1
The Groundwork of Utilitarian Morals: Reconsidering Hare’s Argument for
Utilitarianism
DRAFT: NOT FOR QUOTATION
Peter Singer
In Moral Thinking, R.M. Hare argues that if we think clearly about what we ought to do,
we will be led to a form of utilitarianism. Given Hare‟s prominence in ethics at the time,
and the difficulty of establishing any strong normative conclusions in ethics, this
ambitious claim has received surprisingly scant attention. J. L. Mackie‟s dismissal of the
argument appears to have persuaded many that there is no need for further discussion,
and the general verdict of others who do discuss it is overwhelmingly negative.1 In my
view, the strength of Hare‟s argument has not been sufficiently recognized, and Mackie‟s
seemingly knockdown blow fails to grapple with a central element of it. My goal is to
revive interest in the argument. To do so I shall show how the argument developed out of
Hare‟s early works, indicate what Mackie fails to take into account, and then discuss the
argument‟s strengths and weaknesses. I conclude by comparing Hare‟s method of
arguing for utilitarianism with that of the greatest of the nineteenth century utilitarians,
Henry Sidgwick.
Hare regarded his position as Kantian, in the sense that it is the position Kant should have
reached, on the basis of his understanding of the categorical imperative as requiring
universalizability.2 The Kantian reference of my title should not, however, be taken as
an invitation to go into vexed issues of how Kant should be interpreted. Rather, it is a
response to the assertion – which I understand was the inspiration for this conference
1
Mackie, Ethics, ….., also Michael McDermott, “Hare‟s argument for utilitarianism,” in
Phil Quart. 1983, and several of the essays in Douglas Seanor and N. Fotion, eds., Hare
and Critics .
2
R. M. Hare, 'Could Kant have been a Utilitarian?', Utilitas 5 (1993).
2
session – that Kantians and utilitarians never talk to each other. I hope that Kantians will
join the discussion of the argument I defend below.
The Development of Hare’s Argument: First Phase
Hare was a non-cognitivist. He held that moral judgments do not describe the world, and
in that sense cannot be true or false, as ordinary descriptive statements can be. They are
prescriptions, which means they belong to the broad family of imperatives. In Freedom
and Reason, published in 1963, Hare argues that the freedom we have to make up our
own minds about what is right or wrong lies in the fact that we choose what we shall
prescribe. That choice, however, is not unconstrained. Reason enters the picture both
because there can be logical relations between prescriptions, and because moral
judgments are universalizable. If I say “Shut all the windows!” and “Leave the central
window open!” I contradict myself. I also contradict myself if I say “All tax cheats ought
to be imprisoned” and “I am a tax cheat, but I ought not to be imprisoned.”
Universalizability means that, in Hare‟s words:
One cannot with logical consistency, where a and b are are two individuals, say
that a ought, in a certain situation specified in universal terms without reference
to individuals, to act in a certain way, also specified in universal terms, but that b
ought not to act in a similarly specified way in a similarly specified situation.
This is because in any „ought‟-statement there is implicit a principle which says
that the statement applies to all precisely similar situations. This means that if I
say „This is what ought to be done; but there could be a situation exactly like this
one in its non-moral properties, but in which the corresponding person, who was
exactly like the person who ought to do it in this situation, ought not to do it‟, I
contradict myself.3
Moral judgments, in other words, cannot be based on just any reason. In particular, we
cannot make exceptions for ourselves. If I claim that you ought to give to the poor, I
3
R.M. Hare, “Universal Prescriptivism” in Peter Singer, ed., A Companion to Ethics,
(Blackwell, Oxford, 1991) p. 456. See also Freedom and Reason, p10ff.
3
must either agree that I too ought to give to the poor, or I must find a reason why it is not
the case that I ought to give to the poor, and this reason must be specifiable without
reference to individuals. That reason might be that you are a billionaire and I am
struggling to keep up with my mortgage payments. But it cannot be that I am the one
who will benefit if I keep all my money, whereas you will benefit if you keep all your
money.
