), the
Second Lebanon War, and other military operations in Lebanese and Pales-
an

tinian territories. Like all other Israeli citizens, they have been repeatedly
ic

exposed to Palestinian terror attacks, suicidal bombers, and Iraqi missile attacks
er

during the last 20 years. We believe that such recurrent exposure to war may
Am

have impeded the recovery of many of the traumatized veterans.

The unexpected rise in rates of PTSD casualties 20 years after the war
e

may have two main causes. First, 20 years after the war most of our participants
th

were in midlife. This is a particularly high-risk period for either delayed-onset

by

or reactivated PTSD because it often involves some reduction in work and

other activities and provides an opportunity, welcome or not, to reminisce and
11

review one’s life (Solomon & Mikulincer, 2006). This transition may allow
20

forgotten or suppressed traumatic memories to resurface (Solomon & Prager,

1992). Second, the fourth wave of measurement, 20 years after the war, was
©

carried out in the midst of the second intifada (2003), when the Israeli popu-
ht

lation was exposed to numerous terror attacks and sights of destruction, injuries,
ig

and violent death. This intense exposure of the previously traumatized veter-
yr

ans to harsh political violence may have reactivated the psychological break-
op

down they had experienced 20 years earlier, during or immediately after the
First Lebanon War.
C

THE PSYCHOLOGICAL TOLL OF EXPOSURE TO POLITICAL VIOLENCE 389

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 390

Effects of Repeated Exposure to Combat

Israel’s many wars have both obliged and enabled us to study the impact
of recurrent combat exposure. Because most countries are fortunate enough
not to call up the same soldiers to serve in repeated wars, the available litera-

A)
ture on combat stress has little to offer on this issue. However, studies of other

P
adverse experiences provide three alternative theoretical perspectives on the

(A
potential effects of recurrent exposure to war violence. First, the vulnerability
perspective (Coleman, Butcher, & Carson, 1980) considers repeated exposure

n
io
to stressful events to be a risk factor because it drains a person’s coping

at
resources. Second, the stress inoculation perspective (Epstein, 1983) holds that

ci
repeated exposure to stressful events serves as an “immunizer” because it fos-

so
ters the development of more effective coping strategies and thereby promotes

As
adaptation. Third, the stress resolution hypothesis (Block & Zautra, 1981) pro-
poses that what matters is not so much the person’s repeated exposure to a par-

al
ticular stressful event but how he or she coped with it. According to Block and

ic
Zautra (1981), successful coping leads to a sense of well-being and to increased

og
coping resources, whereas unsuccessful coping leads to increased vulnerability
and distress and an erosion of coping resources. ol
ch
To examine the effects of recurrent exposure to war violence, we asked
our sample of First Lebanon War veterans to indicate whether they had partic-
y
Ps

ipated as combatants in previous Israeli wars and whether they had sustained a
psychological breakdown in each of them (Solomon, Mikulincer, & Jacob,
an

1987). We found that the highest rate of CSR casualties in the First Lebanon
ic

War was among soldiers who had experienced a psychological breakdown in a

er

previous Israeli war (66%); the lowest rate was among soldiers who had fought
Am

in previous Israeli wars without experiencing a psychological breakdown

(44%). Among soldiers with no prior war experience, the rate of CSR casual-
ties fell in between (57%).
e
th

These findings suggest that successful resolution of previous war stress

helps soldiers cope with subsequent battles. But they also indicate that novice
by

soldiers are better off than those who had collapsed in a previous war. Although
11

not every soldier who sustains CSR is doomed to a second breakdown under
similar circumstances, CSR leaves most casualties more vulnerable the second
20

time around (Coleman et al., 1980). We should note that only a minority of
©

CSR casualties go on to fight in subsequent wars. It is a highly select group, hav-

ing been deemed fit to return to active duty by the army and having displayed
ht

personal motivation to do so. If these veterans still display increased suscepti-

ig
yr

bility, other CSR casualties who were not allowed to return to military duties
would presumably be even more vulnerable.
op

The detrimental impact of recurrent exposure to combat violence becomes

C

more apparent when the number of previous wars is taken into account. Among

390 SOLOMON AND GINZBURG

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 391

soldiers with a prior psychological breakdown, CSR rates in the First Lebanon
War increased linearly with the number of prior war experiences: 57% after one
war, 67% after two, and 83% after three. Among soldiers who had fought with-
out a prior psychological breakdown, CSR rates in the First Lebanon War were

A)
curvilinear: Veterans who actively participated in either one or three previous
wars had higher CSR rates (50% and 44%, respectively) than did soldiers who

P
(A
participated in two (33%). Taken together, our research findings suggest that
exposure to military violence scars an individual and weakens his resistance

n
to future combat-related violence. They also suggest that whatever the pos-

io
sible inoculation benefits of successful stress resolution, repeated trauma may

at
eventually break even the hardiest souls.

ci
so
Reactivation of Traumatic Reactions

As
Reactivation of traumatic reactions is a well-known phenomenon. Sur-

al
vivors of many traumatic events, including the Holocaust, combat, and rape,

ic
among others, can experience a reactivation of their response to the original

og
trauma when they are reminded of it (Christenson, Walker, Ross, & Malthie,
1981; Lindemann, 1944). ol
ch
As we have noted, 66% of veterans who suffered from psychological break-
down in the Yom Kippur War showed a recurrent CSR episode in the First
y
Ps

Lebanon War. Studying these cases in a deeper and more detailed way, we
found that they represented two different types of trauma-related reactivation
an

(e.g., Solomon, Garb, Bleich, & Grupper, 1987). Twenty-three percent of these
ic

cases exhibited uncomplicated reactivations or classic reactivations. These vet-

er

erans seemed to have completely recovered from their previous Yom Kippur
Am

War-related psychological breakdown in 1973. They were virtually symptom

free between the wars. The first indication that all was not well came with the
emergence of CSR during the First Lebanon War, which was generally precip-
e
th

itated by a threatening incident directly reminiscent of their Yom Kippur War

experience. The remaining cases (77%) can be more aptly termed exacerbated
by

PTSD. Here, the Yom Kippur War-related psychological breakdown in 1973

11

left more salient scars, and veterans continued to suffer from PTSD symp-
toms of high (51%), medium (9%), and low severity (17%) between the Yom
20

Kippur War and the First Lebanon War. Symptoms became intensified during
©

reserve duty, and the call-up notice to Lebanon provoked considerable antic-
ipatory anxiety. Moreover, these men were so vulnerable to combat-related
ht

violence that their CSR during the First Lebanon War had been triggered by
ig
yr

an incident unrelated to their original war trauma and, in many cases, one that
did not pose a direct or immediate danger.
op

All of these casualties had made great efforts to function effectively in

C

the 9 years between the Yom Kippur War and the First Lebanon War, and

THE PSYCHOLOGICAL TOLL OF EXPOSURE TO POLITICAL VIOLENCE 391

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 392

they had generally been successful. While relying on coping mechanisms such
as repression, suppression, denial, and cognitive avoidance, most of them
married, started families, and held jobs—and some did well professionally.
None were hospitalized. All continued to serve in the reserves, despite the

A)
fact that their symptoms were intensified in the presence of military violence.
Many hid their symptoms from their friends, families, and army commanders.

P
(A
Their recurrent CSR during the First Lebanon War revealed the psy-
chological damage that the first breakdown had created and seemed to deepen

n
the damage. In general, there were more symptoms during the recurrent CSR

io
episode in the First Lebanon War than during the first CSR episode in the

at
Yom Kippur War, and the symptoms were more intense and debilitating. Fur-

ci
thermore, even though some soldiers with prior CSR in the Yom Kippur War

so
participated in the First Lebanon War without further breakdown, the detri-

As
mental effects of the earlier episode were still detectable a decade later
(Solomon, 1993).

al
ic
og
LONG-TERM IMPLICATIONS OF WAR CAPTIVITY:
LESSONS FROM THE YOM KIPPUR WAR ol
y ch

Participation in active combat exposes soldiers to extreme physical and

Ps

mental stress that has pathogenic effects. For one group of soldiers, combat is
but the first step in a traumatic journey. For these soldiers who fall into enemy
an

hands, the war continues even after the shooting stops. These POWs continue
ic

to be exposed to prolonged and often even more extreme traumatic conditions.

er
Am

Stressors of War Captivity

e

At time of physical capture, a combatant is engaged directly and at short

th

range with his enemies, and brutal force is typically used to deprive him of his
by

autonomy (e.g., Avnery, 1982). During captivity, the POW is often held in
unsanitary and uncomfortable conditions and may be deprived of sufficient
11

amounts of food and water (e.g., Hunter, 1993). The POW may even be sub-
20

jected to brutal torture and interrogations, as well as humiliation. Mock execu-

tions may be carried out, and solitary confinement may be used as well.
©

Deprivation of benevolent human interactions enhances the captive’s depend-

ht

ency on his captors. The lack of social support, denial of privacy, and continu-
ig

ous torture and humiliation may cripple a POW’s identity and potentially pave
yr

the way for a breakdown of psychological defenses.

op

The trauma of captivity is unique in the sense that it entails recurrent

exposure to deliberate infliction of extreme physical and mental violence.
C

These stressors are added on to the extreme conditions and hazards that the

392 SOLOMON AND GINZBURG

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 393

POWs experienced during combat. In addition, although the experience of

war may be impersonal, captivity is characterized by personal interactions
between captives and captors, and as such it generates a unique form of con-
trolling and coercive relationship (Herman, 1992). Various methods of con-

A)
trol and coercion are used to deprive the POWs of their sense of autonomy
and replace it with a sense of horror and helplessness.

P
(A
Captors use various techniques to break a captive’s spirit. During the Yom
Kippur War, Israeli POWs were repeatedly exposed to anti-Israeli propaganda,

n
misinformed of the death of Israel’s leaders, and falsely told about the triumph

io
of Arab states over Israel and the occupation of Israel by Arab armies. At times,

at
captives were informed that their homes had been destroyed and their loved

ci
ones and relatives killed (e.g., Avnery, 1982). These were deliberate efforts to

so
exacerbate the POWs’ feelings of loneliness, weakness, and isolation.

