This article was downloaded by: [University of California, San Diego]

On: 01 January 2015, At: 16:42

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number:
1072954 Registered office: Mortimer House, 37-41 Mortimer Street,
London W1T 3JH, UK

Australasian Journal of
Philosophy
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/rajp20

Moral realism and the

sceptical arguments from
disagreement and queerness
a
David O. Brink
a
Cornell University
Published online: 02 Jun 2006.

To cite this article: David O. Brink (1984) Moral realism and the sceptical
arguments from disagreement and queerness, Australasian Journal of Philosophy,
62:2, 111-125, DOI: 10.1080/00048408412341311

To link to this article: http://dx.doi.org/10.1080/00048408412341311

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all
the information (the “Content”) contained in the publications on our
platform. However, Taylor & Francis, our agents, and our licensors
make no representations or warranties whatsoever as to the accuracy,
completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of
the authors, and are not the views of or endorsed by Taylor & Francis.
The accuracy of the Content should not be relied upon and should be
independently verified with primary sources of information. Taylor and
Francis shall not be liable for any losses, actions, claims, proceedings,
demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in
relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,
reselling, loan, sub-licensing, systematic supply, or distribution in any
 form to anyone is expressly forbidden. Terms & Conditions of access
and use can be found at http://www.tandfonline.com/page/terms-and-
conditions
Downloaded by [University of California, San Diego] at 16:42 01 January 2015
 Australasian Journal o f Philosophy
Vol. 62, No. 2; June 1984

MORAL R E A L I S M A N D T H E S C E P T I C A L A R G U M E N T S FROM
D I S A G R E E M E N T A N D QUEERNESS

David O. Brink
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

1. Introduction
The most important kind of challenge to moral realism or moral objectivism
argues that there is a special problem with realism in ethics. I shall defend
moral realism against two influential versions of this challenge recently
formulated by J. L. Mackie in his book Ethics." I n v e n t i n g R i g h t a n d Wrong. 1
According to standards o f argument which Mackie himself sets, neither his
argument from disagreement nor his argument from queerness shows any
special problem for moral realism. Let me explain why.
Moral realism is best explained as a special case of a global realist thesis.
The general thesis c o m m o n to realist claims about a variety of disciplines
is a two part metaphysical claim:
R: (a) there are facts of kind x, and
(b) these facts are logically independent of our evidence, i.e. those beliefs
which are our evidence, for them. 2
Moral realism is then obtained by substituting 'moral' for the variable 'x'.
MR: (a) there are moral facts, and
(b) these facts are logically independent of our evidence, i.e. those
beliefs which are our evidence, for them.
Moral realism claims that there are objective moral facts and implies that
there are true moral propositions.
Moral scepticism is technically an epistemological doctrine and so is
officially neutral with respect to the metaethical thesis of moral realism. Moral
scepticism claims that we have no moral knowledge and this claim is compat-
ible with the existence of objective moral facts and true moral propositions.
But while moral realism and moral scepticism are compatible (we may just
have no cognitive access to moral facts), the standard and most plausible
reason for claiming that we have no moral knowledge is the belief that there
are no moral facts. This must be why Mackie construes moral scepticism
J. L. Mackie, Ethics: Inventing Right and Wrong (New York: Penguin Books, 1977)(here-
inafter E). Mackie further discusses a number of features of these two arguments in Hume's
Moral Theory (Boston: Routledge& Kegan Paul, 1980)(hereinafter HMT) and The Miracle
of Theism (New York: Oxford UniversityPress, 1982)(hereinafter MT). Parenthetical refer-
ences in the text to E, HMT, or MT are to pages in these books.
z Cf. Michael Devitt, 'Dummett's Anti-Realism'Journal of Philosophy 80 (1983), pp. 75-6.
For obvious reasons, the kind of dependence asserted in R is logical, not causal.
111
 112 Moral Realism and the Sceptical Arguments

as an anti-realist thesis. I shall follow Mackie in this and treat moral scepticism
as a denial o f the existence o f objective v a l u e s )
There are two basic kinds o f m o r a l sceptic. The first kind applies general
sceptical considerations to the special case o f morality. O n his view, there
are no m o r a l facts, but neither are there any other objective facts a b o u t the
world. O f course, this first sort o f sceptic is quite radical and has not been
terribly influential as a source o f moral scepticism. The second kind o f moral
sceptic claims that there is a special problem about realism in ethics, a problem
which does n o t afflict realism a b o u t most other disciplines. This clearly has
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

been the m o r e popular and philosophically influential version o f moral scep-

ticism. Mackie is this second kind o f m o r a l sceptic.
As this second kind o f m o r a l sceptic, Mackie complains that belief in the
existence o f objective values is no part o f a plausible realist world-view.
(E: p. 17) Mackie's sceptical arguments, therefore, cannot turn on the appli-
cation o f general sceptical considerations. I f it can be shown that the moral
realist's metaphysical and epistemological commitments are no less plausible
than those of, say, the physical realist, then Mackie's sceptical arguments
will have been answered.
A l t h o u g h it is possible to d e f e n d m o r a l realism against sceptical arguments
without establishing a n y kind o f p r e s u m p t i o n in its favour, there are, as
Mackie recognises, general considerations which require the moral sceptic
to bear a certain b u r d e n o f p r o o f . First, this second version o f m o r a l scepti-
cism concedes a p r e s u m p t i o n in f a v o u r o f m o r a l realism. If, as this second
sort o f scepticism assumes, realism is plausible a b o u t a wide range o f disci-
plines, then there must be some special justification for taking a different
view about the existence and nature o f moral facts. O f course, this establishes
only a very weak presumption in favour o f moral realism, but it is one which
the m o r a l sceptic must rebut.
M o r e o v e r , belief in m o r a l realism is supported by certain features o f our
m o r a l practice. In m o r a l deliberation and m o r a l a r g u m e n t we search for
answers to ot/r moral questions, answers whose correctness we assume to
be independent o f our means o f arriving at them. 4 O f course, this presumption
t o o is defeasible, but this takes some argument. As Mackie claims, moral
scepticism must have the status o f an error theory; it must explain h o w and
why our commitment to the objectivity o f moral values is mistaken. (E: p. 35)
Mackie distinguishes two arguments for the second version o f moral scepti-
cism. The first turns on the a p p a r e n t unresolvability o f m a n y m o r a l disputes
and so is best thought o f as an argument f r o m disagreement, 5 while the second

