Metaphysics Postponed: Liberalism, Pluralism, and Neutrality
| Author | Glen Newey |
| Date | 01 June 1997 |
| DOI | 10.1111/1467-9248.00082 |
| Published date | 01 June 1997 |
| Subject Matter | Article |
Metaphysics Postponed: Liberalism,
Pluralism, and Neutrality
GLEN NEWEY
University of Sussex
Many recent liberal theorists have argued that state neutrality is supported by a
metaphysical thesis about value, namely pluralism, which asserts that there are some
conceptions of the good life which neither form a hierarchy nor represent versions of
a single good. It is however doubtful whether neutrality is supported by pluralism;
indeed, it may in some cases be precluded by it. Arguments for pluralism can, in
many cases, be reconciled with a monistic metaphysics of value, and pluralism itself
fails to support neutrality. This is particularly true of traditional liberal policy
positions such as religious toleration and opposition to censorship, where attention to
diverse conceptions of the good may favour, or demand, non-neutral policies. The
political problems which neutrality addresses arise before we accept the metaphysical
`truth' of pluralism, and often remain even if the parties to a political con¯ict have
false conceptions of value. A sharp question for the pluralist neutralist is why
con¯icting conceptions of the political cannot themselves feature in plural con-
ceptions of the good life. Dispensing with pluralism may not, however, be enough to
rescue neutrality, since the disputes which neutrality was designed to deal with may
not be resolvable neutrally; and more particularly, some of the traditional liberal
policies may be incapable of neutral justi®cation. If so, liberals may ®nd a more
traditional form of non-neutral liberalism more attractive.
This paper criticizes one attempt to provide metaphysical foundations for
liberalism. It examines pluralism as a foundation for neutrality ± the view,
popular in recent liberal theory, that the state should ideally remain impartial
between dierent conceptions of the good life. I will argue that the marriage of
pluralism and neutrality is at best one of convenience and, in the way of such
unions, is liable to end badly. It is doubtful whether neutrality is supported by
pluralism, and in some cases belief in pluralism demands non-neutral policies. I
will further argue that traditional liberal policies, such as a commitment
to toleration, or opposition to censorship, are in any case not well served by a
commitment to neutrality. Arguments purportedly basing neutrality on
pluralism can be reconciled with a monistic metaphysics of value, and the
political problems which neutrality addresses arise before we accept the
metaphysical `truth' of pluralism; these problems usually remain, regardless
of whether the parties to a political con¯ict have false conceptions of value.
Resolving these problems, moreover, will demand resort to some decision
procedure which will eectively be non-neutral, reliant as it will be on a
substantive interpretation of some moral principle such as equal respect or
autonomy. This need not rule out pluralism, either as a doctrine about value or
as an in¯uence on policy, but it is liable to limit its scope more than some
# Political Studies Association 1997. Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main
Street, Malden, MA 02148, USA.
GLEN NEWEY
297
neutralists have thought, or hoped. If so, this is both good and bad news for
liberalism: good insofar as liberalism is not tethered to a contentious doctrine
about the metaphysics of value, bad insofar as neutrality has aspired to either
de facto universality or at least to command universal rational assent. A sharp
question for the pluralist neutralist is why con¯icting conceptions of the
political cannot themselves feature in plural conceptions of the good life.
Dispensing with pluralism may not, however, be enough to rescue neutrality,
since the disputes which neutrality was designed to deal with may not be
resolvable neutrally. If so, a more traditional form of non-neutral liberalism
may be more compelling.
Pluralism and Political Neutrality
This section argues that the resources available to monism are greater than some
pluralists have thought. I do not go so far as to claim that monism is true, or
that pluralism is false: the main aim is to suggest that pluralism may be less
readily justi®ed than some recent writers have thought. This serves two linked
purposes in the ensuing argument. First, it argues that liberals have taken
metaphysical pluralism to be more compelling than it really is: monism about
value has more to be said for it than they have assumed. Second, as a result, if
the argument from metaphysical pluralism to neutrality goes through, its
starting-point is not as uncontroversial as many liberals have thought. I will
however also question the argument itself later. After that, I argue that
neutrality is better supported by a position which, if it deserves the name
`pluralism' at all, makes no metaphysical claim (though, as I will argue, this also
encounters problems).
