The Journal of Political Philosophy: Volume 15, Number 1, 2007, pp.

93–114

The Paradox of Political Representation

David Runciman
Politics, Cambridge University

A T the heart of the idea of political representation lies an intractable puzzle,

which in this article I will call the paradox of presence. This is not a formal
or logical paradox, but a linguistic one, and it has its roots in the etymology of
the term: ‘re-presentation’ implies that something must be present in order to be
‘re-presented’ but also absent in order to be ‘re-presented’. Since things cannot be
present and absent at the same time, there appears to be at the very least a tension
built into the idea of representation, and possibly a fundamental incoherence.
The classic description of this paradox is given by Hanna Pitkin, in The Concept
of Representation (1967) and in a shorter article, published separately but
containing the core of the same argument, called ‘The Paradox of Representation’
(1968).1 Although Pitkin’s work remains the standard English-language account
of the idea of representation, and is still widely cited by analytical political
theorists, her description of the paradox of presence has largely been ignored.
Instead, it is her distinction between ‘substantive’ and ‘descriptive’ representation
that has provided the focus for most subsequent commentators. Meanwhile,
the paradoxes that have captured the attention of Anglo-American political
theorists are the formal paradoxes of democracy. In this context the concept
of representation has usually been treated in narrowly functional terms, and as
subject to the wider imperatives of democratic theory. Only when democracy
itself emerges as paradoxical (for example, through the voter’s paradox) is
representation treated as paradoxical (since the voter’s paradox also applies to
the decision-making processes of representative assemblies).2 In this article I
explore what happens if this line of inference is reversed, and the paradox of
presence is made central to our understanding of democratic politics.
Part of the reason for the neglect of the paradox of presence by political
theorists is that it can appear to be a purely semantic problem, better suited to the
concerns of linguistic than political philosophy. However, as Pitkin makes clear,
the linguistic tensions in the idea of representation have important implications
1
Hanna Pitkin, The Concept of Representation (Berkeley: University of California Press, 1967)
and ‘Commentary: The paradox of representation’, Nomos X, Representation, ed. Roland J. Pennock
and John W. Chapman (New York: Atherton, 1968), pp. 38–42.
2
For a comprehensive recent account of the way democratic paradoxes play out in the setting of
representative institutions, and of the ways in which the paradoxes may be resolved, see Gerry
Mackie, Democracy Defended (Cambridge: Cambridge University Press, 2003).

© 2007 The Author. Journal compilation © 2007 Blackwell Publishing Ltd., 9600 Garsington Road,
Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
94 DAVID RUNCIMAN

for practical politics. Moreover, given the significance of the idea of

representation in the historical development of democracy, these tensions can
help to explain the evolution and functioning of various democratic institutions.
Pitkin argues that the historical evolution of liberal democracy has generated a
relatively stable set of institutional practices within which the tensions inherent
in the idea of representation can be accommodated. But her account is too
reductive, and fails to do justice both to the complexity of the paradox and the
complexity of the democratic institutions that have sought an accommodation
with it. In particular, she cannot fully explain how it is possible to represent ‘the
people’ or ‘the public interest’ in a way that is consistent with her understanding
of the concept of representation. Some democratic political theorists have used
this difficulty with Pitkin’s account to cast doubt on whether the concept of
representation can ever be an appropriate tool for making sense of democracy.
In this article I want to argue that there is a more satisfactory way of dealing
with the paradox of presence, which offers a better fit with the institutional
arrangements of contemporary democracies. This resolution involves adapting a
key feature of Pitkin’s argument – what I call her ‘non-objection criterion’ – to
the representation of groups rather than individuals. Pitkin, in common with
many other liberal political theorists, is deeply suspicious of any account of
representation that allows representatives to act in the name of groups over and
above the objections of their individual members. In this article I suggest that
some of these suspicions are unfounded, and that such an account of
representation is necessary to make sense of a number of otherwise puzzling
features of democratic politics.

I. PRESENCE VS. ABSENCE

The paradox of representation, as understood by Pitkin, arises out of ‘a
fundamental dualism built into the meaning of representation’.3 This dualism is
a consequence of a shift in the sense of the term from its origins in the Latin word
‘repraesentare’, which meant to grant something a literal presence, for example to
pay in cash or to present a new Pope before a crowd. It is essential to the meaning
of the modern conception of ‘representation’ that it does not involve the granting
of a literal presence to something: persons and things represented are not actually
present but, as Piktin puts it, must nevertheless ‘be made present in some sense
while not being present literally or fully in fact’.4 Non-literal presence here means
non-physical presence (represented persons and things are not actually in the
space where they are represented), but it does not follow that non-physical
presence is therefore unreal or insubstantial; rather, persons and things are
granted a kind of artificial presence by the act of representation, which can be as

3
Pitkin, Concept of Representation, p. 9.
4
Pitkin, ‘The paradox of representation’, p. 41.
THE PARADOX OF POLITICAL REPRESENTATION 95

real as the physical kind (the purpose of representation is to enable those who
are literally absent nevertheless to make their presence genuinely felt).5 The
paradoxical character of the concept of representation derives from the
simultaneity of this requirement of presence and non-presence, enabling
contradictory conclusions to be drawn from any given use of the concept. In
politics, it allows for the idea of representation to be identified both with the view
that representatives should take decisions on behalf of their constituents (because
the constituents must be absent for there to be representation at all), and with
the view that voters should issue instructions to their representatives (because
genuine representation also requires their presence in some recognisable form).
This is the source of the ‘mandate/independence’ controversy that has bedevilled
the idea of representation throughout the history of modern politics, and the
paradoxical dualism contained in the term is for Pitkin the reason why this
controversy can never be definitively resolved.6
Nevertheless, in The Concept of Representation, Pitkin sets out one possible
accommodation with the paradoxical requirements of representation that can be
achieved within a broadly liberal democratic understanding of the concept. This
is the view that ‘the substance of the activity of representing seems to consist in
promoting the interests of the represented, in a context where the latter is
conceived as capable of action and judgment, but in such a way that he does not
object to what is done in his name’.7 In this way, Pitkin makes clear that although
the activity of representation is tied to the promotion of interests, it cannot simply
be identified with the representation of interests, because it also has to allow for
a kind of ‘presence’ on the part of those whose interests are represented. This
presence comes from the ability of individuals to object to what is done in their
name. Because this is essentially a negative account of what constitutes the
activity of representation – representation takes place when there is no objection
to what someone does on behalf of someone else – it can cope with the
contradiction implied by the requirement that individuals be both present and
absent when they are represented. The non-objection criterion allows a kind of
latent presence for the represented, such that their silence can be taken as a form
of assent. However, it also means that where that silence is broken, and explicit
objections are voiced, representation starts to break down. It is very hard to
5
The word ‘artificial’ can be ambiguous in this context, since it is capable of conveying both that
something is essentially unreal and also that it is merely a man-made form of reality. So an artificial
leg is not a real leg, but an artificial lake is a real lake. I take the artificial presence generated by the
act of representation to denote what is merely a man-made form of reality.
6
‘It may be, as the notion of paradox in the meaning of representation suggests, that the issue is
usually formulated in such a way that it cannot be answered and will not allow a consistent response’
(Pitkin, ‘The paradox of representation’, p. 42). In a later book, Pitkin uses the example of the
open-endedness in the meaning of the concept of representation to argue that paradoxes of this kind
are best understood as a feature of our use of language, and therefore as one of language’s resources,
rather than as logical flaws in the concepts themselves; see Hanna Pitkin, Wittgenstein and Justice:
The Significance of Ludwig Wittgenstein for Social and Political Thought (Berkeley: University of
California Press, 1972).
7
Pitkin, Concept of Representation, p. 155.
96 DAVID RUNCIMAN