Although Hare‟s words sometimes gave the impression that he understood
universalizability to be a requirement of reason (for example: “to universalize is to give
the reason”4) his considered position leaves room for amoralism, and the amoralist, in his
view, need not be inconsistent or irrational.5 Moral language excludes non-
universalizable judgments, but there is no logical requirement that one use moral
language, or guide one‟s life by universalizable principles. If that is right, the
significance of any conclusions that we may be able to reach by drawing out the
implications of universalizability will be weakened. One can always escape those
conclusions by refusing to make moral judgments. Nevertheless, since many of us do
want to use terms like “ought” and “right,” the possibility of rational amoralism does not
render pointless the exercise of exploring the implications of universalizability, and
seeing what normative conclusions may follow from its application.
When I prescribe something, using moral language, my prescription commits me to a
substantive moral judgment about all relevantly similar cases. This includes hypothetical
cases in which I am in a different position from my actual one. So to make a moral
judgment, I must put myself in the position of the other person affected by my proposed
action – or to be more precise, in the position of all those affected by my action. Whether
I can accept the judgment – that is, whether I can prescribe it universally – will then
depend on whether I could accept it if I had to live the lives of all those affected by the
action.
4
Freedom and Reason, p.5.
5
In Moral Thinking, Hare discusses amoralism, but does not reject it on the grounds that
the amoralist is necessarily involved in a contradiction. Instead he appeals to prudential
considerations as a reason for not being an amoralist. See especially p. 186..
4
That sounds as if universalizable moral judgments will have to be based on considering
the preferences or interests of all affected – which would mean that we can arrive very
rapidly at a form of utilitarianism based on maximizing the satisfaction of preferences or
interests. But Hare did not arrive at utilitarianism so swiftly. In Freedom and Reason he
raises the issue of whether it is wrong for a woman to take a well-paid job undressing
herself a at a strip club. The stripper prefers the work to anything else she can get, and
those in the audience enjoy the performance. As far as interests are concerned, Hare tells
us, “since everybody gets what he or she wants, nobody‟s interests are harmed.”
Nevertheless, that does not settle the moral issue, Hare tells us, because “it is a question
not of interests but of ideals.” Some hold ideas of how a person should behave – or
perhaps, how a woman should behave - that are incompatible with undressing for the
sexual enjoyment of strangers. Hare considers the possibility of confining the moral
terms to questions concerning the impact of our actions on other people‟s interests, but he
rejects such a “terminological fiat” on the grounds that such a restriction “would truncate
moral philosophy by preventing it saying anything about ideals.”6
In the same work, Hare makes a similar statement about a Nazi who asserts that all Jews
should be killed. This prescription commits the Nazi to prescribing that he himself
should be killed, in the hypothetical case in which he discovers that he is a Jew. Since,
we assume, the interests of Jews in continuing to live are greater than the interests of
Nazis in killing them, no one would prefer to live the lives of all those affected by the
prescription that all Jews be killed. Nevertheless, Hare says, a Nazi could be so fanatical
about his belief in racial purity that he would think that he should be killed, even if he
were a Jew. If he accepts that, he is not violating universalizability. Instead, in Hare‟s
words, “the fanatic nails his flag to the content of the ideal, irrespective of its holder.”7
Development of Hare’s Argument: Final Form
6
Freedom and Reason p.147.
7
Freedom and Reason, pp. 144-147.
5
In “Ethical Theory and Utilitarianism,” published in 1976, Hare took the argument an
important step further. He had, he wrote discovered how “to deal in an agreeably clear
way with the problem of the fanatic, who has given me so much trouble in the past.”
Here is how he thought this could be done:
In so far as, in order to prescribe universally, I have to strip away (qua author of
the moral decision) all my present desires etc., I shall have to strip away, among
them, all the ideals that I have; for an ideal is a kind of desire or liking (in the
generic sense in which I am using those terms) … This does not mean that I have
to give up having ideals, nor even that I must stop giving any consideration to my
ideals when I make my moral decisions; it means only that I am not allowed to
take them into account qua author of the moral decision.8
It is, of course, only when I can take my ideals into account qua ideals – that is, as
something other than interests – that ideals can trump interests. If I can only take them
into account as some kind of desire or liking, the content of the ideal no longer matters.