As
The trauma of captivity was often made worse by prisoners’ feelings that
they had failed to meet the heroic standards emphasized by Israeli culture.

al
This ethos requires that prisoners of war refuse to disclose any information,

ic
endure the unbearable pain of interrogation without cracking, and even take

og
their own lives if necessary. This code of conduct is entwined with the ideal
ol
image of the Israeli combatant who is expected to “fight until the end” and
ch
“maintain his honor” during interrogations (Gavriely, 2006). Thus, many
y

Israeli POWs were regarded not only as soldiers who failed in the important
Ps

role of defending their homeland but also as a threat to national security

because they might have disclosed potentially sensitive information. This
an

public notion is mirrored by Israeli military law, which defines surrender to

ic

the enemy when not ordered to do so as a severe offense—a betrayal—that

er

may deserve the death penalty. Moreover, disclosure of secrets while in cap-
Am

tivity is defined as treason, a despicable behavior (Gavriely, 2006). This

intense conflict between personal survival and living up to cultural standards
e

is at the heart of what Avnery (1982) called the captive’s dilemma: staying
th

alive and sustaining criticism and condemnation versus complying with the
by

norm at the cost of losing one’s life. For many ex-POWs, this dilemma pro-
duced feelings of utter failure and unbearable weakness.
11

During the 1973 Yom Kippur War, several hundred Israeli soldiers were
20

captured on the Syrian and Egyptian fronts. POWs held in Egypt were released
relatively quickly, within 4 to 6 weeks. POWs in Syria were held for 8 months.
©

During captivity in Egypt, the prisoners were held in separate cells, whereas in
ht

Syria, after a rigorous interrogation period, POWs were held in groups, each
ig

in a large common room. In both states of captivity, Israeli soldiers were sub-
yr

jected to interrogation and torture designed to break them down mentally.

op

To study the psychological consequences of war captivity, we followed-

up ex-POWs from the Israeli army land forces who were taken captive on either
C

the Egyptian or the Syrian front during the Yom Kippur War, and we compared

THE PSYCHOLOGICAL TOLL OF EXPOSURE TO POLITICAL VIOLENCE 393

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 394

them with a control group of combat veterans who fought on the same fronts
as the ex-POWs during the same war but were not taken captive. Controls were
matched with the ex-POWs in terms of personal and military background. Par-
ticipants in the two groups (i.e., ex-POWs, controls) were interviewed twice:

A)
18 and 30 years after the war (see Solomon & Dekel, 2005, for details).

P
(A
Long-Term Psychological Sequelae of War Captivity

n
Findings from our study reveal that 3 decades after their release from cap-

io
tivity, 23% of former Israeli POWs still met symptom criteria from the Diag-

at
nostic and Statistical Manual of Mental Disorders (4th ed.; American Psychiatric

ci
Association, 1994) for PTSD, as compared with 5% of non-POW controls

so
(Solomon & Dekel, 2005). This figure points to both the resilience of the 77%

As
of former POWs who did not meet PTSD criteria and to the long-lasting psy-
chological damage of captivity to the remaining 23%. The question is why the

al
psychological damage of captivity should be so much more enduring than that

ic
of combat, which is itself pathogenic.

og
Several explanations may be offered. The simplest is based on the spe-
ol
cial hardships of captivity: the torture, humiliation, and isolation that are part
ch
and parcel of war captivity (Mollica et al., 1990). In addition, these hardships
y

are personal (Herman, 1992). That is, the threat of combat to the life and
Ps

physical integrity of the soldier is a relatively impersonal threat; it is directed

toward whomever is in the line of fire and not at any particular soldier. Thus,
an

there is no affront to the soldier’s personhood, even if he is injured. The trauma

ic

of captivity, however, occurs within a personal relationship between the cap-

er

tive and his captors. The special torments of captivity are part of a planned
Am

and concerted effort to “break” a particular person and are intentionally

inflicted on him by persons he gets to know and on whom he is dependent for
e

physical survival.
th

Another explanation for higher PTSD levels among POWs has to do

by

with the uniqueness of the social context of combat captivity. Combatants

are equipped with weapons and protective devices, and they fight alongside
11

commanders and comrades. The powerful stress-reducing effect of unit cohe-

20

sion and social support derived from comrades and commanders is a well-
documented sustaining force for combatants (e.g., Solomon, Mikulincer, &
©

Hobfoll, 1987). However, captivity renders the POW totally isolated and
ht

deprived of any human compassion or support. The severity of captivity may

ig

thus be compounded by isolation and loneliness, leaving a more profound

yr

and enduring traumatic imprint.

op

Still another possible explanation is that POWs adopt strategies that are
useful in captivity, such as suspiciousness and hyperalertness, and then apply
C

them in civilian life later on, where they are often counterproductive. Eberly,

394 SOLOMON AND GINZBURG

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 395

Harkness, and Engdahl (1991) suggested that traumatized POWs can be viewed
as survivors who continue to exhibit patterns of thought, emotion, and behav-
ior that were adaptive during their period of traumatization.
A fourth possible explanation focuses on the compounding of traumatic

A)
experiences. For most POWs, the trauma of captivity follows on the heels of
the trauma of combat. Captivity thus extends the duration of the traumatic

P
(A
experience, further drawing on the soldier’s already depleted coping resources
(Ursano et al., 1996). As is well known, the longer a traumatic experience

n
lasts, the more severe the ensuing psychiatric disorders are likely to be (Hunter,

io
1993). The cumulative damage of multiple traumas is known to be more severe

at
than the damage of a single trauma (Herman, 1992).

ci
so
Trajectories of Posttraumatic Stress Disorder: Changes Over Time

As
The theoretical literature on the longitudinal effects of trauma offers three

al
alternative views on the temporal course of postcaptivity symptomatology. One

ic
view is that time is a healer and that ex-POWs will recover partially or com-

og
pletely with the passage of time (e.g., Engdahl, Speed, Eberly, & Schwartz,
ol
1991; Potts, 1994). A second, different view is that PTSD is a chronic ailment
ch
that will intensify over time because of a natural decline in resilience, perhaps
y

especially in midlife (e.g., Maercker, 1999). This view gains some support from
Ps

a study that documented increased rates of PTSD symptoms over a period of

4 years in a group of community-dwelling former American POWs from World
an

War II and the Korean War (Port, Engdahl, & Frazier, 2001). The third view
ic

is that, barring an initial decline in psychological distress soon after a POW is

er

released from captivity, no clear temporal trajectory is discernable (Zeiss &

Am

Dickman, 1989). In this view, time can either heal or intensify the psycholog-
ical wounds of captivity, depending on an ex-POW’s personality and life expe-
e

riences after release. This view stresses the liability of PTSD symptomatology
th

and the possibility that life events and psychological changes or developments
by

can cause the symptoms to intensify or decline (Buffum & Wolfe, 1995).
Trauma researchers have not yet established which of these three views
11

best describes the long-term effects of war captivity. Too few studies tracing
20

the longitudinal effects of captivity have been carried out, and most have
assessed recovery and other changes in PTSD symptomatology only through
©

retrospective self-reports. In addition, the observed variability in the after-

ht

math of captivity, both between and within groups, is not well understood.
ig

In our longitudinal two-wave study conducted 18 and 30 years after

yr

the Yom Kippur War (Solomon & Dekel, 2005), we found that PTSD fol-
op

lowed a different course among ex-POWs and combat controls. Ex-POWs

were 10 times more likely than controls to experience deterioration in their
C

psychological condition in the 12-year interval between the two assess-

THE PSYCHOLOGICAL TOLL OF EXPOSURE TO POLITICAL VIOLENCE 395

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 396

ments. Almost 20% of ex-POWs who did not meet PTSD criteria 18 years
after their release did meet them at the 30-year assessment. This deteriora-
tion occurred in less than 1% of combat controls. Given the study design,
it is impossible to know whether this 20% rise in ex-POWs’ PTSD rates

A)
reflects reactivation or delayed onset of PTSD. Previous studies reported
delayed-onset PTSD in 11% (Green et al., 1990) and 20% (Wolfe, Erickson,

P
(A
Sharkansky, King, & King, 1999) of various traumatized groups.
POWs also showed a statistically significant increase in endorsement of

n
each of the PTSD symptom clusters (i.e., intrusion, avoidance, hyperarousal).

io
Among controls, in contrast, there was no change in endorsement of the three

at
symptom clusters, along with a downward trend in the endorsement of most

ci
PTSD symptoms. These findings indicate that time exacerbates the detrimen-

so
tal effects of war captivity. This increase in PTSD among ex-POWs is consis-

As
tent with findings of increased PTSD rates and symptom levels over a 4-year
measurement interval among American ex-POWs (Port et al., 2001) but dif-

al
fers from findings of decreased PTSD symptoms over time (Engdahl et al.,

ic
1991). The differences are probably related to the times of measurement in the

og
research studies.
ol
A previous study (Port et al., 2001) found a U-shaped pattern, with high
ch
PTSD rates immediately after captivity, followed by a gradual decline, and
y

then, from midlife onward, a rise in PTSD rates. It is possible that our first
Ps

assessment, taken 18 years after prisoners’ release, fell within the lower part of
the curve, and our second assessment, 12 years later, reflected the rising rates
an

as men aged. It is interesting to note that ex-POWs’ heightened PTSD is sim-

ic

ilar to increases in PTSD rates observed among First Lebanon War veterans
er

(reported earlier in this chapter). In our view, both of them can be explained
Am

either by aging or by the unremitting threat of war and terror in Israel.

Along with the vulnerabilities of ex-POWs found in this study, we should
e

mention the resilience of the non-POW veterans, who had low rates of PTSD
th

both 18 and 30 years after the war (3.8% and 4.8%, respectively). Even though
by

all of them had seen combat and, as noted, most continued to serve in active
reserve and were exposed to the ongoing threat of terror, they did not show any
11

sign of psychopathology. Among ex-POWs, the PTSD rates were considerably

20

higher, but the majority did not meet PTSD criteria at either time of assess-
ment. This high level of resilience in both groups lends further support to the
©

contention that resilience in the face of trauma is more common than psycho-
ht

pathology (American Psychiatric Association, 1994).

ig
yr

Posttraumatic Stress Disorder and Changes in Attachment Orientation

op

Studies of individuals who were subject to repeated abuse in which they

C

were helpless captives under control of their captors suggest that survivors

396 SOLOMON AND GINZBURG

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 397

tend to develop not only PTSD but also a unique form of posttraumatic seque-
lae that penetrates and alters their personality, often referred to as complex
PTSD or disorder of extreme stress not otherwise specified (DESNOS; Herman,
1992; Terr, 1991; van der Kolk, 2002). Herman (1992) suggested that pro-

A)
longed captivity disrupts captives’ personal relationships, which may result in
long-lasting attachment injuries and may be manifested in either anxious or

P
(A
avoidant attachment. Whereas anxious attachment characterizes individuals
who are concerned that their significant others will not be available in times

n
of need and who wish intensely for proximity to and care from others, avoidant

io
attachment characterizes individuals who cannot trust others and who with-

at
draw from intimacy and interdependence (Mikulincer & Shaver, 2007).

ci
To test Herman’s (1992) ideas, we assessed attachment orientations

so
among former POWs and controls 18 and 30 years after the Yom Kippur War,

As
and we examined reciprocal associations between PTSD and attachment ori-
entations over time (Solomon, Dekel, & Mikulincer, 2008). Paralleling the

al
increase in PTSD rates among ex-POWs between the two waves of assessment,

ic
there were increased levels of self-reported anxious and avoidant attachment

og
among ex-POWs during the same period. This trend was evident only in the
ol
ex-POW group; levels of anxious and avoidant attachment were quite stable
ch
in the control group. In addition, increases in both anxious and avoidant
y

attachment were positively associated with increases in PTSD symptoms.