reconcile the overt presence of an individual who objects to the actions of their
representative with the implied absence of the represented on which the concept
of representation also depends.
There are a number of things to be said in favour of Pitkin’s response to the
paradox of presence through the non-objection criterion. It makes explicit that
there can be no straightforwardly substantive notion of representation (that is,
representation understood as acting in another’s best interests), because merely
acting in someone’s interests is not itself sufficient for something to count as
an act of representation.8 For example, a parent who feeds a child every day is
acting in that child’s best interests, but is not representing them in doing so.
Representation must entail some sense that the actions are being performed not
just on behalf of the represented (that is, to promote their best interests) but also
in the name of the represented (that is, giving them a stake in the action itself).
There are a variety of different ways in which it is possible to have a stake in the
actions of another person, not all of which count as instances of representation.
If I simply bear the consequences of an action performed by someone else (for
example, if I am adversely affected by someone else’s mismanagement of a public
good), then I may have a stake in the action without necessarily being represented
by the person performing it.9 Representation implies that the represented do not
merely bear the consequences of another’s action, but have some presence in the
action itself by dint of this fact. In other words, they must be capable of asserting
their stake. The ways in which this stake can be asserted range from strong forms
of control (such as issuing binding instructions to the representatives), through to
somewhat weaker forms of identification (choosing representatives with whom
one shares features in common and therefore with whom one can identify) or
weaker forms of control (choosing the procedure by which representatives are
appointed), through to mere non-objection (having but not utilising the ability to
object). This last comes closest to capturing the way the concept of representation
has evolved in the politics of liberal democratic states. We do not on the whole
believe that political representation depends on individuals having a stake in the
actions of their representatives through having instructed them or being able to
identify with them. (In this sense, the mandate/independence controversy, though

8
Pitkin is often taken to have offered just such a straightforwardly substantive definition of
representation. See, for example, Adam Przeworski, Susan C. Stokes and Bernard Manin ed.
Democracy, Accountability and Representation (Cambridge: Cambridge University Press, 1999), p. 2,
in which Pitkin is quoted saying: ‘This is what we mean by “representation”: acting in the best
interests of the public’. This definition is typical of the way Pitkin’s work has usually been employed
by contemporary political theorists and political scientists, and of the neglect of the role of the
paradox of presence in her account of representation. For a similar reading of Pitkin that neglects her
concern with the paradoxical character of representation, see David Lublin, The Paradox of
Representation: Racial Gerrymandering and Minority Interests in Congress (Princeton, N.J.:
Princeton University Press, 1997).
9
The contemporary use of the term ‘stakeholders’ to describe all those who may be impacted by
a collective project makes this point implicitly: ‘stakeholders’ include individuals who may otherwise
be neglected by focussing on familiar lines of representation (and may for example include individuals
whose relationship to a project is simply their physical proximity to it, as ‘neighbours’).
THE PARADOX OF POLITICAL REPRESENTATION 97

it rumbles on, has come to favour the independence side of the argument).10
Equally, despite the fact that lip-service is often paid to the idea that citizens
should be able to alter the constitutional rules governing the procedure for
choosing representatives, the opportunities for doing this are in practice
extremely rare; and when it happens, it tends to be at the behest of the
representatives themselves. Yet we do feel that representatives whose actions
provoke explicit objections on the part of those they claim to represent must
eventually cease to be their representatives, even if they believe themselves to be
acting in their constituents’ best interests. Thus the most significant and decisive
forms of political control in modern democracies derive not from positive acts
of instruction or identification or procedural initiatives on the part of the
represented, but from the possibility of their objecting to what is being done on
their behalf.
However, Pitkin’s account suffers from an important limitation. Although the
non-objection criterion is a less exacting test of what is to count as representation
than the ‘mandate’ view, it is only less exacting so long as individuals do not
object to what is being done by their representatives. Where they do object, then
their objections are decisive. This means that a great deal of weight must be
placed on the meaningful possibility that individuals will object to, and therefore
disrupt, the actions of their political representatives. Indeed, as a test of political
representation the non-objection criterion is meaningless unless it is assumed that
individuals will sometimes object to what is being done on their behalf. So the
absence of objections cannot be taken as a form of assent if it is the result of
coercion: tyrannical or dictatorial regimes that bully or terrorise their subject
populations into not objecting to the way they are being governed are not
representative in any meaningful sense. But it also follows that representation
must cease whenever the person or thing being represented lacks the ability to
object because of some inherent incapacity of their own. For example, Pitkin does
not believe it makes sense to talk of representing a small child, regardless of the
kinds of actions undertaken on their behalf, so long as the child is incapable of
making its own views known: ‘If we think of him . . . as helpless and incapable
of action, as being taken care of, then we will not speak of representation’.11
Pitkin’s account also excludes the realistic possibility of representing what she
calls ‘abstractions’, all those entities that cannot act in their own right at all. ‘An
abstraction’, she writes, ‘does not have wishes and cannot suddenly rise up and
object to what a representative is doing in its name’.12 Of course, abstractions can
be represented in other ways – in works of art, in symbols and so on – but this
for Pitkin is precisely what distinguishes substantive political representation from
mere symbolism.