3 The simple denial of the existence of moral facts, accompanied by no positive account of
the nature of moral values, is moral nihilism. But Mackie, like most moral sceptics, not only
denies the existence of objective values but also adopts a constructivist or subjectivist position
about the nature of value according to-which we make or choose moral value.
4 See Thomas Nagel, 'The Limits of Objectivity' in S. McMurrin (ed), The Tanner Lectures
on Human Values/(Salt Lake City: University of Utah Press, 1980), p. 100. Cf. E: pp. 35,
48-9; HMT: pp. 34, 70-5, 136.
5 Mackie himself refers to this argument as 'the argument from relativity'. (E: p. 36) But this
label is at least misleading and on a natural reading of 'relativity' begs the question, since
 David O. Brink 113

turns on the mysterious character objective values would seem to have to

have and so represents an argument from queerness. Mackie presses both
of these arguments against moral realism and in favour of moral scepticism
and subjectivism. 6 In what follows, I shall examine and rebut Mackie's argu-
ments from disagreement and queerness; I shall argue that neither argument
establishes any special problem for moral realism. Although these two argu-
ments may not exhaust the arguments for the second version of moral scepti-
cism, they are sufficiently important both historically and philosophically that
successfully rebutting them will go a long way towards defending moral
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

realism.

2. Moral Objectivity
Before discussing the details o f the arguments f r o m disagreement and queer-
ness, we need to establish just which version of moral realism is or need be
in question. In section 1 I described moral realism as the metaethical view
that there are obj ective moral facts. However, in pressing the arguments from
disagreement and queerness, Mackie employs a stronger or more committal
version of moral realism according to which not only are there moral facts
but also these moral facts are objectively prescriptive. (E: pp. 23, 26-7, 29,
40, 42; H M T : pp. 22, 53, 55, 134, 146; MT." pp. 102, 104, 115-16) Indeed,
although both the argument from disagreement and the argument f r o m
queerness apply to my formulation of moral realism, some of the special
appeal of the argument from queerness derives from the assumption that
moral facts would have to be objectively prescriptive. ( E : p p . 40-1;
HMT: p. 61)
In claiming that moral facts would have to be objectively prescriptive,
Mackie is claiming that moral realism requires the truth of internalism.
Internalism is the apriori thesis that the recognition of moral facts itself either
necessarily motivates or necessarily provides reasons for action. Internalism
is an apriori thesis, because its proponents claim that the recognition of moral
facts necessarily motivates or provides reasons for action no matter what
the moral facts turn out to be. We can distinguish motivational internalism
(MI) and reasons internalism (RI): MI holds that it is apriori that the recog-
nition of moral facts itself necessarily motivates the agent to perform the
moral action, while RI claims that it is apriori that the recognition of moral
facts itself necessarily provides the agent with reason to perform the moral
action. Externalism, by contrast, denies both MI and RI. 7

the argument in question alleges that moral relativity (or at least the denial of moral realism)
is the best explanation of the facts about moral disagreement.
6 There is some question about whether Mackie's rejection of moral realism in chapter 1 of
part I of E is consistent with his defence in part I l of a mixed deontological/consequentialist
ethical theory based on considerations of rational self-interest, l shall not pursue this issue
here; it is the main topic of Jonathan Harrison, "Mackie's Moral 'Scepticism'" Philosophy
57 (1982), pp. 173-191.
7 The apriori character of the internalist thesis and the M I / R I distinction are often obscured.
Cf. W. D. Falk, "'Ought' and Motivation" reprinted in W. Sellars and J. Hospers (eds),
Readingsin Ethical Theory (New York: Appleton-Century-Crofts, 1952); William Frankena,
 114 Moral Realism and the Sceptical Arguments

Although Mackie is unclear as between MI and RI, 8 he clearly thinks that

some version o f internalism is required by moral realism. Both MI and RI
make exceptionally strong claims. M I claims t h a t - - w h a t e v e r the moral facts
turn out to be and regardless of the psychological make-up of the a g e n t - the
mere recognition of a moral fact necessarily provides some motivation to
perform the moral action, while RI claims t h a t - w h a t e v e r the moral facts
turn out to be and regardless of the agent's interests or d e s i r e s - t h e mere
recognition of a moral fact necessarily provides the agent with at least some
reason to p e r f o r m the moral action. These claims are quite implausible, and
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

it is unclear why moral realism is committed to either of them.

