Mind Association

Rule-Consequentialism
Author(s): Brad Hooker
Source: Mind, New Series, Vol. 99, No. 393 (Jan., 1990), pp. 67-77
Published by: Oxford University Press on behalf of the Mind Association
Stable URL: http://www.jstor.org/stable/2254891
Accessed: 14-03-2018 18:19 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
http://about.jstor.org/terms

Mind Association, Oxford University Press are collaborating with JSTOR to digitize,
preserve and extend access to Mind

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
 Rule-Consequentialism'

BRAD HOOKER

Suppose that accepting rules is a matter of having certain desires and

dispositions. Now consider the theory that an act is morally right if and
only if it is called for by the set of desires and dispositions the having of
which by everyone would result in at least as good consequences judged
impartially as any other.2 For lack of a better name, we might call this
theory disposition/rule-consequentialism, or just rule-consequentialism for
short. Two crucial features of this theory should be noted. One is that it
assesses the rightness and wrongness of any particular act, not directly in
terms of its consequences, but indirectly in terms of a set of desires,
dispositions, and rules, which is then assessed in terms of the conse-
quences of everyone's having that set.3 The other is that it assesses the
rightness of any given act, not in terms of the desires, dispositions, and
rules which are such that the agent's having them would bring about the
best overall consequences, but rather in terms of the desires, dispositions,
and rules which are such that everyone's having them would bring about the
best overall consequences.4 Let me refer to the set of desires, dispositions,
and rules which are such that everyone's having them would bring about
the best overall consequences as the optimific set.
To those who are attracted to consequentialism but want a moral theory
that accords with at least most of our intuitions, rule-consequentialism

1 This paper is much better than it would otherwise have been because of written comments on
earlier drafts from Penelope Mackie, Mark Overvold, Peter Vallentyne, Anthony Ellis, Mark Nelson,
and Richard Brandt. I am extremely grateful to these people, and to Alan Fuchs, James Griffin, R. M.
Hare, Roger Crisp, David Dyzenhaus, Cheryl Misak, Eldon Soifer, Howard Robinson, Madison
Powers, and Greg Trianosky for helpful discussions about the ideas in the paper. Remaining defects,
however, are my responsibility.
2 I do not want to make any controversial assumptions about what desires and dispositions human
beings are capable of having. I do not want to enter here into debates about how 'plastic' human nature
has been up to now; nor do we need to take up questions about whether trying to 'improve' human
nature via genetic engineering would be advisable.
3 So rule-consequentialism is a kind of indirect consequentialism. It is a mistake to think that
indirect consequentialism assesses not acts but only the rightness and wrongness of other things such
as motives, standing dispositions, rules, and social practices. Indirect consequentialism does assess acts
as well (see B. Williams, 'A Critique of Utilitarianism', in J. J. C. Smart and B. Williams,
Utilitarianism: For and Against, Cambridge, Cambridge University Press, I973, p. I2I).
4 See R. M. Adam's distinction between 'Individualistic motive utilitarianism' and 'Universalistic
motive utilitarianism' ('Motive Utilitarianism', journal of Philosophy, I976, p. 480) and D. Parfit's
distinction between 'Individual Consequentialism' and 'Collective Consequentialism', Reasons and
Persons, Oxford, Clarendon Press, I984, pp. 30-I. There might seem to be something fishy about the
claim that the right act is whatever one would be called for by the desires, dispositions, and rules which
are such that, if every last person had them, the best consequences would result. I shall return to this.

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
68 Brad Hooker

should have considerable initial appeal. In section I of this paper I try to

bring out the attractions of rule-consequentialism by showing how it can
be formulated so as to be safe from the main objections to other
consequentialist theories. Yet, even if the best version of rule-consequenti-
alism is immune to the objections that plague other consequentialist
theories, rule-consequentialism has difficulties of its own one of the most
serious of which is the so-called partial compliance objection.5 In section
II I consider how rule-consequentialists might reply to the partial
compliance objection. Then, in section III I explore the question of
whether rule-consequentialists can avoid the partial compliance objection
without opening up their theory to the objection that it makes unreason-
ably severe demands on individuals.

