<<

. 21
( 38 .)



>>

to perform that action in those circumstances, provided that the agent
were also to have some further naturalistic features. According to Brandt,
for example, these further naturalistic features would include being fully
informed and undergoing a rigorous course of cognitive psychotherapy.19
Is it a necessary consequence of such an account that any given reason
for an agent to perform a certain action in certain speci¬c circumstances
would provide a speci¬c degree of motivation? Take, for example, the reason
one might have to go to France for vacation. The relevant question here is
the following: ˜How motivated would one be to go to France for vacation,
if one were fully informed about what it would be like to go there, and
if one underwent a rigorous course of cognitive psychotherapy?™ It is very
plausible that there is no unique correct answer to this question. This is true
even though, if one somehow became fully informed and went through a
rigorous course of cognitive psychotherapy, it is at least plausible to hold
that one would then have some speci¬c degree of motivation to go to
France for vacation.20

18 Smith (1994), pp. 166ff has argued that if we hold that there are any normative reasons,
then we must also hold that the desires of rational agents will converge. It might seem
that if his argument went through, it would show that the suggestion of this chapter
must be false. But Smith (2002) has recently clari¬ed his position, and it now includes
something he calls ˜disjunctive reasons.™ The existence of these reasons means that even a
fully rational agent, on Smith™s view, might desire that his less than fully rational self do any
one of a number of mutually exclusive actions, and, importantly, that this desire cannot
be understood simply as the conjunction of two (or more) opposing desires of roughly
equivalent strength. Thus, even if Smith™s convergence thesis is true, it does not seem to
exclude, for practical purposes, the view suggested in this chapter. It is a testimony to the
attraction of the uniqueness assumption, that Smith seeks to preserve it by introducing
desires of such a strange and controversial character. Nor does Smith attempt to defend
the postulation of such desires, except by noting that (as they were designed to do) they
result in a more acceptable range of rationally permissible actions.
19 See Brandt (1979), pp. 10“15.
20 Throughout this chapter I grant the ideal motive theorist the assumption that actual people
do have speci¬c degrees of motivations for such actions. I call this assumption into question
elsewhere.

121
Brute Rationality

Consider an analogous question, which we can presume to be asked in
a certain place at a certain time: ˜If I were to drive to Boston starting now,
and didn™t have any problems on the road, what time would I arrive?™ If I
were to drive to Boston, there is no doubt that the time I would arrive there
would be a speci¬c time, at least up to the degree of vagueness involved in
determining when one has arrived in a city. But this does not mean that
there is a speci¬c time that is the correct answer to the question as posed.
Rather, a correct answer is more probably something of the following sort:
˜Between four-thirty and six.™ For the question does not specify the route
or the speed at which I would drive, or how long I would spend at rest
stops, or a host of other details that might make it more credible that there
would be a speci¬c answer. And the question ˜How motivated would one
be to go to France, if one became fully informed and went through a
course of cognitive psychotherapy?™ is also missing the sort of details that
would make it credible that there was a unique answer.
To make the point even starker, consider another question: ˜If I were
to go to Boston, what would be the age of the ¬rst person I talked
to there?™ Now, if I were actually to go to Boston and talk to some-
one there, there is no doubt that the person would have some speci¬c age.
But this does not mean that there is a unique age that is the correct answer
to the question. In fact, plausible correct answers are ˜Who knows?™ or
perhaps ˜Between zero and a hundred and ¬fty.™ Similarly, the fact that
one would have some speci¬c degree of motivation to go to France if one
actually went through cognitive psychotherapy provides no argument at
all in favor of the uniqueness assumption. Admittedly, there may well be
a connection between going through cognitive psychotherapy and one™s
degree of motivation to go to France. But this connection need not be
suf¬ciently strong to ¬x a unique degree of motivation. It may only be
suf¬ciently strong to ¬x a range of possible degrees of motivation. And the
admission that one would have some speci¬c degree of motivation after
the psychotherapy provides no argument whatsoever that the connection
is suf¬ciently strong to ¬x a unique right answer. After all, the ¬rst person
one talked to in Boston would certainly have some speci¬c age. What other
kinds of ages do people have? And yet it is clear that the answer to the
Boston question does not have a unique age as the correct answer, and
that there is no connection between being in Boston and the age of the
¬rst person one would talk to there.
The general point here can perhaps usefully be made in the language of
possible worlds. Consider the question ˜How motivated would one be by

