Scots Philosophical Association

University of St. Andrews

Consequentialist Defences of Liberal Neutrality

Author(s): Simon Caney
Source: The Philosophical Quarterly (1950-), Vol. 41, No. 165 (Oct., 1991), pp. 457-477
Published by: Oxford University Press on behalf of the Scots Philosophical Association and
the University of St. Andrews
Stable URL: https://www.jstor.org/stable/2220080
Accessed: 08-05-2025 21:39 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Scots Philosophical Association, University of St. Andrews, Oxford University Press

are collaborating with JSTOR to digitize, preserve and extend access to The Philosophical
Quarterly (1950-)

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 The Philosophical Quarterly Vol. 41 No. 165
ISSN 0031-8094 $2.00

CONSEQUENTIALIST DEFENCES OF LIBERAL

NEUTRALITY

BY SIMON CANEY

Many contemporary liberals claim that the state should be neutral between
competing conceptions of the good life.' Ronald Dworkin, for example,
asserts that 'political decisions must be ... independent of any particular
conception of the good life, or of what gives value to life'.2 Similarly John
Rawls argues that 'the state is not to do anything intended to favor or
promote any particular comprehensive doctrine rather than another, nor to
give greater assistance to those who pursue it'.3 This conclusion is shared
by many other liberal political philosophers like Ackerman, Feinberg,
Gutmann, Kymlicka, Larmore, Lloyd-Thomas, Rosenblum and
Waldron.4
In this article I examine a particular type of argument that has been
given in defence of neutrality, namely consequentialist arguments. Such
arguments are not stressed by Rawls and Dworkin but they are emphasized
by many other neutralist liberals. In particular I shall consider and
reject (1) Will Kymlicka's 'competitive-evolutionary' argument; (2) Joel
Feinberg and Bruce Ackerman's 'fallibility' argument; (3) Stephen Guest's
'best judge' argument; (4) Feinberg and von Humboldt's 'character
deterioration' argument; and (5) Ackerman's 'importance of doubt'

Exceptions include Joseph Raz, The Morality of Freedom (Oxford: Clarendon Press,
1986); Vinit Haksar, Equality, Liberty and Perfectionism (Oxford: Oxford University Press,
1979).
2 Ronald Dworkin, 'Liberalism', A Matter of Principle (London: Harvard University
Press, 1985), p. 191.
3 John Rawls, 'The Priority of the Right and Ideas of the Good', Philosophy and Public
Affairs, 17 (1988), p. 262.
4 Bruce Ackerman, Social Justice in the Liberal State (New Haven, Connecticut: Yale
University Press, 1980), p. 11; Joel Feinberg, The Moral Limits of the Criminal Law, Volumes
1-4 (Oxford: Oxford University Press, 1984, 1985, 1986, 1988); Amy Gutmann, 'Commun-
itarian Critics of Liberalism', Philosophy and Public Affairs, 15 (1985), p. 313: Will Kymlicka,
Liberalism, Community and Culture (Oxford: Oxford University Press, 1989), pp.76-85, pp.
95-7, p. 96 n.2; Charles Larmore, Patterns of Moral Complexity (Cambridge: Cambridge
University Press, 1987), ch. 3; David Lloyd-Thomas, In Defence of Liberalism (Oxford:
Blackwell, 1988), pp. 1-2: Nancy Rosenblum, 'Introduction', Liberalism and the Moral Life.
(Cambridge: Cambridge University Press, 1989); Jeremy Waldron, 'Autonomy and Perfec-
tionism in Raz's Morality of Freedom', Southern California Law Review, 62 (1989).

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
458 SIMON CANEY
argument. In addition, I evaluate (6) Ackerman and Lloyd-Thomas's
'experimentalist-consequentialist' argument; (7) Larmore, Gutmann and
Rosenblum's modus vivendi argument; and (8) Kymlicka's 'dictatorship of
the articulate' and 'distorted dialogue' arguments. What each of these
arguments has in common is the belief that a neutral state - one that
abstains from using propositions about the good life when making policy -
has better consequences than a perfectionist state.

I. CONCEPTIONS OF NEUTRALITY
Before we appraise the consequentialist defences of the claim that the state
should be neutral between competing conceptions of the good, we should,
however, clarify what exactly is meant by neutrality. In particular it is
worth distinguishing between:

justificatory neutrality: the state is neutral if and only if it does not

make decisions on the basis of any considerations about the good
life.

consequential, neutrality: (equal effects neutrality) the state is

neutral if and only if it has an equal effect on all conceptions of the
good.
consequential,, neutrality: (equally easy neutrality) the state is
neutral if and only if it ensures that all conceptions of the good do
equally well.

Now neutralists invariably defend justificatory neutrality and eschew the

two consequential conceptions of neutrality.5 This is important because
critics of neutrality often foist consequential accounts of neutrality on
neutralists and then argue (plausibly) that such neutrality is unattainable.
Thus Stephen Salkever, for example, argues against the consequentiali
conception of neutrality that 'the liberal theorist's goal of just laws rather
than virtuous people is only apparently neutral, since all laws favor some
ways of life over others'.6 Similarly Michael Perry, Alasdair MacIntyre
and Ian Shapiro reject consequentialij neutrality on the grounds that it is
impossible for the state to act in a way such that all conceptions of the good

5 See for example, Ackerman, Social Justice in the Liberal State, p. 11; Dworkin
'Liberalism'; Will Kymlicka, 'Liberal Individualism and Liberal Neutrality', Ethics, 99 (1989),
pp. 883-5; J. Rawls, A Theory of ustice (London: Oxford University Press, 1971), p. 332, and
'The Priority of the Right and Ideas of the Good', p. 262. See also the references given in
footnote 4 for further evidence that neutralists are committed to justificatory neutrality.
6 '"Lopp'd and Bound": How Liberal Theory Obscures the Goods of Liberal Practices',
in R. Bruce Douglass, Gerald M. Mara and Henry S. Richardson (eds), Liberalism and the Good
(London: Routledge, 1990), p. 174.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 CONSEQUENTIALIST DEFENCES OF LIBERAL NEUTRALITY 459

do equally well.7 But, however powerful these comments are as critiques

of some consequentialist conception of neutrality, they are unpersuasive as
critiques of justificatory neutrality because all the state has to do in order to
satisfy this conception of neutrality is to abstain from using certain types of
reasons.8

