. 17
( 38 .)


into a deliberative community that shares ethical reasons. . . . But the device
can do this only because it is understood not as a device, but as connected
with justi¬cation and with reasons that the agent might have had; and it
can be understood in this way only because, much of the time, it is indeed
connected with those things.39

When Williams says “He ought to have done it, as moral blame uses that
phrase, implies there was reason for him to have done it,” he is not dissenting
from the implication. Rather, his point concerns the consequences of com-
bining the implication with internalism about reasons. The internalist view,
in sense (I) or (II), says that A will have reason to • only if •-ing will serve
a motive, or a desire, in A™s S. By virtue of the implication, it follows that if
•-ing doesn™t serve a motive, or a desire, in A™s S it won™t be true that A ought

38 Or, perhaps, the circumstances of your social context prevent you from seeing it? To what
conditions is the ability relative, if it is to play its role in (III)? That is a question to be
considered, although not here.
39 Williams (1995a), p. 16.
94 John Skorupski

to have •-ed, “as moral blame uses that phrase.” The ¬ction we are then
led into, Williams says, is that of treating A as though he really did have the
relevant ethical reasons. We are led into it because (sometimes, or often)
we want to blame people who “ by these internalist standards “ simply do
not have these ethical reasons. So, to safeguard the implication, we end
up ascribing the reasons to them anyway. Clearly, the same point exactly
can be made by the cognitive internalist. If we insist on ¬nding people like
Tom, who just lack the ability to see that considerations of gratitude have
reason-giving force, blameworthy for their ungrateful acts, we are going to
end up ascribing reasons to them that they just don™t have.
We could put the implication as follows:

(IV) If A has a moral obligation to • then A has reason to •.

However, this is an important point at which the ambiguity in the phrase
“A has reason to •” is relevant. So far we have not been distinguishing it
from “there is reason for A to •.” As we noted in §1, either of these can
refer in an appropriate context to (i) what there is reason for A to do, given
the facts or (ii) to what A is justi¬ed in believing there is reason for A to
do, given what he is justi¬ed in believing to be the facts. If the implication
we are considering is to hold, then we must understand “A has reason to
•” in sense (ii): for moral obligation and responsibility follow what we are
justi¬ed in believing to be the facts, not what the facts actually are. For
example, if I am justi¬ed in believing that you™ve just swallowed a glass of
petrol, I probably have a moral obligation to ring for an ambulance. But
if I have no reason to believe that, then I don™t. That remains true even if
(unknown to me) you have swallowed a glass of petrol.
So let™s now use the phrase “there is reason for A to •” in the epistemi-
cally unrelativised sense (i). Thus, if the building is about to collapse, there
is reason for us to leave, whether or nor there is any reason for us to believe
that it™s about to collapse. Similarly, if you have swallowed a glass of petrol,
there is reason for me to ring for an ambulance. And let™s use “A has reason
to •” in the epistemically relativised sense (ii). “A has reason to •” means
“A has a justi¬cation for believing that there is reason for A to •.”40

40 Some important points to note about having a justi¬cation: (i) A can have a justi¬cation for
believing something whether or not he believes it (ii) if he has a justi¬cation he is able to
recognize the justi¬cation (iii) if he has a justi¬cation for believing but does not believe “
even when the context requires a verdict on his part “ he may be open to criticism of his
rationality, attention, care, and so on. (iv) justi¬cation is relative to his actual circumstances “
not just his epistemic state, but the time at his disposal, the other things he has to do, and
so on.
Internal Reasons and the Scope of Blame

Understood in this way, (IV) is true.41 The connection is traceable to
the constitutive emotions involved in moral blame. Emotions in general
are not free-¬‚oating affects attachable to any belief; they have determinate
intentional contents. For example, it doesn™t make sense to be grateful for
an injury, or to resent genuine hospitality; someone who reacts in these
ways can only have a distorted view of what is being rendered to them. In
similar fashion, it is intrinsic to the intentionality of the emotions involved
in blame that you cannot reasonably blame people for failing to do what
they had no justi¬cation for thinking there was reason to do “ or indeed
for doing what they had a justi¬cation for thinking there was most reason
to do. What else should they have done?
This principle, internal to blame, gives rise to (IV) because moral obli-
gation and blame are connected: if A has a moral obligation to • then A
can be blamed for not •-ing.42 But the principle internal to blame says that
if A is justi¬ed in thinking that what there is most reason for him to do is
to refrain from •-ing then he can™t be blamed for not •-ing. Putting these
two together: if A has a moral obligation to • he cannot have a justi¬cation
for thinking that what there is most reason for him to do is not to •: he
must, rather, have a justi¬cation for thinking that there is most reason for
him to •.43 So he has a reason to •.
Suppose, then, that on the one hand we want everyone to fall within the
scope of blame, whereas, on the other, there are people who just don™t have
certain morally salient reasons. In that case “ if any of the three kinds of
internalism we™ve discussed is correct “ we must either resort to ¬ction or
accept that we cannot have what we want. If the position is as described,
then, by virtue of (IV), we shall want to “recruit” people into having reasons
which they do not have. Going in the other direction, but still relying on
(IV), Williams wants to say that since they have no such reasons they do
not fall within the scope of blame.
At this point, of course, someone might wish to deny (IV). But although
it has been disputed, we have just argued that it is ¬rmly rooted in the