Ps

These findings further highlight the pervasive impact of captivity. In line with
previous studies (van der Kolk, Roth, Pelcovitz, Sunday, & Spinazzola, 2005),
an

they suggest that exposure to repeated interpersonal trauma inflicts long-term

ic

psychological injuries that are not captured by the PTSD diagnosis alone.
er

Clinical observations have suggested that repeated and prolonged trauma

Am

may lead to major personality changes, including significant changes in one’s

identity and one’s behavior in close relationships (Herman, 1992; Terr, 1991).
e

Similarly, harsh and dramatic experiences, such as war captivity, may alter a
th

person’s basic trust in others in a way that undermines the ability to maintain
by

secure attachments. Hence, even ex-POWs who have had secure attachments
may become more anxious and may defensively avoid interpersonal contact fol-
11

lowing exposure to trauma. The fact that changes in attachment orientations

20

were found so many years after the war ended demonstrates the pervasive and
dramatic effects of intentional, human-inflicted victimization.
©
ht

Need for Professional Help and Help Seeking

ig
yr

Finally, our findings indicate that about twice as many ex-POWs as com-
op

bat controls felt that they needed psychotherapy, and about five times as many
ex-POWs as combat controls actually sought and obtained it. The observed
C

rates of seeking psychotherapy and being willing to admit a need for help

THE PSYCHOLOGICAL TOLL OF EXPOSURE TO POLITICAL VIOLENCE 397

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 398

among the ex-POW group were high relative to norms in Israel. They are tes-
timony to the intensity of the distress from which ex-POWs suffer and to the
increasing acceptance in Israeli society of seeking help following traumatic
military experiences (Gavriely, 2006).

A)
The higher rates of both reported need for help and actual help seeking
among ex-POWs may be explained by their greater trauma-related and gen-

P
eral distress and their lower recovery rates. Even those who received treatment

(A
were less prone to recover than combat controls who were treated. The com-

n
plex and prolonged stressors to which they were exposed may have contributed

io
to their intensive, pervasive, and widespread distress (Herman, 1992). The

at
lower rate of recovery among treated POWs than among untreated controls is

ci
further evidence not only of the difficulties in treating trauma but also of the

so
fact that the more massive the trauma, the more damage it causes and the more

As
difficult it is to ameliorate with professional intervention.

al
ic
CONCLUSIONS

og
The studies discussed in this chapter were conducted among Israeli com-
ol
batants and ex-POWs. They were based on longitudinal designs, with assess-
ch
ments conducted 20 and 30 years after exposure to war-related violence.
y

Findings showed that for many combatants and former POWs, exposure to
Ps

war violence, atrocities, and massive acts of destruction can cause severe,
an

long-lasting psychological disorders. Although the war ended and the prison
doors opened years ago, many of the combatants and ex-POWs are still faced
ic

on a daily basis with the pathogenic effects of political aggression. Moreover,

er

for many of these veterans, the picture has become even bleaker with time,
Am

as their mental and physical state has deteriorated.

Although our studies shed light on the enduring toll of political aggres-
e

sion, they also testify to the amazing resilience of many veterans. Further
th

research is needed to determine what differentiates resilient from nonresilient

by

war casualties and to devise more effective therapeutic interventions and pre-
ventive measures.
11

PTSD is the only psychiatric disorder that directly follows exposure to a

20

recognized traumatic stressor. Thus, the best way to prevent combat-induced

psychopathology is to stop making war. We have not yet been able to do that
©

in all of recorded history, so we must continue to rely on our ability to learn

ht

more about causes of psychopathological disorders following traumatic expe-

ig

riences and ways to reduce their prevalence and severity. More thorough and
yr

systematic research on PTSD can increase our understanding of effective ways

op

to help traumatized individuals. Increased awareness of their plight may open

our hearts to these men and chip away at the denial that generally blinds us to
C

their suffering.

398 SOLOMON AND GINZBURG

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 399

REFERENCES

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental

disorders (4th ed.). Washington, DC: Author.

A)
Avnery, A. (1982). Coping and adjustment to war captivity (Unpublished master’s
thesis). Hebrew University, Jerusalem, Israel.

P
(A
Block, M., & Zautra, A. J. (1981). Satisfaction and distress in the community: A test
of the effects of life events. American Journal of Community Psychology, 9, 165–180.

n
doi:10.1007/BF00896365

io
Buffum, M. D., & Wolfe, N. S. (1995). Posttraumatic stress disorder and the World

at
War II veteran: Elderly patients who were in combat or were prisoners of war

ci
may have special health care needs that may not be obvious. Geriatric Nursing,

so
16, 264–270. doi:10.1016/S0197-4572(95)80006-9

As
Christenson, R. M., Walker, J. L., Ross, D. R., & Malthie, A. A. (1981). Reactivation
of traumatic conflicts. The American Journal of Psychiatry, 138, 984–985.

al
Coleman, J. C., Butcher, J. N., & Carson, R. C. (1980). Abnormal psychology and

ic
modern life. Glenview, IL: Scott Foresman.

og
Eberly, R. E., Harkness, A. R., & Engdahl, B. E. (1991). An adaptational view of
ol
trauma response as illustrated by the prisoner of war experience. Journal of
ch
Traumatic Stress, 4, 363–380. doi:10.1002/jts.2490040305
Engdahl, B. E., Speed, N., Eberly, R. E., & Schwartz, J. (1991). Comorbidity of psy-
y
Ps

chiatric disorders and personality profiles of American World War II prisoners

of war. Journal of Nervous and Mental Disease, 179, 181–187. doi:10.1097/
an

00005053-199104000-00001
Epstein, S. (1983). Natural healing processes of the mind: II. Graded stress inoculation
ic

as an inherent coping mechanism. In D. Meichenbaum & M. Jaremko (Eds.),

er

Stress prevention and management: A cognitive behavioral approach (pp. 39–66). New
Am

York, NY: Plenum Press.

Gavriely, D. (2006). Israel’s cultural code of captivity and the personal stories of Yom
e

Kippur War ex-POWs. Armed Forces and Society, 33, 94–105. doi:10.1177/
th

0095327X05282531
by

Green, B. L., Lindy, J. D., Grace, M. C., Gleser, G. C., Leonard, A. C., Korol, M., &
Winget, C. (1990). Buffalo Creek survivors in the second decade: Stability of
11

stress symptoms. American Journal of Orthopsychiatry, 60, 43–54. doi:10.1037/

20

h0079168
Herman, J. L. (1992). Trauma and recovery. New York, NY: Basic Books.
©

Hunter, E. J. (1993). The Vietnam prisoner of war experience. In J. P. Wilson

ht

& B. Raphael (Eds.), International handbook of traumatic stress syndromes

(pp. 297–303). New York, NY: Plenum Press.
ig
yr

Lindemann, E. (1944). Symptomatology and management of acute grief. The Amer-

ican Journal of Psychiatry, 101, 141–148.
op

Maercker, A. (1999). Lifespan psychological aspects of trauma and PTSD: Symptoms

C

and psychosocial impairments. In A. Maercker, M. Schutzwohl, & Z. Solomon

THE PSYCHOLOGICAL TOLL OF EXPOSURE TO POLITICAL VIOLENCE 399

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 400

(Eds.), Posttraumatic stress disorder: A lifespan developmental perspective (pp. 7–42).

Seattle, WA: Hogrefe & Huber.
Mangelsdorff, A. D. (1985). Lessons learned and forgotten: The need for prevention
and mental health interventions in disaster preparedness. Journal of Commu-

A)
nity Psychology, 13, 239–257. doi:10.1002/1520-6629(198507)13:3<239::AID-
JCOP2290130302>3.0.CO;2-T

P
(A
Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynamics,
and change. New York, NY: Guilford Press.

n
io
Mollica, R. F., Wyshak, G., Lavelle, J., Truong, T., Tor, S., & Yang, T. (1990).

at
Assessing symptom change in Southeast Asian refugee survivors of mass vio-
lence and torture. The American Journal of Psychiatry, 147, 83–88.

ci
so
Nadler, A. (1983). Personal characteristics and help-seeking. In J. D. Fisher,
A. Nadler, & B. M. DePaulo (Eds.), New directions in helping: Help-seeking (Vol. 2,

As
pp. 124–151). New York, NY: Academic Press.

al
Port, C. L., Engdahl, B., & Frazier, P. (2001). A longitudinal and retrospective study

ic
of PTSD among older prisoners of war. The American Journal of Psychiatry, 158,

og
1474–1479. doi:10.1176/appi.ajp.158.9.1474
ol
Potts, M. K. (1994). Long-term effects of trauma: Posttraumatic stress among civilian
internees of the Japanese during WW2. Journal of Clinical Psychology, 50, 681–698.
ch
doi:10.1002/1097-4679(199409)50:5<681::AID-JCLP2270500504>3.0.CO;2-3
y
Ps

Solomon, Z. (1993). Combat stress reaction: The enduring toll of war. New York, NY:
Plenum Press.
an

Solomon, Z., & Dekel, R. (2005). Posttraumatic stress disorder among Israeli ex-
ic

prisoners of war 18 and 30 years after release. The Journal of Clinical Psychiatry,
er

66, 1031–1037. doi:10.4088/JCP.v66n0811

Am

Solomon, Z., Dekel, R., & Mikulincer, M. (2008). Complex trauma of war captivity:
A prospective study of attachment and PTSD. Psychological Medicine, 38,
1427–1434. doi:10.1017/S0033291708002808
e
th

Solomon, Z., Garb, R., Bleich, A., & Grupper, D. (1987). Reactivation of combat-
related posttraumatic stress disorder. The American Journal of Psychiatry, 144,
by

51–55.
11

Solomon, Z., & Mikulincer, M. (2006). Trajectories of PTSD: A 20-year longitudi-

20

nal study. The American Journal of Psychiatry, 163, 659–666. doi:10.1176/

appi.ajp.163.4.659
©

Solomon, Z., Mikulincer, M., & Hobfoll, S. (1987). Objective versus subjective
ht

measurement of stress and social support: The case of combat-related reactions.