It is unlikely that the recognition of moral facts necessarily motivates or
provides reasons for action; it is very unlikely that the recognition of moral
facts alone necessarily motivates or provides reasons for action; and the mere
recognition of moral facts almost certainly does not necessarily motivate or
provide reasons for action regardless o f what the moral facts turn out to be.
Whether the recognition of moral facts motivates certainly depends upon
what the moral facts are, and, at least on most plausible moral theories,
whether recognition of these facts motivates is a matter of contingent (even
if deep) psychological fact about the agent. Whether the recognition of moral
facts provides reasons for action depends upon whether the agent has reason
to do what morality requires. But this, of course, depends upon what morality
requires, i.e. upon what the moral facts are, and, at least on standard theories
o f reasons for action, whether recognition of these facts provides reason for
action will depend upon contingent (even if deep) facts about the agent's
desires or interests. So, internalism is false; it is not something which we can
know apriori, i.e. whatever the moral facts turn out to be, that the recognition
o f moral facts alone either necessarily motivates or necessarily provides
reasons for action. 9
It is hard to see why moral realism should be committed to the truth of
internalism. Mackie claims both that moral realists have traditionally been
internalists (E: p. 23) and that internalism is part of c o m m o n sense moral
thinking. (E: p. 35) But both claims seem false and would carry relatively
little weight, even if true. Once we make clear the strength of the internalist
c l a i m - t h a t we know apriori that mere recognition o f moral facts necessarily
motivates or provides reasons for a c t i o n - i t is less clear that the tradition
of moral realism is a tradition of internalism. In particular, although I cannot
argue the claims here, I doubt that Plato, H u m e , or Sidgwick is, as Mackie
claims, an internalist. And, of course, even if m a n y moral philosophers have
thought that internalism is true, it would not follow that they were right.
Nor does common sense moral thinking seem to support belief in internalism;

'Obligationand Motivation in Recent Moral Philosophy'in A. I. Melden (ed), Essaysin Moral

Philosophy (Seattle: Universityof Washington Press, 1958); and Thomas Nagel, The Possibility
of Altruism (Princeton: Princeton University Press, 1970), pp. 7-12. Cf. MT: p. 115.
8 E."pp. 23, 40, 49; HMT: pp. 22, 53: and MT: p. 102 require MI, while MT: p. 115 explicitly
requires RI.
9 Cf. William Frankena, Ethics (Englewood Cliffs, NJ: Prentice-Hall, 1973), pp. 114-16.
 David O. Brink 115

in fact, it seems extremely unlikely that any belief so recherche could be part
of common sense moral thinking. Even if belief in internalism were part of
common sense moral thinking, it would be revisable, especially if it could
be shown that belief in internalism plays a social role such that it would persist
even if mistaken.
So no good reason has been produced for thinking that internalism is true
or for thinking that moral realism requires internalism. This means that the
moral realist can defend externalism. In particular, determination of the
motivational and reason-giving power of moral facts will have to await
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

specifications of the moral facts and of the desires and interests of agents.
In defending moral realism against the arguments from disagreement and
queerness, t will offer what I call a functionalist theory of moral value
according to which moral facts are facts about human well-being and flourish-
ing as a model specification o f moral realist claims. This account illustrates
the kind of justification of morality which the externalist can provide, for
this functionalist theory implies that moral facts will as a matter o f fact at
least typically provide agents with reasons to do the morally correct thing.

3. The Argument from Disagreement

Mackie claims that the best explanation of inter- and intra-societal ethical
disagreement is that there simply are no moral facts, only differences o f atti-
tude, commitment, or decision. (E: pp. 36-7) Of course, disagreement does
not entail scepticism. Mackie recognises that we do not infer from the fact
that there are disagreements in the natural sciences that the natural sciences
are not objective disciplines. Nor do we make what might appear to be the
more modest inference from the fact that there is a specific dispute in some
subject that there is no fact o f the matter on the particular issue in question.
For example, no one concluded from the apparently quite deep disagreement
among astronomers a short while ago about the existence o f black holes that
there was no fact of the matter concerning the existence of black holes.
Mackie's claim is that disagreement in ethics is somehow more fundamental
than disagreement in other disciplines. In particular, realism about a discipline
requires that its disputes be resolvable at least in principle, and, while most
scientific disputes do seem resolvable, many moral disputes do not.
Mackie imagines the moral realist replying that moral disputes are resolv-
able, because deep moral disagreements are not really cases of disagreement.
Rather, they are cases in which 'disputants' apply antecedently shared moral
principles under different empirical conditions. (E: p. 37) The resulting moral
judgments are about different action types, so the 'disagreements' in question
are really only apparent.
Mackie issues two rejoinders to this realist reply. His first rejoinder is that
this realist response commits the realist to (a) claiming that necessity can only
attach to general moral principles and (b) accepting the following counter-
factual: ' . . . if things had been otherwise, quite different sorts of actions
 116 Moral Realism and the Sceptical Arguments

would have been right'. (E: p. 37) (a) and (b), Mackie claims, imply that many
action types will be right or wrong only contingently.
Although this rejoinder does raise some interesting questions about the
modal status of moral facts, it in no way threatens moral realism. First,
certainly some moral facts are contingent, and, even if this realist reply
requires the contingency of some moral facts, this shows nothing about how
many moral facts the realist must regard as contingent. But, secondly and
more importantly, Mackie's modal issue is a red herring. The truth of moral
realism turns on the existence of moral facts, not their modal status.
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