Ideals are then treated simply as another kind of preference. When Hare revisits the
example of the fanatical Nazi, he shows how this works. Now, he says:
…the only sort of fanatic that is going to bother us is the person whose ideals are
so intensely pursued that the weight that has to be given to them, considered
impartially, outbalances the combined weights of all the ideals, desires, likings,
etc. that have to be frustrated in order to achieve them.9
No actual Nazi could possibly have desired the extermination of Jews with sufficient
intensity to outweigh all the desires of the Jews to continue to live, so this kind of fanatic
8
“Ethical Theory and Utilitarianism,” in H.D. Lewis, ed., Contemporary British
Philosophy 4, Allen and Unwin, London, 1976; reprinted in R.M. Hare, Essays in Ethical
Theory, (Clarendon Press, Oxford, 1989). The passage quoted is from that volume,
p.219.
9
Ibid., p.220.
6
is quite fantastic. Since our moral intuitions in fantastic situations are, Hare argues,
unreliable, such fanatics need not trouble us.10
Mackie’s Objection: Sliding Through the Stages of Universalizability
Mackie‟s critique of Hare‟s argument appeared in his Ethics: Inventing Right and Wrong,
published a year after the version of the argument in “Ethical Theory and Utilitarianism”,
but before the fuller version of the argument that was to appear in Moral Thinking.
Mackie distinguishes three different stages of universalizability:
1. The irrelevance of numerical differnces
2. Putting oneself in the other person‟s place
3. Taking (equal) account of different tastes and ideals.
We can grant, Mackie says, that universalizability, in some sense, is part of the meaning
of the moral terms, but if so, this will be no more than the first, or at most the first two, of
these stages of universalization. Hare, however, needs to appeal to the third stage for his
argument against the fanatic to succeed. For if the moral terms allow me, qua author of
the moral decision, to give my own ideals a weight that is independent of the strength
with which I desire them, then we cannot treat moral questions as simply a matter of
weighing up desires or interests. And, Mackie asserts, the moral terms do allow this: they
allow me to assume that my ideals are true and ought to prevail over any mere
aggregation of interests, including my own. Mackie sees this as the fatal weakness in
Hare‟s argument for utilitarianism, and he is not alone in taking that view.
Another’s Sorrow
10
For Hare on the unreliability of moral intuitions in fantastic cases, see his 'What is
Wrong with Slavery', Philosophy. and Public Affairs, vol. 8, reprinted in Peter Singer,
Applied Ethics, (Oxford, Oxford UP, 1986)
7
On several occasions, Hare commented on Mackie‟s objection to his argument, and
invariably denied that he makes use of different stages of universalization. Moral
judgments are, as he wrote in Moral Thinking, “universalizable in only one sense, namely
that they entail identical ethical judgments about all cases that are identical in their
universal properties.” What Mackie refers to as distinct senses or stages of
universalization is, rather, “a progression in the use we make of this single logical
property as we develop our theory of moral reasoning.”11
In support of this claim, Hare points to his argument in the preceding chapter of Moral
Thinking, a chapter headed, in tribute to a poem by William Blake, “Another‟s Sorrow.”
The chapter makes a crucial but neglected point in the overall argument. Hare‟s initial
appeal here is to the general requirement of rationality “to make our moral judgments in
the light of the facts.” If I am prescribing something that will affect no one but myself, it
would be irrational for me to do so without knowing the relevant facts, including the
impact that what I am prescribing will have on my own preferences. Now suppose that I
am making a universalizable prescription. The relevant facts will include the impact my
prescription will have on the preferences of others, since I must be willing to agree to the
prescription even if I am in the position of others affected by my decision. (These are,
after all, situations that are identical to mine in their universal properties.)