One of the most popular objections to utilitarianism is that interpersonal

comparisons of utility are impossible. I assume any plausible version of
rule-consequentialism will have a utilitarian component, and will therefore
need an answer to the objection about interpersonal comparisons. But for
the purposes of this paper I shall simply assume that some acceptable way
of making interpersonal comparisons is possible.6
Another prominent objection to act-utilitarianism is that it does not
allow that fairness or equality in the distribution of benefits and burdens
can be morally required even when a less fair or less egalitarian
distribution would provide greater net welfare.7 But it is possible to
accommodate this objection without abandoning even act-consequential-
ism. For it might be held that, while the right act is still just whatever one
will bring about the best outcome, outcomes are to be ranked in terms of
not only how much well-being they contain but also how equally or fairly
it is distributed.8
A prominent objection to standard kinds of act-consequentialism is that

5 Indeed, R. Brandt-long one of the pre-eminent rule-consequentialists-admits that this

problem is the most important of the problems peculiar to the indirect form of consequentialism. See
his 'Fairness to Indirect Optimific Theories in Ethics', Ethics, I988, pp. 341-60.
6 As consequentialists have long pointed out, it had better be, if we are to be able to conform even
with a common-sense duty of beneficence, such as W. D. Ross's (on which see Ross, The Right and the
Good, Oxford, Clarendon Press, I930, ch. 2). For a recent defence of interpersonal comparisons of
utility, see J. Griffin, Well-being: Its Meaning, Measurement and Moral Importance, Oxford, Clarendon
Press, I986, Part Two.
7 For an account of the main intuitive objections to act-utilitarianism, see S. Scheffler's
'Introduction' to his collection Consequentialism and its Critics, Oxford, Oxford University Press, I988,
pp. I-I3, esp. pp. 2-4.
8 See, for example, T. M. Scanlon, 'Rights, Goals, and Fairness', in Public and Private Morality
ed. S. Hampshire, Cambridge, Cambridge University Press, I978, pp. 93-I I I, especially sect. 2. This
paper is reprinted in Scheffler's collection Consequentzalism and its Critics. See also Scheffler, The
Rejection of Consequentialism, Oxford, Clarendon Press, I982, pp. 26-34, 70-9; and Parfit, Reasons
and Petsons, p. 26.

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
 Rule-Consequentialism 69

they give no direct weight to deontological, agent-relative considerations:

standard kinds of act-consequentialism hold that it is morally right to
harm people, or to ignore one's special obligations to those with whom one
has some special connection, when such acts would bring about even
slightly more good overall.9 Many people confidently believe that it is
morally wrong to commit murder, to torture someone for information, to
frame the innocent, to steal, to break one's promises, to fail to give special
weight to the interests of those with whom one has some special
connection, and so on, even when doing one of these things would produce
somewhat better consequences judged impartially. But here, too, much can be
said in defence of act-consequentialism.'0 Defenders of the theory usually
start by pointing out that the kinds of act in question very rarely produce
the best available outcome. They add that human limitations and biases
are such that we are not accurate calculators of the expected consequences
for everyone-of our alternatives. That is, we frequently do not have
the needed information, time, or the capacity to weight benefits and harms
impartially (for example, most of us are biased in such a way that we tend
to underestimate the harm to others of acts that would benefit us). For
these reasons, a sophisticated act-consequentialism would prescribe that
we inculcate and maintain in ourselves and others both firm dispositions
not to commit some kinds of acts and dispositions to disapprove of others
who do commit them. Indeed, given our psychological limitations, act-
consequentialism may favour our moral dispositions' running so deep that
we could not bring ourselves to do the kinds of act in question even in the
rare cases in which they would bring about the best consequences
impartially calculated. Such might be the dispositions-of those sets that
are psychologically possible-which are such that one's having them
would produce the most good. But all this notwithstanding, act-conse-
quentialists insist that right acts are those which would result in the most
overall good (even if morally good people could not bring themselves to
perform some of these acts). And in making this claim act-consequential-
ism conflicts sharply with the deontological convictions I mentioned."
Rule-consequentialism, however, does not make that problematic claim.
Rule-consequentialiam claims instead that individual acts of murder,
torture, promise-breaking, and so on, can be wrong even when those
particular acts bring about better consequences than any alternative acts
9 Of course, Scheffler himself advances a powerful attack on the idea that deontological
considerations should figure in the most basic, or first, principles of morality (Rejection of
Consequentialism, ch. 4).
" Prominent among those associated with the defence I am about to summarize are H. Sidgwick,
The Methods of Ethics, 7th ed, London, Macmillan, I907, bk. IV, ch. III; R. M. Hare, Moral Thinking,
Oxford, Clarendon Press, i98i, pp. 35-52, I30-59; Parfit, Reasons and Persons, pp. 27-8; and P.
Railton, 'Alienation, Consequentialism, and Morality', Philosophy and Public Affairs, I984,
pp. I34-7I, esp. pp. I53-4, I57-9 (Railton's paper is also reprinted in Scheffler's collection).
" See Conrad Johnson, 'The Authority of the Moral Agent', in Consequentialhsm nand its Critics, ed.
Scheffler, p. 264.

