This article was downloaded by: [University of Sydney]

On: 13 March 2015, At: 15:32

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

International Journal of
Philosophical Studies
Publication details, including instructions for authors
and subscription information:
http://www.tandfonline.com/loi/riph20

In Defence of Morality: A
Response to a Moral Error
Theory
a
Paul Barry
a
La Trobe University, Australia
Published online: 15 Jan 2014.

Click for updates

To cite this article: Paul Barry (2014) In Defence of Morality: A Response to a Moral
Error Theory, International Journal of Philosophical Studies, 22:1, 63-85, DOI:
10.1080/09672559.2013.860613

To link to this article: http://dx.doi.org/10.1080/09672559.2013.860613

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 Sydney] at 15:32 13 March 2015
 International Journal of Philosophical Studies, 2014
Vol. 22, No. 1, 63–85, http://dx.doi.org/10.1080/09672559.2013.860613

In Defence of Morality: A Response to

a Moral Error Theory
Paul Barry
Abstract
This paper responds to Richard Joyce’s argument for a moral error theory. Joyce
claims that our moral discourse purports to speak of something objective in that
it presupposes the existence of non-institutional, categorical reasons for action.
Downloaded by [University of Sydney] at 15:32 13 March 2015

Given this, he argues that a proper vindication of our moral discourse would be
one carried out from a point of view that is objective inasmuch as it is external
to the ‘institution of morality’. And since our moral discourse cannot be vindi-
cated from that external point vantage point, it follows that that discourse is
fundamentally flawed. My critique of Joyce’s argument goes to his appeal to an
external point of view from which to assess the legitimacy of our moral
discourse. I argue that our moral talk is intelligible only in the context of the
understanding we inhabit as moral agents. Hence, the external vantage point that
Joyce identifies is a point of view from which moral claims are deprived of the
conditions in which they make sense. I therefore reject Joyce’s claim that our
moral discourse is conceptually non-institutional: whilst morality is committed to
categorical reasons, these reasons are ‘institutional’ insofar they are intelligible
only from within the institution of morality.
Keywords: error theory; morality; Richard Joyce; objectivity; external point
of view; conditions of intelligibility

I. Introduction
Moral error theorists think that everyday moral thought is discordant with
some aspect of reality, and that therefore moral discourse is fundamentally
flawed. An argument for a moral error theory includes two basic claims. The
first is a conceptual claim about our moral discourse: the error theorist claims
that our moral discourse presupposes some thesis, a commitment to which is
essential to thinking or talking in moral terms. The second is a substantive
claim about that thesis: the error theorist claims that it is false. The conclusion
drawn is that everyday moral thought is hopelessly flawed since it is inescap-
ably committed to a false premise. The error theorist’s attack on morality is of
the same form as the atheist’s attack on the Judeo-Christian belief by way of
the argument from evil. The argument starts by fixing upon a set of commit-
ments that are apparently essential to Judeo-Christian discourse: this discourse

© 2014 Taylor & Francis

 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

is essentially committed to the existence of an all-powerful, all-knowing,

all-loving God. It then goes on to claim that the existence of such a God is
irreconcilable with the evil that exists in the world. Christian discourse is
(allegedly) fundamentally flawed since it is essentially committed to a false
view of the world.
In this paper I offer a defence of morality in response to Richard Joyce’s
argument for a moral error theory. As with John Mackie’s error theory, Joyce
claims that our moral discourse is committed to a kind of ‘objective prescrip-
tivity’ which turns out to be indefensible. Morality purports to speak of some-
thing objective insofar as it is committed to the existence of reasons for action
that are both ‘categorical’ and ‘non-institutional’. Moral reasons claim to be
categorical in that they pretend to carry authority that is independent of an
agent’s desires or interests. And these reasons claim to be non-institutional in
Downloaded by [University of Sydney] at 15:32 13 March 2015

that they pretend to carry authority that is not dependent upon the agent
already occupying a moral point of view. Since morality is conceptually non-
institutional in this way, Joyce argues that a proper vindication of our moral
discourse would be one that affirms the legitimacy of moral reasons from some
external perspective, i.e. from a standpoint that is outside of the ‘institution of
morality’. We need, he says, to ask the ‘external question’ of morality. But
since moral reasons cannot be legitimated from this external position, such rea-
sons are indefensible. Morality is fundamentally flawed in that it is committed
to a type of practical reason for which no rational basis can be found.
In reply, I defend the categorical element that is entailed in our moral dis-
course by way of an examination of the conditions in which the reasons
implied in our moral discourse are intelligible. I argue that these reasons make
sense only in the context of the understanding over which we preside when we
inhabit a moral point of view. The move to the external vantage point from
which Joyce seeks to assess our moral discourse involves a move to a vantage
point from which moral reasons are no longer intelligible. Hence, I agree with
Joyce that moral thought cannot be vindicated from this external position, but
argue that what this really shows is that moral thought is institutional in nature.
My defence of categorical moral requirements consists, therefore, in a rejection
of Joyce’s conceptual claim that moral discourse is committed to
non-institutional reasons.
The paper proceeds as follows. In the following section, I give an account
of Joyce’s argument: I explain the basis for his claim that morality is commit-
ted to the existence of categorical, non-institutional reasons for actions; I
explain the external vantage point from which he seeks to assess the legitimacy
of such reasons; and I explain why he thinks such reasons cannot be legiti-
mated from this external perspective. In Section III, I aim to show that the
external position from which Joyce says morality needs to vindicate itself is a
position that deprives moral reasons of the conditions in which they are
intelligible. What follows, I argue, is that morality is conceptually institutional
in a way that is ignored by Joyce. In Section IV, I try to show that my view of
64
 IN DEFENCE OF MORALITY

morality as conceptually institutional is compatible with the inescapability that

characterizes the authority which morality claims to hold.

II. Joyce’s Argument For a Moral Error Theory

Joyce follows Mackie in viewing ‘objective prescriptivity’ as the feature of our
moral discourse that is both essential to it and problematic. Our moral
discourse seems to aspire to capture something that is independent of us but
which nonetheless bears some essential relation to our behaviour in that it
demands that we act in a certain way. Mackie is convinced that everyday
moral talk is committed to the existence of moral value or moral properties,
where ‘values themselves have been seen as at once prescriptive and objec-
tive’.1 His argument from queerness aims to show that we have good grounds
Downloaded by [University of Sydney] at 15:32 13 March 2015

for thinking that such values do not belong to an objective order of things: ‘If
there were objective values, then they would be entities or qualities or relations
of a very strange sort, utterly different from anything else in the universe.’2
Since our moral talk commits us to the existence of such ‘queer’ properties or
entities – items we have good empirical reasons for thinking are non-existent –
a moral error theory is entailed. Similarly, Joyce’s argument for a moral error
theory centres upon the way that our moral discourse speaks of practical
requirements that emanate not from ourselves but from some external source.
He thinks that objective prescriptivity is a non-negotiable feature of our moral
discourse, but is also indefensible.
The objectivity embodied in moral thought is typically explained in terms of
Kant’s distinction between ‘categorical’ and ‘hypothetical’ imperatives.

Now all imperatives command either hypothetically or categorically. The

former represent the practical necessity of a possible action as a means
to attain something else which one wills (or which it is possible that one
might will). The categorical imperative would be that one which repre-
sented an action as objectively necessary for itself, without any reference
to another end.3

Hypothetical imperatives are those whose legitimacy depends upon one having
the corresponding desire or end, to which the action prescribed is a means.
Hence they evaporate if the end to which they are subservient is absent. The
claim, ‘You should leave home early if you want to arrive for the start of the
game’, loses all authority if it turns out that you don’t want to see the start
the game. Categorical imperatives are those that claim legitimacy not as a
means to the satisfaction of a given desire or end, but in themselves. They
don’t evaporate in the way that hypothetical imperatives do. Morality, it is
generally assumed, deals in imperatives of the categorical variety, and is in this
sense committed to objective prescriptivity.