This much seems difficult to deny. But now comes a more controversial move. Hare
claims that to know what it is like to be someone with a preference of a certain strength is
to take on that preference, if one is in that person‟s position. Hare puts this in terms of an
entailment between two distinct propositions:
11
Moral Thinking, p.108; see also R. M. Hare, 'Rights, Utility and Universalization: a
Reply to John Mackie', in R. Frey, ed., Utility and Rights (Minneapolis, Minn.:
University of Minnesota Press), and reprinted in R.M. Hare, Essays on Political Morality,
Oxford University Press, Oxford, 1989, pp.87-8.
8
(1) I now prefer with strength S that if I were in that situation x should happen,
rather than not;
(2) If I were in that situation, I would prefer with strength S that x should happen
rather than not.
Of these propositions, Hare writes:
What I am claiming is not that these propositions are identical, but that I cannot
know that (2), and what that would be like, without (1) being true, and that this is
a conceptual truth, in the sense of “know” that moral thinking demands.12
Hare hasn‟t put this with his usual precision, because his (1) can be read as an all-things-
considered preference, and as such, it obviously would not follow from my knowing (2).
Even putting aside any consideration of other values or ideals that I may hold – and I
shall come to that shortly – “that situation” may refer to me being in the situation of just
one of many people affected by the action I am contemplating, whereas when I now, qua
author of the moral decision, decide what I prefer, I must take into account the
preferences of all those affected by the prescription. Hare should have said that what is
entailed by my knowing (2) is:
I now have a preference, of strength S, that if I were in that situation, and
everything else was equal, x should happen.
Hare comes closer to this way of putting his point a few pages later:
I cannot know the extent and quality of others‟ sufferings, and in general,
motivations and preferences without having equal motivations with regard to what
should happen to me, were I in their places, with their motivations and
preferences.13
12
Moral Thinking, p.96. I am grateful to Michael Smith for pointing out the significance
of this argument, and thus for providing the inspiration for this paper.
13
Moral Thinking, p.99.
9
This claim seems plausible. If it is, it undercuts Mackie‟s claim that Hare‟s argument
involves a separate stage of universalization, one that requires an additional substantive
moral step to go from the second to the third stage of universalization. Instead, Hare‟s
argument relies on the standard form of universalization that Mackie accepts as part of
moral language. It couples that standard form with two other claims. One is the surely
unimpeachable view that to act rationally is to act with full knowledge of the facts
relevant to your action.14 The other is the more contentious claim to which I now turn:
that there is a link between knowing what it is like to be another person in a certain
situation, and wanting, now, if I were in that situation, what that person wants.15
Ideals, Values and Preferences
Our acceptance of an “other things equal” qualification to what I now prefer, qua author
of the moral decision, leaves open the possibility of resisting Hare‟s utilitarian conclusion
by maintaining that other things are not equal because there are other values involved that
are not simply matters of preference or desire. The objection, in other words, is that even
if I must take on the preferences of all those who are affected by my actions, I am not
required to treat all values as simply a matter of summing up preferences. I can hold,
universalizably, a different kind of value – in Hare‟s terminology, an ideal – that has
weight irrespective of its impact on anyone‟s preferences.
The question is, however, how I can, qua author of the moral decision, justify treating my
values or ideals as something other than one of my preferences. Let‟s return to Hare‟s
example of the well-paid stripper. We assume that her ability to work as a stripper
satisfies her overall preferences, since she enjoys her work, and earns more by stripping
than she could by any other form of employment. We assume also that the paying
customers enjoy watching her perform. Let‟s also, for the sake of the argument, make the
14
It may be objected that sometimes the costs of acquiring full knowledge of the relevant
facts are so great that it is rational to act without that knowledge. We accept this
qualification, but assume that it does not affect the point we are making.
15
Hare correctly notes that the argument applies not only to persons, but to all sentient
beings, but uses “person” to avoid more cumbersome language.. INSERT PAGE NO.
We follow his practice.