Mackie's second rejoinder to the realist reply is simply that some moral
disputes are real disputes. Not all putative moral disagreements can be ex-
plained away as the application of antecedently shared moral principles in
different circumstances. (E: p. 38)
Mackie is right that many moral disputes are genuine, and, if the realist
had no account of these disputes, Mackie would have a strong argument
against moral realism. But the realist can account for moral disputes.
As we have seen, not every apparent moral disagreement is a genuine dis-
pute. But the realist need not maintain even that all genuine moral disputes
are resolvable. He can maintain that some moral disputes have no uniquely
correct answers. Moral ties are possible, and considerations, each of which
is objectively valuable, may be incommensurable. 10 So the moral realist need
only maintain that most genuine moral disputes are resolvable.ll
Indeed, the realist can plausibly maintain that most genuine moral disputes
are in principle resolvable. Mackie's discussion of the realist's reply shows
that Mackie thinks moral disagreement is resolvable if and only if antecedent
agreement on general moral principles obtains. This claim presupposes a
one-way view of moral justification and argument according to which moral
principles justify particular moral judgments but not vice versa. However,
this view of moral justification is defective. As Goodman, Rawls, and other
coherentists have argued, justification proceeds both from general principles
to particular cases and from particular cases to general principles. 12 Just as
agreement about general moral principles may be exploited to resolve dis-
agreement about particular moral cases, so agreement about particular moral
cases may be exploited to resolve disagreement about general moral principles.
Ideally, trade-offs among the various levels of generality of belief will be made
in such a way as to maximise initial commitment, overall consistency, ex-
planatory power, etc. A coherentist model of moral reasoning of this sort

lo Of course, the mere absenceof a single fact or set of facts in virtue of which both considerations
are valuable does not establish incommensurability.
li A realist could maintain that most or even all genuine moral disputes are unresolvable, as
long as he was willingto claim that moral ties and incommensurablevalues occurredfrequently
enough. Althoughthese claimsare compatiblewith his position, reliance on them wouldweaken
his reply to the argument from disagreement.
~2See, e.g., Nelson Goodman, Fact, Fiction, and Forecast (Indianapolis: Hackett, 1979), p. 66;
John Rawls, A Theory of Justice (Cambridge, MA: Harvard UniversityPress, 1971), pp. 20,
46-51; RolfSartorius, Individual Conduct and Social Norms (Encino, CA: Dickenson, 1975),
pp. 31-3;and Norman Daniels, 'Wide ReflectiveEquilibrium and TheoryAcceptancein Ethics'
Journal of Philosophy 76 (1979), pp. 256-282.
 David O. Brink 117

makes it much less plausible that disagreements over moral principles are
in principle unresolvable. 13
Moreover, a great many moral disagreements depend upon disagreements
over the non-moral facts. First, many disagreements over the non-moral facts
result from culpable forms o f ignorance of fact. Often, for moral or non-
moral reasons, at least one disputant culpably fails to assess the non-moral
facts correctly by being insufficiently imaginative in weighing the consequences
for the relevant people of alternative actions or policies. This sort of error
is especially important in moral disputes, since thought experiments (as
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

opposed to actual tests) play such an important part in the assessment of

moral theories. Thought experiments play a larger role in moral methodology
than they do in scientific methodology, at least partly because it is often
(correctly) regarded as immoral to assess moral theories by realising the
relevant counterfactuals.
Secondly, many moral disagreements result from reasonable but nonethe-
less resolvable disagreements over the non-moral facts. The correct answers
to moot moral questions often turn on certain non-moral facts about which
reasonable disagreement is possible and which may in fact be known by no
one. Correct answers to moral questions can turn at least in part upon correct
answers to non-moral questions such as 'What (re)distribution of a certain
class of goods would make the worst-off representative person in a particular
society best-off?', 'Would public ownership of the means o f production in
the United States lead to an increase or decrease in the average standard of
living?', 'What is the correct theory of human personality?', and 'What kind
of life would my severely mentally retarded child lead (if I brought the preg-
nancy to term and raised the child), and how would caring for him affect
my family and me?'. However difficult and controversial these questions are,
the issues which they raise are in principle resolvable. Moral disputes com-
monly do turn on disagreement over issues such as these, and, insofar as
they do, moral disputes are clearly resolvable in principle.
Mackie argues that if moral realism were true, all moral disputes should
be resolvable, and s i n c e m a n y seem irresolvable, he concludes that moral
realism is false. But the moral realist need only claim that most genuine moral
disputes are in principle resolvable. Not all apparent moral disagreements
are genuine, because some apparent moral disputes merely reflect the appli-
cation of antecedently shared moral principles under different circumstances.
Not every genuine moral dispute need be even in principle resolvable, since
moral ties are possible and some objective moral values may be incommen-
surable. Of those genuine moral disputes which the realist is committed to
treating as in principle resolvable, some depend upon antecedent disagreement
over moral principles, while others depend upon disagreement over the non-
moral facts. The realist can claim that antecedent disagreement over moral
principles is in principle resolvable by coherence arguments and that dis-

13Although a coherentist theory of moral truth would be incompatible with moral realism, this
part of the realist's reply requires only a coherence~':theory of moral justification.
 118 Moral Realism and the Sceptical Arguments
agreement over the non-moral facts is always in principle resolvable. 14 The
moral realist gives a plausible enough account o f moral disagreement for us
to say that Mackie has not shouldered the burden of p r o o f for his claim that
the falsity of moral realism is the best explanation of the nature of moral
disagreement.

4. The Argument from Queerness

The rough idea behind the argument f r o m queerness is that objective moral
facts and properties would have to be so different f r o m the sort of natural
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