10
There are of course those who dissent from this view. See for example Anne Phillips, The Politics
of Presence (Oxford: Oxford University Press, 1995).
11
Pitkin, Concept of Representation, p. 155.
12
Ibid.
98 DAVID RUNCIMAN

This limitation on the scope of who or what can be represented is a way of

trying to rescue the practice of representation from the paradox of presence. It
does this by limiting genuine representation to the representation of individuals
who could object but choose not to: this allows them to be present (because their
non-objection is what determines the representation) but also absent (because it
is still non-objection). However, it remains a considerable limitation, and it is
open to a number of serious objections. First, it is not clear that the evolution of
the term does exclude the representation of persons and things incapable of
acting for themselves. Certainly, it is common enough to speak of representing
children, or the unborn, or abstractions like ‘the environment’. Pitkin is
committed to claiming that these are misuses of language, and in many instances
she is right – for example, she is correct to point out that the frequent
identification of legal ‘trusteeship’ as a form of representation is a mistake,
because trustees act in relation to pieces of property, rather than representing the
persons to whom the property will revert.13 However, Pitkin does not consider
the possibility that the non-objection criterion can be met without limiting
representation to those individuals capable of objecting on their own behalf. A
child, for example, might be represented in such a way (say in a court of law) that
a parent could object to what is being done on the child’s behalf. This is, indeed,
what we normally mean when we speak of the representation of children, or of
other incapable entities: the representative may speak on behalf of those who
cannot speak for themselves, so long as a mechanism exists for objecting to what
is being done in their name. Where no such mechanism exists – if a parent
claimed the exclusive right to speak for their child such that no one could object
to what the parent did – then there is a strong case for excluding such actions
from the scope of representation. But such mechanisms do exist, in law and, as
I will discuss later on, in politics.
This difficulty also holds the other way round. It is not clear that
representation must cease in all instances where the person represented objects to
what is being done on their behalf. A court of law may impose a representative
on an individual against that individual’s wishes, if it believes that the person in
question is incapable of acting for himself or herself. This is a complex area
(witness, for instance, the legal disputes over whether Slobodan Milosevic could
be compelled to acquire legal representation at his war crimes trial, against his
wishes to continue his own defence),14 but it seems perverse to exclude this form
13
Ibid, pp. 129–31.
14
Details of some of the arguments are available at: http://hague.bard.edu/. For example, from the
6 July 2004 Order on the Future Conduct of the Trial: ‘However, the right to defend oneself in person
is not absolute . . . there may be circumstances . . . where it is in the interests of justice to appoint
counsel. No such circumstances have, as yet, arisen in this trial. However, as the Trial Chamber has
said, it will keep the position under review. . . . A Trial Chamber has indeed an obligation to ensure
that a trial is fair and expeditious; moreover, where the health of the Accused is in issue, that
obligation takes on special significance . . . [W]hile ensuring that the trial is fair and expeditious, a
Trial Chamber must also ensure that the rights of the accused, as set out in Article 21 of the Statute,
are not infringed’.
THE PARADOX OF POLITICAL REPRESENTATION 99

of legal action from the category of representation when it does occur. What
makes it a form of representation is the capacity of some party (in this case, the
court) to object to what is being done on behalf of the defendant (and if necessary
to remove the representative altogether, if they appear to be acting against the
defendant’s interests). Incapable persons and things can have a presence in the
actions of their representatives, so long as that presence is capable of being
asserted by someone. It does not have to be asserted by the person being
represented themselves.
The question of what is to count as a genuine instance of representation is
evidently a legal and semantic problem as much as it is a political one. There is,
however, a particular political difficulty with Pitkin’s argument. Her
non-objection criterion explicitly confines representation to the representation of
individuals. She identifies ‘the substance of the activity of representation’ with the
representation of the individual who ‘does not object to what is done is his [sic]
name’. A similar presupposition also underlies Pitkin’s contention that ‘normally
a man’s wishes and what is good for him will coincide. Thus if a representative
in fact succeeds in doing what is good for his constituents, normally he should
not then find himself in conflict with their wishes’.15 This line of argument
reinforces Pitkin’s contention that the practice of representation does not need
to become bogged down in the independence/mandate controversy: genuine
representation is possible in the absence of explicit instructions from the voters,
so long as it also coincides with the absence of explicit objections from the voters.
However, the argument depends on seeing the voters as a series of individuals,
capable of expressing their wishes for themselves. What remains unclear is
whether the objections of individuals can be equated with the objections of a
group of constituents as a whole. If not, then it needs to be asked how the
representatives of groups can continue to act in the face of objections of
individual members, that is, in the face of those who are capable of objecting for
themselves.

II. INDIVIDUALS VS. GROUPS

What is the relationship between the objections of individuals to the actions of
their representatives, and the ability of representatives to act in the name of the
groups to which those individuals belong? One possible answer, normatively
attractive to many forms of liberal political theory, is that individuals ought to be
able to exercise a power of veto – actions undertaken in the name of the group
should not be taken to represent individuals who explicitly state their objections.
Certainly this view accords with the strong individualistic bias in many
contemporary uses of the term, such that it is possible for individuals to claim
that they are no longer being represented when their views are not being listened

15
Pitkin, Concept of Representation, p. 156.
100 DAVID RUNCIMAN

to. Indeed, where representation is explicitly the representation of views,

individuals do exercise a veto over the claims of their representatives to represent
them. If one human being purports to represent the views of another, a simple
declaration by the other that ‘Those are not my views’ is sufficient to end any
claims that the representation is genuine. There is nothing the representative can
do in these circumstances to gainsay the objection – even if it is possible for a
representative to argue that they know better than those they represent, it is
nonsensical to claim that they know their views better. This kind of objection can
have implications for various forms of authorisation or legitimacy. For example,
if a lawyer pleads not guilty on behalf of a defendant, but the defendant declares
that they wish to plead guilty, it is very hard to see how the lawyer can plausibly
be said to represent their client, unless authorised to do so by the court (and even
then, it is hard to see how such a fundamental disagreement could be reconciled
with the court’s claim to be acting in the interests of the defendant).
The ability of individuals to express their views, and to have those views heard,
is a central feature of any kind of ‘genuinely’ representative politics: no use of the
term ‘representative’ to describe a political system in which individuals are denied
the ability to express their views would be plausible. Freedom of expression is
thus one of the necessary conditions for any claim regarding representation in
politics to hold true. However, freedom of expression is not the only genuine
feature of representative politics, nor is it necessarily the consideration that
trumps all others. The problem derives from another necessary feature of systems
of representative government, which is majority voting. Inevitably, under any
form of politics in which majorities decide, some individuals will have their
objections overruled. Members of the outvoted minority can say with genuine
conviction that their views are not being represented, particularly if they belong
to permanent minorities or lack competing political outlets through which to
have their views expressed. But individuals whose objections are overruled do not
cease to have representatives, in the sense that they still belong to the constituency
on whose behalf a representative acts. On one level, this is a technical distinction,
because elected representatives occupy certain formal roles in relation to their
constituents, regardless of their views (for example, a constituent may seek advice
or help from ‘their’ representative on a matter of personal concern, regardless of
whether or not they have ever voted for them). The question is whether all
representation that overrides the objections of individuals is merely ‘technical’, or
whether it is genuinely possible to represent groups in the face of the objections
of some of their individual members.
The problem Pitkin has in answering this question is that her original
insistence that only persons capable of action in their own right are capable of
being represented makes the question meaningless, since it allows no meaningful
distinction to be drawn between the group and the views of its individual
members. Yet it is not a meaningless question. Indeed, it is hard to see how we
can make sense of the practice of political representation if we are to exclude the
THE PARADOX OF POLITICAL REPRESENTATION 101