10
perhaps more questionable assumption that no-one‟s non-moral preferences are adversely
affected by such performances – that is, watching her strip shows does not cause men to
harass women sexually, nor to treat them less well in any other respect than they
otherwise would have, and other women do not feel degraded or think less well of
themselves because of her work. We make these assumptions, of course, so that we have
a case in which, if we put ourselves in the position of those whose interests – that is, their
preferences apart from any moral ideals they may hold - are affected by a woman‟s
decision to work as a stripper, there is no objection to such work. But now imagine that
someone we‟ll call Paul accepts all this, but nevertheless urges that strip shows should be
banned because the male desires that a woman satisfies when she undresses for the
enjoyment of strangers are inherently depraved, and we should not satisfy desires that are
depraved, whether or not anyone is aware of, or suffers in any way from, this depravity.
How should we understand such a claim? Paul might be understood as saying that he
would prefer to live in a world in which no one wished to watch a stranger undress. But
while that preference can, and –arguably - should, be counted alongside all the other
preferences of those who would be affected by a decision to permit, or ban, strip shows,
Paul is not, qua author of the moral decision, entitled to give it special weight merely
because it is his preference. That would be a violation of the requirement that we make
similar judgments about situations that are identical in their universal properties. If the
preference to live in a world in which no one wishes to watch another undress were not
his preference, Paul would not be giving it more weight than any other preference.
Paul would no doubt respond that this completely misunderstands his claim. He is not
saying that the desire to watch a stranger undress is wrong because he doesn‟t like it, he
is saying that such desires are wrong because they are depraved. Depravity, he would
maintain, is an objective notion. He might say that this is something we know by direct
intuition, or by revelations about God‟s commands, or perhaps he would say it is to be
discovered by a study of natural law and the proper relationship between sexual desire
and reproduction. In whatever way he grounds his claim to objective truth, he would say
that it would be wrong to desire to watch a stranger undress even if he himself were,
11
horror of horrors, so far sunk in depravity as to be an enthusiastic patron of such
performances.
How are we to understand the claim that Paul‟s view is true? For non-cognitivists like
Hare, moral judgments are neither true nor false. As now understood, Paul‟s claim can
succeed only if some form of moral realism is true. Moreover, this will have to be a form
of moral realism in which the truth of at least some moral claims is independent of
preferences or interests. Whether the appeal is to self-evident intuitions, or God‟s will, or
to natural law, or to any other way of establishing objective truth in morality, to defend it
is a large and difficult task. Moreover it is a task that Mackie himself would have
thought doomed to fail, for Mackie held that, although we speak as if morality were part
of “the fabric of the universe” this is an error, and moral judgments are not really true or
false.
Nevertheless, it may be objected, surely no meta-ethical view should exclude the idea
that some things matter for their own sake. I can commit myself to wanting something to
be the case, irrespective of whether I will desire it at the time, as Paul may commit
himself to banning strip clubs even if he recognizes that as he gets older he may want to
frequent them.) Why does this presuppose moral realism?
Some may suggest that the conservative and perhaps religious overtones of objections to
strip shows loads the example against the defender of the view that moral principles are
independent from desires. (There are also feminist objections to such shows, but these
are less likely to be based on ideals than on claims about the effect the shows have on the
interests of women.) So let‟s redress the balance by taking an example that will find
sympathy with a different range of people. Imagine an ecologist who says: “Failing to
conserve an endangered plant is wrong, even if its extinction would make no difference
to the interests of any sentient being, now or in the future.” To make such statements,
ecologists do not have to believe in objective moral truth. They may be non-cognitivists,
expressing their own attitudes. Their views are nevertheless distinct from those of more
self-centered advocates of protecting endangered species who say: “I want endangered
plants protected because I enjoy finding rare species of plants.” The self-centered
12
environmentalists may not care if, after they die, nothing is done to preserve endangered
species. The ecologist, on the other hand, will care about this. These views are distinct.
A non-cognitivist like Hare can recognize this distinction without construing either of
them as making a statement about how the world is. Hare would regard them as desires,
in the broadest sense of the term – we might also call them attitudes, or preferences. The
question can therefore be asked, of the advocates of each of these views, why their
desires, attitudes or preferences should be given greater weight than the equally strong
and firmly-held desires, attitudes or preferences of anyone else.