II. THE 'COMPETITIVE-EVOLUTIONARY' ARGUMENT

Now that we have reported the conception of neutrality that liberal
neutralists defend, let us examine the arguments. One common con-
sequentialist argument given in defence of justificatory neutrality is
provided by Will Kymlicka, who defends neutrality on the basis of what
might be called the 'competitive-evolutionary' argument. The argument
runs as follows: in a neutral state, conceptions of the good compete and as
they compete the worthy will defeat the unworthy. The process is
analogous to the economic model of perfect competition and as the forms of
the good (products) compete, the best ones will attract the most adherents
(customers) and the worst the least adherents. Hence, neutral institutions
provide a framework for the evolution of the most worthy forms of life. As
Kymlicka argues,

liberal neutrality ... allow[s] each group to pursue and advertise

its way of life, and those ways of life that are unworthy will have
difficulty attracting adherents. Since individuals are free to choose
between competing visions of the good life, liberal neutrality
creates a marketplace of ideas, as it were, and how well a way of life
does in this market depends on the kinds of goods it can offer to
prospective adherents. Hence, under conditions of freedom,
satisfying and valuable ways of life will tend to drive out those which
are unsatisfying. Liberals endorse civil liberties in part precisely
because they make it possible 'that the worth of different modes of
life should be proved practically' (Mill On Liberty, 54).9

7 See A. MacIntyre, Whose Justice? Which Rationality? (London: Duckworth, 1988), p.

345; M. Perry, Morality, Politics and Law: A Bicentennial Essay (Oxford: Oxford University
Press, 1988), p. 67; I. Shapiro, The Evolution of Rights in Liberal Theory (Cambridge:
Cambridge University Press, 1986), p. 285.
8 We should note here that there are many ways in which the state may violate
justificatory neutrality. It might (a) coerce people to lead better lives (henceforth coercive
perfectionism). Or it might (b) give subsidies or tax exemptions to valuable conceptions of the
good (henceforth non-coercive perfectionism). Such non-coercive methods still violate
neutrality if the reason why the state subsidizes certain forms of life is that they are superior to
others.
9 Will Kymlicka, Contemporary Political Philosophy (Oxford: Clarendon, 1991), p. 219.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
460 SIMON CANEY
The claim then is that the truth concerning what lives are worthy flourishes
in, and only in, neutralist societies. It relies upon Millean and Hayekian
claims about the preconditions of the emergence and continued existence
of worthy forms of life. Competition in the cultural marketplace, it affirms,
will lead to 'the decline of the groups that have adhered to the wrong
beliefs'.?1 Neutrality is therefore justified from a consequentialist point of
view as being maximally conducive to the emergence and persistence of
ethically valuable conceptions of the good (and the disappearance of
worthless practices).
The chief problem with this argument lies in the assumption that
valuable forms when confronted with worthless forms will prevail.
Kymlicka simply assumes that the most convincing arguments and those
that attract the most adherents are in fact the most truthful, that 'satisfying
and valuable ways of life will tend to drive out those which are
unsatisfying'." But this confidence in cultural evolution is unwarranted
and we have no reason to think that the most valuable ideals always
triumph, and that truth defeats falsity. Many factors influence the survival
and success of forms of life quite apart from their merit - the
persuasiveness of practitioners of a form of life, peer-group pressure,
manipulation, ignorance, cost - and it is simply too naive to think that the
most marketable are the most valuable. Indeed, sometimes the allure of
worthless ways of life and the strenuous nature of some worthwhile forms
of life may lead to the former driving out the latter. Kymlicka's assumption
that the worthy will triumph over the valueless appears unduly sanguine.

III. THE 'FALLIBILITY' ARGUMENT

A second consequentialist argument in defence of neutrality is given by
Bruce Ackerman and Joel Feinberg, who argue that the state should not
promote conceptions of the good because it is possible that the state will
make mistakes and will promote unattractive conceptions of the good life.12
It has a powerful resonance among liberals who fear that the New Right
will impose fundamentalist Christian values on society, and among liberals

10 F. A. Hayek, The Constitution of Liberty (London: Routledge and Kegan Paul, 1960), p.
36.
" Kymlicka, Contemporary Political Philosophy, p. 219.
12 This argument, it is worth noting, is inconsistent with Ackerman's espousal of a
'sceptical' argument in defence of neutrality. In making this 'fallibility' argument and in
asserting that the state might oppress worthwhile practices and might make ethical errors,
Ackerman is assuming that there are worthwhile practices and justified ethical values. This
assumption is, however, incompatible with his claim elsewhere that there is no ethical truth
(Social Justice in the Liberal State, pp. 11, 386, 369; and 'What is Neutral About Neutrality?',
Ethics, 93 (1983), p. 387).

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 CONSEQUENTIALIST DEFENCES OF LIBERAL NEUTRALITY 461

in England who fear the state's ethical view that homosexuality is an

unworthy form of life. Stated abstractly the argument is as follows:

(1) The state might make mistakes about the good life.'3 This is
because:
(a) politicians, like anyone else, are fallible. So even a
thoughtful well-intentioned statesman may promote policies
which are ethically disastrous;
(b) the experience of powers corrupts and thus politicians are
likely to be less interested in the promotion of what is good
than in their own self-aggrandizement;'4
(c) even if the politician is able to discern ethical worth and
wishes to promote ethical worth the state might still make
mistakes about the good because bureaucrats have a tendency
to influence policy to their own interest.15
(2) If the state might make mistakes about the good life, then it
should not promote any conceptions of the good life.16
(3) Therefore the state should not promote any conceptions of the
good life. It should remain neutral between competing
conceptions of the good life.1

The problem with this argument is, however, that it is too strong. It is also
true that the state might make mistakes about the right and justice: but this
does not lead neutralists to conclude that the state should not promote
justice and rights. There are a great variety of conceptions of justice
(ranging from Nozick's libertarianism to Rawls's two principles; from
Hayek's theory that the market is neither just nor unjust to Dworkin's
claim that justice demands equality of resources) and it is possible that the
state will not select the right one out of all of these. Given this diversity, the
state might make a mistake, and there seems to be no reason to think that
the state is less likely to make a mistake concerning the right as it will
concerning the good. All the factors listed in premise (1) apply to a state
that is required to institute principles of social justice. But then if this is
true, neutralists ought to conclude - if premise (2) is correct - that the state

