Professional Documents
Culture Documents
DOI: 10.1007/s11017-005-3983-y
Springer 2005
ANDREW GLEESON
PETTIT ON CONSEQUENTIALISM
AND UNIVERSALIZABILITY
ABSTRACT. Philip Pettit has argued that universalizability entails consequentialism. I criticise the argument for relying on a question-begging reading of the
impartiality of universalization. A revised form of the argument can be constructed
by relying on preference-satisfaction rationality, rather than on impartiality. But
this revised argument succumbs to an ambiguity in the notion of a preference (or
desire). I compare the revised argument to an earlier argument of Pettits for
consequentialism that appealed to the theoretical virtue of simplicity, and I raise
questions about the force of appeal to notions like simplicity and rationality in
moral argument.
KEY WORDS: consequentialism, desire, impartiality, rationality, universalizability
262
ANDREW GLEESON
Philosophers sympathetic to both universalizability and consequentialism have often suspected some deep connection between
them. Peter Singer believes that the (supposed) impartialist requirement of ethics to consider the interests of all creatures and not merely
ones own, is good reason for treating utilitarianism as presumptively
the best ethical theory.1 R. M. Hare, more ambitious, argues that the
logic of the ordinary moral term ought entails universalizability,
which in turn entails utilitarianism.2 Hares argument has not found
much support. However, recently, another consequentialist with a
strong Australian connection, Philip Pettit, has revived in a different
way, the attempt to derive consequentialism from universalizability.3
This paper is an examination of Pettits argument.
Unlike Hare and Singer, Pettit is not a utilitarian, so his argument
is to that extent less ambitious. He distinguishes between the theory of
the good the theory of what is good or valuable and the theory of
the right the theory of what makes an action the morally right action
to perform. The two theories are closely connected. Indeed in one
place Pettit characterises the theory of the right as a theory about
what individual and institutional agents should do by way of
responding to valuable properties.4 The big issue in the theory of the
right, according to Pettit, is whether we are to promote good generally
or instantiate good in our own actions.5 Or more precisely, whether
good should be promoted at the expense, if necessary, of instantiating
it, or whether instantiation should always take precedence over promotion if the two clash. (I shall follow Pettit in glossing promotion as
the maximization of expected value.) Most of the time instantiating
harmonises with promoting. Among the ways for me best to promote
veracity is by telling the truth myself. But there are cases Pettit calls
them perverse where this is not so. A politician may sometimes
better promote veracity by lying to conceal past lies, preserving a
(illusory) good example. Police may sometimes deter criminals by
breaching the law themselves. And so on.
Those who advocate these breaches of normal duty in order to
promote greater duty compliance are consequentialists. Those who
hold that (at least sometimes) I should instantiate veracity or lawabidingness, even when doing so entails a greater number of lies and
more criminality, are taking a non-consequentialist stance. Pettits
aim is to show that if one universalizes some action one judges right,
then one must agree, in perverse cases, to promote that action at the
expense of oneself instantiating it. That is, one must be a consequentialist.
263
I
Pettits aim is to show that universalizability entails consequentialism, given the unacceptability of a relativistic alternative. The argument is a reductio ad absurdum. Pettit assumes non-consequentialism
and tries to show that (short of taking the relativist escape route) it
cannot be consistently combined with universalizability. Universalizing that is, stipulating or acknowledging that [w]hatever makes it
right that I do O in C makes it right... that any agent do O in C8
leads the non-consequentialist into inconsistency.
The argument begins with this assumption:
Judging that an action is right involves approving of the deed and gives one a
normative reason to prefer it.9
264
ANDREW GLEESON
right. Now, when I universalize I judge that for all other agents, if
they are ever in the relevant circumstances, the right action for each
to perform is the same one as I judged was right for me to perform.
So by the above principle, and in parity with my own case, I must
approve of the action for them too, and acknowledge that I have a
normative reason to prefer that they perform the action. As Pettit
puts it, by universalizing:
I commit myself to there being a normative reason for me to prefer, with any agent
whatsoever, that in C-type circumstances that agent do O.10
The only fair, impartial procedure the only one true to the spirit
of universalizability is to aim at maximising overall compliance.