ig

Journal of Consulting and Clinical Psychology, 55, 577–583. doi:10.1037/0022-

yr

006X.55.4.577
op

Solomon, Z., Mikulincer, M., & Jacob, B. R. (1987). Exposure to recurrent combat
stress: Combat stress reactions among Israeli soldiers in the Lebanon War. Psy-
C

chological Medicine, 17, 433–440. doi:10.1017/S0033291700024995

400 SOLOMON AND GINZBURG

12051-22_CH21-rev2.qxd 5/25/10 12:01 PM Page 401

Solomon, Z., & Prager, E. (1992). Elderly Israeli Holocaust survivors during the
Persian Gulf War: A study of psychological distress. The American Journal of
Psychiatry, 149, 1707–1710.
Terr, L. C. (1991). Childhood traumas: An outline and overview. The American Jour-

A)
nal of Psychiatry, 148, 10–20.
Turner, M. A., Kiernan, M. D., McKechanie, A. G., Finch, P. J. C., McManus, F. B.,

P
(A
& Neal, L. A. (2005). Acute military psychiatric casualties from the war in Iraq.
The British Journal of Psychiatry, 186, 476–479. doi:10.1192/bjp.186.6.476

n
Ursano, R. J., Rundell, J. R., Fragala, M. R., Larson, S. G., Jaccard, J. T., Wain,

io
H. J., . . . Beach, B. L. (1996). The prisoner of war. In R. J. Ursano & A. E.

at
Norwood (Eds.), Emotional aftermath of the Persian Gulf War (pp. 443–476).

ci
Washington, DC: American Psychiatric Press.

so
van der Kolk, B. A. (2002). The assessment and treatment of complex PTSD. In

As
R. Yehuda (Ed.), Treating trauma survivors with PTSD (pp. 127–156). Washington,
DC: American Psychiatric Press.

al
van der Kolk, B. A., Roth, S. H., Pelcovitz, D., Sunday, S., & Spinazzola, J. (2005).

ic
Disorders of extreme stress: The empirical foundation of a complex adaptation

og
to trauma. Journal of Traumatic Stress, 18, 389–399. doi:10.1002/jts.20047
ol
Wolfe, J., Erickson, D. J., Sharkansky, E. J., King, D. W., & King, L. A. (1999). Course
ch
and predictors of posttraumatic stress disorder among Gulf War veterans: A
y

prospective analysis. Journal of Consulting and Clinical Psychology, 67, 520–528.

Ps

doi:10.1037/0022-006X.67.4.520
Zeiss, R. A., & Dickman, H. R. (1989). PTSD 40 years later: Incidence and person-
an

situation correlates in former POWs. Journal of Clinical Psychology, 45, 80–87.

ic

doi:10.1002/1097-4679(198901)45:1<80::AID-JCLP2270450112>3.0.CO;2-V
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

THE PSYCHOLOGICAL TOLL OF EXPOSURE TO POLITICAL VIOLENCE 401

C
op
yr
ig
ht
©
12051-22_CH21-rev2.qxd

20
11
5/25/10

by
th
e
12:01 PM

Am
er
ic
Page 402

an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 403

INDEX

Aarts, H., 234 Aggression. See also specific headings

A)
Aberrant individuals, 25 across life span, 124–126

P
Abnormal populations, 107 and attachment, 246–249

(A
Abu-Aish, Az-a-Din, 306 behavior genetic studies of, 146–147
Abu Ghraib prison, 15 cognitive factors in, 18–19

n
Abuse, in childhood. See Child continuity of. See Continuity of

io
maltreatment aggression

at
Abusive personalities, 262–265 defined, 3–4, 18, 35, 211

ci
Action tendencies, 316 development of, 156–157
displaced, 224–225

so
Activation, of power system, 76–79
Adaptation, 58–60 forms of, 151–153

As
Adaptationist program, 53 between groups of people, 26–27
Additive genetic influences, 149 heritability of, 176–177

al
ADHD (attention-deficit hyperactivity influence of government actions on,

ic
disorder), 157 27–29

og
Adolescents, 125–126, 169–170 operationalizations of, 144
aggression prevention/intervention power vs., 75
strategies for, 212
ol
reduction of, 29
ch
reinterpretation of, 373–377
externalizing spectrum in, 93–95
and revenge, 221–222
y

lower order traits in, 94–95

Ps

in reward system, 113

narcissism in, 208
and self-views, 209–211
respect for parents in, 283–284
social–cognitive models of, 19, 165
an

self-esteem as cause of aggression in,

in societies, 17
206–207
ic

Aggression-prone individuals, 29
violent intimate relationships of, 175
er

Aggression Questionnaire, 78
Adoption studies, 144
Aggressive action, 61
Am

Adults Aggressive antisocial behavior, 153–155

attachment in, 243–244 Aggressive behavior
attachment styles of, 262–263
e

in adolescents, 125–126, 169–170

th

externalizing spectrum in, 91–93 historical views of, 4

low self-esteem in, 204–205 and parenting behavior, 126–127,
by

Advantaged groups 133–137

attitudes toward, 338 in social neuroscience perspective,
11

power of, 340 112–114

20

and resource distribution, 335–336 Aggressive retaliation

Affect anger and hatred in, 316–321
©

as internal state, 21 as last resort, 321

negative, 266, 354 Aggressive scripts, 111–112
ht

regulation of, 353 Alcohol consumption, 44

ig

victim expectations of, 377–378 Alcohol dependence, 97

yr

Affective routes, 36 Alpha 2, 178

op

Aftselakhis, 196 Altruism, reciprocal, 231–232

Agents of change, 335 Altruistic punishment, 229
C

Age periods, critical. See Critical age American Theocracy (K. Phillips), 268
periods Amygdala activation, 172, 175, 176

403
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 404

Anderson, C. A., 4, 24 Appraisal tendency framework,

Anger, 53–67 320–321
in aggressive retaliation, 316–321 Approach/inhibition theory of power,
and apologies, 61, 64 250–251
and arguments, 62 Approach orientation, 250–252

A)
and attachment, 244–246, 260 Arguments, 62, 65
Armageddon, 268

P
behavioral responses to, 59–60,
Arousal, 21

(A
62–63
brain regions for, 109–114 Arousal routes, 36
Arriaga, X. B., 376

n
cognitive model of, 57

io
control of, 262 Assertion behavioral system. See Power
system

at
and cost imposition, 60–61, 63
Assertiveness Questionnaire, 78

ci
and evolutionary biology of conflict,
Attachment, 241–254, 259–270

so
54–57
and abusive personalities, 264–265
and face, 64
in adulthood, 243–244

As
features of, 60–64 and aggression, 246–249
functional, 245 and anger, 244–246

al
and hatred, 322–324 in approach/inhibition theory of

ic
insults as cause of, 62, 64–65 power, 250–251

og
and intent, 61, 63–64 attachment threat, 266–269
in intergroup aggression, 315–316, in children’s psychological
320–326 ol development, 355–356
ch
neurophysiological locality of, 63 dyadic processes in, 265–266
y

in power-system activation, 76 empirical studies on, 262–264

Ps

recalibrational theory of, 58–60, neurobiology of, 261–262

63–66 neuropsychological aspect of, 261
an

short-term, 316 and power, 249–253

in social neuroscience perspective, “primitive” attachment threat,
ic

109–110 260–261
er

and status, 64 religion as symbolic form of, 268

Am

welfare tradeoff ratios of, 57–58 theory of, 242–243

“Anger of despair,” 245, 246 Attachment and Loss (J. Bowlby), 72, 76
“Anger of hope,” 245, 246 Attachment anxiety, 245–247, 262–263
e

Angry rumination, 109–110 Attachment avoidance, 252–253,

th

Antisocial behavior 262–263, 358

Attachment behavioral system, 242
by

in adolescents, 125–126
activation of, 73
aggressive, 153–155
in combat, 267
11

attachment insecurity in, 247–248

fear-based activation of, 260
candidate genes for, 157–158
20

Attachment disruption, 266

defined, 17 Attachment figures
familial influences on, 143–144
©

and anger, 260

nonaggressive, 153–155 in attachment theory, 242
ht

Anxiety in behavioral systems, 73

ig

attachment-related, 242, 243, perceived unavailability of, 243–244

yr

245–246 Attachment insecurity

battlefield, 386
op

and anger, 246

Anxious attachment, 247 in antisocial behavior, 247–248
C

Apologies, 61, 64 and attachment figure unavailability,

Appraisal-based model, 319–320 243–244

404 INDEX
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 405

and death, 269 Baer, B. A., 78

forms of, 261–262 Baker, L. A., 152
in intergroup aggression, 248–249 Balance theory, 371–372
in intimate partner abuse, 247, Baldwin, Stanley, 297
263–264 Bartels, M., 156

A)
and memory, 357–358 Bartholomew, K., 265
Battlefield violence, 386

P
with primary caregivers, 355–356
Baumeister, R. F., 204, 207

(A
severe and early, 265
Attachment orientation, 396–397 Behavior(s)
aggressive. See Aggressive behavior

n
Attachment relationships, 355–356

io
Attachment security antisocial. See Antisocial behavior
autonomy-inhibiting, 285

at
and attachment styles, 242–243
changes in, 343–344

ci
broaden-and-build cycle in, 243
conditioning of, 132–133

so
and parent–child respectfulness,
disrespectful, 289
283–285
effects of combat stress reaction on,

As
and power, 75–76 387
Attachment style externalizing. See Externalizing

al
in adults, 262–263 behaviors

ic
and attachment theory, 242–243 functional definitions of, 223–224

og
in intimate relationships, 265–266 hyperactivated power-oriented, 77–78
and mortality salience, 269 measurement of, 210
power in, 251 ol
normal ranges of, 107–108
ch
security in, 242–243 of parents, 131–137
y