facts and properties for which we do have evidence that we have good a
posteriori reason to reject moral realism. 15 (E: pp. 38-42; M T : pp. 115-16)
As I said in section 2, the argument f r o m queerness is supposed to tell es-
pecially against the existence of moral facts conceived of as being objectively
prescriptive. (E: pp. 40-1; HMT: p. 61) I claimed that in committing realism
to objective prescriptivity Mackie is claiming that moral realism requires
internalism. But I argued that internalism is implausible and that Mackie
produces no good reason for committing realism to internalism. Instead, the
realist can defend externalism; determination of whether agents have reason
or motive to be moral will depend upon the content of morality and facts
about agents. In explaining why objective values are not queer, I will offer
a model specification of moral realism, which, together with plausible em-
pirical assumptions, implies that agents generally do have reasons to be moral.
There are two limbs to the argument f r o m queerness: one metaphysical,
one epistemological. (E: p. 38) I turn to the metaphysical branch of the
argument first. Mackie thinks that moral realism is a metaphysically queer
doctrine, because he believes that moral facts or properties would have to
be ontologically simple or independent. (E: p. 38) The assumption is that
moral properties would have to be sui generis, that is, ontologically indepen-
dent of natural properties with which we are familiar. Although it is not
inconceivable that there should be sui generis moral properties, we have very
good a posteriori evidence for the truth of materialism and for the falsity
of ontological pluralism.
However, Mackie's crucial assumption that moral facts and properties
would have to be suigeneris is false; moral realism does not require ontological
pluralism. The moral realist has at least two options on the assumption that
materialism is true: he can claim that moral properties are identical with
certain physical properties, or he can claim that moral properties supervene
upon certain physical properties. Because moral properties and their instances
could be realised in non-physical as well as a variety of physical ways, neither
moral properties nor their instances should be identified with physical proper-
ties or their instances. 16 For this reason, it is best for the moral realist to
14Actually, the argument from disagreement presupposes that disagreement over non-moral
facts is always in principle resolvable.
15As Mackie himself observes (MT: p. 116), the queerness argument is a posteriori and not
apriori as R. M. Hare, Moral Thinking (NewYork: Oxford UniversityPress, 1981), pp. 82-6,
insists.
16Cf. Saul Kripke, 'Identity and Necessity'reprinted in S. Schwartz (ed), Naming, Necessity,
 David O. Brink 119

claim that m o r a l properties supervene u p o n physical properties.

Mackie recognises the realist's claim about the supervenience o f moral facts
and properties on physical facts and properties but claims that the alleged
supervenient relation is also metaphysically queer:
Another way o f bringing out this queerness is to ask about anything that
is supposed to have some objective m o r a l quality, how this is linked with
its natural features. W h a t is the connection between the natural fact that
an action is a case o f deliberate c r u e l t y - say, causing pain just for f u n -
and the m o r a l fact that it is wrong? It c a n n o t be an entailment, a logical
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

or semantic necessity. Yet it is n o t merely that the two features occur

together. The wrongness must s o m e h o w be 'consequential' or 'super-
venient'; it is w r o n g because it is a piece o f deliberate cruelty. But just
what in the world, is signified by this 'because'? (E: p. 4 t )
Although I do not think that Mackie has really motivated a metaphysical
worry about .moral supervenience, I shall defend m o r a l realism against the
charge o f metaphysical queerness by a d o p t i n g the strategy which Mackie
mentions o f finding partners in g u i l t - a l t h o u g h once it is clear what sort o f
company the realist is keeping it would only be perverse to regard them as
partners in guilt. ~7 1 shall argue that the supervenient relation which the realist
claims obtains between mo~'al properties and natural or physical properties
is neither u n c o m m o n nor mysterious.
Although it is an interesting question what the precise relation is between
property identity and supervenience, it is fairly clear that one property can
supervene u p o n another without those two properties being identical. TM A
supervenient relation obtains between two properties or sets o f properties
just in case the one p r o p e r t y or set o f properties is causally realised by the
other property or set o f properties; the former p r o p e r t y or set o f properties
is the supervening property or set o f properties, and the latter property or
set of properties is the base p r o p e r t y or set o f properties. Supervenience
implies that no change can occur in the supervening property without a change
occurring in the base property, but it also asserts a claim o f ontological
dependence. 19 Assuming, as M~ckie does, that materialism is true, all proper-

and Natural Kinds (Ithaca, NY: Cornell University Press, 1977), pp. 76, 98-9 and Naming
and Necessity (Cambridge, MA: Harvard University Press, 1980), p. 148; and Richard Boyd,
'Materialism without Reductionism: What Physicalism Does Not Entail' in N. Block (ed),
Readings in Philosophy of Psychology I (Cambridge, MA: Harvard University Press, 1980).
17Also, recall that Mackie advocates a selective and not a general kind of scepticism.
18Cf. Boyd, 'Materialism without Reductionism: What Physicalism Does Not Entail'; and
Jaegwon Kim, 'Causality, Identity, and Supervenience in the Mind-Body Problem' Midwest
Studies in Philosophy IV (1979), pp. 31-49.
19See Henry Sidgwick, The Methods of Ethics, seventh edition (New York: Macmillan, 1907),
pp. 379-80; G. E. Moore, Philosophical Studies (Boston: Routledge & Kegan Paul, 1922),
pp. 260-l; C. D. Broad, Five Types of Ethical Theory (Boston: Routledge & Kegan Paul,
1930), p. 223; W. D. Ross, The Right and the Good (New York: Oxford University Press,
1930), pp. 79, 89, 115, 155; Jaegwon Kim, 'Causality, Identity, and Supervenience in the
Mind-BodyProblem' and 'Supervenienceand NomologicalIncommensurables'American Philo-
sophical Quarterly 15 (1978), pp. 149-156;and Donald Davidson, 'Mental Events' reprinted
in Donald Davidson, Essays on Actions and Events (New York: Oxford University Press,
 120 Moral Realism and the Sceptical Arguments

ties u l t i m a t e l y supervene o n m a t e r i a l or physical base properties. 2° Physical

properties are basic t h e n i n the sense that all other properties are n o t h i n g
over a n d a b o v e physical properties. Biological, social, psychological, and
m o r a l properties are all realised physically; they are simply different kinds
o f c o m b i n a t i o n s a n d a r r a n g e m e n t s o f m a t t e r which h a n g together
explanatorily. 21
Supervenience is a r e l a t i o n o f causal c o n s t i t u t i o n or dependence. There
is n o t h i n g strange a n d certainly n o t h i n g u n i q u e a b o u t the supervenience of
moral properties o n physical properties. A s s u m i n g materialism is true, mental
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