65
 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

But in itself, showing morality to be committed to some kind categoricity is

not enough to show that there is anything problematic in our moral discourse.
Both Joyce and Mackie acknowledge that there are systems of rules that are
impersonal or categorical in the sense that their applicability does not depend
upon one having a desire or an interest in complying with them.4 Mackie talks
of the rules of chess. The queen rule does not cease to apply if at some point
in a game I feel like breaking it, or can gain some advantage from doing so.
Similarly, the applicability of the grammatical rules of a given language are
not held hostage to the whims of a particular speaker. The rule, ‘One should
use the “simple present” when making a general statement of fact’ remains
valid, even if someone really wants to use the present continuous when making
such a statement. These rules are categorical imperatives in that they share the
feature of non-evaporability that Joyce and Mackie take to be essential to
Downloaded by [University of Sydney] at 15:32 13 March 2015

moral imperatives. But since nobody thinks there is anything problematic about
the rules of chess or those of the English language, there must be something
extra present in moral thought if the imperatives entailed therein are to be
shown to be uniquely objectionable.
Joyce accounts for this ‘something extra’ by contrasting ‘institutional’
categorical imperatives with ‘non-institutional’ categorical imperatives.5 Institu-
tional categorical imperatives are those whose authority is underwritten by one
already being situated within the institution to which those imperatives belong.
The imperatives of chess are institutional in this sense: they apply categori-
cally, but only when we speak from within the institution of chess. If someone
is not interested in playing chess, then of course those imperatives have no
authority for her. Non-institutional categorical imperatives are those that claim
authority over each of us irrespective of how we are situated. Unlike, say, the
imperatives of chess, non-institutional categorical imperatives purport to exert
an authority whose legitimacy transcends any particular institutional
framework. Morality, Joyce argues, is conceptually committed to categorical
imperatives that are non-institutional. ‘In the moral case, we are not content to
admit that our claim that there is a reason to refrain from killing is merely a
permissible way of speaking from a perspective that endorses the dictates of
morality.’6 The basis for claiming that moral imperatives purport to be non-
institutional has to do with the inescapability that is contained in our moral
language. If I say, ‘You ought not to torture children’, then the applicability of
this claim does not depend upon your first having signed up to morality. We
think that these moral demands are valid no matter what. They don’t evaporate
if we learn that you really enjoy torturing children, or that you never consented
to a moral code that includes this prohibition in the first place. When we speak
morally, Joyce argues, we make an appeal to reasons that are ‘objective’ or
‘real’ in that they are not merely embedded in an institution, but exert a kind
of authority that cannot legitimately be ignored.
It is this commitment to non-institutional categorical imperatives that Joyce
believes renders all moral claims false. Whilst we want to imbue our moral
66
 IN DEFENCE OF MORALITY

prescriptions with inescapable force, it turns out that there is no rational basis
for doing so. Thus, his argument for a moral error theory looks like this:

(1) Our moral talk is conceptually committed to non-institutional

categorical imperatives (conceptual step).
(2) In fact, non-institutional categorical imperatives are indefensible
(substantive step).
(3) Hence, our moral talk is fundamentally flawed, and all moral claims
are false (conclusion).7

What of the substantive step in this argument? At this point Joyce’s error the-
ory deviates from Mackie’s in an important way. As noted above, Mackie’s
strategy is to show that our moral discourse commits us to the existence of a
Downloaded by [University of Sydney] at 15:32 13 March 2015

certain kind of entity or property, and that there are good grounds to think that
such items have no place in the physical universe. He argues that ordinary
users of moral language are committed to a certain view of the world: the
question as to whether this view of the world is accurate is answered by how
well it squares with the picture of reality supplied by the physical sciences.
The difference in Joyce’s version of an error theory stems from the fact that
he focuses not on properties or entities, but on the kinds of reasons to which
our moral talk commits us. Non-institutional categorical imperatives entail a
commitment to reasons for action that are likewise categorical and non-institu-
tional. Since these reasons are non-institutional, we should be able to affirm
their legitimacy from a point of view that is external to the institution of
morality. Borrowing a phrase from Rudolf Carnap, he says that we need to ask
the ‘external question’ of morality. ‘There has to be some way of stepping out-
side of the institution of morality and asking “Does so-and-so really have a
reason to ϕ?” “Is it really the case that she ought to ϕ?”’8 Raising the external
question against morality is necessary if we are to avoid begging the question
against the moral sceptic. If we seek to judge the veracity of our moral claims
from a moral perspective, then we assume the legitimacy of our moral dis-
course rather than establishing this for someone who is not yet convinced.
‘Internal to the framework in question, confirming or disconfirming the truth
of these propositions is a trivial matter. But traditionally philosophers have
interested themselves in the external question – the issue of the adequacy of
the framework itself’ (original emphasis).9 Since the present enquiry goes to
the adequacy of our moral discourse, the external question, Joyce insists, is the
one in need of an answer.
An external vantage point from which to pose this question is available,
Joyce argues, because morality is a framework of thought that is ‘nested’
within a broader framework – that of practical rationality – ‘and the adequacy
of the former as a whole [can be] appraised in terms provided by the broader
one’.10 Morality is wholly encompassed by practical rationality, according to
Joyce, insofar as our moral reasoning is necessarily committed to the existence
67
 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

of practical reasons, but the reverse is not the case: practical reasoning does
not commit one to the existence of moral reasons. In this way, the framework
of practical rationality supplies the external vantage point from which the
claims of morality can be assessed. The analogy that Joyce draws between
moral discourse and phlogiston discourse is helpful in clarifying his basic strat-
egy.11 Phlogiston discourse involved a commitment to a type of chemical sub-
stance, so the legitimacy of that discourse was assessable from a perspective
that accepted the existence of chemical substances in general, but was uncom-
mitted as to the existence of the particular substance in question. Analogously,
our moral discourse involves a commitment to a type of practical reason, so
the legitimacy of that discourse, Joyce says, can be assessed from a perspective
that accepts the existence of practical reasons, but is uncommitted as to the
existence of the particular type of reason to which moral thought is committed.
Downloaded by [University of Sydney] at 15:32 13 March 2015

‘Outside morality we know very well what reasons are: most conspicuously,
we have a normative framework of giving and accepting reasons for acting
known as “practical rationality”’ (original emphasis).12 So, where for Mackie
the challenge for defenders of morality is to show that moral value or proper-
ties really are part of ‘the fabric of the world’,13 for Joyce what needs to be
shown is that morality can be vindicated from another agential perspective –
one that is divorced from moral thought and practice.
The question, for Joyce, is whether the non-institutional categorical reasons
that are (allegedly) an essential feature of our moral discourse can be legiti-
mated from this external perspective. He says that they cannot. He outlines
and defends an instrumental account of practical rationality, according to which
practical reasons are essentially hypothetical in nature.14 ‘Practical rationality
[…] yields only hypothetical imperatives, and therefore cannot be appealed to
as a way of vindicating “moral inescapability.”’15 What follows is the claim
that ‘non-institutional categorical imperatives are indefensible’: the substantive
step of Joyce’s error theory.
My critique of Joyce’s argument consists in a defence of the categorical
element that is embodied in moral thought. I agree that our moral discourse
purports to speak of something objective in that it is committed to categorical
requirements, but argue that there need not be anything problematic about this
feature of everyday moral talk. It might therefore seem natural to suppose that
my argument involves an acceptance of the conceptual step of Joyce’s argu-
ment, but a rejection of the substantive step. This is particularly so given that
much of what I have to say pertains to the strategy that Joyce employs in
attempting to execute that substantive step, namely the attempt to question
morality from an external vantage point.
Nonetheless, I frame my critique of Joyce’s error theory as a challenge to
the conceptual step in his argument. Whilst I agree with Joyce that moral
thought purports to speak of something objective, I think that he overstates the
objectivity presupposed therein. Specifically, I argue that while our moral
discourse is committed to categorical requirements, these requirements are, in
68
 IN DEFENCE OF MORALITY