possibility that groups can be represented in their own right. If it is accepted that
groups can act through the wishes of the majority of their members, then it would
follow that the outvoted minority are still being represented as members of the
group, despite their objections to what is being done in their name. On the other
hand, if it is insisted that groups cannot act for themselves, and are merely made
up of the separate actions of their individual members, then representation can
never be determined by majority decision, and must stand or fall on an individual
by individual basis. This seems implausible, both conceptually and politically.
Take, for example, the case of the recent Iraq war, in which elected governments
went to war despite the vociferous objections of large numbers of individuals. In
one respect, those individuals who took part in demonstrations against the war
(compare the full version of Pitkin’s non-objection criterion, which is to ‘rise up
and object’) were making it clear that they no longer believed themselves to be
represented by the actions of their governments. One of the slogans of the
anti-war movement – ‘Not in My Name’ – made this point explicitly: individuals
were seeking to sever the ties of representation by asserting their non-presence in
the actions of their representatives. However, the governments that went to war
could contend that this claim does not sever the ties of representation. That is
because they did not act ‘in the name’ of any specified individuals. Rather, they
acted in the name of a group of individuals – ‘the people’ – or even, in the British
case, a kind of abstraction – ‘the state’ (or more technically, but also somewhat
absurdly, ‘the Crown’).16 The capacity to act in the name of a group presupposes
the ability to represent the group despite the objections of some and perhaps
many individuals to the way their interests are being promoted. This is something
that representative governments have always claimed to be able to do.
If representation is determined by the ability of individuals to object to what
is being done in their name, then it is hard to see how any claim to representation
that overrides the objections of individuals in the name of the group can stand up;
in other words, it is hard to see how the objectors can continue to be represented.
Yet this conclusion need not follow if one of the premises on which Pitkin’s
account depends is abandoned. This is the insistence that representation can only
apply to entities capable of objecting in their own right. If representation is taken
to include the possibility of representing abstractions, on whose behalf others can
rise up and object, then it is possible to understand the representation of the
people as something genuine – or at least as genuine as any other form of
representation. Even if the people itself cannot object to what is being done in its
name, individuals can object on its behalf; moreover, if sufficient numbers of
individuals object on its behalf, then it becomes impossible for a government to
continue to claim to represent the people. If nothing else, such an account of
representation is able to make sense of what is otherwise a deeply puzzling
16
On the juristic and philosophical history of this absurdity, see F. W. Maitland, ‘Corporation sole’
and ‘Crown as Corporation’, Maitland: State, Trust and Corporation, ed. David Runciman and
Magnus Ryan (Cambridge: Cambridge University Press, 2004).
102 DAVID RUNCIMAN

feature of democratic politics, which is the practice of general elections. When a

government is voted out of office following such elections, many individuals will
have voted for the defeated party, yet we do not say that these individuals
altogether cease to be represented by the new government that replaces their
preferred choice. This is because we do not believe that governments simply
represent individuals and their choices; they also represent the people as a whole,
and cease to represent the people as a whole only when sufficient numbers of
individuals object to what is being done in the people’s name. Elections are the
best way of registering such objections, but not the only way (a government’s
claim to represent the people can be undermined by any overwhelming and
unambiguous expression of public dissent, though because these must be
overwhelming and unambiguous, they are also rare – as the Iraq war showed).
What representative politics requires is the capacity of individuals to object to
what is being done by representatives in whosever name those representatives
claim to be acting. It does not require that individuals should only ever be able to
object to what is being done on their own behalf. Indeed, such a requirement
makes representative politics impossible.

III. REPRESENTATION VS. DEMOCRACY

The danger of such an account is that it appears to introduce an unwarranted
ambiguity and instability into all forms of democratic politics. It allows that the
objections that individuals may make on their own behalf to what their
representatives do can be trumped by the claim of their representatives to speak
on behalf of the groups to which those individuals belong. At the same time, it
insists that the criterion by which acts of genuine representation are judged
remains the non-objection of individuals to what is being done in the group’s
name. How do we distinguish between the objections which do and the
objections which do not terminate a representative’s claims to represent? This is
not an easy question to answer, and the answer may vary, as I make clear below.
But having a variable answer is better than having no answer at all. Pitkin accepts
that the claims of politicians to represent certain groups, including the people
as a whole, do sometimes trump the objections of individuals to what is being
done in their name. She writes: ‘The representative system must look after the
public interest and be responsive to public opinion, except in so far as
non-responsiveness can be justified in terms of the public interest’.17 But she does
not explain how such non-responsiveness can be justified in terms of her concept
of representation, which identifies genuine representation with non-objection.
In the end, Pitkin is forced to concede that she cannot offer a satisfactory
explanation of what is meant by the representation of the people as a whole,
except to say that it is ‘a very general, abstract, almost metaphorical idea . . . the

17
Pitkin, Concept of Representation, p. 224.
THE PARADOX OF POLITICAL REPRESENTATION 103

people of a nation are present in the actions of its government in complex ways’.18
This has led some critics to point to the fundamental ‘incoherence’ in Pitkin’s
account of representation: as Adam Przeworski has argued, there is apparently no
way of reconciling the freedom for manoeuvre representatives need to act in the
public interest with the presumption that they should not act against the wishes
of their individual constituents.19
However, the conclusion Przeworski goes on to draw from this gap in Pitkin’s
account is too strong. He suggests that in consequence representation itself
should be seen as an inescapably incoherent idea, and that there is no way to
make sense of the idea of representatives acting in the interests of the people
within the confines of democratic theory (since once interests are separated out
from opinions, representatives cannot act in the public interest without being
undemocratic, and they cannot be consistently democratic without going against
the public interest). Przeworski concludes that the best we can hope for is a
minimal defence of democracy, as the best system for ensuring non-violent
transfers of power.20 Anything more – anything more ‘representative’ – is just an
illusion. But this is to assume that all the alternatives to Pitkin’s attempted
resolution of the paradox of presence through her strongly individualistic
non-objection criterion are equally unpalatable. It is true that many of the
alternatives are unpalatable. For example, one way out of the paradox would be
to discount the objections of individuals to what their representatives are doing
in the name of the public, on the grounds that the public has a mind of its own,
separate from the minds of its individual members. Pitkin herself accepts that an
abstraction might be capable of being represented in its own right, even on her
non-objection criterion, so long as it is conceived as capable of action in its
own right: ‘If we think of the abstraction as acting through its representative,
present in his activity, animating and directing what he does, then we will
speak of representation’.21 But she also assumes that there is no coherent way
of explaining how an abstraction can be present in the activities of its
representatives in this way without ceasing to be an abstraction. If it is to remain
an abstraction, then it hard to know how it can object to what is being done in
its name. One way out of this difficulty might be to argue that groups are not
abstractions precisely because they act through majority decisions, and that the
group can object for itself through the voice of the majority. But this simply
restates the problem rather than resolving it. If a group can only act through a
majority of its members because it is assumed that the majority can be said to