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
7o Brad Hooker

would have. For rule-consequentialism makes the rightness and wrongness

of particular acts, not a matter of the consequences of those individual acts,
but rather a matter of conformity with that set of fairly general rules whose
acceptance by (more or less) everyone would have the best conse-
quences. 12 And, this acceptance of fairly general rules forbidding murde
torture, promise-breaking, and so on, would clearly have better conse-
quences than everyone's accepting fairly general rules permitting such
acts. (This point could just as easily be put in terms of dispositions rather
than rules.)
Consider now one more objection to act-utilitarianism-that it is
unreasonably demanding, construing as duties what one would have
thought were supererogatory self-sacrifices. 13 To fully appreciate this
objection we need to keep in mind the following three things: (i) money
and other material goods usually have diminishing marginal utility; (2)
each dollar can buy vastly more food in, for example, Ethiopia than it can
in a 'First-World' country; and (3) other relatively well-off people will not
give much. We must thus accept that it would be utility-maximizing, and
thus optimific according to most versions of act-consequentialism, if I gave
away most of my material goods to the appropriate charities. I must, of
course, take into account the effects of my present actions on my future

12 The rules in question must of course be 'fairly general' because rule-consequentialism collapses
into extensional equivalence with act-consequentialism if the rules are allowed to be infinitely specific.
Brandt mentions this problem ('Indirect Optimific Theories', p. 347). The question of just how
universal the acceptance of the mode code must be is one I return to below.
13 This objection is discussed in many places. Sidgwick acknowledges that utilitarianism 'seems to
require a more comprehensive and unceasing subordination of self-interest to the common good' than
common-sense morality does (Methods of Ethics, p. 87, italics added). See also ibid., pp. 492, 499.
Other discussions of this objection include: K. Baier, The Moral Point of View, Ithaca, Cornell
University Press, I958, pp. 203-4; Scheffler, 'Introduction' to Consequentialism and its Critics,
pp. 3-4; G. Harman, The Nature of Morality, New York, Oxford University Press, I977, p. I57-62;
F. Feldman, Introductory Ethics, Englewood Cliffs, Prentice-Hall, I978, ch. 4; R. Brandt, A Theory of
the Good and the Right, Oxford, Clarendon Press, I979, p. 276; J. Hospers, Human Conduct: Problems
of Ethics, 2nd edn, New York, Harcourt Brace Jovanovich, i982, pp. i62-5; D. Brock, 'Utilitarianism
and Aiding Others', in The Limits of Utlihtarianism, ed. H. Miller and W. Williams, Minneapolis,
University of Minnesota Press, i982; Thomas Carson, 'Utilitarianism and World Poverty', in Miller
and Williams; Parfit, Reasons and Persons, pp. 30-I; and B. Williams, Ethics and the Limits of
Philosophy, Cambridge, Mass., Harvard University Press, i985, p. 77. And consider the following
passage from Railton, a defender of act-consequentialism: [J]ust how demanding or disruptive it [act-
consequentialism] would be for an individual is a function-as it arguably should be-of how bad the
state of the world is, how others typically act, what institutions exist, and how much that individual is
capable of doing. If wealth were more equitably distributed, if political systems were less repressive
and more responsive to the needs of their citizens, and if people were more generally prepared to
accept certain responsibilities, then individuals' everyday lives would not have to be constantly
disrupted for the sake of the good' (p. i6i). So act-consequentialism holds that in the real world, where
wealth is not equitably distributed, where political systems are repressive and unresponsive to the
needs of their citizens, and where most others are not more generally prepared to accept certain
responsibilities, most of your acts will not be morally right unless your everyday life is constantly
disrupted for the sake of the good? And one of the main ways in which act-consequentialism can
become not constantly disruptive is for a much wider group of people to make sacrifices for the sake of
those in need?