states supervene o n physical states, yet few t h i n k that m e n t a l states are meta-
physically queer (and those t h a t do do n o t t h i n k that supervenience makes
t h e m queer). Social facts such as u n e m p l o y m e n t , inflation, a n d exploitation
supervene u p o n physical facts, yet n o one supposes that social facts are meta-
physically queer. Biological states such as being a n o r g a n i s m supervene on
physical states, yet no one supposes that organisms are queer entities. Macro-
scopic material objects such as tables supervene o n micro-scopic physical
particles, yet n o one supposes that tables are queer entities. I n short, it is
difficult to see h o w the realist's use o f supervenience in e x p l a i n i n g the re-
l a t i o n s h i p b e t w e e n m o r a l a n d physical properties makes his p o s i t i o n queer.
Moral properties are not ontologically simple or independent; b u t then neither
are m e n t a l states, social facts, biological states, or m a c r o - s c o p i c material

1980), p. 214. If, as these writers seem to suggest, the sufficiencyof a change in the supervening
property for a change in the base property were both a necessary and sufficient condition
for supervenience, then supervenience would be compatible with epiphenomenalism. (An
epiphenomenalist construal of supervenience might explain why Moore, Ross, and Broad,
who are all non-naturalists and think that moral properties are sui generis, nonetheless claim
that moral properties supervene on natural properties.) But the realist should construe super-
venienceas a relation of ontologicaldependenceand so should count the sufficiencyof a change
in the supervening property for a change in the base property as only a necessary and not
a sufficient condition for supervenience.
20Supervenience is a transitive relation.
21Although Mackie does not press Moore's open question argument or any of its cognatesagainst
moral realism, it is worth pointing out that property superveniencedoes not require any kind
of syntactic or linguistic reductionism. Just as property identity does not require property
predicate synonymy, so property superveniencedoes not require synonymy or meaning impli-
cations between supervening property predicates and base property predicates. For instance,
whether or not human pains supervene on C-fiber firings, the truth of this claim does not
depend upon whether there are synonymy relations or meaning implications between 'human
pain' and 'C-fiber firing'. Thus, although biological, social, psychological,and moral properties
all supervene on physical properties, biological, social, psychological, and moral terms need
not be definable in the language of particle physics. This explains how moral realism can be
true even if there are no reductive definitionsof moral terms. Cf. Richard Boyd, 'Materialism
without Reductionism: Non-Humean Causation and the Evidence for Physicalism'in Richard
Boyd, The Physical Basis of Mind (Cambridge, MA: Harvard UniversityPress, forthcoming);
Jerry Fodor, 'The Special Sciences, or the Disunity of Science as a Working Hypothesis'
reprinted in N. Block; Geoffrrey Hellman and Frank Thompson, 'Physicalism: Ontology,
Determination, and Reduction' Journal of Philosophy 72 (1975), pp. 551-564; and Hilary
Putnam, 'On Properties' reprinted in Hilary Putnam, Mathematics, Matter, andMethocl (New
York: Cambridge University Press, 1979), p. 312, Reason Truth, and History (New York:
Cambridge University Press, 1981), pp. 84-5,207, 'Possibility and Necessity', pp. 53-5, and
'Beyond Historicism', p. 291, both in Hilary Putnam, Realism and Reason (New York:
Cambridge University Press, 1983).
 David O. Brink 121

objects. It is unlikely that moral properties are identical with physical proper-
ties; moral properties could have been realised non-materially. But there is
every reason to believe that in the actual world moral properties, like other
natural properties, are realised materially.
This realist account of supervenience discharges any explanatory obligation
which the argument f r o m metaphysical queerness imposes. The details of
the way in which moral properties supervene upon other natural properties
are worked out differently by different moral theories. Determination of which
account of moral supervenience is best will depend upon determination of
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

which moral theory provides the best account o f all our beliefs, both moral
and non-moral. Although I obviously cannot do here what is needed to defend
a particular account of moral supervenience, I will now offer a m o d e l speci-
fication of the moral realist's metaphysical claims.
When trying to determine the way in which moral properties supervene
upon other natural properties, one might start by looking at plausible theories
about other kinds of properties. Functional theories provide plausible
accounts of a wide variety o f kinds of properties; the nature of biological,
psychological, social, and economic properties is profitably viewed in func-
tional terms. Consider functionalist theories of mind as an example. Although
functionalism is not without its critics, 2z it is fair to say that there are no
rival theories in the philosophy of mind today. 23 What is essential to any
particular mental state type, according to functionalism, is the causal role
which that mental state plays in the activities which are characteristic of the
organism as a whole. Mental states are identified and distinguished from other
mental states in terms of the causal relations which they bear to sensory inputs,
behavioural outputs, and other mental states. To take a hoary example,
functionalist theories of mind claim that pain is identified and distingished
from other mental states by virtue of its tendency to result from tissue damage,
to produce an injury-avoidance desire, and to issue in the appropriate injury-
avoidance behaviour. The physical states which realise this functional state
are the physical states upon which pain supervenes.
Similarly, the moral realist might claim that moral properties are functional
properties. He might claim that what is essential to moral properties is the
causal role which they play in the characteristic activities of h u m a n
organisms, z4 In particular, the realist might claim that moral properties are
2zSee, e.g., Ned Block, 'Troubles with Functionalism' reprinted in N. Block.
23The functionalist literature is quite extensive; I rely principally upon the following: David
Armstrong, A Materialist Theory of Mind (Boston: Routledge & Kegan Paul, 1968); Ned
Block, 'What is Functionalism?'in N. Block; Ned Blockand Jerry Fodor, 'What Psychological
States Are Not' reprinted in N. Block; Richard Boyd, 'Materialism without Reductionism:
What Physicalism Does Not Entail'; Austen Clark, Psychological Models and Neural
Mechanisms (New York: Oxford University Press, 1980); Jerry Fodor, 'Materialism' in
D. Rosenthal (ed), Materialism and the Mind-Body Problem (Englewood Cliffs, NJ:
Prentice-Hall, 1971); Gilbert Harman, Thought (Princeton: Princeton UniversityPress, 1973);
HilaryPutnam, Mind, Language, and Reality (NewYork: CambridgeUniversityPress, 1975),
chapters 14, 16, 18, 20-2; and K. V. Wilkes, Physicalism (Boston: Routledge & Kegan Paul,
1978).
24A functionalist moral realist might claim that moral properties are properties which play a
certain role in the activities which are characteristic of sentient organisms.
 122 Moral Realism and the Sceptical Arguments