41 So I may have a moral obligation to • even though there is no reason for me to •. For
example, in the case in which I am justi¬ed in believing that you™ve swallowed a glass of
petrol even though you haven™t, I have a moral obligation to send for an ambulance but
there is (“in fact”) no reason for me to do so. Nonetheless I have reason to do so.
42 Again, there are complications, turning in this case on how exactly one works in the possibil-
ity of extenuating circumstances: so assume for simplicity that no extenuating circumstances
43 Because there™s at least one thing A has most reason to do. The argument actually warrants
a conclusion stronger than (IV): if A has a moral obligation to • then A has most reason
to •. And that is often what is meant when it is said that morality is categorical. However,
only the weaker (IV) is needed for present purposes.
96 John Skorupski

intentionality of the emotions which are involved in blame.44 So we need
rather to ask why it should be thought that some people don™t have some
(morally relevant) reasons. It makes a difference whether the argument to
this conclusion starts from (I) or (II) “ or (III). Quite clearly, if (IV) is
combined with (II) some striking consequences will follow.45 By (II) there
is no reason for A to • unless A has some desire the satisfaction of which
will be served by •-ing, and by (IV), if A doesn™t have a reason to • then
A has no moral obligation to •. So A has a moral obligation to • only if A
has a justi¬cation for thinking that •-ing will serve the satisfaction of some
desire that he, A, has. But we don™t think that: we think that A can have a
moral obligation to • whether or not A has any reason to think that •-ing
will satisfy any of A™s existing desires.
Williams ingeniously attempts to mitigate the force of this point by
invoking what he calls a “proleptic” theory of blame. It appeals to the
“desire to be respected by people whom, in turn, one respects.” Blaming a
person who has that desire but otherwise has no motivation to avoid some
particular moral wrong is “as it were, a proleptic invocation of a reason.”46
It makes it true that he has a reason to avoid it, in virtue of his desire to
avoid blame when it comes from people he respects. For example, even if
I don™t have a desire to thank you for your good turn, I nevertheless have
reason to do so, because I desire the respect of respectable people, and I
won™t get it if I don™t thank you.
Unquestionably the desire to have the respect of people one respects is
a pretty important element in people™s psychology; but considered as a res-
ponse to the speci¬c dif¬culty we are at present considering Williams™
appeal to it is unconvincing. First, what is the basis of the respect that is
sought? If it is to be relevant to moral motivation, then it has to be that I res-
pect you as a good judge of when I deserve blame. (If I seek your respect be-
cause I respect you as a good judge of what™s cool or macho, say, that doesn™t
necessarily give me anything much related to a moral motivation “ on the

44 It can be disputed from a number of angles. There are, for example, accounts of practical
rationality (such as instrumentalism) which accord no automatic rational force to require-
ments of morality. This is what Stephen Darwall calls “morality/reasons” externalism in
Darwall, Gibbard, and Railton (1997), p. 306. Externalists of this kind have to break the
connection between blameworthiness and doing something you had reason not to do. In
a different way, some contemporary consequentialists about morality must reject (IV). For
they argue that one has a moral obligation to do that which there is, in fact, most reason to
do. (These consequentialists could still hold that acting against moral obligation is blame-
worthy only when the agent can tell he is transgressing, but they would have to deny the
view discussed below, that if one has a moral obligation one can tell that one has.)
45 Similar points would apply to the combination of (I) and (IV).
46 Williams (1995), p. 41. In this discussion, Williams seems to revert to a Humean framework.
Internal Reasons and the Scope of Blame

contrary, this may cut across moral motivation.) However, if I believe (II)
and (IV), and think I can tell what reasons I have, then when I ¬nd that
you™re blaming me for doing something I had no reason not to do I will
have to revise my assessment of you as a good judge of blameworthiness. So
proleptic blame will get leverage on what reasons I have only if I disbelieve
(II) or (IV). Second, what about the hard cases who do not care about oth-
ers™ moral respect? Williams says that these are beyond blame.47 But this is
a mistaken criterion of moral agency. It is not the desire to earn respect that
makes one a moral agent, but “ as the cognitive internalist rightly says “ the
capacity to recognize a morally salient reason as a reason.
Williams might reply that all of this only serves to highlight the ¬ctions
built into our moral practice. But it™s at least as convincing to say that it
highlights the falsity of (II). We saw that Williams provides no convincing
argument for (II); we also saw that a much more plausible case can be
made from the requirement of effectiveness that he places on reasons to
(III). Moreover (III), as against (II), ¬ts better with our views about when
a person simply falls outside the scope of blame. And (III), as we shall now
see, has the sort of implications for the scope of moral blame that Williams
wants to draw.