And if that turns out to require that I do not perform O in C, that my
desire that I do so be frustrated, then so be it. The point is that one is
now a promoter of value, not an instantiater. The non-consequentialist doctrine (of whatever version) if combined with universalizability, is self-refuting.
265
II
My principal objection to Pettits argument is that universality and
impartiality are themselves contested notions contested between the
consequentialist and the non-consequentialist and that Pettits
argument begs the question by assuming the consequentialist reading.
There are two spirits of universalizability and Pettit simply chooses
the one that suits his case. The point is similar to one made by Tim
Chappell, though he puts it in different terms.12
Pettit says that the universalizing non-consequentialist, as a result
of his universalizing, has normative reasons (one for each agent X,
including himself) for desiring that X should do O in C. Granted. If I
commend O-ing in C to others; if I approve of others doing O in C; if I
prescribe to others that they should O in C; then, surely, if I am
rational, I will desire that others O in C. But desire is one thing, duty
266
ANDREW GLEESON
another, and the two can come apart. It is one thing to desire, for
moral reasons, general compliance with O-ing in C, another to have a
duty to take action designed to secure that compliance. It is also one
thing not to desire my duty be done more (or less) strongly than
I desire anyone elses be done, on the ground of impartiality; but
another to think, on the same ground, that I have a duty to promote
others dutifulness, and especially that I have a duty to promote
others dutifulness at the expense of doing my own (first-order) duty.
The former, weaker duty may be resisted on the ground of minding
my own business, the latter, stronger duty on the ground that it is the
very antithesis of impartiality: trying to get others to do their duties
when I dont do it myself is the very paradigm of an unjust and
hypocritical, self-serving partisanship.
(The non-consequentialist will no doubt also add that if [in an
example of Pettits own] I neglect my own children in order to spend
time campaigning for parental responsibility, I wrong my children,
sacricing them for a greater cause; while on the other hand, asymmetrically, it is not true that if I forgo the opportunity to improve
overall compliance with good parenting in order to be a good parent
to my own children, that I have wronged or mistreated those children
of other parents who would have beneted from a successful goodparenting campaign. They have been wronged by their own parents,
but not by me. It does seem to me that consequentialism here is far
removed from moral reality. Neither the unfortunate children of the
other parents, nor those neglectful parents themselves, could blame
me [not even in thought alone] for looking after my own children.
Can any consequentialist, outside the sheltered world of the academic
seminar room, gainsay that?)
Pettit can contend that, at a higher level, impartiality is preserved:
if I can (and should) neglect my (apparent) duty in order to maximise
general dutifulness, then so can (and should) you. I am claiming no
special privilege for myself. But in saying this he appeals to a conception of impartiality, which assumes his conclusion that promotion
is the right attitude to adopt to values. There is available an alternative
conception (an alternative spirit of universalizability), the one I appealed to just above, which takes the view that values are to be
instantiated, even at the cost of overall promotion. This alternative
insists that I do not neglect my contribution (so to speak) just because,
as a result of the neglectfulness of some others, I can improve overall
compliance with duty by not doing mine. In appealing to the higher
level of impartiality Pettit begs the question.
267
Pettit may respond that he has already shown that his argument
is insulated against this criticism. He writes that his argument
abstracts from other relevant values that would argue against satisfying either the self-directed preference or the more general preference.13 It is of course the latter that we are concerned with here.
Pettit is saying that his argument abstracts away from values that
exhort us to mind our own affairs, that would resist our taking action
concerned with others compliance, including general compliance
with a duty. He has in mind values that require one to be, he says, a
respectful fellow-citizen, colleague, or friend, which recommend
against moralistic intrusion into others affairs: e.g. values of respect
for privacy and autonomy. What does Pettit mean in saying that
his argument abstracts from these values? Something like this
I take it: that when no other such values are relevant (or on the
assumption they are not) what justification can we have for taking
other than a strictly neutral approach to the remaining duties of each
agent, seeking to maximise overall compliance? The problem of
course is that other such values frequently are relevant. One cannot
just wish them away. Moreover they include that conception of
impartiality which says it is hypocritical and unjust to forgo ones
own duty in the name of general duty-promotion and this value is
always relevant according to the non-consequentialist, so there is no
abstracting away from it without begging the question at issue.
