important possibilities as to what might be going on. The Ô¬Ārst is that
we are right to regard the consideration as not being a reason. Then our
desire, and the related action, may well be bizarre.36 But this need not be
because we lack a necessary normative judgment that would have made
the action intelligible. It may be because we are motivated by something
that is not a good reason, and that is in fact a strange source of motivation.
The second possibility is that we are motivated by a good reason that we
actively judge not to be a good reason. This is bizarre also. But in this case
what is left unexplained is the fact that we regard a consideration as not
being a good reason, when it is one. This is especially bizarre if we are
actually motivated by the reason.
But surely, one might say, if an agent regards a certain end as of great
value, then he has a reason to pursue it. Doesn‚Ä™t this show that norma-
tive judgments provide reasons for action?37 No. Dancy himself provides
the form of argument with which to deal with this sort of claim. His
presentation of the argument comes in response to the parallel claim that
having a desire must give one a reason, since we sometimes give advice to
people, based on knowledge of their desires, even though we think they
have no reason for those desires. Dancy‚Ä™s example involves a couple who
have decided to insulate their house so well that no sound can penetrate
from the outside.38 Suppose one holds that there is no reason to have this
desire, or to undertake this project. One might still advise these people
to buy a certain brand of insulation, rather than another, given their end.
This seems to show that a bare desire can ground a reason.
36 ‚Ä˜May well be bizarre,‚Ä™ rather than ‚Ä˜will be bizarre,‚Ä™ because of common motives like
revenge, envy, and hatred, which can make actions intelligible without seeming, even to
the agent, to provide a normative reason in their favor.
37 The view suggested here, which we might call ‚Ä˜the constitutive judgment thesis,‚Ä™ seems to
add the following claim to the judgment thesis: the normative judgments that are required
for rational action are not merely a necessary condition for such action, but somehow
contribute to the reasons for which the agent acts. This stronger view should not be
attributed to Dancy. It is presented here because it may appeal to other philosophers, and
because Dancy himself provides the model for an argument against it. Moreover, even
if normative judgments do sometimes contribute to an agent‚Ä™s reasons for action, this does
nothing to support the claim that normative judgments are always present in every case of
intentional or rational action.
38 Dancy (2000), p. 34.
Dancy‚Ä™s response to the above argument is to disallow the relevant
detachment in the following form of argument:
You ought, if e is your end, to pursue e in way w.
e is your end.
So you ought to pursue e in way w.39
The ‚Ä˜ought‚Ä™ in the Ô¬Ārst premise governs the whole of the conditional, and
not merely the consequent. As a result, modus ponens certainly does not
warrant the conclusion. And Dancy‚Ä™s claim is that no other inference rule
does either. Rather, the point of the Ô¬Ārst premise is to assert that there is
a certain kind of irrationality involved in having e as an end and failing to
pursue it in way w: a kind of irrationality that goes beyond merely having
e as an end in the Ô¬Ārst place. One does not have an extra reason, in virtue
of one‚Ä™s desire for e. Rather, one merely has an additional way in which
to be irrational: taking some other means to e than w.40
It should be clear how to modify Dancy‚Ä™s argument to deal with the
above suggestion that normative judgments can give reasons. Suppose that
someone regards switching radios on as a very good thing. If this is the case,
then there is something irrational going on when this person tries to turn
radios on by blowing on them. Because of this irrationality, one might be
tempted to make the conditional claim that if a person regards switching
radios on as good, then he has a reason to use his Ô¬Āngers to switch them
on. In a particular case, in which someone actually does make the odd
judgment that switching radios on is very good, one might be tempted to
use this conditional claim to infer that the person has a reason to use his
Ô¬Āngers to switch radios on. But this latter conclusion is as unwarranted
as the claim that the couple in Dancy‚Ä™s example have a reason to buy a
certain type of insulation. Regarding a nonreason as a reason does not
convert it into a reason. Nor does it give one a reason to pursue the things
‚Ä˜favored‚Ä™ by the nonreason. Rather, there is a form of irrationality that
consists in regarding something as a reason (rightly or wrongly), and then
failing to act in a manner that is consistent with this judgment. This does
not mean that one is acting rationally if one does act in a manner that is
39 Dancy (2000), p. 43.
40 For additional arguments against the same sort of detachment, see Greenspan (1975),
pp. 271‚Ä“74; Hare (1971), pp. 85‚Ä“89; Darwall (1983), pp. 15‚Ä“16 and 46‚Ä“48; Broome
(1999), pp. 409‚Ä“11 and 415‚Ä“17. Darwall points out that the detachment is at least more
plausible when the relevant ends are intended, and not merely desired, but both Darwall
and Broome hold that even understood in this way, the detachment is not valid.
consistent with one‚Ä™s wrong judgment. It only means one is avoiding a
Joseph Raz tries to provide some argument for a crucial assumption that
Dancy makes but does not defend: that a normative judgment is part of
the ‚Ä˜in the light of‚Ä™ relation.42 It will be easiest to discuss Raz if we Ô¬Ārst
make more explicit one of the main views that he is attacking, and that
this chapter is defending: the view that what makes a desire or choice
intelligible (but not necessarily rational) is a matter of brute regularity.