18
Ibid, p. 235. Pitkin concedes that there must necessarily be a gap between this vague goal and
‘some fairly concrete, practical and historically traditional institutions intended to serve such an
outcome’.
19
Adam Przeworski, ‘Minimalist conceptions of democracy’, Democracy’s value, ed. Ian Shapiro
and Casiano Hacker-Cordon (Cambridge: Cambridge University Press, 1999), pp. 23–50 at pp. 31–3.
20
For a recent summary of some of the implications of this position, see Adam Przeworski,
‘Institutions matter?’, Government and Opposition, 4 (2004), 527–40.
21
Pitkin, Concept of Representation, p. 155.
104 DAVID RUNCIMAN

represent the group as a whole, then this is the beginning of a regress, since the
ability of the majority to represent must depend on the possibility of individuals
objecting to that representation. On the other hand, if the majority does not
merely represent the group but in some sense constitutes it, then it is hard to see
what weight individual objections have at all. The same difficulty arises in
relation to decisions about the procedure for objecting to the actions of
representatives. If a group decides to adopt a procedure that makes the voice of
the majority decisive, then that procedural decision will either have to be the
unanimous choice of all members, giving everyone a veto, or it will be taken on
behalf of the group by a majority of its members, which begs the question of how
to treat the group’s relationship to those individuals who object to the procedural
decision.22 Groups that act in their own right, over and above the objections of
their individual members, must be assumed to have a mind of their own. The
problem is how to conceive this apparently metaphysical category in a way which
is consistent with the assumptions of the non-objection criterion.
This does not mean, however, that there is no alternative between an
incoherently individualistic theory of group representation and an unacceptably
illiberal one. The solution is to find a way of establishing a separate identity for
group ‘persons’ without attributing metaphysical qualities to them. There is more
than one way in which this can be done. For example, groups can acquire an
identity in their own right by the process of ‘collectivising reason’, an approach
that has been suggested by Philip Pettit as a solution to the problem of giving
groups their own personality without attributing to them a separate Geist-like
quality.23 This form of group identity is achieved by the expedient of adopting a
premise-based approach to collective decision-making, such that decisions follow
from majority approval of separate premises, rather than the majority preferences
of separate individuals. The result is that a group may legitimately reach a
decision that is not the preference of a majority of its members, even though
the decision remains consequent upon member preferences, tabulated on a
premise-by-premise basis. For Pettit, this is sufficient to afford such groups their
own personality (albeit ‘of a crude, bounded and bloodless variety’)24 and to
require of them the same standards of reasoning that we expect of other types of
persons, above all some consistency in their judgments across time. The problem
with such a solution for the purpose of making sense of political representation,

22
It is clear that individuals in modern states do often object to the procedure of majority-decision
making, so that it is not possible to assume unanimity on this score. Indeed, this is often the point at
issue in the objections of minorities to the actions of governments that discriminate against them: they
object not merely to the discrimination itself but to the fact that only the objections of the majority
could have forestalled it, which sets the bar for many minorities at an unacceptably high level. A
recent example of this is the objections of many fox-hunters to the ban imposed on their sport by the
British Labour government in 2004, which rested not merely on the fact that they objected to the ban
but also on the fact that they objected to the majoritarian presuppositions that lay behind it.
23
Philip Pettit, ‘Groups with minds of their own’, Socializing Metaphysics, ed. F. Schmitt (New
York: Rowman & Littlefield, 2003), pp. 167–93.
24
Ibid, p. 188.
THE PARADOX OF POLITICAL REPRESENTATION 105

however, is that this form of reasoning is not available to all groups, and Pettit
attaches some stringent conditions to the groups to which it does apply. He
argues that individuals ought still to be in a position to exercise some power of
veto over collectivised decisions, whether by exercising some right of exit from
the group, or by asserting their own right to refuse to act in the name of group
decisions. ‘Natural persons have an inescapable priority’, Pettit says, and
therefore ‘it will be up to the natural person to decide whether or not to cede
place to the institutional’. It must always be left up to individuals to decide
‘whether to act in their own name or in the name of the collective’.25
These conditions make it difficult to see how such a standard of collectivised
reason could hold for groups on the scale of modern states. States do not allow
a comparable right of veto to their individual members. Equally, as Pettit
recognises, the publics of such states do not exhibit the kind of consistency in
their decision-making over time that is a necessary condition of collectivised
reason. Public opinion is notoriously fickle, and in the absence of the institutional
mechanisms needed to collectivise its views on a premise-by-premise basis, the
public must remain ‘a non-person in the relevant sense’.26 But this does not mean
that the public cannot be conceived as a person at all. What it does suggest is that
in order to acquire an identity in their own right, states and their publics must
rely on the institution of representation, rather than collectivised reason. In
this sense, Pettit’s account of group reason does not offer an explanation of
representative politics, but an alternative to it, applicable to the kinds of
enterprises that can operate on the smaller scale and within the constraints
that collectivised reason requires. Pettit’s terminology is not the language of
representation, but of co-operative action. In fact, individual citizens only act ‘in
the name of the collective’ (that is, the state) when they perform certain specified
roles on its behalf (as soldiers, civil servants, diplomats and so on); otherwise, the
state acts for them. In this sense, modern states are not co-operative enterprises,
and their decision-making processes are not collective ones. Pettit, in his account
of collectivised reason, accepts that there may be circumstances in which even
small, co-operative groups need certain officials to direct the action of the
collective in order to help it reason in a premise-based way. He speaks of the
need for ‘plenipotentiaries [to] resist various irrationalities’.27 But these
plenipotentiaries, whatever their powers, operate as facilitators for the
collectivisation of group preferences into a rational form. The ultimate decisions
remain those of the group. By contrast, the plenipotentiaries of modern
representative democracies are not simply facilitators for collective
decision-making. Their decisions are taken in the name of the collective, and on
its behalf.