If this question has an answer, then we need to hear and assess it. But in the absence of
any claim to objective truth in ethics, one cannot say that the moral ideals one holds
should be given greater weight because they are true. Nor, for reasons I have offered
elsewhere, do I accept the approach to normative ethics that takes such ideals as data, and
then assumes we must choose the normative theory that best matches these and other
considered moral jdugments in reflective equilibrium.16 If we put aside both objective
truth and reflective equilibrium, however, it is hard to see what reasons a person could
offer for giving greater weight to her desires, attitudes or preferences than the equally
strong and firmly-held desires, attitudes or preferences of anyone else. It seems that all
such a person could say is: “It is the moral view that I hold.” Universalization, however,
does not allow me to give more weight to moral views because they are the ones that I
have. That being so, the deep ecologist‟s views should not be given more weight than
anyone else‟s equally strong preferences.17
Can we dispose of all non-utilitarian moral principles in the same manner, thus leaving
the field clear for a maximizing form of preference utilitarianism? Such a claim would
be rash, for it would attempt to shortcut all arguments for contrary positions. Instead,
each such argument should be considered on its merits. Imagine, for instance, a choice
between two actions, one of which leads to world A, in which by chance, some have very
16
See 'Sidgwick and Reflective Equilibrium', The Monist, vol. 58 (July 1974), pp. 490-
517, and 'Ethics and Intuitions', The Journal of Ethics, vol 9, no. 3-4 (October 2005), pp.
331-352.
17
For a similar response, see R.M. Hare, “Comments” in Douglas Seanor and N. Fotion,
eds., Hare and Critics, pp.270-272.
13
high welfare and others have very low welfare, while the other action leads to world B, in
which welfare varies not at all, or only on the basis of desert. Suppose that the average
welfare in world A is slightly higher than in world B. It seems that if we are
universalizing in Hare‟s way, we would prefer world A, on the grounds that if I were to
live the lives of all those in world A, my total life would be slightly preferable to my total
life if I were to live the lives of all those in world B? If someone prefers B, can this
egalitarian (or perhaps prioritarian) preference be treated as merely that person‟s
preference, to be given no more weight than any other preference? Perhaps it can, but
any such attempt will face the familiar objection that it fails to take seriously “the
separateness of persons,” so the success of argument for maximizing overall welfare will
depend on its ability to rebut that objection. I will not attempt that rebuttal here. There
may also be other arguments against maximizing to consider, so this paper will leave
open the question of whether Hare‟s argument leads to a pure maximizing utilitarianism,
or is compatible with a form of welfarism that accepts some distributive constraints on
maximization.
An Objective Morality?
Is Hare, as I have presented him, trying to have it both ways on objective moral truth? To
some it has seemed that he rejects the possibility of objective truth in ethics while at the
same time developing a line of argument for the conclusion that the only objectively true
ethical view is a form of utilitarianism. No matter how strong Hare‟s argument may
seem, that can‟t be right.
Hare‟s response to this charge was to insist on the distinction between a formal constraint
on what can count as a moral judgment, and a substantive ethical principle. Moral
judgments are prescriptions, and the notion of an objective prescription was, he
sometimes said, one he did not understand. So objective moral judgments remain out.
Nevertheless, if we make moral judgments, we make them under the constraint of
universalizability, and this should lead us to agreed conclusions about what we are able to
prescribe morally. Universalizability, properly understood, will lead us to agree on what
we are willing to prescribe – as long as we also agee on all the relevant facts. These facts
14
includes the facts about the preferences of everyone affected by our acts, and that
includes our own preferences.