13 Joel Feinberg, The Moral Limits of the Criminal Law, Volume 4: Harmless Wrongdoing
(Oxford: Oxford University Press, 1988), pp. 310-11.
14 Ackerman, Social Justice in the Liberal State, pp. 12, 362; 'What is Neutral About
Neutrality?', p. 387.
15 Ackerman, Social Justice in the Liberal State, p. 363.
16 Ackerman, Social Justice in the Liberal State, p. 369; Feinberg, The Moral Limits of the
Criminal Law, Volume 4: Harmless Wrongdoing, pp. 286, 311.
17 Ackerman Social Justice in the Liberal State, p. 369; Feinberg, The Moral Limits of the
Criminal Law, Volume 4: Harmless Wrongdoing. pp. xx, and 311.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
462 SIMON CANEY
should abstain from instituting principles of justice. Or, they must explain
why, on the one hand, the fact that the state might make mistakes over the
right does not entail that it should abstain from instituting the right, while,
on the other hand, that fact that the state might make mistakes over the
good does entail that it should abstain from instituting the good. Why this
asymmetry between the right and the good? Neutralists thus face a
dilemma:
either: the fact that the state might make a mistake X-ing entails that it
should not X, in which case neutralists must accept that the state should
not institute principles of justice - a wildly counterintuitive claim that no
neutralist would wish to endorse;
or: the fact that the state might make a mistake X-ing does not entail that
it should not X, in which case the fact that the state might make a mistake in
instituting principles of the good does not entail that the state should not
promote the good life.
A further problem with the 'fallibility' argument should be noted. The
argument asserts, recall, that because a perfectionist state might act in a way
which damages the well-being of its citizens, therefore the state should
eschew perfectionism. The central point is that there are costs to a state that
tries to promote worthwhile lives. But this is unpersuasive because an
analogous point may be made against a neutral state. This arises from the
fact that Ackerman and Feinberg claim only that the state might make
mistakes about the good life. In doing so they recognize the possibility that
the state might get it right. They do not say that the state will always get it
wrong - this would indeed be an unbelievably pessimistic assumption.'8
But if the state can make good decisions about the good life, then there are
clearly costs to a neutral state. In those cases where it can discern
worthwhile forms of life, and could therefore improve its citizens' lives by
promoting those valuable conceptions of the good, to take a neutral stance
clearly hampers the well-being of its citizens. The fact that a perfectionist
state might act in a way that damages the well-being of its citizens does not
therefore support a neutral state because it is true of a neutral state that it
too might act in a way that hampered the well-being of its citizens.
Feinberg and Ackerman's 'fallibility argument' against perfectionism
thus fails. It is, of course, relevant that the state might make mistakes, but
the correct conclusion is not that the state should abstain from perfectionist
policies, rather that one should guard as diligently as possible against
abuse. Moreover, Feinberg and Ackerman's point fails to engage with
perfectionists who think that the state ought to promote well-being - not
just what politicians think well-being involves. Perfectionists assert not that
the state may promote what its officials think are important conceptions of

18 Ackerman, Social Justice in the Liberal State, p. 363.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 CONSEQUENTIALIST DEFENCES OF LIBERAL NEUTRALITY 463

the good (even if they are wrong), just that they should promote what is
good. So the state feared by Feinberg and Ackerman in which the state
makes calamitous decisions about the good would also be one feared by
perfectionists like Plato, Aristotle and Raz.

IV: THE 'BEST JUDGE' ARGUMENT

A further unsatisfactory feature of the last argument is that it makes no
comparative assessment of the ability of the state and the ability of
individuals to ensure the promotion of the good. While it draws our
attention to the failings of the state, this cannot fully convince anyone of the
desirability of neutrality without considering whether individuals are even
worse. This is what the next consequentialist argument tries to do, and
Stephen Guest argues - in what I have called the 'best judge' argument -
that the state should not try to influence individual assessments of the good
life because individuals are the best judges of what is in their own interest.
Individuals, the argument goes, know better what is good for them than
does the state, and therefore the latter is not 'to be trusted with the task of
determining which and what lives count as worthwhile'.l9 The state should
neither coerce people to engage in worthwhile activities, nor non-
coercively promote the good life because in both cases individuals are
better equipped to make their lives go better, and whenever states have
tried to promote conceptions of the good they 'have got it wrong about
which individuals and which life-styles have a right to concern and
respect'.2
But this argument relies on overly optimistic assumptions about (i) the
ability of individuals to discern ethically worthwhile goals; and (ii) the
ability of individuals to institute their principles of the good life.
Let us take the first point first. Individuals are in many cases not the best
judges of what is best for them for several reasons:
(a) epistemological problems: individuals often lack perfect knowledge of
valuable ways of life and are unaware of valuable and important
conceptions of the good. Consider, for example, a case in which there are
individuals who would find Zola's novels rich and rewarding (more
rewarding than anything else they have come across), but who would never
encounter Zola's novels without state assistance. In such a case a
perfectionist state which introduced people to Zola (on the grounds that
Zola's novels are profound and moving) would enable individuals to lead
more rewarding lives. A state engaged in non-coercive perfectionism could,