25
Ibid, p. 190.
26
Ibid, p. 186.
27
Ibid, p. 189.
106 DAVID RUNCIMAN

The account I have offered of group representation in this article seeks to do

justice to this basic fact of modern political life – that states are not co-operative
enterprises – while taking seriously the kind of group identity championed by
Pettit: that is, an identity which groups have independent of the views of their
individual members, yet one which remains subject to the views that their
individual members take of the activities of the group. In the case of the state, this
can be achieved by allowing a distinction to be drawn between the public on
whose behalf political representatives act, and the public whose opinions of the
actions of those representatives determine whether or not they can plausibly
claim to be representing the people as a whole. The first of these publics is an
abstraction in Pitkin’s terms – an incapable object at the mercy of the competing
claims of its representatives and the separate individuals who retain the right to
object to what is being done in its name. The second public – the public of ‘public
opinion’ – is not an abstraction as such, since it is constituted by agents capable
of acting in their own right. But it is still a group, and it will require some form
of direction or institutional control in order to voice its opinions effectively.
Public opinion, as Pettit points out, is unlikely to be either consistent or coherent
over time. This does not mean, however, that the public cannot be represented in
a consistent or coherent way. Because the abstraction on whose behalf political
representatives act cannot speak for itself, it can only acquire a separate personal
identity by being represented as though it possessed such an identity, which
means it must be represented in a broadly consistent way. This sort of consistency
is what we expect of the legal representatives of abstractions and other incapable
entities, such as small children. Such expectations of consistency are also
widespread at all levels of politics – no national government can repudiate all the
commitments of its predecessors and expect to be taken seriously. Nevertheless,
it is the objections or otherwise of the active public, and not the expectation of
consistency over time on the part of an inactive ‘people’, that is the final arbiter
of political representation. Public opinion may turn decisively against a
government seeking to uphold earlier commitments undertaken in the name of
the people – for example, many democratic governments have been forced to
repudiate their public debts because of the objections of large sections of their
populations to the privations honouring those debts would entail. In these
circumstances, the non-objection criterion trumps the claims of rational
consistency. Not all the twists and turns of public opinion constitute decisive
objections to the claims of representatives to act in the name of the people –
indeed, most do not, which is what allows consistent action to be undertaken on
behalf of the public over time. But some objections are unanswerable, and
without the possibility of such objections, political representation would not be
meaningful at all.
On this account, it is clear that individuals do not possess a veto over the
claims of their governments to represent them. As a result, political
representation is best understood not in the language of veto but of competition.
THE PARADOX OF POLITICAL REPRESENTATION 107

Objections to the actions of representatives can prove decisive when they

constitute a plausibly competing claim to speak in the name of the person or thing
being represented. A parent who appoints a representative to act on behalf of a
child cannot reasonably expect to retain a right of veto over everything the
representative does; if they do retain such a right, it makes more sense to think of
the parent and not the child being represented. But a parent can object to what
a representative does by asserting a competing claim to speak on behalf of the
child, trumping the claims of the representative to act in the child’s name.
Likewise, political representatives who act in the name of the people can have
that claim trumped by the competing claims of other representatives. There are
no natural claims here, comparable to the claims of a parent to speak for their
child.28 Moreover, in modern representative democracies, the competition still
revolves around electoral cycles, with their complex and somewhat contingent
rules and conventions governing what is to count as a decisive objection. But
there can be other forms of competition and other kinds of objections, up to and
including the kinds of objections voiced by crowds of protestors. For these
objections to count, they must be voiced not on behalf of the crowd, speaking for
themselves, but by a crowd making a claim to speak in the name of the people.
An alternative slogan to ‘Not in My Name’ for gatherings of protestors objecting
to the Iraq war was ‘Not in Our Name’ (this was the slogan of sections of the US
anti-war movement). There are two ways of understanding such a slogan: first, as
an act of collective solidarity in the face of government indifference (that is, you
may not act on behalf of any of us gathered here); second, as a competing claim
to speak for the people as a whole (that is, we gathered here represent the wider
public, on whose behalf we are voicing our objections). It is difficult, but by no
means impossible, to imagine the circumstances in which the second kind of
claim might prove decisive, whereas the first claim is merely expressive of the
personal views of the individuals who articulate it. If the second claim holds, it
will be because the objection assumes a form comparable to the claims of other
popular representatives (including elected ones) to be acting in the name of the
people.29
To view political representation as a competition for the right to speak on
behalf of an inanimate ‘people’ has strong echoes of the Schumpeterian
conception of democracy. But this is not a strictly Schumpeterian account because
the Schumpeterian conception of democracy is designed, in the words of one of
28
Although some classic theorists of political representation, including Thomas Hobbes
surprisingly, have tried to push the analogy between the parent/child relationship and the relationship
between individual citizens and the public as a whole; see Quentin Skinner, ‘Hobbes and the purely
artificial person of the state’, Visions of Politics (Cambridge: Cambridge University Press, 2002), vol.
3.
29
This is one of the reasons why elected assemblies are so reluctant to allow popular
demonstrations to take place in the same place and on the same day that the assembly itself is deciding
on matters of importance: the British government allowed mass demonstrations to take place in
London against the Iraq war, but did not permit mass protests at Westminster on the days of the
crucial House of Commons votes on the war.
108 DAVID RUNCIMAN

its most prominent current champions, to ‘elide’ the language of representation,

which is seen as ‘inescapably suspect’.30 This suspicion of representation derives
from the paradox of presence: as Ian Shapiro writes, the language of
representation seems to offer an unpalatable choice between the unacceptable
‘elitism’ of representatives who ignore the wishes of their constituents, and the
unstable pandering of representatives who respond to the vicissitudes of public
opinion.31 In this article I have tried to show that the language of representation
does not have to be reduced to this sort of choice, and that it can accommodate
a broadly coherent conception of politics that is still a recognisably democratic
one. Nor is this simply a minimalist conception of democracy: representative
politics, although it reduces the focus of political activity to the competing claims
of representatives to act in the name of the people, does not reduce the public to
a passive role. Individual members of the public remain the arbiters of political
representation. In order to fulfil that role they need access to the kind of
information that enables them to form judgments about their representatives, and
access to the kind of outlets that enable them to pass such judgments as they
make. The language of democratic ‘consumer sovereignty’, which Shapiro offers
as the Schumpeterian alternative to the language of representation, emphasises
the need for competitive incentives to ensure free flows of information as well as
opportunities for dissent. But the language of representation can provide for the
same sort of emphasis,32 while also making sense of a distinctive feature of
modern political life, which is that elected politicians do not simply offer a choice
of wares to discerning consumers; they also promise to represent them. These
promises are not meaningless because political representation offers a uniquely
flexible form of collective activity, in which actions are undertaken in the name
of the abstract group, but assessed by a collection of separate individuals.