Does this agreement on what we are willing to prescribe make the moral judgment on
which we agree objectively true? Not really. We are stll each uttering prescriptions. The
point is that these are the only universalizable prescriptions to which beings like us, in
these circumstances, can assent. But a tiny element of choice still remains. To see this
tiny element writ large, imagine that I am a utility monster and my preferences are
billions of times as intense as anyone else‟s preferences. Then my preferences would
count much more heavily in the summation of all preferences – not because they were
mine, but because they are so intense. Since I can to some extent choose what I prefer –
as I could choose right now if I were asked if I prefer to continue working on this paper
or to go for a walk – if I were a utility monster I could, by making such a choice, be able
to reach a different conclusion about what ought to be done. In that sense, prior to my
decision about what I prefer, there is no objectively true answer to the question “what
ought I to do?” But if this is true for a utility monster it is also true, in a greatly
attenuated sense, for anyone able to decide what to prefer, because everyone must include
his or her own preferences in the total to be weighed up before deciding what to do. In
theory, my own decision about what I prefer could change the judgment about what
ought to be done. In reality, since I am not a utility monster, in any decision involving a
large number of people, the likelihood of my own choice about what I prefer playing such
a pivotal role is small.
The Status of Universalizability
Hare‟s argument rests on the claim that when we think about what we ought, morally, to
do, we can only accept prescriptions we can universalize. What is the status of this
claim? If, as Hare consistently argued, it is a feature of moral language, it remains open
to use to refrain from using moral language. Freed of the need to universalize, we can,
without inconsistency or irrationality, decide what to do in ways that give more weight to
our own preferences (including our ideals) than we give to the preferences of others.
Whether we will suffer any personal loss by turning away from morality as a guide to
15
how to live is unclear. Hare argues that to bring up our children in this way – as
amoralists – would not be in their interests. He is less clear whether it would be in the
interests of an adult to live as an amoralist.18 And in any case, the considerations he
offers against bringing up your children as amoralists assume that the choice is either to
universalize fully, in his sense, or to live as an amoralist, in the everyday sense of that
term. He does not consider whether it would be contrary to the interests of your children
to bring them up to adhere to a set of general principles, like honesty, truthfulness,
helping and not harming others, and so on, while also acting on some ideals whether
about strip shows or about preserving endangered species, that one follows merely for the
– admittedly non-universalizable – reason that they are, say, “the ideals our family
holds.”
Is that all there is to be said? It may seem to be a weakness of Hare‟s approach that it has
the implication that the person who does not universalize and does not use moral
language is not guilty of a more fundamental mistake. Can we not even say that such a
person is acting contrary to the demands of reason? If not, is this not a major weakness
of Hare‟s view, and perhaps of any view that seeks to reach a normative conclusion by
relying on an analysis of moral language?
If it is a weakness, it should not be linked with Hare‟s linguistic approach to philosophy,
for it is shared by the greatest of the nineteenth century utilitarians, Henry Sidgwick, who
was no linguistic philosopher. Sidgwick held that: “There are certain absolute practical
principles, the truth of which, when they are explicitly stated, is manifest…” Among
these principles is a version of universalizability that he puts like this: “whatever action
any of us judges to be right for himself, he implicitly judges to be right for all similar
persons in similar circumstances.”19 Combining this principle with a notion of what is an
individual‟s “good on the whole,” Sidgwick arrives at a further “self-evident principle”
that he refers to as the “axiom of rational benevolence”:
18
I discuss this question in How Are We to Live? Prometheus, Buffalo, NY, 1995.
19
Henry Sidgwick, The Methods of Ethics, 7th edition, p.379. The reference to this as
“the Kantian maxim” is from p. xvii.
16
… the good of any one individual is of no more importance, from the point of
view (if I may say so) of the Universe, than the good of any other; unless, that is,
there are special grounds for believing that more good is likely to be realised in
the one case than in the other.