19 Stephen Guest, 'Neutrality as the Basis for Liberalism - A Response to Bellamy', in

Richard Bellamy (ed.), Liberalism and Recent Legal and Social Philosophy (Stuttgart: Franz
Steiner Verlag Wiesbaden GMBH, 1989), pp. 48, 49.
20 Guest, 'Neutrality as the Basis for Liberalism - A Response to Bellamy', p. 48.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
464 SIMON CANEY
therefore, overcome the epistemological shortcomings of individuals by
subsidizing and advertising valuable ways of life which will otherwise be
overlooked. Indeed, if it is true that individuals are the best judges of their
own ethical interests - in the sense that when confronted with a valuable
option they will choose it - this supports rather than undermines such non-
coercive perfectionism. The recognition of the ability of individuals to
discern value combined with the realization that individuals often lack
adequate information thus commits us not to neutrality but to a
perfectionist state committed to introducing people to valuable modes of
life which they would not otherwise encounter.21 (If they were poor
choosers and unlikely to choose ethically worthwhile conceptions of the
good, then there would be no point in putting a valuable option before
them.)
(b) evaluative problems: second, even if we grant individuals perfect
knowledge of all worthy options, it is still questionable whether they will
always be better judges of their own interests than the state. This
contention is supported by several reasons. (i) One reason given in support
of the claim that individuals always know their ethical interests best can be
seen to be unfounded. Mill thought that individuals (given adequate
knowledge) were always the best judges of their own interests because each
individual had a unique essence, an individual quiddity utterly distinct
from others' personalities. But as communitarians have rightly reminded
us, individuals are shaped by their communities, and hence one can, by
drawing on our knowledge of the interests of others, make a reasonable
assessment of an individual's ethical interests.22 Our commonality means
that individuals do not have a privileged access to their own well-being and,
therefore, there are some occasions in which others know what is better for
them. (ii) This conclusion is strengthened by the observation that
individuals are notoriously bad at making probabilistic assessments. As
Robert Goodin notes, individuals 'are liable to all sorts of biases, in
assessing such probabilities, growing out of judgemental short-cuts that are
now well-mapped by psychologists and decision theorists. Perhaps the
most worrisome is the tendency to assume "it will never happen to me",
even when the odds are decisively against that proposition'.23 (iii) Third,
people often suffer from temporal problems and there are occasions in
which we irrationally prefer a lesser present good to a far greater future

21 Such non-coercive action is incompatible with justificatory neutrality because the state is
making policy on the basis that some conceptions of the good are superior to others. It supports some
conceptions of the good on the basis that they are valuable (reading Zola) and avoids supporting
other less valuable conceptions (like pornography say) precisely because they are less valuable.
22 For a further disucssion of this point see my 'Liberalism and Communitarianism: A
Misconceived Debate', Political Studies (1992).
23 Robert Goodin, 'Liberalism and the Best-Judge Principle', Political Studies, XXXVIII
(1990), p. 189.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 CONSEQUENTIALIST DEFENCES OF LIBERAL NEUTRALITY 465

good. (iv) Moreover, in many cases we recognize that an individual has

chosen an unworthy way of life and we think those who dedicate their
entire lives to necrophilia, heroin addiction, counting blades of grass,
Russian roulette or watching television all day every day are poor judges of
ethical well-being - worse choosers than the state.24 Individuals can be
poor choosers of what is valuable - worse than the state - in which case
there is a role for a perfectionist state.
Even if we disregard these previous points and assert that individuals
always know better than the state does what is good for them, this still does
not entail neutrality. It does not because even if individuals realize that a
particular conception (ballet, for example) is admirable, they may not be
able to translate their wishes into practice. While they might know what is
good for them, they might not be able to ensure that they can practise this
valuable form of life. This might occur for several reasons:
(a) akrasia: individuals might suffer from weakness of will, and while
they may rightly judge that a life devoted to pornography is unworthy, they
may simply be unable to prevent themselves from reading it.
(b) collective action problems: the citizens of a city might each individually
realize that a library would be valuable but they face an assurance dilemma
and each will only contribute to the construction of a library if each of them
can be assured that others will pay their part.25 In such a scenario a state is
justified in resolving the dilemma and providing the library. That
individuals are adept at discovering what is in their interest does not of
itself ensure that they will be able to guarantee the flourishing of that ideal.
For these reasons, then, the (disputable) ability of individuals always to
judge their ethical interests better than the state does not entail a
commitment to neutrality.

V: THE 'CHARACTER DETERIORATION' ARGUMENT

Even if individuals are not always the best judges of their own ethical
interests this does not necessarily entail a rejection of neutrality and an
endorsement of perfectionism. Von Humboldt, for example, argues, in
what might be called the 'character deterioration' argument, that even if an
individual is a worse judge with respect to some particular option (X-ing)
the state should still desist from guiding the individual to refrain from

24 Like those who present this argument I assume here that a person's interests can be
defined independently of his or her current conception of those interests. See Mill, On Liberty,
Collected Works, Volume XVIII (Toronto: University of Toronto, 1977) pp. 215, 261, and ch.
IV. Good reasons for doing so are given by James Griffin in Well-Being: Its Meaning,
Measurement and Moral Importance (Oxford: Clarendon, 1986), pp. 7-17.
25 They might of course suffer from a prisoner's dilemma in which case the optimal result
will also not be achieved.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
466 SIMON CANEY
X-ing.26 He argues that, if the state intervenes in the interest of the
individual in one area, then the individual will become increasingly
dependent on the state for guidance in other areas. He will not judge
matters for himself, will not evaluate his options, nor create new ones
because he knows that the state will come to his assistance. Thus the
consequence of any form of perfectionism is the deterioration of the
character of the individual, and perfectionism generates a tendency for
individuals to become even worse judges of their own ethical interests.
Moreover, von Humboldt argues, this is as true of non-coercive measures
as it is of coercive brands of perfectionism.27 Both undermine a person's
ability to think for himself. Thus a neutral state would have better
consequences than a perfectionist state in the sense that it will not engender
this apathy, dependence and inability to judge one's own interests.
These considerations do, I think, provide some support for a general
commitment to liberty. They do not, however, generate a wholesale
commitment to neutrality. First, von Humboldt's argument, while it has
some force against coercive perfectionism, seems implausible as a critique
of non-coercive perfectionism. It is perhaps true that, if the state judges in
one area (sexual relations, for example) that the individual's judgement is
wrong, then that individual will lack confidence in his judgement in other
areas too and will defer to the state. The ability of the state to make
judgements on behalf of the individual may, it is true, lead to a
deterioration in a person's general character. But when the state subsidizes
several forms of life on the grounds that they are worthwhile, it does not
make a judgement on the individual's behalf but makes other options
available. So a state that non-coercively promotes worthwhile activities
does not usurp the role of the individual and does not therefore lead to the
deterioration of his character. The individual still has to make a choice and
hence the state's actions are unlikely to encourage him to rely upon the
state's judgement. Non-coercive perfectionism does not make an in-
dividual's decision for him and hence von Humboldt's argument that a
perfectionist state, in making the individual's decision for him, encourages
a worsening of his deliberative powers is false.28