an important sense, institutional. Hence, I reject the conceptual step in Joyce’s

argument: that morality is committed to non-institutional categorical
imperatives. The basis for this rejection concerns the conditions in which moral
prescriptions are intelligible. I argue that the reasons entailed in our moral
discourse are intelligible only from a moral perspective, and so those reasons
claim authority for those who are already situated within the so-called ‘institu-
tion of morality’. Given this, there is no way of stepping back and holding
morality ‘up to the light’, as it were, so as to assess the legitimacy of its
claims. The external question that Joyce wants to ask of morality helps
illustrate this point, in that moral reasons cease to make sense from such a
perspective. This, at least, is what I plan to argue.
Although quite different in detail, my critique of Joyce has something in
common with John McDowell’s critique of Mackie’s error theory.16 Whilst
Downloaded by [University of Sydney] at 15:32 13 March 2015

agreeing with Mackie that ‘ordinary evaluative thought presents itself as a mat-
ter of sensitivity to aspects of the world’,17 McDowell argues that Mackie mis-
characterizes the objectivity thus implied. McDowell effectively offers an
alternative account of what it takes for something to belong to the ‘world’, and
claims that value satisfies those conditions. Central to his case against Mackie
is the view that moral thought becomes unintelligible on Mackie’s analysis. A
key difference between my argument and McDowell’s corresponds to the
already noted difference between Joyce’s error theory and Mackie’s. Joyce
does not try to argue that morality would need to be vindicated from a natural
scientific worldview, but instead argues that, if the claims of morality are legiti-
mate, we should be able to affirm the legitimacy of those claims from some
other (non-moral) agential perspective. My claim against Joyce is that moral
reasoning becomes unintelligible from the agential standpoint he wants to
adopt, and that this is seemingly implicit in Joyce’s own argument.
My critique also bears some proximity to the one offered by Stephen Finlay
in his paper, ‘The Error in The Error Theory’, and then again in his reply to
Joyce’s reply, ‘Errors upon Errors: A Reply to Joyce’.18 An important differ-
ence between his critique and my own is that Finlay takes an essentially defen-
sive approach, aiming to show that the evidence to which Joyce appeals in
support of his view that moral discourse presupposes the kind of objectivity in
question can be explained at least as well without attributing this presupposi-
tion to moral thought. My approach, in contrast, is an offensive one. I argue
that a proper understanding of our moral concepts reveals that they are not
committed to objectivity in the way that Joyce alleges. Specifically, such an
understanding shows moral prescriptions to be institutional.
I make this case in the following section. In the final section I aim to show
that viewing moral claims as essentially institutional is compatible with the
inescapability that is embodied in ordinary moral thought.

69
 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

III. Objectivity and an External Point Of View

Joyce’s argument is premised on the idea that we can move to the external
vantage point that he identifies in order to assess the legitimacy of moral
claims, without losing a grasp of the meaning of those claims (otherwise that
which is to be assessed slips from view). Hence, the point of view of practical
rationality has, Joyce thinks, two things going for it. It is a point of view from
which we are uncommitted to an acceptance of the claims of morality. But it is
also one from which we can make sense of those claims, namely through the
‘terms provided’ by the broader framework of practical rationality.
It is essentially this last claim that I want to challenge. Our moral concepts,
such as the concept of moral obligation that Joyce scrutinizes, are constituted
by the understanding over which we preside as moral agents i.e. as participants
Downloaded by [University of Sydney] at 15:32 13 March 2015

in the institution of morality. To move to a perspective that is altogether sepa-

rate from that which we inhabit as moral agents is to take up a position from
which the claims of morality are no longer intelligible. It is in this sense that I
take moral claims to be ‘institutional’. Since they only make sense from within
the institution of morality, these claims are limited to one’s participation in this
institution.
Consider a normative framework that is uncontroversially institutional – the
game of chess. There are rules that are constitutive of chess. The rule that
states that one must not move a pawn backwards is one; the rule that states
that the objective is to capture your opponent’s king is another, and so on. To
say that these rules are ‘constitutive’ is just to say that the game of chess does
not exist without such rules.19 To play chess just is to be responsive to these
rules. These rules imply reasons, and since the rules are constitutive of the
activity in question, the reasons are unavoidable for anyone engaged in this
activity.20 What I am doing, when playing a game of chess, is only describable
with reference to those reasons, insofar making my actions intelligible to some-
one else would require an appeal to the reasons that are entailed in those rules.
Were someone to ask me what I am doing at any given point in a game, an
informative answer would have to be something like, ‘I am moving my
bishop’, or ‘I am trying to trap my opponent’s queen’.
These reasons are external to the agent in a straightforward sense. To play a
game of chess just is to be answerable to standards that are not of one’s choos-
ing. Of course generally you only find yourself playing a game if you want to
play and have chosen to do so. But granted that you are playing chess, you are
subject to standards that are independent of your desires and choices. This is
because, as noted, the very activity is only intelligible on the basis of these
external or impersonal standards. Hence, participating in that activity entails
that one is answerable to external or impersonal standards, and thus open to
criticism by others on the basis of those standards. All of this is just to say that
the reasons in question are categorical: they are not simply hedged on the
desires or aims of a given agent.

70
 IN DEFENCE OF MORALITY

Are these reasons real? I don’t want to get distracted by a semantic

discussion of what it means for a reason to be ‘real’, but we can agree that
these reasons enjoy all the reality that they claim to enjoy. They are capable of
doing two things that are generally asked of reasons: they are capable of
explaining and justifying an agent’s actions. As well as explaining my
behaviour to others, the rules of chess, at least on some occasions, serve as
good (justificatory) reasons for action. Were someone to ask me why I didn’t
capture my opponent’s queen, the response ‘it is illegal to move a pawn back-
wards’ is, all things being equal, a good justification for my action (or in this
case, my inaction).
In chess we have a normative framework that is committed to the existence
of a kind of practical reason – call these ‘chess reasons’. Thus, if we are to
follow Joyce, this framework is susceptible to the same external treatment to
Downloaded by [University of Sydney] at 15:32 13 March 2015

which Joyce wants to subject morality. Outside of chess we know very well
what reasons are: most conspicuously, we have a normative framework of
giving and accepting reasons for acting known as practical rationality. Hence
we can move to the broader framework of practical rationality, from which to
judge the legitimacy of chess reasons. This would involve saying something
like, ‘I know that speaking from within the institution of chess I have a reason
not to move my pawn backwards, but does this reason hold when I take up a
position outside of this institution? Do I really have a reason not to move my
pawn backwards?’
The answer, of course, is ‘No’. But what follows from this? Arriving at this
answer is certainly not a case of having gotten a clearer view of things, and
thus having found chess reasons to be illusory in some way. To move to this
external perspective is, rather, to have taken up a position from which chess
reasons are no longer intelligible. Talk of these reasons makes sense only
within the context of one’s involvement in the game, and in the context of the
understanding that belongs to that involvement. Chess reasons are indeed a
type of practical reason, but they exist only for a certain type of agent: a chess
player. Since being a chess player implicates one in a grasp of chess reasons,
such reasons only exist for those who already have some grasp of them. In this
way, there is no way of stepping outside of the institution of chess, and hold-
ing onto an understanding of those reasons. What the external question shows
is what is necessary or indispensable to the intelligibility of the reasons in
question. Since chess reasons no longer make sense in the absence of the
understanding involved in playing a game of chess, we can say that this
understanding forms part of the meaning of the normative concepts employed
in discourse on chess. Chess reasons are only properly articulable from within
the institution of chess. It is precisely this feature that makes the normative
framework ‘institutional’: its claims are only intelligible in the context of the
understanding over which one presides as a participant in that institution.
Now turn to the question of the legitimacy or reality of the reasons entailed
in our moral discourse. An important point to note about Joyce’s argument is
71
 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