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
 Rule-Consequentialism 71

capacity to give. In the light of that consideration, I should keep whatever

proportion of my income and possessions is necessary for me to continue
earning so that I can maximize the amount I can give over the course of
my whole life. 4 But, presumably, what I am allowed to keep for myself is
still very little. And many of us may on reflection think that it would be
morally unreasonable to demand this level of self-sacrifice for the sake of
others.15 This thought is not inconsistent with the realization that
morality can from time to time require significant self-sacrifice for the sake
of others; nor does it oppose the claim that giving most of what one has to
the needy is both permissible and extremely praiseworthy. But most of us
are quite confident that such self-sacrifice is supererogatory-that is, not
something morality requires of us.
I admit that there is something unsavoury about objecting to a moral
theory because of the severity of its requirements. One might be tempted
to think that the demandingness objection will appeal to people whose
self-interest is clouding their moral judgement, that it will appeal to
people who have a lot to lose from a strong requirement to aid others.16
But it would be unfair to dismiss the objection on such grounds to do
so would be to find the objection 'guilty by association'. Furthermore,
even after we acknowledge that the demandingness objection may appeal
to some disreputable characters, the objection retains considerable
force.
I have been discussing the act-utilitarian requirement to keep making
sacrifices for others until further sacrifices would result in less overall
welfare in the long run. Consider now an alternative and less demanding
principle of aid. This principle is that we are required to come to the aid of
others as long as the benefit to them is very great in comparison to the
sacrifice to us.'7 Even this less demanding principle makes heavy demands
on those of us with spare money: most of us would have to sacrifice most of
our welfare in helping others before we reached the point at which the
sacrifice to us would no longer be very much smaller than the benefits

14 See P. Singer, Practical Ethics, Cambridge, Cambridge University Press, I979, p. I63.
15 Scheffler and T. Nagel usefully set out a number of different ways moral theorists might respond to
a very demanding moral conception (Scheffler, 'Morality's Demands and their Limits', Journal of
Philosophy, I986, pp. 53I-7; Nagel, The View from Nowhere, New York, Oxford University Press,
I986, ch. X; but cf. Griffin, Well-Being, pp. I27-62, I95-206, 246-5I, 302-7; Railton, pp. I69-70
n. 42; and D. Brink, 'Utilitarian Morality and the Personal Point of View', Journal of Philosophy, I986,
pp. 432-8). Particularly important is the following contrast. When confronted with an account of
morality which pictures it as extremely demanding, some of us would respond by rejecting that
account of morality's content. Another response is to accept that account of its content but then limit
morality's rational authority. Though there are other possible responses as well, the idea I am drawing
on in this paper is that any account of the content of morality which makes it constantly and
relentlessly demanding is ultimately mistaken. The purpose of the second half of this note is to
acknowledge that not everyone will find that idea seductive.
16 See Carson, 'Utilitarianism and World Poverty', p. 243.
17 A principle Sidgwick thinks is part of 'common-sense morality' and even, 'broadly speaking,
unquestionable' (Methods, p. 348-9; see also pp. 253, 26I).