26 See also Mill, On Liberty, p. 305; Joel Feinberg, 'The Interest in Liberty on the Scales',
Rights, Justice and the Bounds of Liberty: Essays in Social Philosophy (Princeton: Princeton
University Press, 1980); and Feinberg, The Moral Limits of the Criminal Law, Volume 1: Harm
to Others (1984), p. 211.
27 Von Humboldt explicitly makes this point in The Sphere and Duties of Government
(London: John Chapman, 1854), pp. 18, 20-1,44-5, 65-7, 71. Thus he writes: 'however it may
accomplish this, - whether directly or indirectly by law, or by means of authority, rewards, and
other encouragements attractive to the citizen - or lastly, by merely recommending its
propositions to his attention by arguments, - it will always deviate very far from the best
method of instruction' (The Sphere and Duties of Government, p. 24).
28 Linked to this is von Humboldt's false assumption that perfectionists are necessarily

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 CONSEQUENTIALIST DEFENCES OF LIBERAL NEUTRALITY 467

Second, it may often be the case that a neutral state rather than a
perfectionist state will lead to character deterioration. Thus one might
argue as follows:

(i) the ability to judge one's critical interests and to deliberate

requires a variety of conceptions of the good. Without such
variety individuals will have less scope for deliberation and
without such practice their critical faculties will become
debilitated.
(ii) A neutral state cannot guarantee that the sufficient variety
required for the enhancement of a person's critical skills
obtains.
(iii) Therefore, the state may have to promote conceptions of the
good in order to ensure that variety exists, and this requires
that it must promote conceptions of the good.

Thus perfectionism, far from causing a deterioration of one's deliberative

abilities, may be required to enhance and facilitate these deliberative
faculties.
A third defect in von Humboldt's argument lies in the fact that often one
cannot make a judgement about a particular practice without experience of
that practice. This means that unless individuals are given experience of a
particular practice they will never be able to judge it properly. Mill made
exactly this point about education and he concluded that individuals
should be coerced to experience education in order to enable them to be
better judges.29 So a commitment to bettering someone's deliberative
faculties may require compulsion.

VI: THE 'IMPORTANCE OF DOUBT' ARGUMENT

One might, however, still defend neutrality in a consequentialist fashion on
the grounds that a worthwhile life involves the availability of worthless
options. Mill urged this in his defence of freedom of expression and argued
that without the existence of bad beliefs the truth became dead dogma. And
Ackerman has relied upon the same argument in his defence of neutrality.

monists who reject moral pluralism. He does not therefore distinguish between
monistic perfectionism: there is a single correct conception of the good and it is the duty of the
state to promote this good,
and
pluralistic perfectionism: there is no single correct conception of the good and there are many
worthy ways of living. It is the duty of the state to promote these various diverse ways of life.
29 See his Principles of Political Economy, Collected Works, Volumes II and III (Toronto:
University of Toronto, 1965), Book V, ch. XI, ?8.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
468 SIMON CANEY
A life can only go well, he claims, if individuals have the possibility of
making mistakes and this requires in turn that the state does not ban
unworthy forms of life.30 The argument at work is analogous to the
theological argument that persons were given freewill to do both good and
evil because without the possibility of evil one cannot be truly good.
Genuine goodness requires the possibility that one has the possibility to do
wrong but in fact one rejects the evil option. Similarly, Ackerman argues,
genuine ethical worth requires that one is open to doubt and that one
therefore chooses worthy options for the right reasons. This requires in
turn that the state desist from banning worthless options - hence neutrality
is justified. No form of the good life should be banned on the grounds that it
is unworthy because even unworthy conceptions of the good serve a role.
Thus the state should be neutral between conceptions of the good.
The limitations of this argument are, however, clear. First, it has no
force against a state that non-coercively promotes valuable forms of life,
and even if the existence of unworthy options serves a useful function, this
entails only that the state should allow sufficient numbers of worthless
options to exist. The argument will at most show that the state should
not ban unworthy options - it is inapplicable as an argument against
encouraging certain ways of life on the grounds that they are valuable.
Second, it does not entail a complete ban on coercive perfectionism.
What it dictates is that there should be a sufficient number of unworthy
options to contest the worthy options and that the elimination of a single
unworthy option may still leave plenty of other unworthy options.
Consider a case where an agent has the choice between worthy options, a, b,
c, d, and also the unworthy options e, f, g, h, i, j and k. Now, even if the
existence of some valueless options serves some role, it is difficult to see why
the state cannot eliminate one worthless option. After all, the agent in
question will still be able to adopt six worthless options. In other words,
even if the agent should be presented with some valueless conceptions of
the good, this does not entail that he should face an unlimited set of bad
conceptions of the good.
A more radical defect in Ackerman's argument is that the intuition
underlying the argument (and which gives the argument the plausibility it
possesses) does not seem to support Ackerman's conclusion. Ackerman and
Mill's intuition seems to be this: to live a valuable life it must be true that

(i) the actions one performs and the ideals one professes are
worthy;

30 See Ackerman 'What is Neutral About Neutrality?', p. 387; and Social Justice in the
Liberal State, pp. 365-7.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 CONSEQUENTIALIST DEFENCES OF LIBERAL NEUTRALITY 469

(ii) one values the ideals one performs and the ideals one professes
for the right reasons (and not simply out of habit or
socialization), where one believes something for the right
reasons when one is aware of other ideals and where one
critically evaluates them and makes one's decision on the basis
of an assessment of the comparative merits of several options.

Acceptance of (ii), Ackerman then claims, entails the acceptance of the

following claim:

(iii) In order to believe something for the right reasons one must be
confronted with unworthy options.

But (iii) does not follow from (ii). What follows from (ii) and the emphasis
on making a deliberated choice is that one reject other options. It does not
entail that these be unworthy. This means that the requisite deliberation
could be attained by an agent who is faced by a plurality of valuable but
incompatible and competing ideals. The existence of worthless options, as
such, is not necessary for the right-reasoned choice that Ackerman and Mill
emphasize.
Ackerman's 'necessity of error and importance of doubt' argument thus
fails to justify neutrality because (i) it is inapplicable with respect to non-
coercive perfectionism; (ii) it does not entail a complete ban on coercive
perfectionism; and (iii) the existence of unworthy options is not a
precondition of a right-reasoned choice.