IV. CONCLUSION: TWO OBJECTIONS

I want to conclude by considering two possible objections to the attempted
resolution of the paradox of presence that I have offered here. The first concerns
its implications for questions of collective responsibility. By turning the group
into an abstraction, this account of political representation may make it possible
for individuals to escape their share of responsibility for what is done in the name
of the group. Certainly, it is a model that shares some features in common with
the ‘limited liability’ model of corporate responsibility: in both cases the existence
of a corporate entity which is entirely distinct from the individual members of the

30
Ian Shapiro, The State of Democratic Theory (Princeton, N.J.: Princeton University Press, 2003),
p. 58.
31
Ibid.
32
On the importance of ‘informational access’ for the viability of representative politics, see John
Dunn, ‘Situating political accountability’, Democracy, Accountability and Representation, ed.
Przeworski et al., pp. 340–2.
THE PARADOX OF POLITICAL REPRESENTATION 109

group allows individuals to distance themselves from the consequences of group

activity by attributing the burden of responsibility to the impersonal corporate
entity.33 If political representation is not undertaken in the name of individual
citizens, but of the group, then it is hard to see why individual citizens should be
responsible for what political representatives do. Such a model contrasts sharply
with recent attempts by political theorists, most notably David Miller, to identify
forms of collective responsibility which distribute responsibility for a group’s
actions among all its individual members, even if those actions are undertaken
on the group’s behalf by others, or in contradiction to the personal views of some
of the members concerned. On these accounts, personal responsibility derives
from the contribution, either explicit or implied, of every individual member
to the group’s activities. For this reason, strong attributions of personal
responsibility make best sense for collectives of a certain character: they should
be either what Miller calls ‘like-minded groups’ (whose members share certain
aims and outlooks) or ‘co-operative practices’ (whose members share benefits
from activities in which they all participate).34 As I have suggested already, it is far
from clear that associations on the scale of modern states fit either of these
categories. But equally significantly, associations that do fit these categories do
not need to employ the language of representation to make sense of the collective
actions of the group. Indeed, groups in which personal responsibility can be
attributed to all members must discount the paradox of presence by assuming
that the presence of members in the actions of the group, even if merely implied,
trumps the absence of individuals from the performance of those actions. Groups
which rely on the concept of representation cannot escape the paradox of
presence so easily, because representation implies that presence must co-exist
with a form of absence, instead of simply trumping it. Thus the fact that modern
states have proved themselves so dependent on the concept of representation to
make sense of their collective identities is one reason why states do not readily
correspond with the ‘like-minded’ or ‘co-operative’ group-type.
On my account of the non-objection criterion, states do not fit these strong
models of collective responsibility. Nevertheless, two potentially signficant
conclusions follow from this account for the question of individual responsibility

33
See David Runciman, ‘Is the state a corporation?’, Government and Opposition, 35 (2000),
90–104.
34
David Miller, ‘Holding nations responsible’, Ethics, 114 (2004), 240–68 at p. 249. Miller himself
is primarily interested in the responsibilities of nations rather than those of states, because he believes
that state responsibility may allow individuals to escape their personal responsibilities: ‘So why do we
need to consider national responsibility at all? One reason is that if we divorce state responsibility
from national responsibility, then it becomes difficult to show how individual people can share in the
responsibility to compensate those whom the state they belong to has harmed, whereas if we treat
states as acting on behalf of nations, such collective responsibility will be easier to establish’.
However, he does not explain in any detail what he means by ‘states acting on behalf of nations’,
which is the point at issue in my discussion here. I consider some of these questions in more detail in
David Runciman, ‘Moral responsibility and the problem of representing the state’, Can Institutions
Have Responsibilities? Collective Moral Agency and International Relations, ed. Toni Erskine
(Basingstoke: Palgrave Macmillan, 2003), pp. 41–51.
110 DAVID RUNCIMAN

for actions undertaken in the name of the state. First, individual objections to the
actions of political representatives are not sufficient to enable the individuals
concerned to escape responsibility for those actions. ‘Not in my name’ does not,
as we have seen, provide a get-out. Second, because actions are not undertaken
in the name of individuals but of the state, the relationship between individual
citizens and political representatives does not determine the distribution of
responsibilities for collective actions. Rather, the burden of responsibility is
determined by the relationship between individual citizens and the impersonal
entity of the state. This relationship is one of membership rather than
representation: individual citizens don’t represent the state and the state doesn’t
represent individual citizens; instead, the individuals are members of the state,
and as such the burdens that they share as individuals are determined by the
conditions of their membership (for example, when political representatives raise
taxes in the name of the state, individual citizens are bound to pay those taxes by
the terms of their membership of the state, not because they were individually
represented, and therefore implicated, in the decision to raise taxes). What this
means is that there is a triangular relationship at the heart of modern politics:
individuals pass judgments on politicians; politicians act in the name of the
impersonal state; membership of the state determines the personal responsibilities
of individuals.
Each of these separate relationships is significant in its own right. But what is
most significant for the question of collective responsibility is that none of the
relationships predetermines the content of any of the others. The objections or
non-objections of individuals to the actions of government neither exculpate nor
implicate the individuals concerned. Certainly it does not follow from the fact
individuals cannot abdicate their responsibility for collective actions by their
personal objections, that they must all share equally in any collective
responsibility. For example, the antithesis of the view that ‘Not in My Name’
exculpates individuals from collective responsibility for the Iraq war is the view
that individual citizens are legitimate targets for terrorist reprisals for the war, on
the grounds that in a democracy all individuals are equally implicated in
government actions by their failure to prevent those actions by changing the
government.35 This very strong interpretation of the implications of the
non-objection criterion is in no way endorsed by the argument I have put forward
here. It is true that on my account the responsibility for any failure to object to
what government does in the name of the state belongs to individual citizens. But
that responsibility remains distinct from any responsibility for the actions

35
This is essentially Osama Bin Laden’s position, as set out in his ‘Letter to the American
People’ of November 24, 2002; available at: http://www.globalsecurity.org/security/
library/report/2002/021120-ubl.htm (accessed 10 July 2006). The possibility of making sense of
this very strong notion of democratic accountability within the terms of the Western
political tradition has been highlighted by Richard Tuck, ‘Democracy and Terrorism’ (mimeo.,
Cambridge 2003).
THE PARADOX OF POLITICAL REPRESENTATION 111

themselves, which attaches to the state. How the responsibility that attaches to
the state is then distributed among individual citizens cannot be read off the
relationship between individuals and their political representatives. There is no
chain of responsibility that follows the chain of representation. Instead, the chain
of responsibility that runs from the impersonal state to the individual citizen
provides the basis on which individuals ultimately pass judgment on acts of
political representation. These judgments do not predetermine the distribution
of collective responsibilities. What they do determine is the outcome of the
competition between rival claimants to be the representatives of the people, and
the outcome of that competition ultimately determines the extent of the personal
burdens that individual citizens will have to bear. In other words, political
representation does not in itself guarantee that the burdens will be either heavy
or light. How heavy they are is a function of the terms of individual citizenship
in the state in question. Political representation guarantees that there exists the
means to object to the burdens imposed, by objecting in the name of the state to
what its political representatives are doing in its name. The objections of
individual citizens to what is being done in their own names do not carry the
same weight, and in some circumstances they do not carry any weight at all.36
This open-ended, indeterminate account of collective political responsibility
accords with the open-ended, indeterminate understanding of collective political
responsibility that runs through the history of modern states. There are no fixed
answers. How burdens are distributed depends upon the interplay of three
relationships: between public opinion and government action; between
government action and the state in whose name it is undertaken; and between the
state that bears the responsibility and the individual citizens among whom the
burdens of responsibility are allocated. Each of these relationships is potentially
complex, and their interrelationship is potentially more complex still. However,
this suggests a second line of objection to the account of political representation
I have offered: that it is a very state-centric account, and ignores others forms of
political representation which provide clearer and more immediate chains of
accountability. I have explained why I think strong models of collective
responsibility do not fit the workings of the state, though they may fit other
groups with more specific goals or operating on a smaller scale than the state.
But I have simply assumed that the workings of the state are central to an
understanding of the concept of political representation. It is clear that the
concept of political representation has been shaped by its integral role in the