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
72 Brad Hooker

produced for them.'8 And it might well be thought that a requirement

that one sacrifice most of one's own welfare for the sake of strangers whose
suffering is not one's fault is unreasonably demanding, particularly when
most others in a position to help are not doing so.
Indeed, one of the notable initial attractions of rule-consequentialism is
that it-unlike both act-consequentialism and the more modest principle
of aid just mentioned-calls for an amount of self-sacrifice that is not
unreasonable.'9 If each relatively well-off person contributed some
relatively small percentage of his or her income to famine relief, there
would be enough to feed the world. Perhaps ten per cent from each of the
well-off would be enough. Though giving that much may seem difficult
enough for most of us, the demand that we do so does not seem
unreasonab'le. Rule-consequentialism therefore seems to escape the de-
mandingness objection. If so, it is acquitted of all of the main charges
against act-utilitarianism.
I do not mean to suggest that there are not other powerful reasons for
finding rule-consequentialism the most appealing form of consequential-
ism. One might well think that reflection on the nature of morality
suggests that it must serve as a public system of principles to which we can
appeal to justify our behaviour to one another, and that rule-consequenti-
alism does, but not act-consequentialism does not, do justice to this
insight.20 Or one might well think that rule-consequentialism does justice
to the importance of the common question 'what if everyone did that?' in a
way that act-consequentialism cannot. I also do not mean to suggest that
the only way of reacting to the excessive demands of act-consequentialism
is to join up with the rule-consequentialists.2' But those are matters I do
not want to explore here.
18 This claim is similar to one of the central points of J. Fishkin's tightly argued Limits of
Obligation, New Haven, Yale University Press, I982, a book with profound implications. See also
Nagel's observation that, given existing world circumstances, not just act-utilitarianism but any
morality with a substantial impersonal component will make voracious demands ( View from Nowhere,
p. I 92).
1' See Parfit, Reasons and Persons, p. 3I. Cf. Singer, Practical Ethics, ch. 8, esp. pp. i8o-i; and
Fishkin, Limits of Obligation, pp. I62-3.
20 The literature on the 'publicity condition' is voluminous. For a sampling, see Sidgwick,
Methods, pp. 489-go: K. Baier, Moral Point of View, p. I98; A. Donagan, 'Is There a Credible Form
of Utilitarianism?' in Contemporary Utilitarianism, ed. M. Bayles, Garden City, NY, Doubleday & Co.,
I968, pp. I87-202, p. I94; J. Rawls, A Theory of Justice, Cambridge, Mass., Harvard University Press,
I97I, pp. I30 n. 5, I33, i8i, 582; J. Hospers, Human Conduct, Problems of Ethics, New York, Harcourt
Brace Jovanovich, Inc., I972, pp. 3I4-I5; Brandt, Theory of the Good and the Right, pp. I73-4, and
'Indirect Optimific Theories', p. 348; Parfit, Reasons and Persons, sect. I7; Williams, Ethics and the
Limits of Philosophy, pp. IOI-2, I08-9.
21 A different response is Scheffler's advocacy of an agent-relative prerogative to give one's own
proiects and commitments somewhat greater weight than those of others. See his Rejection of
Consequentialism. Scheffler's theory is not yet fully worked out, but the difficulties with it (on which see
S. Kagan, 'Does Consequentialism Demand Too Much? Recent Work on the Limits of Obligation',
Philosophy and Public Affairs, I984, pp. 239-54, esp. p. 25I) are serious enough to keep us from
dropping work on rule-consequentialism. Another interesting idea is that act-consequentialism turns
out not to be unreasonably demanding once we take into consideration the point that every criterion of

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
 Rule-Consequentialism 73

II

What is the partial compliance objection to rule-consequentialism? It is

that following the moral code that would be optimal in a world in which
everyone accepted it can be (in Brandt's words) 'counterproductive or
useless' in the real world where there is actually only partial social
acceptance of that code.22 To use an example Brandt cites, suppose that
what would produce optimal results would be for everyone to act in a race-
blind way. But suppose also that you are surrounded by people with fierce
racial prejudices. It is easy to imagine that your acting in a race-blind way
in front of these rabid racists would have very bad consequences.
On the simplest rule-consequentialist view, the best set of rules, desires,
and dispositions is whatever set is such that, if absolutely everyone
accepted those rules and had those desires and dispositions, the results
would be better than if absolutely everyone accepted other rules and had
other desires and dispositions. But then how can it make sense to object
that some dispositions which would produce the most good if everyone had
them would be counter-productive or useless in situations where not
everyone had them? It is clearly illogical to imagine both that everyone
accepts a certain code of rules or has a certain set of dispositions and that at
the same time some do not accept that code and do not have those
dispositions. Brandt provides a neat solution to this problem: he defines
the optimal moral code as the one that would result in the most good if it
were accepted by 'all except those whose agreement is precluded by a
description of some moral-problem situations'. 23
But there remain problems with how to interpret the partial compliance
objection to rule-consequentialism. Consider three possibilities:

(i) The objection might be that you might sometimes be required by

the rules whose currency would produce the most good to do
something that, because others are not complying, would produce
slightly worse consequences viewed impartially.

right and wrong must be sensitive to a requirement of psychological realism, i.e., act-consequentialism
cannot demand more self-sacrifice than human nature can deliver. I am grateful to James Griffin for
this suggestion, but I do not have room here to explore it here.
22 Brandt, 'Fairness to Indirect Optimific Theories in Ethics', p. 357. See also his Theory of the Good
and the Right, pp. 297-9, and his 'Problems of Contemporary Utilitarianism: Real and Alleged', in
Ethical Theory in the Last Quarter of the Twentieth Century, ed. N. E. Bowie, Indianapolis, Hackett
Publishing Co., I983, pp. 99-I02. I should mention that I will usually follow common practice in
contrasting full and partial compliance. But, since the best versions of rule-consequentialism (including
Brandt's) are concerned with acceptance-utility rather than compliance-utility, this terminology will be
less than perfectly accurate. The distinction between acceptance-utility and compliance-utility is
important because the consequences of everyone's accepting some code (and being motivated
accordingly) may involve more than the consequences of their acts of compliance with that code. (On
this, see Williams, 'Critique of Utilitarianism', pp. I I9-30; Adams, 'Motive Utilitarianism', pp. 467-8I,
p. 470; S. Blackburn, 'Errors and the Phenomenology of Value', in Morality and Objectivity, A Tribute to
f. L. Mackie, ed. T. Honderich, London, Routledge and Kegan Paul, I985, p. 2I n. I2.)
23 'Fairness to Indirect Optimific Theories in Ethics', p. 342; see also p. 358.

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
74 Brad Hooker

(ii) Or the objection might be that you might sometimes be required

by the rules to do something that, because others are not
complying, would be harmful or inconvenient to you and benefi
the very people who are not complying.
(iii) Or the objection might be that you might sometimes be required by
these rules to do something that, because others are not complying,
would produce very much worse consequences viewed impartially.

The first of these interpretations can, I think, be dismissed. The claim

that one should follow certain rules even when breaking them would
produce slightly more good does not seem counter-intuitive (except to
hard-line act-consequentialists).24
So consider now the second interpretation of the objection that,
sometimes, because others are not complying, my following simple rules
would be both harmful to me and beneficial to the very people who are not
complying. The idea that the simple rules provide the criterion of rightness
and wrongness for such cases does offend against our sentiments about
fairness.25 But it is pretty obvious how rule-consequentialists can defuse
this objection: they can agree it is not wrong to refuse to follow certain rules
in one's dealings with those who do not reciprocate. The moral rules that
would produce the most good in our partial compliance world would have
provisions written into them designed to give incentives to those who are
possible beneficiaries of our rule-following to do their part in following the
rules. Such provisions would permit us to ignore the ordinarily appropriate
rules when we are dealing with those who refuse to follow those rules.
The third interpretation of the objection seems the most powerful.
Rule-consequentialism is less than credible if it claims one is required
always to follow certain rules, even when not following these rules would
prevent very much worse consequences for others. (Since from now on I
shall ignore the first two ways of construing the partial compliance
objection, I shall henceforth refer to the objection construed in the third
way as simply the partial compliance objection.) Now, can rule-conNe-
quentialists provide a convincing reply to the objection?
Brandt articulates what would seem to be the natural reply for the rule-
consequentialist to make. He suggests that rule-consequentialism would
prescribe our having a set of moral motivations consisting of

24 See the famous discussion in Ross, Right and the Good, pp. 34-5. The claim made in my text
must not be misconstrued as a reply to the following common objection against rule-consequentialism:
given that rule-consequentialism says the point of following moral rules is to optimize, its refusal to
permit us to break the rules when doing so would be optimific makes the theory internally incoherent.
I do not have room here to go into detail about this well-worn objection, which Brandt does take up in
'Fairness to Indirect Optimific Theories' on pp. 353-7, and to which I believe that there are
persuasive indirect consequentialist replies. But I am grateful to Anthony Ellis for pointing out that I
need to distinguish the idea I discuss in the text from this objection.
25 For a recent attempt to deal with fairness by someone sympathetic to utilitarianism, see Griffin,
Well-being, ch. X.