VII: THE 'EXPERIMENTALIST-CONSEQUENTIALIST' ARGUMENT

The principle of justificatory neutrality has, however, been defended in
other ways and Bruce Ackerman and David Lloyd-Thomas provide an
'experimentalist-consequentialist' defence of justificatory neutrality.31
They argue as follows:

(1) Our most important interest is in leading a life based upon

correct views about the good life.32
(2) We can only know what is good in life if we experiment. As
Lloyd-Thomas argues, 'The attempt to discover what is
worthwhile depends on experimentation with different
" Lloyd-Thomas makes these arguments in In Defence of Liberalism and 'Rights,
Consequences, and Mill on Liberty', in A. Phillips Griffiths (ed.), Of Liberty (Cambridge:
Cambridge University Press, 1983).
32 Lloyd-Thomas, In Defence of Liberalism, p. 2.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
470 SIMON CANEY
"modes of existence".'33 Or, as Ackerman puts it, 'you can
only learn anything true about the good when you are free to
experiment'.34
(3) Therefore the state ought to ensure that the conditions for
experimenting with different forms of life exist (by (1) and (2)).
(4) Only a neutral state can ensure that the conditions for
experimentation with different forms of life are satisfied -
'liberal rights are conditions for having more reasonable views
about what activities are worth while'.35 It alone is 'favourable
to experimentation with new modes of existence'.36
(5) Therefore the state ought to be neutral between conceptions of
the good life (by (3) and (4)).37

This 'experimentalist-consequentialist' argument is, however, unper-

suasive for two reasons. First, premise (2) is too strong and may be
criticized on two grounds.
(i) It is not a necessary precondition of finding a valuable form of life
that one experiments - one might select a valuable form of life without
experimenting with a variety of ways of life. A brilliant gymnast or
ballerina may know nothing of other forms of life and may never
experiment - and yet it is difficult to conclude that they cannot therefore
know what is good in life. Experimentation is thus not a necessary
precondition of the good life.
(ii) Indeed enjoyment of some valuable forms of life precludes
experimentation. A commitment to experimenting with various forms of
life rules out certain valuable and worthy forms of life, namely, those
involving long-term (maybe even lifelong) commitments and attachments.
For example, to be a good ballet dancer or a brilliant decathlete precludes
experimenting with other forms of life and require complete dedication to
one form of life alone. The same might be said about being married or
being a scholar, a politician, a civil servant, an athlete, or a musician.
These objections to premise (2) are not, however, insurmountable and
Lloyd-Thomas might recast the second premise as follows:

(2a) It is true of many forms of life that we can only know if they
are good if we experiment with them.

Thus adapted, the second premise eschews the claim that experimentation
is a necessary precondition of living every good life, but recognizes that in

3 Lloyd-Thomas, In Defence of Liberalism, p. 2.

3 Ackerman, Social Justice in the Liberal State, p. 11.
35 Lloyd-Thomas, In Defence of Liberalism, p. 38; see also pp. 36, 37.
36 Lloyd-Thomas, 'Rights, Consequences, and Mill on Liberty', p. 178.
37 Lloyd-Thomas, In Defence of Liberalism, pp. 1-2.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 CONSEQUENTIALIST DEFENCES OF LIBERAL NEUTRALITY 471

many cases experimentation facilitates the attainment of a worthwhile life.

In this way Lloyd-Thomas may avoid the first objection made to his
argument.
But even if Ackerman and Lloyd-Thomas's argument for neutrality
avoids the first objection, it faces other more serious problems, and premise
(4) is dubious. It is questionable whether a commitment to experi-
mentation necessarily entails a commitment to neutrality. Indeed, if
experimentation is our goal this would seem to entail that those individuals
who would not otherwise experiment should do so and should be
compelled to do so. To see this consider someone, X, who is by disposition
not inclined to experiment and is happy being a musician. Consequently,
under a neutral political system he is not compelled to experience other
diverse forms of life (like being an artist or political philosopher or
mountaineer) and his experiences and the extent to which he experiments
are limited. To ensure that X experiments with various forms of life it
would seem that he should be compelled to experience other conceptions
of the good - without this he will not experiment. And if he does not
experiment, then, according to the 'experimentalist-consequentialist'
argument, it is unlikely that he will discover what is worthwhile in life.
Thus a serious commitment to experimentation would seem to justify the
compulsion of the unadventurous to be experimentative. Premise (3) does
not therefore preclude coercive perfectionism.
Nor does it preclude non-coercive perfectionism - and this is my third
objection to Ackerman and Lloyd-Thomas's 'experimentalist-consequen-
tialist' argument. If one is committed to providing the conditions necessary
for experimentation, it is difficult to see why that entails preventing non-
coercive perfectionism which sponsors a variety of conceptions of the good
on the grounds that these are valuable. Indeed the more forms of life
available then surely the greater possibilities for experimentation.

VIII: THE MOD US VI VENDI ARGUMENT

In the event of the failure of the last five consequentialist arguments in
defence of neutrality, one might want to defend neutrality on more
pragmatic than philosophical grounds. This approach is taken by several
contemporary neutralist liberals who have defended neutrality on the basis
of what they call the modus vivendi argument. They argue that stability is
desirable but that, because principles of the good life are controversial,
stability will only be assured if the state is neutral between competing
conceptions of the good life. This argument is advanced by Charles
Larmore and was later adopted by Gutmann and Rosenblum.38 Formally

38 See Larmore's 'Review of Liberalism and the Limits of Justice', Journal of Philosophy, 80

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
472 SIMON CANEY
stated the argument is as follows:

(1) If there are fundamental disagreements concerning con-

ceptions of the good, then the only way to ensure stability and
'civil peace'39 is for the state to be neutral between competing
conceptions.
(2) There are fundamental disagreements concerning conceptions
of the good and modern societies are characterized by what
Rawls calls 'the fact of pluralism'.
(3) Therefore, the state can only assure civil peace if it is neutral
between competing conceptions of the good. Were the state to
promote one conception of the good life, there would be 'civil
war among communities each bound by a singular moral
purpose'.4
(4) The state should ensure the existence of civil peace.
(5) Therefore, the state ought to be neutral between competing
conceptions of the good life. 'Neutrality', for Larmore, is
therefore 'simply a means of accommodation'41 and it is a
pragmatic device designed to ensure maximum stability.