that while he calls upon an external approach in order to assess the legitimacy
of our moral concepts (i.e. in order to execute the substantive step of his error
theory), this external approach shapes the way that he accounts for those
concepts: it shapes the way he handles the conceptual step of his error theory.
In his account of the essential commitments of moral discourse, he focuses
almost exclusively on what we would say to or about someone who has done
something immoral. The aim, in doing so, is to capture the impersonality of
the standards to which moral thought is committed: we would want to say to
the wrongdoer that he ought not to have done what he did, that he had a rea-
son not to, etc., and our moral approbation is in no way hedged upon the
wrongdoer having certain desires or interests that were frustrated by his
immoral act. The perspective taken here is both third-personal and after-the-
fact, and it is not hard to see how this approach animates the analogy with
Downloaded by [University of Sydney] at 15:32 13 March 2015

phlogiston. The suggestion is that users of moral concepts postulate obligations

and reasons in much the same way that natural scientists once postulated the
existence of phlogiston, or in the way that a doctor might postulate an irregular
growth on a patient’s brain. The trouble is that this encourages us to think of
moral agency as something that we observe from an external viewpoint, and to
forget the crucial fact that this is an activity that we participate in. As such
there are conditions that are essential to this participation, and these conditions
are, naturally enough, constitutive also of the concepts employed in moral dis-
course.
A moral act is not simply a set of physical movements that causes a certain
outcome, and that happens to either comply or fail to comply with some set of
objective moral rules or standards, no more than a move in chess involves
merely altering the spatial location of a small piece of wood. Viewing it in this
way implies a merely contingent or external relation between moral actions
and those moral standards. A moral act does not just happen to conform to
moral standards, as one’s building a house might just happen to conform to a
set of building regulations. Moral behaviour just is behaviour that is guided by
moral norms: moral acts are carried out with understanding, and this under-
standing includes a recognition of the moral standards that are at issue. This
feature is essential to an act counting as a ‘moral’ one. This is not meant as an
endorsement of the view that Kant is often thought to have held, namely that
an act is only of moral worth if it is done for the sake of the moral law. Many
rightly recoil from this idea to the extent that it denies that acts born of feel-
ings of sympathy, such as an act of spontaneous kindness, have moral worth.
But while such an act is not carried out with the explicit aim of conforming to
some moral rule, it is performed with at least an implicit sense of the moral
significance of the situation. Showing affection to a friend is not like scratch-
ing an itch: it is not as though I am acting here on a brute impulse. The key
difference is that in an act of affection I act in light of a sense that there is
something morally at issue, however difficult it might be to give a clear articu-
lation of what this is. Even in these more spontaneous acts, the agent has some
72
 IN DEFENCE OF MORALITY

handle on the moral standards in play such that he recognizes that he is in

some way called upon to perform the action. Were a grasp of these standards
altogether absent, we simply wouldn’t be dealing with an exercise of moral
agency.
Moreover, this recognition of moral standards is internally related to an
understanding of ourselves. If I say something about the amount of oxygen in
the room, then I make a claim about the world that bears no essential reference
to my being an agent of a certain kind. If, on the other hand, I claim that such
and such is morally required of me, then I clearly do make such a reference.
Because now I am not simply stating a neutral fact about the world: the claim
necessarily involves a reference to myself, since I am claiming that the per-
ceived obligation lies on me, or others like me. This self-understanding is part
of the conditions of intelligibility for a sense that a given act is morally
Downloaded by [University of Sydney] at 15:32 13 March 2015

required of me: my thinking in terms of moral obligation is possible only

because I see myself as a ‘moral respondent’ – as being answerable to the
claims of morality. The same of course is true whenever we lay an obligation
on others. If I say, ‘You ought to be upfront with me’, then this claim reposes
upon my sense that you are not only capable of recognizing the importance of
honesty, but also more generally that you understand yourself to be answerable
to the moral claims that others may raise upon you. Again, this is not to sug-
gest that this is explicit in our moral thinking. For the most part it is not: by
and large moral discussion remains focused upon the content of particular
moral claims, and this understanding of ourselves remains in the background.
Indeed it is the sheer pervasiveness and obviousness of this self-understanding
that leads to its remaining merely implicit in our moral talk. Nonetheless, this
self-understanding belongs to the very concept of moral obligation, inasmuch
as it is indispensable to the intelligibility of that concept.21 As with chess rea-
sons, the authority that moral reasons claim upon us make sense in the context
of the understanding we inhabit as moral agents, and this includes a grasp of
those reasons.
We can return now to the external question that Joyce raises against moral-
ity. Recall that Joyce thinks that we can raise this question, since morality is
‘nested’ within the broader framework of practical rationality. Thus he thinks
that we can articulate the claims of morality from this broader perspective in a
way that preserves the meaning of those claims. It is exactly this that I take to
be mistaken. Moral reasons are institutional in nature in that those reasons are
only intelligible from a moral perspective. In this sense, there is no way of
stepping out of the institution of morality and appraising it from an external
perspective. All that we can say is that the reasons to which we are committed
when we engage in moral discourse cease to make sense from that external
perspective.
In one sense Joyce’s own argument implicates him in a recognition of this
point. In Mackie’s error theory, the notion that there exist values or moral
properties that are both objective in an absolute sense – independent of human
73
 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

thought and behaviour – while also being prescriptive in the sense of making
practical claims upon us, is taken to be coherent. Hence, the substantive step
of Mackie’s argument for an error theory consists in the empirical claim that
moral properties are not objective in this way. The implication is that these
properties have no place in the actual world, but might well enough belong to
other possible worlds. Joyce’s argument contrasts with Mackie’s in that the
substantive step consists not in an empirical claim, but in the conceptual claim
that morality’s commitment to reasons that are categorical and non-institutional
is incoherent. There is, in Joyce’s words, ‘no sense to be made of such rea-
sons’,22 since practical rationality as such (separate from any particular institu-
tion) ‘yields only hypothetical imperatives’.23 On this point we agree: moral
reasons cease to make sense when removed from the context in which we
appeal to and respond to those reasons. Stepping away from the understanding
Downloaded by [University of Sydney] at 15:32 13 March 2015