Sidgwick notes that this axiom is “required as a rational basis for the Utilitarian system”
Like Hare‟s argument, it leads to a utilitarian conclusion. But Sidgwick appears to go
further than Hare when he adds:
And it is evident to me that as a rational being I am bound to aim at good
generally,--so far as it is attainable by my efforts,--not merely at a particular part
of it.20
In the concluding chapter of The Methods of Ethics, however, it becomes clear that this is
not Sidgwick‟s final word, for he acknowledges that the process by which one might
convince an egoist to take the universal good, rather than his own good, as the ultimate
end of conduct, “requires that the Egoist should affirm, implicitly or explicitly, that his
own greatest happiness is not merely the rational ultimate end for himself, but a part of
Universal Good: and he may avoid the proof of Utilitarianism by declining to affirm
this.”21 In other words just as Hare‟s amoralist can escape universalizability by declining
to use moral language, so Sidgwick‟s egoist can avoid the necessity of admitting that he
should aim at universal good by insisting that his own good is merely his own rational
ultimate end, and not a part of good on the whole. If the egoist takes his stand there, he
cannot be shown to be making a mistake:
It would be contrary to Common Sense to deny that the distinction between any
one individual and any other is real and fundamental, and that consequently "I"
am concerned with the quality of my existence as an individual in a sense,
fundamentally important, in which I am not concerned with the quality of the
existence of other individuals: and this being so, I do not see how it can be proved
20
Henry Sidgwick, The Methods of Ethics, 7th edition, p.382.
21
ME 497-8.
17
that this distinction is not to be taken as fundamental in determining the ultimate
end of rational action for an individual.
This leads Sidgwick to his pessimistic conclusion, which he stated most forcefully at the
end of the first edition of The Methods of Ethics (although his views on this matter were
more circumspectly stated in later editions of The Methods, they remained substantially
unaltered):
Hence the whole system of our beliefs as to the intrinsic reasonableness of
conduct must fall, without a hypothesis unverifiable by experience reconciling the
Individual with the Universal Reason, without a belief, in some form or other, that
the moral order which we see imperfectly realized in this actual world is yet
actually perfect. If we reject this belief, we may perhaps still find in the non-moral
universe an adequate object for the Speculative Reason, capable of being in some
sense ultimately understood. But the Cosmos of Duty is thus really reduced to a
Chaos: and the prolonged effort of the human intellect to frame a perfect ideal of
rational conduct is seen to have been fore-doomed to inevitable failure.22
So when it comes to demonstrating the irrationality of amoralism, Sidgwick and Hare are
at one. Fervently as we may wish that it were possible to say that the amoralist is making
some kind of mistake, this peak remains unconquered, and not only by utilitarians. In the
most widely read and admired sections of The Groundwork of the Metaphysic of Morals
– that is, the first two sections - Kant comes no closer to achieving it than Hare. He
argues merely that “if duty is not to be a vain delusion and a chimerical notion,” then
“simple conformity to law in general” must motivate our will.23 It is only in the third
section that Kant offers an argument to show that reason can motivate our will, that the
categorical imperative is rationally required, and hence that duty is not a vain delusion.
Yet today this is the least read and discussed section of The Groundwork – whether
because contemporary Kantians regard it as a failure, or because they do not wish to
22
The Methods of Ethics, 1st ed, p.473.
23
Immanuel Kant, Groundwork of the Metaphysic of Morals, trans T.K. Abbott, sec. I
(first published 1785).
18
commit themselves to its metaphysical presuppositions, I am not sure. Recent writers in
the broadly Kantian tradition, such as John Rawls and T.M. Scanlon, face the same
problem as Hare: they attempt to show that their normative conclusions follow for those
who are prepared to act only on those principles that they would choose from behind a
“veil of ignorance,” or those that others cannot reasonably reject, but in my view they fail
to show that we would be irrational if we chose not to so limit the principles on which we
act.24 The difficulty of putting ethics on a rational foundation is not a problem specific to
utilitarianism: it is a problem that no philosopher of any persuasion has been able to
resolve.
24
See John Rawls, A Theory of Justice, Harvard University Press, Cambridge, Mass.,
1971 and T.M. Scanlon, What We Owe To Each Other, Harvard University Press,
Cambridge, Mass., 2000. For one specific way in which we may rationally reject the
conditions on our choice of moral principles suggested by Rawls and Scanlon (as well as
by Brad Hooker and Bernard Gert) see Katarzyna de Lazari-Radek and Peter Singer,
“Secrecy in Consequentialism: A Defence of Esoteric Morality,” Ratio, forthcoming,
March 2010. For the reasons Sidgwick offers when considering the egoist, more far-
reaching rejections would also seem rationally defensible.