Because this will be the main view discussed in the present section, the
focus will be on intelligible action, rather than on rational action. But
Raz does not endorse the judgment thesis because he thinks that rational
action, as a distinct subclass of intelligible action, requires the making of
normative judgments. Rather, he endorses the judgment thesis because it
holds that all intelligible action requires such judgments. Because of this,
the arguments that follow bear directly on the judgment thesis.
In order to engage in argument against Raz‚Ä™s claim, the relevant notion
of intelligibility needs to be identiÔ¬Āed. After all, even the most psycholog-
ically opaque behavior may (at some future point) be perfectly intelligible
neurophysiologically. And even within a psychological theory of action,
once we stipulate that the agent has certain completely bizarre basic desires,
we may still regard the resulting action as intelligible, in a sense, if we can
see that it comes about through the normal operations of the mechanisms
that the theory of action postulates. What, then, is the sort of intelligibility
Raz has in mind? Raz cannot simply stipulate that action is intelligible ‚Ä“ in
the relevant sense ‚Ä“ only if we can see what the agent took to be of value
in the action, or only if we can see what the agent took to be reasons in
its favor. Characterizing intelligibility in this way would beg the question
in Raz‚Ä™s favor, since we cannot be expected to see what the agent took
to be of value in the action unless the agent did in fact take something in
the action to be of value. And it is one purpose of this book to defend the
claim that intelligible actions need not involve the agent‚Ä™s making any such
judgment. I think the relevant sense of intelligibility is the following: an
action is intelligible if there is a story about what motivated the action ‚Ä“ a
41 In this connection, see Lawrence (1995), pp. 137‚Ä“38.
42 See Raz (1999b), pp. 22‚Ä“45. Raz even uses the phrase ‚Ä˜in the light of‚Ä™ (p. 24).
story that would answer the question ‚Ä˜What was the motivation behind
that action?‚Ä™ in a way that would not leave us puzzled. This is extremely
close to Raz‚Ä™s own characterization of ‚Ä˜typical intentional actions,‚Ä™ which
he characterizes as:
actions about which their agents have a story to tell (i.e., actions manifesting an
internal viewpoint about what one is doing, or is about to do), a story which
explains why one acted as one did . . . a story which shows what about the
situation or the action made it, the action, an intelligible object of choice for the
agent, given who he is and how he saw things at the time.43
This characterization does not mention normative judgments at all, and
thus can be taken as common ground between Raz‚Ä™s view and my own.
In making the claim that brute regularity can make desire or choice
intelligible, the word ‚Ä˜regularity‚Ä™ is not meant to suggest anything that
could, even in principle, be formulated in terms of an exceptionless law,
or indeed in terms of a law of any sort. Rather, it is only meant to describe
events (or connections between events) that are sufÔ¬Āciently common that
we are completely unsurprised when they happen. For example, it is a
brute regularity that humans are motivated by the prospect of food, sex,
rest, intellectual stimulation, the novel, and so on. It is because we are so
familiar with these common kinds of motivations that we are unsurprised
by them. And because we are unsurprised when people act from these
sorts of motives, we are not puzzled by stories that explain their actions in
terms of these sorts of motives. Raz opposes this view. He thinks that such
brute regularities do nothing to make desire, or action, intelligible. Once
this view is rejected, Raz needs to Ô¬Ānd something else that can make an
action intelligible. What he offers is the fact that the agent sees certain
considerations as providing reasons to perform it. In his own words, he
that the central type of human action is intentional action; that intentional action
is action for a reason; and that reasons are facts in virtue of which those actions are
good in some respect and to some degree.44
One might uncharitably read these claims as suggesting that intentional
action is always action undertaken because of some fact in virtue of which
the action actually is good in some respect. But this is too strong a reading.
43 Raz (1999b), p. 24. That this is a characterization of intentional rather than intelligible action
does not matter, for it is clear from the passage, and surrounding remarks, that Raz is not
making any signiÔ¬Ācant distinction between the two at that point in the book.
44 Raz (1999b), p. 23.
Raz certainly allows that intentional action can be undertaken based on
the mistaken judgment that something is a reason. Such actions are not
counterexamples to his view, Raz writes, ‚Äúfor in the eyes of their agents
they are good.‚ÄĚ45 This implies that on his view intentional action only
requires that the agent make an appropriate normative judgment, whether
correct or incorrect. But such a view is too strong: intentional action does
not require any normative judgments at all.
Raz himself supplies three examples of psychological intelligibility that
is the result of nothing but contingent regularity, although, unsurprisingly,
he does not offer the examples under that description. The Ô¬Ārst example
is that it is intelligible that hunger makes concentration difÔ¬Ācult.46 One
might object to the use of this example against Raz, claiming that he only
intends it as an instance of an intelligible neuropsychological explanation.