Attachment theory and personality traits, 89–90, 94–96

Ps

application of, 241 relatedness-inhibiting, 285

attachment style in, 242–243 resource-holding power, 76
an

behavioral systems in, 71–72 social factors in, 206–207

Attachment threat, 266–269 stability of, 156–157
ic

Attention, 166–167 Behavioral dysregulation, 354–355

er

Attention-deficit hyperactivity disorder Behavioral economics, 226–227

Am

(ADHD), 157 Behavioral phenotypes, 129–130

Attitudes Behavioral responses, 59–60, 62–63
and behavior changes, 343–344 Behavioral scripts, 111
e

and intergroup contact, 341–342 Behavioral signatures

th

and mortality salience, 300 and culture, 189

in CuPS approach, 188–189
by

and respect, 278, 288–289

in different cultures, 191–192
supportive of violence, 26
and ideals of culture, 191
11

toward advantaged group, 338

Behavioral systems perspective, 72–74
Automatic systems, 21
20

Behavior genetic studies, 145–151

Automatization, of complex of aggression, 146–147
judgments, 19
©

models for, 149–150

Autonomy-inhibiting behaviors, 285 Beliefs
ht

Avoidance, 242–243 normative, 128

ig

Avoidant attachment. See Attachment supportive of violence, 26

yr

avoidance Berkowitz, L., 111

Avoidant individuals, 247, 248
op

Bias
Awareness, of inequality, 338, 340, 341 of advantaged group members, 340
C

Axelrod, R., 231 in intergenerational aggression

Ayers, E., 190 studies, 126–127

INDEX 405
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 406

Bifactor model, 92–93 power system in, 77

Biological factors, 134. See also Genetic psychological development. See Psy-
and environmental influences chological development in
Block, M., 390 children
Blumer, H., 335 respect in, 282–285

A)
Boden, J. M., 206 and transmission of aggression,
Boivin, M., 151 126–127

P
Boomsma, D. I., 156 Child maltreatment

(A
Borderline personality, 264–265 in children’s psychological develop-
Boundaries, group, 339 ment, 352–358

n
io
Bowlby, J., 71–73, 76, 241–242, and coping strategies, 358
244–245, 260 and externalizing/internalizing

at
Brain activity behaviors, 354

ci
with anger, 63 operationalization of, 359

so
effects of environmental variables and posttraumatic stress disorder,
on, 176 356–357

As
in social information processing, 172 and resilience, 351
Brain anatomy, 106 sexual abuse, 357–358

al
Brain lesions, 107 and social information processing,

ic
Brain regions, 105, 109–114. See also 173–174

og
Prefrontal cortex (PFC) Children
Branden, Nathaniel, 204 aggression continuity in, 125
Brendgen, M., 151, 152
ol
aggression prevention/intervention
ch
Broaden-and-build cycle, 243 strategies for, 212
y

Buckley, K. E., 24 conditioning of, 132–133

Ps

Bugental, D. B., 284 externalizing spectrum in, 93–95

Bullying, 286 fostering respect in, 289–290
an

Bushman, B. J., 4, 207 lower order traits in, 94–95

Buss, A. H., 78 memory in, 351, 357–358
ic

narcissism in, 208

er

Callous–unemotional traits, 95 physical abuse of, 353–354

Am

Candidate genes, 157–158 respect for parents in, 282

Canetti-Nisim, D., 324 self-esteem as cause of aggression in,
Captive’s dilemma, 393 206–207
e

Caring, 288–289 CHRM2, 97

th

Carnagey, N. L., 24 Chronic posttraumatic stress disorder,

Carpenter, J. P., 229 388–389
by

Carver, C. S., 250 Cicchetti, D., 354–355

Caspi, A., 98 Clausewitz, C. V., 297, 303
11

Cassidy, J., 265 Climate, school, 285–286

20

Causal direction, 168 Clutton-Brock, T. H., 225

Change, agents of, 335 Cognition, 18–19, 21, 223
©

Changes over time, 395–396 Cognitive appraisal, 316, 321–322

Chicchetti, D., 354 Cognitive dissonance theory, 372
ht

Childhood development Cognitive effects

ig

and aggressive parenting, 133–137 of child maltreatment, 357–358

yr

contextual influences in, 130–133 of combat stress reaction, 387

information processing, 168–169
op

Cognitive-information processes, 128

in maltreated children, 173–174 Cognitive model of anger, 57
C

neurological, 260–261 Cognitive neoassociationistic model of

and peer social rejection, 174–175 aggression, 111

406 INDEX
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 407

Cognitive neoassociation theory, Cost–benefit transactions, 57–59

111–112 Covert narcissists, 208
Cognitive routes, 36 Criminality, transient, 267
Cognitive systems, 59–60 Critical age periods
Collective action, 336–338 for environmental influences,

A)
Collectivistic cultures, 214 130–131
Columbia County Longitudinal Study, in intergenerational transmission of

P
124, 133 aggression, 133–137

(A
Combat, 385, 390–391 in self-views, 213–214

n
Combat fatigue, 386 Cross-generational continuity, of

io
Combat-induced disorders, 386–392 aggression. See Continuity of

at
Combat stress, 267 aggression
Combat stress reaction (CSR), 386–392 Cross-group friendships, 338–339

ci
Combat veterans. See Veterans CSR (combat stress reaction), 386–392

so
Commitment level, 371–373, 376–377 C-system

As
Commonality-focused contact, in angry rumination, 109–110
338–339, 344 and brain regions, 113

al
Compassionate teachings, 310 control and emotion regulation
Complex posttraumatic stress disorder, mechanisms of, 114

ic
397 in social neuroscience perspective,

og
Compulsive self-reliance, 244 107–108
Conditioning, 132–133 ol
Cues, 55–56
ch
Conduct problems, 169 Cultural approach, 187–188
Conflict resolution, 333 Cultural context, 196–198
y

Conformity, 287 Cultural dictates, 188–189

Ps

Conscientiousness, 90 Cultural logic rejection, 197–198

Consensus, 301–303 Cultural rejectionism, 196–197
an

Consistency theories, 371–372 Cultural superiority, 299–300

ic

Constructive tension, 343 Cultural syndromes

er

Contact theory, 338 in CuPS approach, 188–189

Context, cultural, 196–198 laboratory experiments on, 191–196
Am

Contextual factors Cultural worldview, 310

in aggressive behavior, 206–207 Culture × person × situation (CuPS)
e

in childhood development, 130–133 approach, 187–198

th

and genetic influences, 129–130 and cultural rejectionism, 196–197

with intergroup relations, 334 cultural syndromes and behavioral
by

timing of, 130–131 signatures in, 188–189

in war captivity, 394, 395 experiments on, 191–196
11

Contingent respect, 278–282, 289 honor, dignity, and face in, 189–191
20

Continuity of aggression, 123–138 rejecting cultural logic, 197–198

across life span, 124–126 Cultures
©

intergenerational, 126–127 of dignity, 189–190

longitudinal study on, 133–137 of face, 190–194
ht

psychological processes in, 128–133 of honor, 190

ig

Control, psychological, 284 parent–child respectfulness in,

yr

Control mechanisms, 114–116 284–285

Convergent validity, 81–82 universality of anger across, 62–63
op

Coping strategies Cumberland, A., 132

C

and child maltreatment, 358 CuPS approach. See Culture × person ×

deactivating, 244 situation approach

INDEX 407
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 408

dACC. See Dorsal anterior cingulate Discrimination, perceived, 337

cortex Disorder of extreme stress not otherwise
Daniel, E., 287 specified (DESNOS), 397
Deactivating strategies, 74, 244 Disorganized/disoriented (Type D)
Deactivation, of power system, 78–84 attachment, 261–262, 355

A)
Death Displaced aggression, 224–225
mental structures for denial of, Dispositional influences, 89–99

P
298–301 environmental and situational

(A
reminders of, 304 factors in, 95–97

n
and terror management, 268–269 and externalizing spectrum, 90–95

io
Decentration, 279 genetic research on externalizing

at
Decision making, 114, 305 phenomena, 97–98
Defensive exclusion, 357 trait approach to, 89–90

ci
Delinquents, 248 Disrespectful behavior, 289

so
Democratic societies Disruption, of attachment, 266

As
public support for aggressive Dissociation, 261–262
retaliation in, 318 Dissolution of relationships, 377–378

al
support for military action in, 302 Distributive justice, 281
Dependence, emotional, 379–380 Dixon, J., 342

ic
Deprivation, 337 Dodge, K. A., 354

og
Desensitization, 112 Domestic violence. See Intimate partner
DESNOS (disorder of extreme stress ol violence; Partner aggression
ch
not otherwise specified), 397 Dominance hierarchies, 56–57
Destructive aggression, 246–249 Dorsal anterior cingulate cortex
y

Deterrence (dACC)
Ps

and revenge costs, 230 in brain anatomy, 106

third-party, 228 and negative emotional states, 111
an

value of, 233 as neural alarm system, 115

ic

of withholding of benefits, 228–229 with violent video games, 112

er

Development, childhood. See Childhood Dorsolateral prefrontal cortex, 106

development “Doughnut theory” (narcissistic
Am

DG (dictator game), 227 aggression), 210–211

Diamond, L. M., 246 Dovidio, J. F., 340
e

Diamond, S. R., 227 Durrheim, K., 342

th

Diathesis–stress perspective, 95 Dyadic impellors, 40

Dick, D. M., 97, 178 Dyadic inhibitors, 41
by

Dictates of culture, 188–189 Dyadic negotiation, 226

Dictator game (DG), 227 Dyadic processes, 265–266
11

Dignity cultures Dyadic triggers, 37

20

cultural syndromes in, 189–191 Dysfunctional parenting, 213–214

and reciprocity, 192–194
©

Dionne, G., 151 Early aggressiveness, 125

Direct deterrence, 225–227 Early infancy, 260
ht

Disadvantaged groups Echo effect, 231

ig

collective action on part of, 337–338 Egalitarian actions, 343–344

yr

harmony as maladaptive for, 343 Ego threat

and resource distribution, 335–336 as cause of aggression, 207–208
op

Discontinuity, intergenerational intervention techniques with,

C

environmental, 130–131 212–213

Discriminant validity, 82–83 and narcissism, 43–44

408 INDEX
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 409

Ein-Dor, T., 304 to contextual influences, 130–131

Eisenberg, N., 132 to intimate partner violence, 175
Emotion(s) Externalizing behaviors
in deactivation of power system, and child maltreatment, 354
82–83 genetic and environmental factors

A)
individual and group-based, 320 in, 97–98
in intergroup aggression, 316–319 heritability of, 98

P
in intergroup conflict, 315–316 and personality traits, 94–96

(A
negative, 111 psychobiological factors in, 96
and respect, 288–289

n
Externalizing spectrum

io
in support for aggressive retaliation, dispositional influences in, 90–95
317–321

at
genetic research on, 97–98
Emotional decision making, 305 Extraversion, 90

ci
Emotional dependence, 379–380

so
Emotional sentiment, 318–319 Face
Emotion regulation

As
cultures of, 190–194
angry rumination as form of, 110 in CuPS approach, 189–191
in children’s psychological develop-

al
as feature of anger, 64
ment, 353–355

ic
restoration of, 61–62
C-system mechanisms of, 114 Facial expressions, 353–354

og
observational learning processes in, Factor structure (PBSS instrument),
132
in social neuroscience perspective,
ol 79–80
ch
Fairness, 338, 341
114–116
Familial influences, 143–144
y