that is embodied in moral thought and practice simply deprives those reasons
of the conditions in which they are intelligible.
There is, therefore, a role for an external point of view of the kind that
Joyce seeks to invoke, though it is rather different to the role he wants it to
play. Joyce tries to move to this external viewpoint in the hope of getting a
clearer, more objective view of reality, so as to judge how well our moral
thinking squares with this view. It pays to remember that the issue here con-
cerns the meaning of our moral concepts. One way to identify what is essential
to our moral concepts is to show that our moral thinking ceases to make sense
when abstracted from the conditions in which that thinking is employed. It
makes no sense to think that moral obligations might be applicable in the
absence of agents who have some grasp of those obligations, and who are thus
able to respond appropriately to them. The external position helps to clarify
what our moral thinking amounts to by highlighting the conditions of intelligi-
bility for moral thought and action. As Peter Winch puts it, ‘The only legiti-
mate use of such a Verfremdungseffect is to draw attention to the familiar and
obvious, not to show that it is dispensible from our understanding’ (original
emphasis).24
The key problem has to do with Joyce’s view that practical rationality
provides a meta-normative perspective from which any particular normative
system can be assessed, as the language of a given framework can be reliably
translated into the terms available to one who is a mere practical agent.
Practical rationality is therefore taken to be common or neutral ground between
the moral sceptic and defenders of moral discourse. He thinks this because
practical rationality, as he says, is not just another normative framework among
others. Whilst one can always question whether he should follow the conclu-
sions of his moral reasoning, he cannot intelligibly question whether he ought
to follow the conclusions of his practical reasoning. ‘Even to ask the question
“Why should I be interested in practical rationality?” is to ask for a reason.
Thus even to question practical rationality is to evince allegiance to it’ (origi-
nal emphasis).25 In one sense this is true. What is not true, however, is that the
74
 IN DEFENCE OF MORALITY

conclusions of practical rationality as Joyce conceives of it cannot be ques-

tioned or challenged by a rational agent. The substantive conception of practi-
cal rationality to which Joyce appeals to is an instrumental one, where one has
a reason to act in a given way just when one has some end – a desire or an
interest, actual or ideal – that is furthered by the act. If, when one speaks of
practical rationality, one has this substantive conception in mind, then it is
clearly false to say one cannot intelligibly question its conclusions. This is,
after all, what we often do when we reason along moral or for that matter,
chess, lines. We say, for example, ‘I really want that money, but it doesn’t
belong to me, so…’, or ‘I really want to take his queen, but I’m not allowed
to move my pawn backwards, so…’
I do not mean to suggest here that Joyce offers nothing by way of a defence
of his favoured instrumental conception of practical rationality. The point is
Downloaded by [University of Sydney] at 15:32 13 March 2015

that this conception is not a neutral one that can be used to arbitrate between
the claims of the moralist and the moral sceptic. Joyce’s idea seems to be that
since we all accept that we each have reasons that are entailed by our having
certain desires or interests, an instrumental conception of practical rationality
can be used as the starting point for this discussion.26 The task for defenders
of morality is to work their way from this starting point to an eventual vindica-
tion of the reasons to which our moral talk is committed. He is certainly right
to say this cannot be done, but a defender of morality need not accept that
challenge. Indeed the fact that participating in a given institutional normative
framework is unintelligible by simple appeal to instrumental rationality only
shows up the limits of such a conception in giving an account of our practical
reasoning.
Joyce says that ‘an institution […] is something which, by its very nature,
may be sensibly questioned from the outside’.27 If the argument that I have
presented is to be believed, then the opposite turns out to be true. The pre-
scriptive claims of a given normative framework are not intelligible from a
point of view that is external to that framework, and so cannot be questioned
from such a perspective. That Joyce does raise questions like, ‘I know that
morality requires me to do such and such, but why should I follow the rules of
morality?’ suggests that something other than external questioning (as Joyce
conceives of it) is going on here. When one is situated within a given a norma-
tive framework, there remain normative considerations that are independent of
that framework, such that one can, having recognized the institutional reasons
in play, intelligibly ask the ‘all things considered’ question as to what she
ought to do. On this score, Joyce seems to want to say that, when it comes to
morality, the ‘all things considered’ question is (according to everyday moral
thought) answered just as soon as we have identified a moral reason.28 In con-
trast, we are always open to the possibility that, whatever the rules or reasons
of chess or etiquette or gladiatorial combat, one may identify an ‘all things
considered’ reason to do otherwise. So, for instance, if a sadist has forced me
to play to a game of chess, and the only way of avoiding torture or death is to
75
 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

win the game by breaking the rules (assuming I can get away with it), we
would all agree that I ought to break the rules, i.e. that I have an ‘all things
considered’ reason to do so.
Undoubtedly there is a distinction to be drawn between moral reasons and
the reasons that belong to other institutional frameworks such as chess, and it
is related to the peculiar sort of inescapability that Joyce rightly thinks charac-
terizes moral discourse. Still, I think he misses the mark in the way that he
seeks to account for this distinction. At any given time there are moral reasons
upon which I might act: rather than spending a few dollars on a coffee I might
donate that money to a charity. But only an unusually moralistic conception of
morality would have it that I am obligated to act on each and every moral rea-
son wherever possible: such a view is certainly not a feature of ordinary moral
thought. Joyce might reply by saying that his interest is in cases where we are
Downloaded by [University of Sydney] at 15:32 13 March 2015

obligated, and not in those cases where one might act on moral reasons with-
out being (morally) required to do so. There are two things to say in response.
First, even in cases of explicit obligations, say in the proscription against steal-
ing, there are qualifications relating to one’s self-interest that ordinary moral
thinkers are prepared to recognize. We are inclined to say that stealing is
wrong, but that if one needs to steal a loaf of bread to stop from starving to
death, then he is permitted to do so. Second, the very distinction between
supererogatory and obligatory action is only intelligible on the basis that our
moral discourse is sensitive to, and recognizes the legitimacy of, considerations
of a non-moral nature.
I have, in this section, been trying to show that moral concepts are
institutional in nature, by examining the conditions in which these concepts are
intelligible. In the following section I aim to show that viewing moral claims
as institutional is compatible with an acceptance of the kind of inescapable
authority that morality claims to hold.

IV. Institutionality and Inescapability

Joyce’s view that morality is conceptually non-institutional is based on the type
of inescapability that is essential to moral thought. There is no question that
moral claims purport to carry a kind of authority that is distinct from other nor-
mative frameworks that are straightforwardly institutional, such as the rules of
chess or the rules of gladiatorial combat. Here, the rules in question cease to
claim any authority over my behaviour just as soon as I choose not to play. As
Joyce says, this is not how we think of morality. ‘Morality purports to have
more authority than this – it is not something that a person may escape (in
David Wiggin’s words) “by simply flying the skull and cross-bones and
renouncing altogether the aim of belonging to the moral community.”’29 We
are, if you like, moral conscripts, not volunteers.

76
 IN DEFENCE OF MORALITY

The alleged incompatibility of institutionality and inescapability hinges on

an understanding of what it means for a normative framework to be ‘institu-
tional’. For Joyce this means that its authority is underwritten by our choosing
to participate in that framework, such that that authority evaporates if we opt
out of that institution. Institutional norms are altogether non-binding, since the
authority of these norms is ‘not objective, but is constituted by our choosing
or deciding to think in a certain way’ (emphasis added).30 Certainly if by
‘institutional’ one means that a given normative framework can simply be
opted out of, then morality is not conceptually institutional, since this feature
is clearly incompatible with the inescapability implied in moral thought. Yet
this not what I had in mind when argued in the previous section that moral
thinking is institutional. All that I meant is that the object of our moral dis-
course is constituted by the understanding of those who engage in that dis-
Downloaded by [University of Sydney] at 15:32 13 March 2015