But in fact the example is chosen to elucidate the meaning of ‚Ä˜intelligible‚Ä™
in the claim that morality is intelligible, so it cannot be interpreted in
this benign way. Now, what is it that makes it intelligible that hunger
makes concentration difÔ¬Ācult? As Raz points out, we are all sufÔ¬Āciently
familiar with hunger and its effects when we are trying to concentrate
to understand what someone means when they explain that the reason
they could not concentrate was that they had not eaten in many hours.
But this is simply appeal to a brute regularity, and to our familiarity with
it. Had evolution taken another twist, it might have been that hunger
actually sharpened one‚Ä™s abilities to concentrate. Had that been the case,
then we would understand why someone would be pleased that they were
hungry before an exam, rather than unhappy about it. One might try
to deny that brute regularity is what is doing the work in this case. For
example, one could claim that the intelligibility here depends on nothing
more than the connection between concentration and distraction, and
between distraction and unbidden sensations, i.e. ones whose occurrence
is not, or is not entirely, under our volitional control. But why should
the occurrence of unbidden sensations constitute a distraction, and not a
spur to greater concentration? Again, the answer is simply that this is a
brute regularity in human beings. One should not be fooled by the word
45 Raz (1999b), p. 25. One might take this qualiÔ¬Ācation to indicate that Raz holds a dis-
junctive view according to which all intentional actions are done either for reasons in
the sense of ‚Ä˜facts in virtue of which the action is good,‚Ä™ or for reasons in the sense of
‚Ä˜considerations that are good in the eyes of the agent.‚Ä™ This does not seem to be Raz‚Ä™s
view. But if it were, the argument of this chapter would go equally well against it, since
an agent can act for a reason that meets neither of these considerations.
46 Raz (1999b), p. 174.
‚Ä˜distraction‚Ä™ into thinking that there is a conceptual connection between
unbidden sensations and a diminished ability to concentrate, even if a
good deÔ¬Ānition of ‚Ä˜distraction‚Ä™ is ‚Ä˜difÔ¬Āculty in concentrating caused by
unbidden sensations.‚Ä™ For the presence of a word in the language can
reÔ¬‚ect a contingent regularity in human nature: ‚Ä˜grief‚Ä™ and ‚Ä˜jealousy‚Ä™ are
good examples of this. Admittedly, this entire discussion concerns the
intelligibility of an inability to concentrate. This is quite different from
the intelligibility of a desire or an action. But if brute regularity can make
the effect of hunger psychologically intelligible, it is at least more plausible
that the same sort of regularity can make desire and action intelligible also.
The second example of brute regularity contributing to psychological
intelligibility is the case of contrariness: acting because reasons point in the
opposite direction. Raz admits that some intentional action is the result
of contrariness, and that such action is more or less intelligible.47 How
is it rendered intelligible? ‚Äú[C]ontrariness is an established psychological
phenomenon.‚ÄĚ48 Although Raz does not acknowledge it, the appeal here
boils down to brute regularity. Perhaps it is true, as Raz notes, that we
cannot understand contrariness except as a degenerate case: a rebellion
against the normal case of being motivated by reasons. But even if this
is true, it also remains true that it is the brute regularity of contrariness
that makes contrary action intelligible, to the extent that it is. What else
could it be? It cannot just be the fact that contrariness is a relatively simple
variation on the standard case. For we could imagine a whole host of other
such variations. If those invented variations were not actually part of the
standard human repertoire, then appeal to such a variation, even if it could
be established to have taken place in some odd agent, would not render
that agent‚Ä™s action intelligible in any way. We understand contrariness as we
understand that hunger makes concentration difÔ¬Ācult: through experience
of the brute regularities in which it is manifested.
A Ô¬Ānal example of choice being rendered intelligible by brute motiva-
tional regularity involves the role of personality and character in choice.
Raz has the view that the available reasons typically underdetermine
47 Raz holds that such actions are not completely intelligible. For they are irrational, and
he holds that irrational actions cannot be completely intelligible. But he also admits that
irrational actions can be ‚Ä˜understandable‚Ä™ in a fairly robust sense: we know what they feel
like, we can predict them, they may not even appear irrational from the agent‚Ä™s perspective.
In the light of these admissions, his assertion that all irrational actions are unintelligible
to some degree seems theoretically motivated. See Raz (1999b), p. 35.
48 Raz (1999b), p. 33.
choice, leaving a wide Ô¬Āeld of options, all of which are rationally eligible.49
This view plays a major role in his account of the nature of the will, of the
explanation of moral supererogation, and in other places. The point here
is not to dispute the underdetermination claim, which is extremely plau-
sible. Rather, the question is the following: if the available reasons do not
determine a choice between A and B, how is the choice of A over B to be