England, Lynndie, 15–16

Fearful attachment, 262, 264
Ps

Environmental influences. See Genetic

Fehr, E., 227, 229
and environmental influences
Felson, R., 61
an

Equality, social, 334–339

“Felt power,” 250
Ethnic groups, 284–285
ic

“Felt security,” 73, 250

Ethno–political conflict. See Political
er

violence Fergusson, D. M., 206

Festinger, L., 372
Am

Evolutionary and cultural impellors, 40

Evolutionary and cultural inhibitors, 41 FFM (five-factor model), 90, 91
Evolutionary perspective Field studies, 210
e

on aggression as strategy, 114 Fighting ability, 55–56

th

on anger and conflict, 54–57 Filial piety, 282

on forgiveness, 229–235 Financial dependence, 379
by

on revenge, 221–229, 232–235 First Lebanon War, 388, 390–392

Exacerbated posttraumatic stress Fitness budget, 54–55
11

syndrome, 391 5-HT, 106, 172, 175, 178–179

20

Executive functions, 177 Five-factor model (FFM), 90, 91

Existentialist perspective, 301–307 Florian, V., 269
©

and consensus, 301–303 fMRI (functional magnetic resonance

inevitability of violence in, 306–307 imaging), 108
ht

and justice, 303–306 Fonagy, P., 248

ig

Expectations, 26 Forgiveness, 229–235

yr

Expected utility theory, 303 evolution of, 229–232

Expected value, 233–234 functional definition of, 230–231
op

Exposure and revenge/vengeance, 232–234

C

to combat, 390–391 selection pressures for, 231–232

to conflict-related events, 319–320 Formidability, 55–56

INDEX 409
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 410

Forsythe, R., 227 on intergenerational aggression,

Fredrickson, B. L., 243 130–131
Freud, S., 71–72 model-fitting analyses of, 149–150
Frontal lobe functioning, 44 in social information processing,
Frontal regions, of brain, 107 176–179

A)
Functional anger, 245 Genetic polymorphisms, 177–178
Goal(s)

P
Functional magnetic resonance imaging
of behavioral systems, 73–74

(A
(fMRI), 108
Functional theories, 66 power as, 75–77, 250–251

n
Fundamental attribution error, 23–24 regulatory, 244

io
Goal selection, 167

at
GABA (gamma-aminobutyric acid), Government actions, 27–29
Greenberg, M., 212

ci
178
Group-based emotions, 320

so
GABRA 2, 178
Gächter, S., 229 Group-based hierarchy

As
Gage, Phineas, 107 advantaged and disadvantaged
Galinsky, A. D., 250 groups in, 335–336

al
GAM. See General aggression model stability of, 338
Group boundaries, 339

ic
Gamma-aminobutyric acid (GABA), 178

og
Gaza War, 306
Gene-by-environment interaction Haberstick, B. C., 156
effects, 178–180 ol
Halperin, E., 324
ch
Handbook of Attachment (J. Cassidy and
General aggression model (GAM), 4,
P. R. Shaver), 265
y

15–30
Harmony, 191, 333–334, 343
Ps

and aggression between groups of

Hatred
people, 26–27
in aggressive retaliation, 316–321
an

basic model of, 18–20

and anger, 322–324
decision-making processes in, 114
ic

in intergroup aggression, 320–326

and government actions, 27–29
er

for outgroup, 316

and I3 theory, 36 Heart rate, 171–172
Am

perpetrators of violence in, 25–26 Heider, F., 61, 371

for reducing aggression and Help seeking
violence, 29
e

among soldiers, 397–398

of single episode cycle, 20–23
th

of partner aggression victims, 379

terms defined in, 17–18 Heritability. See also Genetic and envi-
by

and violence escalation cycle, 23–25 ronmental influences

Genetic and environmental influences, of aggression, 176–177
11

129–130, 143–159 of externalizing behaviors, 98

additive and nonadditive, 149
20

Hewitt, J. K., 156

on alcohol dependence, 97 Hezbollah, 309
on antisocial behavior, 153–156
©

Hicks, A. M., 246

behavior genetic studies of, 145–151 Higgins, E. T., 250
ht

on brain activity, 176 Higher order model, 92–93

ig

candidate genes in, 157–158 Hirschberger, G., 304, 309

yr

in development of aggression, Hitler, Adolf, 203

156–157, 176–179 Holy wars, 267–268
op

in dispositional influences, 95–97 Honor, 189–191

C

on externalizing behaviors, 97–98 Honor cultures, 190, 192–196

and forms of aggression, 151–153 Horowitz, J. L., 227

410 INDEX
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 411

Horowitz, L. M., 78 Instigating triggers, 35, 36

Horwood, L., 206 in I3 Theory, 37–39
Hostile attribution, 165 and impelling forces, 43–46
Hostile cues, 354 and inhibiting forces, 44–45
Hostile intent, 167 Institutionalized violence, 25

A)
Hudziak, J. J., 156 Insults, 62, 64–65
Integrity, 194–196

P
Huesmann, L. R., 128
Intent

(A
Humiliation, 208
Humility, 191 of adversary, 308–309

n
Hyperactivated power-oriented behavior, as feature of anger, 63–64

io
77–78 and forgiveness, 233

at
Hyperactivating strategies, 73, 244 interpretation of, 167
and likelihood of anger, 61

ci
Hyperactivation (power system), 79–84
Intentional action, 61

so
I3 Theory, 35–48 Interaction effects, 43–46

As
components of, 35–36 Interdependence theory, 372–373
future research, directions for, 46–48 Intergenerational aggression, 126–127.

al
impelling forces in, 39–40 See also Continuity of aggression
Intergroup aggression, 315–327

ic
inhibiting forces in, 40–41
anger and hatred in, 315–316,

og
instigating triggers in, 37–39
interaction effects of, 43–46 320–326
main effects in, 41–43 ol
appraisal-based model for, 319–320
ch
attachment insecurity in, 248–249
risk factors, 46–47
emotions and sentiments in,
y

Iacono, W. G., 98
316–319
Ps

IDF (Israel Defense Forces), 385–386

and general aggression model, 26–27
Imitation, of parent behavior, 131–132
necessary preconditions for, 320
an

Immediate appraisal processes, 21–22

Intergroup attitudes, 300
Impelling forces, 35, 36
ic

Intergroup comparisons, 337

in I3 Theory, 39–40
er

Intergroup conflict, 299–300

and inhibiting forces, 45–46
Intergroup contact
Am

and instigating triggers, 43–44 and attitudes, 341–342

Impellors, 40 and change for equality, 338–339
Impulsivity, 114, 265
e

research on, 339–343

Incentives (revenge systems), 222–224
th

Intergroup relations, 333–345

Individual differences, 187–189 future directions for, 343–344
by

Individual emotion, 320 intergroup contact research,

Infant development, 260–261 339–343
11

Information-processing model, 128–133 and social change, 338–339

Information processing patterns, 170
20

and social inequality, 334–339

Ingroup, 321 tension and harmony in, 333–334
Inhibiting forces, 35, 36
©

Internalizing behaviors, 354

in I3 Theory, 40–41 Internal states, 21, 22
ht

and impelling forces, 45–46 Internal working models, 73, 242

ig

and instigating forces, 45–46 Interventions

yr

and instigating triggers, 44–45 effective strategies for, 212–213

Inputs, 36 as experimental evidence, 169–170
op

Insecure attachment. See Attachment general aggression model for, 29

C

insecurity I3 theory in, 47

Instigating forces, 45–46 for partner aggression, 378–379

INDEX 411
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 412

Intimate partner aggression, 371–372 Kernis, Michael, 212

Intimate partner violence (IPV), 175, Keyes, M., 98
247, 260, 262–266. See also Khoury-Kassabri, M., 287
Intimate relationships Kim, J. E., 354
Intimate relationships. See also Partner Kin, 231

A)
aggression Knafo, A., 287
abusive men in, 260 Knowledge structures, 19

P
adolescents in, 175 Krueger, R. F., 98

(A
attachment anxiety in, 246 Kusché, C., 212

n
attachment styles in, 262–263

io
expectations in, 373–375 Laboratory studies, 210

at
perceptions of aggression in, Landau, M. J., 302
370–371 Lansford, J. E., 174

ci
respect in, 288 Learning processes, observational,

so
in young adulthood, 170–171 131–132

As
Intrinsic welfare trade ratios, 58 Lebanon Wars
Inventory of Personal Problems, 78–79 First, 388, 390–392

al
IPV. See Intimate partner violence Second, 309, 389
IRT (item response theory) model, Legrand, L. N., 98

ic
91–92 Lerner, J. S., 320

og
Israel Defense Forces (IDF), 385–386 Levinson, A., 248
Israeli experience ol
Lieberman, M. D., 108
ch
attachment orientation of soldiers Life course, aggression across. See
in, 397 Continuity of aggression
y

combat stress reaction casualties in, Life-experience factors, 176

Ps

388–389 Limbic regions, 107

military and political violence in, Limbic system, 172
an

385–386 List, J. A., 227

ic

prisoners of war in, 393–398 Liu, J., 152

er

psychological breakdown in soldiers, Longitudinal studies, 133–136

390–391 Long-term conflict, 320
Am

solution to conflict in, 297 Long-term emotional sentiments, 320

Israeli–Palestinian conflict, 16, 28–29 Long-term hatred, 316
e

Item response theory (IRT) model, Long-term observational learning

th

91–92 processes, 131–132

Iterated prisoner’s dilemma game, 226 Lopez, F. G., 263
by

Love, 288–289
Jacobson, K. C., 152 Lower order traits, 94–95
11

James, W., 316 Low self-esteem

20

Janoff-Bulman, R., 279 in adults, 204–205

JUPI (Justice, Utility, and Peace Inven- as cause of aggression, 203–207, 211
©

tory), 304–305 in children, 206–207

Justice, 281, 303–306, 309 Lubensky, M., 342
ht

Justice, Utility, and Peace Inventory Lyons-Ruth, K., 262

ig

(JUPI), 304–305
yr

Madrid bombings, 306–307

Kahn, Irene, 301 MAI (Siegel Multidimensional Anger
op

Karremans, J. C., 234 Inventory), 262

C

Kaufmann, H., 61 Main effects (I3 Theory), 41–43

Keltner, D., 320 MAO-A. See Monoamine oxidase-A

412 INDEX
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 413

Masculine identity, 389 MPFC. See Medial prefrontal cortex

Mathews, V. P., 115 MS. See Mortality salience
Matthews, P. H., 229 Mullen, B., 148
Mauricio, A. M., 263 My Lai massacre, 267
Mawson, A. W., 267