those which bear upon the maintenance and flourishing of human organisms.
Maintenance and flourishing presumably consist in necessary conditions for
survival, other needs associated with basic well-being, wants of various sorts,
and distinctively human capacities. People, actions, policies, states of affairs,
etc. will bear good-making moral properties just insofar as they contribute
to the satisfaction of these needs, wants, and capacities. People, actions,
policies, states of affairs, etc. will bear bad-making moral properties just
insofar as they fail to promote or interfere with the satisfaction of these needs,
wants, and capacities. 25 The physical states which contribute to or interfere
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

with the satisfaction of these needs, wants, and capacities are the physical
states upon which, on this functionalist theory, moral properties ultimately
supervene.
Although I cannot and do not need to defend here this functionalist model,
it is worth pointing out how this model addresses two issues of concern to
Mackie, namely, the justifiability of morality and the decidability of moral
disputes. In section 2 1 argued that internalism is implausible and that deter-
mination of whether agents have motivation or reason to be moral depends
upon the content of morality and facts about agents. If this functionalist
account of moral value which I have proposed as a realist model is plausible,
then there is reason to think that moral facts will at least typically provide
agents with reasons for action. Everyone has reason to promote his own
well-being, and everyone has reason to p r o m o t e the well-being o f others at
least to the extent that his own well-being is tied up with theirs. Presumably,
any plausible theory of h u m a n needs, wants, and capacities will show that
the satisfaction of these desiderata for any given individual will depend to
a large extent on the well-being of others. People have needs and desires for
friendship and love and for the benefits o f cooperative activity; they also
have capacities for sympathy, benevolence, and social intercourse. In order
to satisfy these social needs, desires, and capacities, agents must develop and
maintain stable social dispositions, and this means that they will often have
reason to benefit o t h e r s e v e n when they do not otherwise benefit by their
action. So, although there m a y be cases in which maintaining or promoting
h u m a n well-being involves no benefit to the agent, there is good reason to
suppose that h u m a n well-being and agent well-being will by and large
coincide. As this functionalist theory of value illustrates, externalism allows
a strong justification o f morality.
This functionalist theory of moral value also helps to explain the nature
o f moral disagreement. C o m m o n sense and attention to the argument from
disagreement tell us that moral disputes can be extremely difficult to resolve.
This functionalist specification of moral realism explains why many moral
disputes which are in principle resolvable are nonetheless so difficult to resolve
even under favourable conditions. Because facts about h u m a n well-being and

25 When suitably developed, this functionalist theory of moral value might be quite similar in
content to the moral theory which Mackie himself defends in part II of E. However, Mackie
and I would still disagree about the status of this theory. I am suggesting that it might be
true; fie is presumably doing something like recommending the adoption of his theory.
 David O. Brink 123

flourishing depend at least in part upon facts in such complex and contro-
versial empirical disciplines as economics, social theory, and psychology, even
disputants who share something like the functionalist theory o f value and
are well informed will often disagree about what morality requires.
In addition to the metaphysical complaint about 'what in the world' a
supervenient relation is, Mackie lodges an epistemological complaint about
how we could know when the appropriate superyenient relation obtains.
(E: p. 41) We may know that certain natural facts or facts under a non-moral
description obtain, but how do we know or go about finding out whether
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

these physical facts realise any moral facts and, if so, which? Mackie claims
that we could gain this kind of moral knowledge only if we had special facul-
ties for the perception of moral facts of the sort ethical intuitionism ensures.
But, Mackie argues, although moral intuitionism could have been true, there
are good aposteriori grounds for believing that no such faculties exist. There-
fore, barring the cognitive inaccessibility of moral facts, moral realism must
be false. (E" pp. 38-9)
The epistemological belief that moral realism is committed to intuitionism
rests at least in Mackie's case on the mistaken metaphysical assumption that
moral values would have to be ontologically sui generis. If and only if moral
facts were queer kinds of entities would we need some special faculty for
cognitive access to them. But the realist denies that moral facts are sui generis;
moral facts supervene on natural facts. One goes about discovering which
natural facts moral facts supervene on by appeal to moral theories. (Of course,
appeal to a particular moral theory is justified o n l y if that theory coheres
well with other moral and non-moral beliefs we hold.) For example, if the
functionalist account of moral value sketched above can be defended, then
we do know how to set about ascertaining which if any moral facts supervene
on a particular set of natural facts. We ascertain whether the natural facts
in question contribute to, interfere with, or are neutral with respect to the
maintenance and promotion of human well-being. Granted, in many cases
this will be no easy task, since completion of the task will depend in part
upon answers to controversial empirical questions in such fields as economics,
social theory, and psychology. But all this shows is that moral knowledge
is sometimes hard to come by, not that it is queer or mysterious.
Mackie might complain that both acceptance and application of moral
theories must be guided by other moral commitments. Not only does accept-
ance of the functionalist theory of value depend upon its coherence with,
among other things, other moral beliefs, but also the findings o f such disci-
plines as economics, sociology, and psychology cannot fully determine the
extension of 'human well-being and flourishing'. Even if the special sciences
can tell us something about human needs, wants, and capacities, and the
effective ways of realising them, these sciences cannot rank these components
of the good or adjudicate conflicts among them. Some irreducibly normative
questions must be answered in determining what constitutes human well-being
and flourishing.
But if the fact that some or all of our moral judgments are theory-dependent
 124 Moral Realism and the Sceptical Arguments