Arguments that aim to justify neutrality from pluralism are only persuasive
insofar as pluralism itself is justi®ed. Pluralism as a metaphysical doctrine has
been more often assumed than argued for, or else has been con¯ated with a
distinct doctrine often going under the same name, which makes a sociological
or anthropological rather than a metaphysical claim. I set this out below.
Following Larmore, we can adopt the following de®nition of pluralism as a
claim about the metaphysics of value.
VP: There are many viable conceptions of the good life that neither
represent dierent versions of some single, homogeneous good nor fall into
any discernible hierarchy.1
Space does not permit any detailed examination of arguments for this
metaphysical form of pluralism. It is, however, worth remarking that where
arguments have been presented, they have often been unconvincing ± in part-
icular, by failing to give sucient reasons for rejecting monism. For example,
in The Morality of Freedom2 Raz claims that incommensurability supports
1 C. Larmore, Patterns of Moral Complexity (Cambridge, Cambridge University Press, 1987),
p. 23.
2 J. Raz, The Morality of Freedom (Oxford, Oxford University Press, 1986), p. 398. Raz oers a
dierent set of formulations in his `Autonomy, Toleration, and the Harm Principle' in S. Mendus
(ed.), Justifying Toleration (Cambridge, Cambridge University Press, 1988) pp. 155±75. There he
de®nes what he calls `weak' moral pluralism as the claim that `there are various forms and styles of
life which exemplify dierent virtues and which are incompatible'; forms of life are incompatible `if,
given reasonable assumptions about human nature, they cannot normally be exempli®ed in the
# Political Studies Association, 1997
298
Metaphysics Postponed
pluralism, where for Raz the mark of incommensurability between values is as
follows: the value realized by a good A is incommensurable with that realized by
another good B, if it is neither the case that A is better than B, nor that B is
better than A, and there is some further good C which, when added to A,
improves it, even though it remains untrue (i.e. false, or neither true nor false)
that the combination of A and C is better than B. So we apparently have a form
of non-linearity for the relation between the relevant measures of value; I will
call this R-incommensurability.
This structure is, however, consistent with a monistic system, such as that of
ethical hedonism. The hedonist may be a realist about divisibility, but simply
deny that we can, at the margin, attain the epistemic resolution required for
full linearity; nonetheless we can, as in Sorites-type comparisons, assert on
procedural grounds that C is a tangible increment to A (e.g. if C and A represent
non-denumerable and distinct but homogeneous volumes of a single good). A
second objection lies in this thought: that the way in which A (and A C) is
good may be R-incommensurably dierent from the way in which B is. That
may mean that the prospects for any comprehensive attempts to measure value
are poor, but nothing in monism as such commits it to a fully transitive system
of measurement. To take Raz's own example: the hedonist may be unsure
whether reading a book is better (sc. more pleasurable) than going for a walk,
and remain unsure even when reading the book is improved by being
accompanied by a glass of whisky. The hedonist may think that no available
calibration is suciently ®ne-grained to decide between the options; or that
while the goodness of both activities consisted solely in their being pleasurable,
the ways in which they were pleasurable were R-incommensurably dierent.
Sometimes it is argued that where desires con¯ict, this must be because they
involve con¯icting values. But desires may con¯ict only contingently, and this
by itself does not entail the metaphysical claim expressed by VP. Indeed, desires
which are (under an appropriate description) content-identical may con¯ict, as
when you and I both desire the last slice of cake (as such). It is not plausible to
regard such con¯icts as involving plural values.
There are, to be sure, non-contingent con¯icts of desires, as Plato recognises
in Republic Book IV3 ± at least, if `non-contingent' means that there can be
con¯icts arising from opposed orectic attitudes directed towards the same aspect
of the same object. From this Plato tries to establish the tripartite division of the
soul. But even non-contingently con¯icting desires do not compel us to
...
To continue reading
Request your trial