A)
McGue, M., 98 Narcissism
Medial prefrontal cortex (mPFC), 106, in children, 208

P
110, 113 developmental origins of, 213–214

(A
Mediational approach, 188 and ego threat, 43–44

n
Media violence as self-view, 207–209

io
in social neuroscience perspective, Narcissistic Personality Inventory

at
111–112 (NPI), 207
and top-down control mechanisms, Natural selection, 54, 66. See also

ci
115–116 Evolutionary perspective

so
Memory, 351, 357–358 Neale, M. C., 156

As
Men Negative affect, 266, 354
abusive, 260, 264–265 Negative emotional states, 111

al
abusive personality in, 262–263 Negative outcome expectancies, 45
attempts to save face by, 61–62 Neoassociation theory, 4, 111–112

ic
socialization of, 125 Neural alarm system, 115

og
Mental disorders, 90–91 Neural processes, 171–173
Mental processes, 105 ol
Neurobiology, 261–262
ch
Mental representation, 167 Neurophysiology, 63, 261
Middle East conflict, 311 Neuroticism, 90
y

Mikulincer, M., 243, 248–249, 262, 269, Neurotransmitters, 106, 172

Ps

388 Nonadditive genetic influences, 149

Military violence Nonaggressive antisocial behavior,
an

and attachment, 266–267 153–155

ic

in Israeli experience, 385–386 Noncontingent punishment, 226

er

Miller, D. T., 281 Nondemocratic societies, 302

Model-fitting analyses, 149–150 Nonnormative collective action, 336
Am

Molecular genetic studies, 97–98 Normal behavior, 107–108

Monitored welfare trade ratios, 58 Normal narcissism, 207
e

Monoamine oxidase-A (MAO-A), 98, Normative beliefs, 128

th

172, 175, 177–178 Normative collective action, 336

Moods, 318 Normative parameters (power system),
by

Moral righteousness, 299–300 74–77

Mortality, 298–301, 304, 307–308 Norms
11

Mortality salience (MS) with intimate relationship aggression,

20

and adversary intent, 308–309 369–370

and attachment style, 269 violation of, 196–197
©

and consensus, 302–303 NPI (Narcissistic Personality Inventory),

impact of, 304–305 207
ht

and inevitability of violence, 307

ig

and perceived vulnerability, Objectification, 252–253

yr

308–309 Observational learning processes,

in support for violence, 301 131–132
op

in terror management research, 299, Openness to experience, 90

C

300 Organized violence, 301–303

Motives, 303–305 Outcome expectancies, 45

INDEX 413
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 414

Outcomes Perceptions, victim. See Victim

with general aggression model, 21–22 perceptions
in I3 theory, 36 Perry, M., 78
Outgroup homogeneity, 24–25 Personal impellors, 40
Overt aggression, 150–153 Personal inhibitors, 41

A)
Overt narcissists, 208 Personal inputs, 20, 22
Personality traits

P
Palestine Liberation Organization on aggressive behavior, 89–90

(A
(PLO), 298 and externalizing behaviors, 95–96

n
Palestinian rebellion, 16, 297, 389. in externalizing behaviors, 94–95

io
See also Israeli–Palestinian conflict facets of, 90

at
Parameters (power system), 74–77 stability of, 124
Parent–child relationship Person inputs, 36

ci
disorganized attachment in, 261–262 Pérusse, D., 151

so
power of parent in, 284 PFC. See Prefrontal cortex

As
respect in, 282–285 Phenotypes, behavior, 129–130
and teacher–student relationship, Phillips, K., 268

al
286–287 Physical abuse, 133–137, 353–354
transmission of aggression in, Physical aggression

ic
126–127, 133–137 interpretation of, 372

og
Parents and norms, 369–370
aggressive behavior of, 133–137 ol
psychological aggression vs.,
ch
and dysfunctional parenting, 370–371
213–214 tolerance of, in intimate relation-
y

imitation of, 131–132 ships, 375

Ps

power of, in parent–child relation- Physical proclivity, 45

ship, 284 Pitt-Rivers, J., 190
an

and respect for children, 282–284 PLO (Palestine Liberation Organization),

ic

Parker, G. A., 225 298

er

Partner aggression, 367–380 Political violence, 297–311, 385–398

and breakup aftermath, 377–378 combat-induced disorders from,
Am

defined, 368–369 386–392

interventions for, 378–379 existential concerns, 301–307
e

perceptions of, 369–371 and intergroup conflict, 299–300

th

reinterpreting the past in, 373–377 in Israeli experience, 385–386

victim perceptions of, 371–375 moving beyond, 310–311
by

Partner violence, 170–171 self-protection as moderator of,

Pathogenic relational environments, 352 307–309
11

PBSS. See Power Behavioral System and terror management theory,

20

Scale 298–300
PCC (posterior cingulate cortex), war captivity, 392–398
©

111–112 Polymorphisms, 177–178

Peers Posterior cingulate cortex (PCC),
ht

bullying by, 286 111–112

ig

rejection by, 174–175 Posttraumatic stress disorder (PTSD)

yr

respect for, 286 and child maltreatment, 356–357

violence of, 170–171 chronic, 388–389
op

Perceived discrimination, 337 from combat trauma, 387–391,

C

Perceived fairness, 338, 341 394–398

Perceived vulnerability, 308–309 complex, 397

414 INDEX
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 415

Power Psychological aggression

of advantaged groups, 340 in intimate relationships, 368–369,
and aggression, 4 375
in approach orientation, 251–252 perceptions of, 373
and attachment, 75–76, 249–253 physical aggression vs., 370–371

A)
defined, 75 and victim well-being, 379
Psychological breakdown, 390–391

P
as goal, 250–251
Psychological control, 284

(A
and objectification, 252–253
in parent–child relationship, 284 Psychological development in children,

n
sense of, 75 351–360

io
Power behavioral system. See Power attachment in, 355–356

at
system emotion regulation in, 353–355
and maltreatment, 352–358

ci
Power Behavioral System Scale (PBSS),
memory in, 357–358

so
79–84
Powerless individuals, 78 methodological considerations with,

As
Power-oriented behavior, 77–78 358–359
Power system, 71–85 in pathogenic relational environ-

al
activation and functioning of, ments, 352

ic
77–79 and psychopathology, 356–357
Psychological processes, 128–133

og
and behavioral systems perspective,
Psychological sequelae, 394–395
72–74
hyperactivation and deactivation of, ol
Psychology of revenge, 223
ch
Psychopathology, 91, 356–357
79–84
Psychophysiology, 171–173
y

normative parameters of, 74–77

Psychosomatic effects, 387
Ps

POWs (prisoners of war), 392–398

Psychotherapy, 397–398
Pratto, F., 340
P300 component, 112
an

Precursor genes, 158

PTSD. See Posttraumatic stress disorder
Predictive validity, 84
ic

Public goods games, 228–229

Preemptive violence, 308
er

Public support
Prefrontal cortex (PFC) for aggressive retaliation, 316–317,
Am

importance of, 106 321

in information processing, 172 for intergroup aggression, 323–326
in self-control, 114–115
e

for military action, 324–326

Prevention orientation, 250
th

for organized violence, 301–303

Primary caregivers, 243, 355–356 Punishment, 225–227, 229
by

Primary strategy, 73 Pyszczynski, T., 304

“Primitive” attachment threat, 260–261
11

Prisoner’s dilemma, 226 Race relations, 335–336

Prisoners of war (POWs), 392–398
20

Raine, A., 152

Proactive aggression, 151–152 Rathus, S. A., 78
Procedural justice, 281
©

Rational decision making, 305

Process dissociation paradigms, 47 Reactivation, of traumatic reactions,
ht

Prolonged trauma, 397 391–392

ig

Promoting Alternative Thinking Reactive aggression, 151–152

yr

Strategies program, 212 Realistic group-conflict theory, 336

Promotion focus, 77 Reappraisal processes, 21–22
op

Promotion orientation, 250 Recalibrational theory, 53–54

C

Provocation, 44, 45, 109 Receptor genes, 158

Psychobiology, 96 Reciprocal altruism, 231–232

INDEX 415
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 416

Reciprocity “Revenge as disease” concept, 222

in cultures, 192–194 Reward system, 113
of negative affect, 266 Righteousness, 299–300
reputation for, 190 Risk factors
of respect, 287 for aggressive behaviors, 35

A)
Reconciliation, 230–232 in I3 theory categories, 46–47
RED (response evaluation and decision), impelling forces as, 39–40

P
167–168 inhibiting forces as, 40–41

(A
Reflective processes, 109–110 instigating triggers as, 37–39
Romantic relationships. See Intimate

n
Regulatory goals, 244

io
Relatedness-inhibiting behaviors, 285 relationships
Rosenberg, S. E., 78

at
Relational aggression, 150–153
Relationships Rumination, angry, 109–110

ci
commitment in, 45, 380 Rusbult, C. E., 376

so
dissolution of, 377–378
Saguy, T., 340, 341

As
expected value of, 233–234
intimate. See Intimate relationships Savin, N. E., 227
Schmitz, S., 156

al
security in, 310–311
Relationship specificity, 170–171 Schools, respect in, 285–288

ic
Relationship value, 231 Schore, A. N., 261

og
Relative deprivation, 337 Scripts
aggressive, 111–112
Religion, 267–268, 310 ol
behavioral, 111
ch
Repeated trauma, 397
in information-processing model, 128
Reputation, 190, 228
y

Script theory, 4, 111–112

Resilience, 351
Ps

Secondary attachment strategies, 242

Resource-holding power, 76
Second Lebanon War, 309, 389
Resources
an

Secure attachment. See Attachment

and behavioral system strategies,
security
ic

74–75
Sefton, M., 227
er

conflicts over, 55 Selection pressures

control of, 335
Am

in conflicts of interest, 54–57

in reappraisal processes, 21–22 for forgiveness, 231–232
Respect, 277–290 for revenge, 225–229
e

and attitudes/emotions, 278, Self-affirmation theory, 212–213

th

288–289 Self-compassion, 214–215

fostering, 289–290 Self-control, 114–115
by

in parent–child relations, 282–285 Self-esteem

in schools, 285–288 of adolescents, 206–207
11

unconditional, 278–282 in aggression prevention programs,

20

Response–decision phase, 167 212

Response evaluation and decision as cause of aggression, 203–207
©

(RED), 167–168 of children, 206–207

Response tendencies, 322–323 and narcissism, 208, 210–211
ht

Resting heart rate, 171–172 Self-protection, 307–309

ig

Revenge, 221–230, 232–235 Self-regulatory strength, 42–44

yr

costs of, 224–230 Self-reliance, 244

evolution of, 222–229 Self-report studies, 81–82, 205
op

and forgiveness, 232–234 Self-respect

C

functional definition of, 222–225 and aggression, 281–282

selection pressures for, 225–229 in children, 289

416 INDEX
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 417

Self-views, 203–215 Social constructions, 299, 302

and aggression, 209–211 Social context
clinical implications of, 211–213 of intergroup relations, 334
critical age periods in, 213–214 of war captivity, 394, 395
future research, directions for, Social cues, 171–172