in this way is supposed to present a genuine epistemological problem for the

moral realist which is not simply the result of applying general sceptical con-
siderations to the case of morality, Mackie must claim that theory-dependence
is a feature peculiar to moral methodology. Is this claim at all plausible?
Here, as before, the moral realist can find quite respectable partners 'in
guilt'. It is a commonplace in the philosophy of science that scientific
methodology is profoundly theory-dependent. Assessments o f theoretical
simplicity and theory confirmation as well as standards of experimental design
and instrument improvement require appeal to the best available background
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

theories in the relevant disciplines. For example, in theory confirmation there

is an ineliminable comparative component. Theories count as well confirmed
only if they have been tested against relevant rivals, and determination of
which alternative theories are relevant or worth considering requires appeal
to background bodies of accepted theory. Acceptance of normal scientific
observations and judgments, as well as application of general methodological
principles, is also theory-laden. For example, judgments about the acidity
or alkalinity of a substance which are based on the results of litmus paper
tests pre-suppose belief in the normality of the test conditions and acceptance
of the relevant chemical theories explaining how litmus paper detects p H and
how p H reflects acidity and alkalinity.
The fact that scientific method is heavily theory-dependent shows that
science and ethics are on a par in being theory-dependent. Thus, the fact
that moral commitments must be appealed to in the acceptance and appli-
cation of moral theories poses no special epistemological problem for moral
realism. O f course, although most of us do not draw non-realist conclusions
from the theory-dependence of scientific method, one may wonder how the
profoundly theory-dependent methodologies in science and ethics can be
discovery procedures. The answer is that theory-dependent methodologies
are discovery procedures just in case a sufficient number of background
theories in the disciplines in question are approximately true. 26 And I have
been arguing that Mackie has provided no good reason for doubting that
some of our moral background theories are approximately true.
Mackie might respond that the moral and scientific cases are not in fact
on a par and that there is reason to doubt the approximate truth o f our moral
theories, because while there is a good deal of consensus about the truth of
the scientific theories appealed to, say, in the making of p H judgments, there is
a notable lack of consensus about which moral theories to appeal to in making
moral judgments. There are at least three reasons, however, for dismissing
this response. First, this response probably overstates both the degree of
consensus about which scientific theories are correct ,and the degree of dis-
agreement about which moral theories are correct. 27 Secondly, the response
26 Boyd argues that because of the profound theory-dependenceof scientific methodology, the
instrumental reliability of scientific method can only be explained by assuming the truth of
scientific realism. See Richard Boyd, 'Realism, Underdetermination, and a Causal Theory
of Evidence'Nous 7 (1973), pp. 1-12and 'Scientific Realism and Naturalistic Epistemology'
PSA 1980 Volume 2 (East Lansing, MI" Philosophy of Science Association, 1981).
27 Cf. Alan Gewirth, "Positive 'Ethics' and Normative 'Science' "reprinted in J. Thomson and
G. Dworkin (eds), Ethics (New York: Harper & Row, 1968).
 David O. Brink 125

probably also overstates the a m o u n t of antecedent agreement necessary to

reach eventual moral agreement. Finally, this response just raises f r o m a
different perspective the argument f r o m disagreement, and we saw that the
moral realist has a plausible account o f moral disputes.
These considerations show that moral realism is committed to nothing
metaphysically or epistemologically queer. The realist holds that moral facts
supervene upon other natural facts and that moral knowledge is acquired
in the same theory-dependent way that other knowledge is. Moral realism
is plausible enough both metaphysically and epistemologically to allow us
Downloaded by [University of California, San Diego] at 16:42 01 January 2015

to say that Mackie has again failed to shoulder the burden of proof.

5. Conclusion
Mackie follows an important sceptical tradition in attempting to show that
there is a special problem about realism in ethics. He recognises that it is
the sceptic who bears the burden of proof but claims that his arguments from
disagreement and queerness satisfy this burden. I argued, however, that
neither argument provides good reason for disbelieving moral realism; cer-
tainly neither argument successfully bears the sceptic's burden of proof. The
moral realist has various resources with which to account for moral disputes,
and neither his account of the supervenience of moral facts nor his account
of the theory-dependence of moral knowledge is queer or u n c o m m o n . I also
introduced and developed a functionalist theory of moral value according
to which moral facts are facts about h u m a n well-being and flourishing.
Although the truth of this functionalist theory is n o t essential to the defence
of moral realism, it does provide a plausible model for a realist program
in ethics. Mackie's arguments from disagreement and queerness do not
exhaust the sceptical challenges to moral realism. But both arguments are
sufficiently important that by successfully rebutting them we have gone a long
way towards defending moral realism. 28

Cornell University Received March 1983

2s I would like to thank Tom Arner, Richard Boyd, Norman Dahl, T. H. Irwin, David Lyons,
John McDowell, Alan Sidelle, Nicholas Sturgeon, and readers for the Australasian Journal
of Philosophy for helpful comments on earlier versions of this paper.