Suppose A lacks the ability to recognize X as a reason to •. So by (III) X
is not a reason for A to •. In that case, even if he is justi¬ed in thinking
that X obtains, that won™t justify him in thinking that he has reason to •. In
other words, he won™t have that reason. And if a person just doesn™t have the
appropriate morally salient reasons, then by (IV) that person doesn™t have
the appropriate moral obligations.
However this argument assumes that if X is not a reason for A to •
then A can™t be justi¬ed in thinking that it is. Yet might not A be justi¬ed in
thinking that it is, simply on the basis of what others tell him, even though
he can™t see it for himself (and, even though, therefore it is false that it is)?
The issues raised by this question call for more discussion than is possible
here. In general, obviously, I can have a reason to do something simply on
the basis that trustworthy people tell me that there is reason for me to do
it. (They might be sworn to secrecy as to the facts and can™t tell me why.)
The argument must be that in morally salient cases justi¬cation requires

47 Williams (1995), p. 43.
98 John Skorupski

personal insight, or at least the capacity for it. In these cases you must be
able to see the relevant reasons for yourself.
Cognitive internalism, at least about moral obligation and about morally
salient reasons, has been a driving element in modern conceptions of moral
agency. It is a central feature of what Kant and Hegel respectively named
the “autonomy” or “subjective freedom” of moral agency; in emphasising
it they were in¬‚uentially capturing an already in¬‚uential modern idea, con-
nected for example with increasingly accepted views about the signi¬cance
of conscience. The idea is that you have a moral obligation to act in a con-
crete situation only if you can tell for yourself that you have; more fully,
only if you have reasons to act that you can acknowledge for yourself as
being of such kind and strength as to render you blameworthy if you fail
to respond to them. You don™t have to be told. Moral agency is responsibility,
the capacity to respond spontaneously to morally salient reasons. It carries
answerability, accountability, in its train: if you can respond but fail to do so
you are answerable for that and can be held to account. Yet those only are
morally responsible who can recognize morally salient reasons and their
strength in generating moral obligation. A moral agent is an agent who has
conscience: the capacity for subjective, free or spontaneous, insight into
moral requirements.
The interesting point about (IV), which connects moral obligations with
reasons, is the way in which it links this “subjective” principle of responsi-
bility to the general doctrine about rationality and reasons stated in (III).
Moral obligations are open to spontaneous personal insight because reasons
in general are.
This doctrine of moral agency, Hegel™s “moral point of view,” is no
¬ction, nor, as we have seen, does Williams claim it to be.48 To reject it
would be to give up a conception of morality which so far at least remains
basic to modern ethics “ that morality is self-governance:

a man must possess a personal knowledge of the distinction between good
and evil in general: ethical and moral principles shall not merely lay their
claim on him as external laws and precepts of authority to be obeyed, but
have their assent, recognition, or even justi¬cation in his heart, sentiment,
conscience, intelligence, etc.49

48 As Hegel puts it, “the moral point of view . . . takes the shape of the right of the subjective
will. In accordance with this right, the will can recognize something or be something only
in so far as the thing is its own, and in so far as the will is present to itself in it as subjectivity,”
Hegel (1991a), section 107.
49 Hegel (1991b), section 503. Hegel thought that the modern ethical life which he favoured
would go beyond or supplement the “merely” moral point of view, but he also thought that
Internal Reasons and the Scope of Blame

Now this “modern” point of view, as stated so far, can allow that if you fail
to have the capacity of moral insight in some respects then you are not an
accountable agent in those respects. If Tom just can™t “get it” about grati-
tude then he is not blameworthy for his failure to live up to the moral norms
of gratitude. Fiction only begins to enter in when a distinct and substantial
ideal of modern ethical consciousness enters onto the scene: the demand
that every human being should be “recognized” as a comprehensively respon-
sible “ answerable, accountable “ moral agent, falling under the moral law.
By (IV) this means that every human being is fully capable of recognizing all
moral reasons, seeing them for him or herself. This powerful drive in mod-
ern ethical thought credits everyone with comprehensive moral agency “ at
the philosophical limit, with equal and absolute moral agency. Or at least
everyone after a few exceptional and, for most social and political purposes,
ignorable cases have been set aside.
But this may be false. Various gradients of moral agency are empirically
possible, in a diversity of respects, and we cannot say a priori how people
will be distributed across them. Suppose then that there are people who just
don™t get, can™t see, certain moral obligations. If we want to blame them for
not observing those obligations in what they do, we shall have to say “
contrary to the combined force of (IV) and (III) “ that they have reason to
observe them. Williams makes an interesting comment about that:


. 17
( 38 .)