A)
213–215 Social dominance theory, 336
and instigating triggers, 43–44 Social equality, 334–339

P
low self-esteem in adults, 204–205 Social factors, 125, 206–207

(A
low self-esteem in children, 206–207 Social inequality

n
narcissism, 207–209 and intergroup relations, 334–339

io
Sensitization, 353 research on, 338–343

at
Sentiments Social information processing, 165–180
emotions vs., 318–319 characteristic styles of, 165–166

ci
in intergroup aggression, 316–319 and child maltreatment, 173–174

so
September 11th attacks, 267, 306–307 environmental factors in, 176–179

As
Sequential assessment game, 56 genetic factors in, 176–179
Sequential prisoner’s dilemma game, 226 mechanisms of, 166–168, 179–180

al
Serotonin, 106, 158 neural and psychophysiological
Sex differences, 152 processes in, 171–173

ic
Sexual abuse, 357–358 patterns of, 168–171

og
Shameful experiences, 208 and peer social rejection, 174–175
Shaver, P. R., 243, 248–249, 265 ol
Social interaction theory, 4
ch
Shell shock, 386 Social learning, 4, 16, 111–112
Shields, A., 354–355 Social neuroscience perspective,
y

Short-term anger, 316 105–116

Ps

Short-term observational learning abnormal populations, 107

processes, 131–132 aggressive behavior in, 112–114
an

Siegel, D. J., 261 anatomy, 106

ic

Siegel, M., 262 and anger, 109–110

er

Siegel Multidimensional Anger Inventory and cognitive neoassociation theory,

(MAI), 262 111–112
Am

Signals, 223–224 control and emotion regulation

Signatures, 188–189 mechanisms in, 114–116
e

Single episode cycle dual-process framework, 108

th

in general aggression model, 22–23 and media violence, 111–112

general aggression model of, 20–23 research in, 109–114
by

Single nucleotide polymorphisms and script theory, 111–112

(SNPs), 97 and social learning, 111–112
11

Situational factors, 95–97 X- and C-systems in, 107–108

20

Situational impellors, 40 Social pressure, 250–251

Situational inhibitors, 41 Social rejection, 41–42
©

Situational specificity, 170 Social signals, 281

Situation inputs, 20, 22, 36 Soldiers
ht

Skitka, L. J., 324 code of conduct for captured, 393

ig

SNPs (single nucleotide polymor- psychological breakdown in,

yr

phisms), 97 390–391
Social attention-holding, 76 Solomon, Z., 388
op

Social change, 338–339 Spectrums of psychopathology, 91

C

Social–cognitive processes, 19, 165 Spinrad, T. L., 132

Social–cognitive theory, 16 Sroufe, L. A., 250

INDEX 417
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 418

Stability Twin studies

of aggressive behavior, 156–157, on genetic and environmental
211–212 influences, 144–145
of PBSS instrument, 80 on heritability of aggression,
in personality, 95–96 176–177

A)
Status, 64 Type D (disorganized/disoriented)
Steinberg, L., 282 attachment, 261–262, 355

P
Strength model of self-regulation, 42

(A
Stress, 386–392 Ubeida, Marwan Abu, 15

n
Stress inoculation perspective, 390 Ultimatum game (UG), 227

io
Stressors, 392–394 Unconditional respect, 278–282

at
Stress resolution hypothesis, 390 Universalism, 287
Structural inequality, 335–339 Universality, of anger, 66

ci
Substance-dependence syndromes, 97 Ureno, G., 78

so
Suicidal terrorism, 15, 302 U.S. War on Terrorism, 28

As
Suicide, 266 Utility motives, 304–305, 309
Support, public. See Public support

al
Symbolic immortality, 269 Vagal tone, 246
Syndromes (CuPS approach), 188–189 van Beijsterveldt, C. E., 156

ic
van der Valk, J. C., 156

og
Tausch, N., 340 Vengeance, 232–234
Taylor, S. P., 113 ol
Ventral prefrontal cortex, 106
ch
Teachers, 286–287 Verhulst, F. C., 156
Tedeschi, J., 61 Veterans, 388, 389, 395–396
y

Tein, J. Y., 263 Victimization, peer, 286

Ps

Tension, 333–334, 343 Victim perceptions

Terrorism, 15, 25, 301, 302, 306–307. of breakup aftermath, 377–378
an

See also Political violence and commitment level to relation-

ic

Terror management theory (TMT) ship, 376–377

er

attachment security in, 268–269 of partner aggression, 371–375

and political violence, 298–300 Video games, 112
Am

Testosterone, 42 Villasenor, L., 78

Third-party deterrence, 228 Violence
e

Third-party triggers, 37, 39 and aggression, 4

th

Threatened egotism, 207–208, 212–213 beliefs, attitudes, and expectations

TMT. See Terror management theory supportive of, 26
by

Toth, S. L., 354 defined, 18

Traits. See Personality traits inevitability of, 306–307
11

Transient criminality, 267 influence of government actions on,

20

Transporter genes, 158 27–29

Trauma. See also Posttraumatic stress institutionalized, 25
©

disorder intimate partner, 175, 247, 260,

of captivity, 391–394 262–266
ht

repeated and prolonged, 397 media, 111–112, 115–116

ig

Tredoux, C., 342 military, 266–267, 385–386

yr

Triggering events, 23 motives for, 303–305

Trust, 233 moving beyond, 310–311
op

Trust game, 227 organized, 301–303

C

Trustworthiness, 194–196 perpetrators of, 25–26

Turning points, 131 political. See Political violence

418 INDEX
12051-23_Index-rev.qxd 5/25/10 12:02 PM Page 419

precursors of, 25–26 War captivity, 392–398

preemptive, 308 War neurosis, 386
as rational option, 303 Warring social groups, 249
reduction of, 29 Waters, E., 250
and revenge, 221–222 Welch, D., 303

A)
in societies, 17 Welfare tradeoff ratios (WTR), 57–60
support for extreme forms of, Well-being, 378–380

P
308–309 Werther, A., 279

(A
in war captivity, 392–393 White, T. L., 250

n
Violence escalation cycle, 23–25 Withholding of benefits, 228–229

io
Violent video games, 112 Women, 125, 263

at
Vitaro, F., 151 Wright, S. C., 342
Vulnerability, 308–309, 390 WTR. See Welfare tradeoff ratios

ci
so
War X- and C-systems model, 107–108

As
aftermath of, 387–392 X-system, 108, 109, 112–114
dichotomizing effects of, 301

al
effects of, on citizens, 27–29 Yamagishi, T., 229
prevalence rates of combat stress Yom Kippur War, 391, 393–394

ic
reaction in, 387 Young, S. E., 156

og
psychological wounds of, 385
public support for, 301–302 ol
Zautra, A. J., 390
ch
undermining of consensus on, 306 Zigler, E., 351
y
Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

INDEX 419
C
op
yr
ig
ht
©
12051-23_Index-rev.qxd

20
11
5/25/10

by
th
e
12:02 PM

Am
er
ic
Page 420

an
Ps
y ch
ol
og
ic
al
As
so
ci
at
io
n
(A
P A)
12051-24_Editor-rev.qxd 5/25/10 12:02 PM Page 421

ABOUT THE EDITORS

PA)
(A
n
io
at
ci
so
As
al
ic
og
ol
y ch
Ps
an

Phillip R. Shaver, PhD, a social and personality psychologist, is Distinguished

ic

Professor of Psychology at the University of California, Davis. Before moving

er

there, he served on the faculties of Columbia University, New York University,

Am

University of Denver, and State University of New York at Buffalo. He has

coauthored and coedited numerous books, including In Search of Intimacy;
e

Measures of Personality and Social Psychological Attitudes; Measures of Political

th

Attitudes; Handbook of Attachment: Theory, Research, and Clinical Applications;

by

and Attachment in Adulthood: Structure, Dynamics, and Change, and has published
over 200 scholarly journal articles and book chapters. Dr. Shaver’s research
11

focuses on attachment, human motivation and emotion, close relationships,

20

personality development, and the effects of meditation on behavior and the

brain. He is a member of the editorial boards of Attachment and Human Devel-
©

opment, Personal Relationships, the Journal of Personality and Social Psychology,

ht

and Emotion, and has served on grant review panels for the National Insti-
ig

tutes of Health and the National Science Foundation. He has been executive
yr

officer of the Society of Experimental Social Psychology and is a fellow of

op

both the American Psychological Association and the Association for Psy-
C

chological Science. Dr. Shaver received a Distinguished Career Award from

421
12051-24_Editor-rev.qxd 5/25/10 12:02 PM Page 422

the International Association for Relationship Research and has served as

president of that organization.

Mario Mikulincer, PhD, is professor of psychology and dean of the New

A)
School of Psychology at the Interdisciplinary Center in Herzliya, Israel. He
has published three books—Human Helplessness: A Coping Perspective;

P
(A
Dynamics of Romantic Love: Attachment, Caregiving, and Sex; and Attachment
in Adulthood: Structure, Dynamics, and Change—and over 280 scholarly journal

n
articles and book chapters. Dr. Mikulincer’s main research interests are attach-

io
ment theory, terror management theory, personality processes in interpersonal

at
relationships, coping with stress and trauma, grief-related processes, and pro-

ci
social motives and behavior. He is a member of the editorial boards of several

so
scientific journals, including the Journal of Personality and Social Psychology,

As
Psychological Inquiry, and Personality and Social Psychology Review, and has served
as associate editor of two journals, the Journal of Personality and Social Psychology

al
and Personal Relationships. Recently, he was elected to serve as chief editor of

ic
the Journal of Social and Personal Relationships. He is a fellow of the Society

og
for Personality and Social Psychology and the Association for Psychological
ol
Sciences. He received the EMET Prize in Social Science for his contributions
ch
to psychology and the Berscheid–Hatfield Award for Distinguished Mid-Career
y

Achievement from the International Association for Relationship Research.

Ps
an
ic
er
Am
e
th
by
11
20
©
ht
ig
yr
op
C

422 ABOUT THE EDITORS