The moral is that Pettit can avail himself of a duty to be concerned with general compliance at the expense of ones own, only if
he introduces such a duty as an independent premise. Such a premise
may seem plausible enough, but it is not forced upon us in unrestricted form. As Timothy Chappell in effect points out, it is not the
only plausible premise requiring one to be concerned with others
performance of their duties. Another is that one should promote
general dutifulness in so far as one can do so without violating any
value (or in a more restricted version, without failing to perform
ones own duty to O in C). Clearly the first proposed premise is a
consequentialist principle, while Chappells is a non-consequentialist
principle. What universalizability commits you to, depends on
which premise you adopt. Given the availability of Chappells
premise, if Pettit adopts the first he is simply begging the question.
Of course, he might advance some other compelling reasons for
preferring the first, but these will simply be reasons for consequentialism that operate quite independently of any considerations
about universalizability.
268
ANDREW GLEESON
Chappells point and mine are essentially the same, made in different ways and placed in different contexts. So I shall call the
objection the Chappell objection.
III
The Chappell objection is decisive against Pettits argument in its
present form. But it is possible to reformulate the argument to surmount the Chappell objection.
I have granted that Pettits argument as it stands does establish
that, if I universalize, and if I am rational in the sense of continent, I
shall have desires, one for each agent X, that X perform O in C. My
act of universalization gives me reason for these desires, and a continent agents desires conform to what he or she has most reason to
desire. (We are assuming there are no countervailing reasons to those
generated by the universalizability.) I also granted that these desires
would be equally strong. Why? Because of their pedigree. Each arises
from a reason, which arises from what we might call some rightmaking facts. Those right making facts, ex hypothesi, are the same
for each agent in C. So they generate the same kind and degree of
rightness for each O. That rightness in turn generates reasons for
desire that will, in turn, be of equal force. And finally, the rational
(continent) agent is one who will, for any reason of force R, have a
corresponding desire of appropriate strength. Since every reason is of
force R, the same force, each of the agents desires will be of the same
strength. So the rational agent will have a desire in respect of each
agent that he does O in C and these desires are of equal strength.
This is the point at which Pettit question-beggingly appeals to the
impartial spirit of universalizability. My suggestion is that instead of
appealing to a principle of impartiality, he appeal to an apparently
plausible principle of rationality: not rationality in the sense of continence ensuring the conformity of desires to reasons, but in the sense
of continence ensuring the conformity of actions to desires, where
that is understood as action which maximises the agents desire
(preference) satisfaction. If this familiar principle is granted, and
concentrating just on those desires concerning the duty to O in C,
given the equal strength of these desires, it follows that the rational
agent (in this stronger sense) will act neutrally to promote the maximum satisfaction of those desires. Which is to say he will promote
overall duty compliance, if necessary at the expense of doing his own
duty to O in C. In short, he will be a consequentialist.
269
270
ANDREW GLEESON
271
I want to emphasise that the revised version of Pettits argument is entirely my invention and not at all to be blamed on
Pettit. Moreover, I am not attributing to him the view, specically,
that there are possible circumstances in which one person should
be killed to save more. Especially since his argument concerns
cases in which overall compliance with duty (and not just any good
effect, like saving lives) is improved by not performing duty. Still,
one can imagine someone for whom lying (to ones dearest friend,
who has put all his trust in you) to improve overall veracity is a
terrible violation of his conscience. We might feel more confident
we can find fault (in our hearts) with such a man on a theoretical
basis than in the killing case. But can we be sure of that from the
philosophical armchair? Can we be sure our philosophical theory
will survive the encounter with someone for whom, as he may
explain himself, untruthfulness (and perhaps especially untruthfulness in the name of truthfulness) dishonours those on whose behalf
the lies are told as much as those to whom they are told; and for
whom such untruthfulness is (in its own way) no more a possibility
than leaping over the moon.
Even if the moral philosopher thinks he can answer yes to these
questions, this resistance to theoretical considerations is itself one of
the most marked features of morality, and any account of morality
must at least address it.