36
For example, a personal objection to continuing to service the national debt on the grounds that
the individual citizen concerned did not want to borrow the money in the first place is an empty
gesture, since public debt does not attach to any named individuals, but to the state as a whole. The
only meaningful objection is to the representatives who undertook the debt; whether replacing them
with different representatives will release individual citizens from the burden of servicing the debt
does not depend on who objected, but on what actions the new representatives undertake in the name
of the state (above all, on whether or not they repudiate the debt itself).
112 DAVID RUNCIMAN

emergence and development of the modern nation state.37 What is less clear is
that nation states will remain the primary focus of political life, now and in the
future. States are increasingly subject to the demands of international law, which
imposes various constraints on the actions their governments can undertake in
their name. Meanwhile, non-state associations are coming to play an ever greater
role in political life, notably at the level of international governance. Both of these
developments have implications for the concept of political representation. I
do not have the space to go into the question of what an enhanced role for
international law might do to our understanding of the concept of representation,
except to say it is unlikely to diminish the need for it to be able to accommodate
complexity. But one thing that is clear is that many of the constraints imposed on
states by international law depend upon establishing a distinct identity for the
state itself, over and above the identities of its individual members and corporate
agents.38 This highlights one of the difficulties of holding groups responsible for
their actions: the more distinct the identity of the group, the greater the possibility
of separating out the group’s responsibilities from those of the individuals who
make it up. In the case of states, this means that although the freedom of action
of governments may be constrained by international law, the central importance
of establishing a separate identity for the state will remain. The model of political
representation that I have offered here can deal with this outcome, because it is
effectively a model of corporate responsibility. What remains to be seen is what
impact the internationalisation of law may have on the role of domestic public
opinion in voicing decisive objections to the actions of political representatives.
It is possible that such a role may be increasingly assumed by international courts
instead.
It is the other potential threat to the continuing centrality of the state that I
want to consider here. An alternative to the account I have offered here is one
that argues for a variety of sites of political representation other than the state,
corresponding to the range of non-state bodies that have the potential to
represent an individual’s interests. A version of this argument has recently been
put forward by Andrew Kuper, in which he draws explicitly on Pitkin’s account
to provide a conception of what he calls ‘representation as responsiveness’ that
fits the institutional arrangements of the 21st century.39 Kuper accepts Pitkin’s
version of the non-objection criterion, such that representation must be assessed
according to whether or not an individual objects to what is being done in his or
her name: ‘responsiveness’ is the requirement that these objections be taken
seriously. But Kuper bypasses the difficulties of applying such a strongly

37
See Bernard Manin, The Principles of Representative Government (Cambridge: Cambridge
University Press, 1997).
38
See James Crawford, The International Law Commission’s Articles on State Responsibility.
Introduction, Text and Commentaries (Cambridge: Cambridge University Press, 2002), esp. ch. 2,
‘Attribution of conduct to a State’.
39
Andrew Kuper, Democracy Beyond Borders. Justice and Representation in Global Institutions
(Oxford: Oxford University Press, 2004).
THE PARADOX OF POLITICAL REPRESENTATION 113

individualistic test to the collective activities of the state by championing a range

of different institutional devices through which individual objections can be
voiced – ombudsmen, advocacy agencies, international pressure groups and so
on. The idea is to ensure that individuals should have some outlet for their
objections beyond the narrow confines of the state. The consequence would be to
move away from a preoccupation with the relationship between individuals and
their political representatives (that is, those who have been voted into office as a
result of general elections), and towards a conception of representation that asks
whether an individual is being represented within ‘the system as a whole’. This is
an essentially functional account of representation and it shares features in
common with earlier functional accounts, particularly the doctrine of political
pluralism that briefly flourished during the first decades of the last century.40
It takes a straightforwardly individualistic conception of representation –
representation as the promotion of a person’s interests in the absence of their
objections – and introduces variety by extending its scope to the different arenas
in which a person might find their interests either being advanced or thwarted:
economic, social, environmental as well as political. It assumes that our interests
cannot be satisfactorily contained at the level of the state, and that there are
obvious advantages to providing different representatives for the different
interests a person may have. This variety is an attempt to increase the chances
that an individual’s objections to the actions of their political representatives will
be heard, and acted upon, by representatives of some other kind.
Functional representation is often presented (as it is more or less presented by
Kuper) as a costless advance on narrower forms of national electoral politics.
However, like earlier twentieth century versions of political pluralism, a scheme
that prioritises variety over uniformity in political representation is not without
costs. It relies heavily on two assumptions: first, that serious conflicts will not
arise between different representatives claiming to act in the name of the same
individual (for example, one representing him or her as a consumer, the other as
a citizen); second, that individuals will not need an outlet for their objections to
the system of representation as a whole. In other words, what may be lost is the
capacity to impose coherence on the system by allowing for its representation as
a coherent system. This can only be achieved by allowing for variety within the
concept of representation – so that systems, or groups, can be represented as well
as individuals. As I have tried to show, this does not mean abandoning the
non-objection criterion, but it does mean adapting it, and introducing variety
into the concept itself, rather than simply trying to apply a rigidly individualistic
concept in a variety of different settings. The paradox of presence suggests that
some variety in our understanding of representation is unavoidable. There is no
single answer to what is to count as ‘genuine’ representation, even though we

40
On the details and fate of this movement, see David Runciman, Pluralism and the Personality of
the State (Cambridge: Cambridge University Press, 1997).
114 DAVID RUNCIMAN

may know in any given case what makes a claim to represent a bogus one. The
question is whether it possible to find a single location in which bogus claims to
representation can be exposed without eliminating the variety on which the
concept of representation also depends. Political representation at the level of the
state does this by allowing claims to represent the people as a whole to be subject
to definitive forms of objection which trump the objections of individuals to the
ways they are represented in other capacities. Kuper, and other cosmopolitan
theorists, are right to point out that the representation of the people should not
be excessively constrained by the contingent historical circumstances of national
politics in which such claims were first put to the test. But it would be wrong to
imagine that the basic structure of popular representation in electoral politics can
be either replaced or even merely supplemented by functional representation
without losing much of its clarity of purpose. There is no reason to think that
such clarity can only be achieved at the level of national politics. But there is also
no reason to suppose that any form of political representation can get by without
the kind of coherence that comes from enabling the state to be represented in its
own right. In the end, coherence of this kind can only be achieved by an
accommodation with the paradox of presence.