This argument in defence of neutrality is, however, unconvincing and it

fails to justify neutrality for (at least) five reasons. First, it overlooks the fact
that there exist, at the moment, many perfectionist states which continue to
survive and do not appear to be on the verge of crisis. The states of
Britain,42 Eire, France, Italy, and Spain,43 for example, all take a stand on
what conceptions of the good are worthwhile and yet none of these appears
to be, for this reason, stricken with civil war nor on the verge of internal
strife. Neutrality is not, therefore, necessary for civil peace.

(1983), and Patterns of Moral Complexity. See too Gutmann, 'Communitarian Critics of
Liberalism', p. 313; Rosenblum, 'Introduction' to Liberalism and the Good Life (Cambridge
Mass.: Harvard University Press, 1989), p. 6. According to Chandran Kukathas the same
argument is given by Hayek (Hayek and Modern Liberalism (Oxford: Oxford University Press,
1989), p. 222, and pp. 226-7).
39 Larmore, Patterns of Moral Complexity, p. 60.
0 Rosenblum, Liberalism and the Good Life, p. 6.
41 Larmore, Patterns of Moral Complexity, p. 74.
42 In Britain, for example, the state is quite clearly perfectionist and it overtly supports the
Church of England, and makes legislation on the grounds that homosexuality and lesbianism
are less valuable than heterosexuality. Moreover, the late Greater London Council and the
present London boroughs have clearly constructed policy on the basis of the worth of some
unorthodox forms of life.
43 These four countries each possesses a state which makes policy on the basis that
Catholicism is true and hence each takes a distinctively Catholic approach to issues of abortion
and divorce.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 CONSEQUENTIALIST DEFENCES OF LIBERAL NEUTRALITY 473

Moreover - and this is my second point - sometimes perfectionism is

required to ensure stability. This is true of Iran and other fundamentalist
countries where one can be sure that the state can only attain stability if it
eschews neutrality and makes policy on the basis that Islam provides the
best account of the good life. In these communities it is not only true that
stability is not jeopardized by perfectionism: it is also true that stability
requires perfectionism and that a neutral state would be unstable.
The first two objections have thus shown that it is empirically false that
neutrality is required for stability.We may now see that the modus vivendi
argument is empirically false because it relies on three false assumptions.
First, it assumes- without providing any empirical support for the claim-
that there is no overlap between communities as to the content of the good
life. It does not even consider the possibility that several cultural
communities share core ethical ideals and hence that one may have a
perfectionist state that commands the positive approval of all.44
Second, it assumes that, in a perfectionist state which promotes the
ethical values of one community above the values of others, the members of
those cultural communities whose conception of the good is not promoted
are so committed to that good that this commitment overrides all other
commitments they may have. Atheists in a state which promotes Christian
values may dislike the state's perfectionism but may also desire a degree of
economic welfare, stability, etc., and rank these considerations higher than
their desire to live in a non-Christian state. For Larmore's argument to
work he must show not only that communities disagree but that they
disagree so much that if the state favours the ethical values of one com-
munity, the members of the community whose values are not promoted
will disregard all other considerations and instigate rebellion.
Third, the modus vivendi argument overlooks the fact that even if people
violently disagree as to what is valuable in life, and even if this is their
most important commitment they may still - because of collective action
problems - not be disruptive or rebellious. They may not instigate insur-
rection simply because they face a prisoner's dilemma or an assurance
dilemma.

4 Recently, Larmore has assumed that a state is not perfectionist even if it promotes a
conception of the good, if that conception of the good is noncontroversial ('Political
Liberalism', Political Theory, 18 (1990), p. 358, n.3). But this definition of neutrality is
radically at odds with Rawls, Dworkin, Ackerman, Kymlicka and all other neutralists'
conception of neutrality. See for example, Ackerman, Social Justice in the Liberal State, p. 11;
Kymlicka, 'Liberal Individualism and Liberal Neutrality', pp. 883-5; Dworkin 'Liberalism';
Rawls, A Theory of Justice, p. 332, and 'The Priority of the Right and Ideas of the Good', p.
262. All the latter state that the state should not make policy on the basis of some conception of
the good, no matter how widely shared that good is.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
474 SIMON CANEY
IX: THE 'DICTATORSHIP' AND 'DIALOGUE' ARGUMENTS
Neutrality is not therefore necessary for civil peace. Before we reject all
consequentialist defences of neutrality we ought to consider two final
consequentialist arguments given by Will Kymlicka in defence of
justificatory neutrality - what might be called the 'dictatorship of the ar-
ticulate' and the 'distorted dialogue' arguments. Kymlicka relies upon
these to show that 'state perfectionism would have undesirable con-
sequences for our society'.45 His arguments are as follows:

The 'Dictatorship of the Articulate' argument

First state perfectionism raises the prospect of a dictatorship of the

articulate and would unavoidably penalize those individuals who
are inarticulate. But being articulate, in our society, is not simply
an individual variable. There are many cultural disadvantaged
groups whose beliefs and aspirations are not understood by the
majority. Recent immigrants are an obvious example whose
disadvantage is partly unavoidable. But there are also groups
which have been deliberately excluded from the mainstream of
American society, and whose cultural disadvantage reflects pre-
judice and insensitivity.46

The 'Distorted Dialogue' argument

State perfectionism would also affect the kinds of arguments given.