course. Nor is the ‘opt-out’ feature that Joyce has in mind a necessary
entailment when one recognizes a normative framework as institutional in the
sense that I have. Some normative institutions may carry authority that is con-
tingent upon our choices – e.g. those of chess and of gladiatorial combat – but
it does not follow that all institutional authority is contingent in this way.
Morality may be an institution, but it is an institution of a rather peculiar
kind, one to which we are related in an altogether different manner than the
way one may be related to the game of chess. It is not as though one chooses
her way into a moral life. Rather, our involvement in morality is so basic to
our form of agency as to be essential to it. Were one to operate outside of what
we are calling the ‘institution of morality’, he would be leading a life that is
radically different from the one that the rest of us lead. Moral standards are
‘inescapable’ in a rather differ sense of the term than that discussed so far:
these standards are indispensable to a proper understanding of what you are
doing. I draw here upon the work of Charles Taylor, in particular the argu-
ments contained in Sources of the Self where he emphasizes this sort of inesca-
pability.31 The question that Taylor raises is whether one could conceivably
live one’s life without morality as a framework for understanding the world
and his interactions with it. This is not a question of whether one is able to
imagine oneself flouting moral norms at every turn. The question is, rather,
whether you can make sense of your experience and action without appeal to
the sorts of distinctions between right and wrong, good and bad and so on, dis-
tinctions which we aim to articulate through moral language. The answer, Tay-
lor rightly argues, is surely ‘No’: such distinctions are genuinely indispensable
in this sense. The inability to do without them, he says:

is not like the inability to stop blinking when someone waves a fist in
your face, or your incapacity to contain your irritation at Uncle George
sucking his dentures, even though you know it’s irrational. It means
rather that you need these terms to make best sense of what you’re
doing. By the same token these terms are indispensable to the kind of

77
 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

explanation and understanding of self and others that is interwoven with

these life uses: assessing his conduct, grasping her motivation, coming to
see what you were really about all these years, etc.32

The point here goes to the task of rendering intelligible our thought and action.
And doing so requires recourse to moral considerations.
Basic aspects of human thought and action embody a sense that there is
something morally at issue, i.e. these aspects present themselves as responses
to external considerations which merit those responses. And with this comes
the possibility of going wrong: that is, that in particular cases we might fall
short of the standards that are indispensable to the intelligibility of those
aspects of our agency. This is most clear when considering our interactions
with other people. We respond to others when they address us, say by calling
Downloaded by [University of Sydney] at 15:32 13 March 2015

out our name, or at least we feel we had better have a good reason not to.33
We hold doors open for others, we slam on the brakes if a toddler wanders
onto the road, we do favours for others and sometimes apologise when we
cannot and so on. These ways of responding to other people are not simply
brute reactions to external stimuli. In each case we are responding to a sense
that other people lay a demand on us that we had better live up to, i.e. that we
would be failing if we didn’t respond in these ways. This is not to suggest that
the ways in which we interact with others are automatically self-validating.
Indeed, something like the opposite is true. Our responses to other people are
only intelligible on the basis that there are standards that are independent of
those responses – standards concerning what we owe to others – such that
these responses leave us open to potential criticism.
Moral reasons are constitutive of basic features of our experience and behav-
iour. This point gets lost in Joyce’s discussion, again because of the external
perspective from which he seeks to conduct the discussion. His focus is upon
after-the-fact portioning of blame, i.e. on what we would say to or about some-
one who has done something we take to be obviously wrong. The examples
that he uses are typically courtroom scenarios, where he says that we (as users
of moral concepts) want to insist that even the most recalcitrant of criminals
had a reason to comply with a given set of moral requirements. Morality is
treated as something that we observe from position that is external to it. This
sort of legislative conception of morality occludes the key point. For those of
us who participate in those basic modes of interpersonal relation, moral reasons
are indispensable to an understanding of what we are doing.
Still, since it is a key point of focus in Joyce’s argument that morality is
conceptually non-institutional, I want to respond to some of his claims about
our practices of portioning blame to others. In his account of these practices,
the discussion centres upon the attitudes that ‘we’ as basically moral people
take to those who are especially immoral, or who have behaved in a particular
immoral fashion.34 Joyce aims to capture the ‘“feel” of moral authority’ by

78
 IN DEFENCE OF MORALITY

making clear the emphatic manner of the condemnation we bring to bear on

such others.35 The authority thus elucidated is allegedly inconsistent with a
view that our moral condemnation is ‘merely’ (as Joyce would have it),
grounded in a given framework of moral assessment. Of course the question of
the moral attitudes that we adopt towards those who are seriously immoral is a
particularly complicated one, and so I do not intend here to offer anything like
an extensive discussion of these attitudes. I will simply use the remainder of
this paper to respond to Joyce’s reasons for thinking that those attitudes are
consistent only with a non-institutional reading of moral authority.
Joyce invites us to consider a criminal who has raped, murdered, tortured,
etc., and who insists that he sincerely wanted to do what he did, and that his
actions did not frustrate any other desires or interests. Suppose that the crimi-
nal has been caught, and now faces a judge. And suppose that in condemning
Downloaded by [University of Sydney] at 15:32 13 March 2015

the criminal and his actions, the judge makes explicit the fact his condemna-
tion is delivered from within a given framework by saying, ‘What you did was
wrong (odious, contemptible), according to our moral standards’, or ‘from our
moral point of view’. In making this qualification, Joyce argues, the judge
would be saying not too much but too little. This would be scandalous, he
says, as it would imply that the wrongness of the actions was merely relative
to our assessment of them rather than something belonging to the actions
themselves. ‘We, the audience, have our attention turned also towards our-
selves, in saying which I mean that both our emotive response and our sense
of the justificatory source for our feelings shift to encompass ourselves.’36 The
worry is that the force of the Judge’s condemnation is diminished in a way that
we would find intolerable when it is made explicit that his assessment draws
its authority from a given moral framework. The alleged weakening of the
force of the condemnation is inadequate to the emphatic appeal to moral
authority we are inclined to make when confronted with really bad people.
‘How much sounder might we sleep at night, were we confident that we
hanged the criminal because he did something objectively wrong, as opposed
to acknowledging that we hanged him because he did something we found
wrong.’37
The contrast drawn in this last quote seems to suggest that once we recog-
nize that a particular moral judgement is grounded in a given moral outlook,
then there must be an arbitrariness to that judgement such that nothing can be
said in its favour over judgements made by those who do not fully share that
outlook. To be sure, if we condemn an action, there is no getting around the
fact that we have formed a view – necessarily ours – that the act was wrong.
But this does not entail that our judgement is arbitrary. To come to the view
that a given act was wrong and should be condemned, just is to think that that
action was objectively (in one sense) wrong, i.e. that the act merits our con-
demnation. I suspect that Joyce’s view is that if we acknowledge that a particu-
lar judgement was made ‘in accordance with our moral standards’, then this
judgement becomes contingent upon our endorsement of those standards, and
79
 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

that this endorsement is ultimately arbitrary. But one need not accept this. By
the time one is in a position to endorse a given set of moral norms, this
endorsement is made on the basis of a sense that this is appropriate: that those
norms warrant our allegiance. This is to say that such an endorsement is made
from a moral point of view, and draws upon a moral sense of things that is not
simply chosen.
It is not true, therefore, that making explicit the fact that moral condemna-
tion is delivered from a given moral point of view scandalously draws one’s
attention away from the immoral actions and towards ourselves. This does not
follow, any more than an evaluative claim that is clearly institutional draws
our attention explicitly to our involvement in that institution. If I say that a
given move in a game of chess was illegal, then I am inviting you to focus on
that move, and on the rule that it violates. I am hardly directing your attention
Downloaded by [University of Sydney] at 15:32 13 March 2015