Of course, these considerations are highly controversial. However,
I do not need to rely on them to undermine the revised version of
Pettits argument. The problem for the argument lies in an ambiguity
in talk of desires, or of preferences. (For reasons of space I shall
largely ignore preferences.) I take it that in ordinary parlance desires
(hunger, lust, a craving for money or success) are to be contrasted
with other springs of action: anger, fear, being startled, cringing,
smiling, or ducking to avoid a missile. None of these are purely
automatic behaviours (we can control them to some degree) but
neither are they desires in the ordinary sense. Not only are these
sources of action distinct from desire, but they can all prompt us to
act against our desires. Two very pertinent examples of this are what I
would describe as being claimed by sympathy or by conscience. I
want to ignore my duty, but conscience says no. I want to ignore the
bad-tempered, filth-ridden patient crying for my attention, but my
heart melts at his pathetic importunity. And so on. By contrast with
ordinary speech philosophers, economists, and others often use
desire (and preference) as terms-of-art referring to all sources of
272
ANDREW GLEESON
273
searching than the earlier considerations in this section, which challenged just that. I believe that challenge interest in which goes well
beyond issues about consequentialism is well-founded. The concept
of rationality is one that has suffered at the hands of philosophers
trying to put more weight on it than it can bear. Even so, I would not
want to deny important connections between morality and rationality
entirely. For one thing, I think there are external reasons (in Bernard
Williams sense: ones not dependent on the agents psychology) not
of the sort that require moral behaviour on pain of irrationality, but
of a sort which require one to recognize the intelligibility of certain
moral reactions (of sympathy or conscience) on pain of insanity or
losing ones reason (better, of losing what it is that binds human
beings together in an un-rationalisable nexus of mutual recognition).
For another, I think that internalist element of rationality that we call
continence, self-control, or strength-of-will is important. In the continent person, his impulses will conform to morality as he understands it (which may not be something abstractly intellectual it may
be what his conscience demands). This is still some distance from the
Socratic thought that if we understand the Good as it really is, then
we cannot but love it with all our heart and mind, but it is part of the
story. If so, then maybe my discussion of the Pettit argument has not
been entirely academic.
ACKNOWLEDGEMENTS
I am grateful for comments on an earlier version of this paper from
Rodney Allen, Greg Bognar and Peter Woolcock. I would like to thank
the Philosophy School at the Australian Catholic University for
granting me leave-without-pay in 2004, making it possible to nish this
article, and the Philosophy Department at Flinders University for
providing a convivial atmosphere in which to do so.
NOTES
1
Peter Singer, Practical Ethics (Cambridge: Cambridge University Press, 1979), pp.
813.
2
R. M. Hare, Moral Thinking: Its Levels, Method and Point (Oxford: Clarendon
Press, 1981) and his Universal Prescriptivism, in A Companion to Ethics, ed. Peter
Singer (Oxford: Blackwell, 1993), pp. 451463.
274
ANDREW GLEESON
REFERENCES
Chappell, Tim. A Way Out of Pettits Dilemma. The Philosophical Quarterly 51
(2001): 9599.
Hare, R.M. Moral Thinking: Its Levels, Method and Point. Oxford: Clarendon Press,
1981.
Hare, R.M. Universal Prescriptivism. In A Companion to Ethics. Edited by Peter
Singer. 451463. Oxford: Blackwell, 1993.
Harris, John The Survival Lottery. In Killing and Letting Die. Edited by Bonnie
Steinbock & Alastair Norcross. 257265. New York: Fordham University Press,
1994.
Pettit, Philip. Consequentialism. In A Companion to Ethics. Edited by Peter
Singer. 230240. Oxford: Blackwell, 1993.
Pettit, Philip The Consequentialist Perspective. In Baron, Marcia W., Pettit, Philip
and Slote, Michael. Three Methods of Ethics. 92174. Oxford: Blackwell, 1997.
275
Discipline of Philosophy
Philosophy DP400
School of Humanities
University of Adelaide
South Australia 5000
Australia
E-mail: andrew.gleeson@adelaide.edu.au