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
 Rule-Consequentialism 75

(i) the standing motivations corresponding to the usual simple moral

rules (that is, a standing desire to treat others fairly, another
standing desire not to hurt others, another not to steal, another not
to break one's promises, another not to lie, and so on),
(2) a standing desire to prevent great harm, and
(3) a standing desire to promote the wider acceptance of the optimific
rules.26

Rule-consequentialists would presumably say that an act is morally wrong

if it is one that this set motivations would oppose. We might summarize
this reply to the partial compliance objection as follows: by including in
their favoured code a particularly strong requirement that one prevent
great harm, rule-consequentialists can escape having to maintain that it is
morally right to stick to the (normally optimific) rules in those situations in
which our doing so would result in very much worse consequences. (For
the purposes of the rest of this paper, however, we can focus on (i) and (2)
above and ignore (3). This is so because in at least many situations the
desire to prevent great harm would not need the help of the desire to
promote wider acceptance of the optimific rules in order to prevail in a
battle with the motivations corresponding to the simple rules. And those
are the cases I shall be focusing on.)

III

Unfortunately, relying on a strong principle of preventing harm

deal with the partial compliance problem threatens to make rule-conse-
quentialism excessively demanding. Suppose that, of the various possible
rules about coming to others' aid, the rule whose acceptance by absolutely
everyone relatively well off would produce the most good requires one to
donate a tenth of one's income to famine relief. Suppose that this is what I
have just done. Knowing that most others in a position to donate are not
complying with that requirement, I am now trying to decide whether I
must donate more. There are still people dying who would be saved if I
gave more. My doing so would thus prevent serious harm to others. Given

26 Actually this is not exactly what Brandt presents. He says that his proposal would lead us to
think that in the example from South Africa in which the agent is being watched by rabid racists 'the
requirement to treat people equally now may be weaker than the prohibition on causing great harm,
but the requirement to do what one can to improve matters will be stronger than simple capitulation to
the rules subscribed to by the majority' ('Indirect Optimific Theories', p. 359). This is confusing
because it suggests that the optimal moral code would include an intrinsic motivation to capitulate to
the rules subscribed to by the majority-which must be a mistake. To be sure, we might accept the idea
that attempts at moral reform may be most successful if they try to build on, rather than completely
overturn, existing rules. (On this idea, see Sidgwick, Methods of Ethics, pp. 467-7I, 473-6, 480-4;
Griffin, Well-being, pp. 206, 302; Brandt, Theory of the Good and the Right, p. 290, and 'Indirect
Optimific Theories', pp. 350, 356-7.) But to accept that attempts at moral reform may be most
successful if they try to build on existing rules is not to accept the idea that people should be motivated
to capitulate to whatever moral rules the majority accepts.

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
76 Brad Hooker

these conditions, Brandt's proposal would require me to give more of my

income. And even if I give another tenth, there would be other people
whom I would save by giving more. So yet again serious harm would be
avoided if I kept giving. In fact, I would have to make myself quite badly
off before I myself would be so poor that the further sacrifices I could
make would be too little to save others in desperate need from great harm.
And so the code in question seems to require self-sacrifices up to that
point. But to require self-sacrifices all the way up to this point seems
unreasonably demanding, especially when others in a position to help are
not doing their share.27
So my argument might be put like this. The partial compliance objection
to rule-utilitarianism is extremely important-after all, we live in a partial
compliance world. Rule-consequentialism can be rescued from the partial
compliance objection by bringing in a strong requirement that one prevent
great harm. But this move seems to pull rule-consequentialism out of the
mouth of one objection only to throw it into the mouth of another.
What might rule-consequentialists say in reply? They might try to claim
that the optimific set of rules and dispositions would permit me to depart
from a given rule in order to prevent great harm, but would not require me
to do so. This reply might sound good when we think about the famine
relief cases, but we can see it cannot be right when we think about the
other partial compliance cases, the cases in which the agent's own welfare
is not in play. To give an intuitively acceptable answer for the cases in
which the agent's own welfare is not in play, rule-consequentialists must
say that the optimific code requires the agent not to do what would result
in great harm. And it would seem that the most natural way for rule-
consequentialists to accomplish this is to maintain that we should obey the
code of rules that would be optimific if everyone complied with it, except
when our following that code would result in great harm because of others'
non-compliance, and that in those cases we should do what would prevent
that harm. But, as I have explained, this answer will make rule-
consequentialism terribly demanding in the famine relief cases.
Here is a more promising reply that rule-consequentialists might try.
Rule-consequentialists favour the code of rules whose currency would
produce the most good. One of the factors counted in the cost-benefit
assessment of any proposed code is what we might call its maintenance
costs, that is, the costs of sustaining people's commitment to it and of
teaching it to the young. Furthermore, many rule-consequentialists
(Brandt among them) hold that wrong acts are those forbidden by that
moral code whose general currency among human beings with their natural
biases and limitations would produce the best consequences.28 And, given
27 For the suggestion that rule-utilitarianism will, under not far-fetched conditions, describe as
duties acts which surely are merely suipererogatory, see Donagan, 'Is There a Credible Form of
Utilitarianism?' pp. I94-6.
28 See 'Indirect Optimific Theories', pp. 346-7, 349-50; and 'Problems of Contemporary