to my own subjective feelings or in any explicit way to the fact that you and I
are playing a game of chess. Viewing moral claims as institutional need not
entail a significant shift in their ‘justificatory source’ as Joyce suggests. The
basis for our condemnation of a given criminal act still draws on consider-
ations about suffering and well-being, about the value of human life, the
significance of dignity, and so on. The judge’s condemning an action with
explicit reference to ‘our moral standards’ needn’t draw protest, so long as she
is understood to be pointing to ideals that are central to our moral outlook, and
not to the fact that those ideals are ours.38 Joyce may be correct to say that we
would want this suffix eliminated, but this is because we don’t want the judge
to wrongly imply that the considerations upon which her condemnation is
based – those concerning the value of human life, for instance – are idiosyn-
crasies of our particular moral standpoint to which everyone else may be
altogether impervious.
Still, viewing moral judgement as institutional in nature does seem to entail
a contingency in the authority of moral claims: it has the implication that the
moral claims have no authority over those to whom moral considerations are
entirely alien. And Joyce thinks that this contingency does not square with the
emphatic manner in which we assert moral requirements, and in which we con-
demn those who violate them, since through our moral discourse ‘we employ a
notion of what “must be done” with which we purport to transcend the
contingent circumstances or constitution of those to whom we apply it’.39
Hence, ‘when it is made plain that this condemnation is (only) “by our moral
standards” then our feelings may be affected by discomfiting awareness of the
contingency of our judgment’ (original emphasis).40
But it seems to me to be simply wrong to say that the moral claims that we
raise upon others are irresponsive to a sense of how the addressee of those
claims is situated. We do alter our moral attitudes towards others when we
learn that moral considerations play either no role, or only a very attenuated
role, in their thought and behaviour. We do not treat children who have not yet
reached an age of moral maturity as we do adults. Nor are individuals who are
80
 IN DEFENCE OF MORALITY

deemed to be acutely pathological treated as the rest of us would be for a

given act: legal conventions that enable such individuals to plead ‘insanity’
rather than ‘guilty’ embody a recognition that the moral censure that we would
normally draw is misplaced in these cases. The standard patterns of moral
comportment towards others are significantly altered when we learn that we
are dealing with an agent to whom moral concerns are altogether alien.
Joyce denies this, insisting that our moral censure takes no prisoners: we
want to apply this censure to acts that violate moral norms no matter what,
irrespective of the agent’s moral orientation (or lack thereof). This position is
made to appear more plausible by the rather artificial way in which he handles
examples of thoroughly non-moral actors. We are presented with a criminal
who has committed murder or torture, and then takes an entirely cavalier atti-
tude towards the whole thing, assuring us that he wanted to do what he did,
Downloaded by [University of Sydney] at 15:32 13 March 2015

that doing so did not frustrate any of this desires or interests, that he doesn’t
mind being punished etc. I call Joyce’s treatment of these examples ‘artificial’
in that the figure is viewed as a mere statistical outlier, as someone who is
more or less like us, though admittedly with a set of desires that diverge signif-
icantly from the norm.41 But if we are genuinely dealing with someone who
‘cares not a jot for morality’, then this person is not like us at all. Anyone
who altogether lacks an operative sense of the contrast between right and
wrong action, and who has no regard at all for other persons, inhabits a form
of life that is a long way removed from that which the rest of us live. My sug-
gestion is that a recognition of this brings with it an alteration in the way that
we view such individuals, and in the moral attitudes we take towards them.
Still, the claim that moral censure is misplaced when dealing with the radi-
cally amoral is hardly uncontroversial. Some will no doubt want to join Joyce
in insisting that morality necessitates our approbation even (or perhaps espe-
cially) when confronted with individuals who are utterly unresponsive to moral
considerations. But the question now is, ‘What hinges on this?’ To the extent
that my analysis entails that this condemnation is misplaced, it turns out that I,
like Joyce, am an error theorist with respect to the very particular species of
moral discourse that involves morally condemning such individuals. But it
helps to remember that moral censure directed at the radically amoral makes
up only a very marginal sector of all moral talk. Focusing on the attitudes that
we might be inclined to take to such individuals, as Joyce does, is a significant
weakness, given that the error theorist’s aim is to capture something that is
essential to all moral discourse.42 If this is to provide a basis for a more gen-
eral moral error theory – if it is to show that all moral claims are false – Joyce
needs to (and indeed does) maintain that this disregard for the moral situation
of the condemned is not merely a feature of this very small set of moral
claims, but is an essential feature of all of our moral talk. But, again, this just
seems implausible. So much of our moral discourse only makes sense when
understood as a genuine conversation between moral actors. This includes
more everyday modes of moral censure, since part of the point of this is to
81
 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

enter into a dialogue with the wrongdoer in the hope of bringing him to a rec-
ognition of what he has done. Yet often our everyday moral talk has little to
do with ascriptions of blame, but consists rather in a kind of negotiation with
others as to how a given situation ought to be dealt with, or as to what an
appropriate moral characterization of the situation might look like. In many
cases these discussions are arrested by disagreement and argument, but even
here the conversation partners agree that there is something at issue about
which some kind of consensus might at least conceivably be arrived at. Far
from these modes of speech being oblivious to the moral positioning of one’s
interlocutor, they very much assume that he occupies a shared moral space
with the speaker.

La Trobe University, Australia

Downloaded by [University of Sydney] at 15:32 13 March 2015

Acknowledgements
I am grateful to Jack Reynolds, Norva Lo, Janna Thompson, Daniel Nellor and
Steven Churchill for their valuable feedback on earlier drafts of this paper. I
would also like to thank the La Trobe University Meta-ethics reading group, in
particular Ricky Sebold, Stan Wood and David Rowe, for points raised in
response to the general argument made here. I am also very grateful to two
anonymous referees for the particularly helpful comments on an earlier draft.

Notes
1 Mackie, Ethics: Inventing Right and Wrong, p. 23.
2 Ibid., p. 38.
3 Kant, Foundations of the Metaphysics of Morals, Sec. II.
4 See Joyce’s discussion of the rules of gladiatorial combat in The Myth of Morality,
pp. 34–7; and Mackie, Ethics, pp. 80–82.
5 See The Myth of Morality, pp. 34–42. Here Joyce uses the term ‘weak categorical
imperative’ rather than ‘institutional categorical imperative’, and ‘strong categorical
imperative’ rather than ‘non-institutional categorical imperative’. Joyce draws the
distinction between institutional and non-institutional categorical imperatives with
reference to Philippa Foot’s discussion of such systems of rules in her paper ‘Moral-
ity as a System of Hypothetical Imperatives’, pp. 305–16. See also Joyce’s ‘The
Error in “The Error in the Error Theory”’, pp. 519–34, for an account of the distinc-
tion: pp. 523–5.
6 Joyce, The Myth of Morality, p. 41.
7 I’ve taken the wording of this argument from Joyce’s ‘The Error in “The Error in
the Error Theory”’, p. 526.
8 Joyce, The Myth of Morality, p. 45.
9 Ibid. p. 46.
10 Ibid. p. 47.
11 The concept ‘phlogiston’ entered into the explanations of seventeenth- and
eighteenth-century natural scientists concerning combustion. It was thought that
phlogiston was a substance or an element that was stored in bodies and released

82
 IN DEFENCE OF MORALITY

during combustion. Phlogiston theory had been abandoned by the end of the eigh-
teenth century, when a consensus had been reached that there is no such substance.
See ibid., pp. 2–5, for Joyce’s discussion of phlogiston discourse.
12 Joyce, The Myth of Morality, p. 53.
13 Mackie, Ethics, p. 15.
14 See Joyce, The Myth of Morality, Chs 3–5.
15 Ibid., p. 51.
16 See McDowell, ‘Aesthetic Value, Objectivity and the Fabric of the World’, and
‘Value and Secondary Qualities’, both in his Mind, Value & Reality, pp. 112–30,
and pp. 131–50, respectively.
17 McDowell, ‘Value and Secondary Qualities’, p. 131.
18 Finlay, ‘The Error in the Error Theory’, pp. 347–69; and ‘Errors upon Errors’,
pp. 535–47. Joyce replies to Finlay in ‘The Error in “The Error in The Error The-
ory”’. Much of the exchange between Finlay and Joyce consists in claim and coun-
ter-claim of misrepresentation. I have endeavoured to ensure that my sketch of
Downloaded by [University of Sydney] at 15:32 13 March 2015