Minority groups whose values conflict with those of the majority
often put a high value on the integrity of their practices and aim at
gaining adherents from within the majority slowly, one by one.
But where there is state perfectionism, the minority must
immediately describe their practices in such a way to be most
palatable to the majority, even if that misdescribes the real
meaning and value of the practice... There would be an
inevitable tendency for minorities to describe and debate con-
ceptions of the good in terms of dominant values, which then
reinforces the cultural conservatism of the dominant group
itself.47

The first argument claims that perfectionism would have disastrous

consequences because the articulate and persuasive would be better
45Kymlicka, 'Liberal Individualism and Liberal Neutrality', p. 900.
46 Ibid.
47 Kymlicka, 'Liberal Individualism and Liberal Neutrality', p. 901.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 CONSEQUENTIALIST DEFENCES OF LIBERAL NEUTRALITY 475

equipped to persuade the state to support their form of life. Inarticulate

cultural communities would thus be penalized. The second argument, on
the other hand, claims that minority cultures will be penalized because they
may only win state support for their form of life if they couch their form of
life in the terms of the dominant value systems. That is, they must
relinquish what is distinctive about their culture and employ the values of
the majority. Do these provide good reasons for concluding that the state
should not make policy on the basis of information concerning the good
life? Do Kymlicka's consequentialist arguments entail justificatory neut-
rality?
I think that they do not. Considering the 'dictatorship of the articulate'
argument, first, as Kymlicka notes, it is because of unjust educational,
economic and political treatment that certain communities are more
articulate than others and hence able to argue more convincingly than
others.48 Given this, the appropriate response is surely to rectify the
inequality in articulation and hence the impotence of certain minority
cultures by removing injustice. If injustice is the source of the problem,
then surely what we must do is eradicate injustice rather than abstain from
perfectionism.
Moreover, the 'distorted dialogue' argument is also unconvincing and
applies only to certain types of perfectionism, in particular that where
perfectionist policy making is controlled by a majority. Kymlicka states
that he assumes 'that the public ranking of the value of different ways of life
which a perfectionist state appeals to would be arrived at through the
collective political deliberation of citizens'.49 The phrase 'collective
political deliberation' is quite vague, but it is evident, however, in
Kymlicka's argument that he means 'decision by the majority'. Thus he
writes that

Minority groups whose values conflict with those of the majority

often put a high value on the integrity of their practices and aim at
gaining adherents from within the majority slowly, one by one.
But where there is state perfectionism, the minority must
immediately describe their practices in such a way to be most
palatable to the majority.53

But while this might show the problems faced by majoritarian per-
fectionists, it fails to show why one cannot have non-majoritarian
" Robert Goodin provides empirical support of how inarticulacy of certain minorities
results from unjust educational institutions and draws on the work of Paulo Freire and Basil
Bernstein. See R. Goodin, Manipulatory Politics (London: Yale University Press, 1980), pp.
77-81.
49 Kymlicka, 'Liberal Individualism and Liberal Neutrality', p. 900.
'" Kymlicka, 'Liberal Individualism and Liberal Neutrality', p. 901.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
476 SIMON CANEY
perfectionism. Kymlicka considers this but argues that the only other
alternative is to put policy-making power in the hands of an unaccount-
able elite - an option that is hardly attractive. But this overlooks other
possibilities which avoid the problems of unaccountable elites and
majoritarian biases. One might, for example, argue that state subsidies
devoted to supporting worthy forms of life should be allocated on a
consociational process where a decision is passed only if it is supported by
all of the cultural communities. Were state subsidies for valuable forms of
life to be allocated according to a principle of unanimity or two-thirds
majority, a dominant group would only be able to allocate money to itself
with the consent of a minority. Thus the minority will be able to bargain for
funds for its way of life. This, in fact, is the institutional system adopted in
many countries which include several minorities.51 One does not have to
enunciate the finer details of the correct institutional system required to
see, then, that there are non-majoritarian and non-elitist forms of
perfectionism which may overcome the problem of distorted dialogue.
In short, the worries Kymlicka raises are important ones and any
perfectionist should take them into account, but they do not entail a
complete repudiation of perfectionism. Caution, not abstinence, is the
conclusion entailed by Kymlicka's arguments.

X. CONCLUSION
This brings us to the end of the review of consequentialist arguments for
neutrality, and we have seen that the consequentialist project has failed to
defend neutrality. The competition between ideals in a neutral state is not
guaranteed to generate the establishment of the most valuable practices.
Nor is the fallibility of those with power itself a reason to eschew neutrality.
In addition, it is questionable whether neutrality can be justified either on
the grounds that unworthy conceptions serve an important function or on
the basis that individuals are always better judges of their interests than the
state. And it is questionable whether perfectionism leads to a withering of
our deliberative powers. Furthermore, our interest in acquiring well-
founded views about the good life, even if it requires experimentation, does
not entail neutrality and certainly does not support a ban on non-coercive
perfectionism. Finally, the more pragmatic arguments also fail to justify

51 See Arend Lijphart's Democracy in Plural Societies (London: Yale University Press,
1977), in particular pp. 25-103; Arend Lijphart, The Politics of Accommodation (Berkeley:
University of California Press, 1968); Ronald Rogowski, Rational Legitimacy: A Theory of
Political Support (Princeton: Princeton University Press, 1974), pp. 77-142. In fact British
experience has shown that one can have a perfectionist state which is not biased towards
majority conceptions of the good. The Greater London Council, for example, was notorious for
subsidizing otherwise disadvantaged ways of life.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms
 CONSEQUENTIALIST DEFENCES OF LIBERAL NEUTRALITY 477

neutrality - a modus vivendi can be attained without neutrality and state

perfectionism need not fall prey to dictatorship of the articulate or distorted
dialogue. Neutrality cannot be grounded on consequentialist lines.
Indeed - on a more positive note - the conclusions established in this
article provide us with some reason to endorse perfectionism on the
grounds that it has better consequences and in particular that it best
enables people to lead worthwhile lives. If(1), as the neutralists discussed
above aver, the political system should be designed to produce the best
consequences and in particular to enable people to live rewarding lives;52
and if(2), as I have argued in this article, a perfectionist state is better than a
neutral state at enabling people to lead rewarding lives, then this gives us
good reason to embrace perfectionism. Perfectionism can therefore be
grounded in a consequentialist way. Of course, in the face of these
arguments neutralists may choose to reject (1): maybe the state should
observe other principles which override any duty it might have to enable
people to lead worthwhile lives. So, much more argument is needed before
we can accept the perfectionist argument sketched above. In the meantime,
however, it is fair to conclude that the attempts considered in this article to
justify neutrality on consequentialist grounds have proved unsuccessful.53

Nuffield College, Oxford

52 The first six arguments addressed in this article, recall, all claimed that a neutral state is
preferable to a perfectionist state because it best enabled individuals to lead valuable lives.
53 I would like to thank G. A. Cohen, David Miller, an anonymous referee and the Editorial
Chairman for their helpful comments.

This content downloaded from 147.8.31.43 on Thu, 08 May 2025 21:39:50 UTC
All use subject to https://about.jstor.org/terms