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms
 Rule-Consequentialism 77

natural human selfishness, etc., it might well be true that an extremely

demanding morality could be successfully taught to, and sustained in,
people only at great cost.29 Therefore, the currency of a somewhat less
demanding morality might have better consequences, all things con-
sidered. That is, the moral code whose currency would result in the
greatest good overall might be less demanding than we would have
thought if we had forgotten about maintenance costs.
The question now is: even after we take maintenance costs into account,
would the optimific set of rules nevertheless be unreasonably demanding?
It is hard to be sure. Part of the reason for this is that our sense of what
counts as being unreasonably demanding is somewhat vague (though
determinate enough to license the charges of excessive demandingness
made earlier in this paper). But equally important is that we are uncertain
which rule of the alternative possible ones about coming to the aid of
others is such that its acceptance by everyone (except those whose
acceptance is precluded by the description of the problem situation in
question) would produce the most good. In other words, we do not know
how demanding the optimific rule about coming to the aid of others would
be.30 Just where the line is between reasonable and unreasonable
demandingness and just how demanding the optimific rule would be are
questions that require further work. But at least we can throw down the
gauntlet, challenging rule-consequentialists to show that the rule about
coming to the aid of others which is such that its general acceptance would
produce better consequences on the whole than the general acceptance of
any alternative would not be in conflict with our fairly confident convic-
tions about what is above and beyond the call of duty.3'

Department of Philosophy, 9I5 W. Franklin Street BRAD HOOKER

Virginia Commonwealth University
Richmond, VA 23284- 2025 USA

Utilitarianism', p. 98. See also J. L. Mackie's remark that in devising a moral code 'we are to take men
as they are and moral laws as they might be' (Ethics: Inventing Right and Wrong, Harmondsworth,
Penguin Books, I977, p. I33).
29 If at all! As Brandt observes in his book, 'What these rules [of obligation] may require is limite
by the strain of self-interest in everyone, and [by] the specific desires and aversions bound to develo
in nearly everyone . . .' (p. 287).
30 As Brandt concedes, 'the optimific indirect theory does run into complications when we try to
work out the details in any realistic way' ('Indirect Optimific Theories', p. 36o).
31 Cf. J. Arthur, 'Equality, Entitlements, and the Distribution of Income', in Applying Ethics, ed. J.
Olen and V. Barry, Belmont, CA, Wadsworth Inc., i989, pp. 362-72.
Brandt's Theory of the Good and the Right contains a powerful and celebrated attack on the idea that
we should judge moral theories by how much they agree with our intuitions (see op. cit., ch. I sect. 3).
But see also N. Daniels, 'Can Cognitive Psychotherapy Reconcile Reason and Desire?', Ethics, i983,
pp. 772-85, esp. pp. 778-8i. My own opinion is that, while we must heed Sidgwick's warning that 'it
cannot be denied that any strong sentiment, however purely subjective, is apt to transform itself into
the semblance of an intuition' (Methods, p. 339) considered judgements might nevertheless play a
legitimate role in the testing of moral theories.

This content downloaded from 108.61.242.12 on Wed, 14 Mar 2018 18:19:57 UTC
All use subject to http://about.jstor.org/terms