Joyce’s position avoids the distortions that he alleges against Finlay, mostly by stay-
ing close to the language that Joyce himself uses in his reply. (The interpretive
issues between them largely relate to Finlay’s choice of the term ‘absolute’ rather
than ‘objective’ to describe the kind of authority that Mackie and Joyce believe that
our moral discourse erroneously presupposes. I have gone with the term ‘objective’
here, hopefully steering clear of any controversy in my reading of Joyce’s argu-
ment.)
19 See Charles Taylor’s discussion of ‘constitutive’ rules in, ‘The Validity of Transcen-
dental Arguments’ in his Philosophical Arguments, pp. 20–33. (He discusses the
rules of chess specifically on p. 29.)
20 That such reasons are constitutive of the kind of behaviour in question is a point
that is (or at least appears to be) missed in Joyce’s discussion. In discussing his
own example of institutional rules – the rules of gladiatorial combat – he says that
it is ‘just a case of there being some set of rules which someone from the outside is
imposing on the gladiator’, and then a little later that ‘the difference between a
strong, Kantian categorical imperative and a weak, institutional imperative, is that
the former purports to bring with it a reason for action, while the latter does not’
(original emphasis): The Myth of Morality, pp. 36, 37.
21 I have drawn this point from Charles Taylor’s discussion of ‘subject-referring emo-
tions’ in Human Agency and Language, pp. 57–8.
22 Joyce, The Myth of Morality, p. 42.
23 Ibid., p. 51. The implication is that the claims of morality are necessarily (rather
than just contingently) false: and that there no possible worlds in which those
claims might be true. This marks an important point of disanalogy between the
external assessment of our moral discourse and an external assessment of phlogiston
discourse: the natural scientists who discovered that phlogiston was non-existent
were, of course, making an empirical claim about the actual world. This feature is
merely implicit in Joyce’s argument: at no stage does he make this explicit. None-
theless, the implication, it seems to me, is unavoidable. The substantive step in
Joyce’s error theory addresses what it means to have a reason for action. According
to his analysis, practical reasons are by their very nature tied to the agent’s ends.
24 Winch, The Idea of a Social Science and its Relation to Philosophy, p. 111.
25 Joyce, The Myth of Morality, p. 49.
26 See, for example, ibid., p. 124: Joyce argues that we all accept that there is such a
thing as instrumental deliberation. ‘It is the person who wants to go beyond this
picture (as opposed to developing its details) who has the explaining to do’.
27 Ibid., p. 49.

83
 INTERNATIONAL JOURNAL OF PHILOSOPHICAL STUDIES

28 This is one way of understanding what Joyce means when he says that when claim-
ing that one is morally obligated to do something, we appeal to what we think is a
‘“real” reason […] – one that he cannot legitimately ignore’ (original emphasis):
Ibid., p. 41.
29 Joyce, The Myth of Morality, p. 42.
30 Joyce, ‘The Error in “The Error in The Error Theory”’, p. 522. This is a direct
quote from Mackie, Ethics, p. 16.
31 Taylor, Sources of the Self. See Part 1: ‘Identity and the Good’, and in particular
Chapter 1: ‘Inescapable Frameworks’.
32 Ibid., p. 59.
33 I have borrowed this example from Christine Korsgaard’s Sources of Normativity,
p. 140.
34 See, for example, his discussion of ‘Gyges’ – the figure from Plato’s Republic – in
The Myth of Morality, p. 32; and in ‘The Error in “The Error in the Error Theory”’,
p. 524.
35 Joyce, ‘The Error in “The Error in the Error Theory”’, p. 525.
Downloaded by [University of Sydney] at 15:32 13 March 2015

36 Ibid., p. 527.
37 Ibid., p. 528.
38 This response to Joyce approximates that made by Finlay in ‘Errors upon Errors: A
Reply to Joyce’, Accepting some kind of relativity in moral judgement does not, as
Joyce suggests, commit us to saying that moral claims are indexed to particular
agents. Whilst moral claims are ‘relativised to our standards or ends, […] this is to
be read de re, not de dicto; i.e. if my relevant moral end is E then my moral claim
is to be interpreted as ought-relative-to-E, not as ought-relative-to-my-ends’ (p.
542).
39 Joyce, The Myth of Morality, p. 32.
40 Joyce, ‘The Error in “The Error in the Error Theory”’, p. 527.
41 See, for example, The Myth of Morality, p. 32: ‘Such criminals with such desires
are statistically rare, but they are surely possible’.
42 My complaint here is a particular variant of a more general problem with moral
error theory identified by Simon Kirchin in ‘A Tension in the Moral Error Theory’,
in Joyce and Kirchin (eds), A World Without Values, pp. 167–82. ‘In order to make
their position convincing moral error theorists must alight on a particular idea and
argue that it is a crucial commitment of everyday moral thought and language. Fur-
ther, in order to be sure that they can convict a commitment of error, error theorists
might need to specify that commitment so as to rule out alternative defensible
understandings. The danger is that the more one specifies the formulation of a com-
mitment, the more it is likely that the commitment is less than crucial to many peo-
ple’s everyday thought and language’ (p. 167).

References
Finlay, Stephen (2008) ‘The Error in the Error Theory’, Australasian Journal of
Philosophy 86(3): 347–369.
——— (2011) ‘Errors upon Errors’, Australasian Journal of Philosophy 89(3): 535–
547.
Foot, Phillipa (1972) ‘Morality as a System of Hypothetical Imperatives’, Philosophical
Review 81(2): 305–16.
Joyce, Richard (2001) The Myth of Morality, Cambridge: Cambridge University Press.
——— (2011) ‘The Error in ‘The Error in the Error Theory’’, Australasian Journal of
Philosophy 89(3): 519–534.

84
 IN DEFENCE OF MORALITY

Kant, Immanuel (2002) [1785] Groundwork for the Metaphysics of Morals, New
Haven: Yale University Press.
Kirchin, Simon (2010) ‘A Tension in the Moral Error Theory’, in Richard Joyce &
Simon Kirchin (eds). A World Without Values: Essays on John Mackie’s Error
Theory, Dordrecht: Springer, 167–182.
Korsgaard, Christine (1996) Sources of Normativity, Cambridge: Cambridge University
Press.
Mackie, J. L. (1977) Ethics: Inventing Right and Wrong, Harmondsworth, UK: Penguin.
McDowell, John (1998a) ‘Aesthetic Value, Objectivity and the Fabric of the World’, in
Mind, Value and Reality, Cambridge, MA: Harvard University Press, 112–30.
——— (1998b) ‘Value and Secondary Qualities’, in Mind, Value and Reality,
Cambridge, MA: Harvard University Press, 131–50.
Taylor, Charles (1985) ‘Self-Interpreting Animals’, in Human Agency and Language,
Cambridge: Cambridge University Press, pp. 45–76.
——— (1989) Sources of the Self, Cambridge: Cambridge University Press.
——— (1995) ‘The Validity of Transcendental Arguments’, Philosophical Arguments,
Downloaded by [University of Sydney] at 15:32 13 March 2015

Cambridge, MA: Harvard University Press, pp. 20–33.

Winch, Peter (2008) [1958] The Idea of a Social Science and its Relation to Philosophy,
London: Routledge.

85