122
The relation to other views

the prospect of a trip to France, if one became fully informed about it, and
went through cognitive psychotherapy?™ In using a framework of possible
worlds to answer this question, we ¬rst need to consider all the accessible
worlds in which the antecedent is true: those in which one becomes fully
informed and goes through cognitive psychotherapy. Then, looking at the
closest of these antecedent worlds, we read off the answer to the question
by seeing how motivated one is, in those worlds, to go to France. Because
there are a great number of ways in which to become fully informed, and
because there are a great number of forms that cognitive psychotherapy
can take, it is plausible that there will be a number of closest antecedent
worlds. If this is the case, there will be no unique possible world from
which we can read off a unique right answer. This is true despite the fact
that in each of the relevant worlds, one does have some speci¬c degree of
motivation. It is this last fact that stands behind the misleading truth of the
claim that if one were to go through cognitive psychotherapy, one would
have a speci¬c (as opposed to a nonspeci¬c) degree of motivation to go to
France. But the same sort of fact stands behind the truth of the claim that
if one went to Boston, the ¬rst person one spoke with would have some
speci¬c (as opposed to a nonspeci¬c) age. For, on this sort of analysis of
counterfactuals, anything that is true in all of the closest antecedent worlds
forms the basis for a true counterfactual claim. The fact remains that the
speci¬c degree of motivation in any one of these closest possible worlds
may well differ from the speci¬c degree in any other.
It may seem as though the preceding objection to the uniqueness
assumption is simply petty quibbling, and that a naturalistic ideal motive
account could deal with it by using phrases such as ˜other things being
equal™ or ˜in normal circumstances.™ But this is false. As Connie Rosati
has pointed out, there are many ways in which one might become fully
informed.21 And there are also many forms that cognitive psychotherapy
might take. The degree of motivation that a given consideration would
generate plausibly depends in a signi¬cant way upon these. For example,
it has been shown that one™s degree of altruistic motivation can depend in
a remarkably signi¬cant way upon whether or not one has recently found
a dime in a payphone.22 As a result, the naturalistic ideal motive theorist
who wants to keep the uniqueness assumption is in the unhappy position
of having to claim that there is a speci¬c privileged form of cognitive
psychotherapy, and a speci¬c privileged way of becoming fully informed,

21 22
Rosati (1995), p. 309. Isen and Levin (1972) referenced in Doris (1998), p. 504.

123
Brute Rationality

such that the degree of motivation produced by them, and by them alone,
is to be regarded as giving the normative strength of a reason.23 Moreover,
one point of naturalistic ideal motive accounts is to avoid dependence
on any prior normative notions. As a result of this, theorists like Brandt
cannot specify the acceptable forms of cognitive psychotherapy and full
information in terms of acceptable psychological results. In fact, this is
the basis of a more destructive objection to naturalistic ideal motive the-
ories. For nothing can guarantee that, after becoming fully informed and
after a full course of cognitive psychotherapy “ even of some privileged
sort “ some agents might not persist in having desires that are universally
regarded as irrational. These might be the result of brain tumors, chemical
depression, stroke, or some other physiological condition. Unfortunately
for the naturalistic ideal motive theorist, there does not seem to be much
reason to believe that there is a general naturalistic criterion that captures all
such potential causes of ineradicable irrational desires. Such a general cri-
terion would need to appeal to a normative concept such as ˜harm,™ or
would need to specify certain objects of desire as irrational, which would
make reference to cognitive psychotherapy an idle wheel in the account of
rational desire.
It seems then that, given the wide range of processes that could count
as cognitive psychotherapy, naturalistic ideal motive theorists also should
acknowledge the existence of a range of possible degrees of motivation
that a given consideration might produce even in an ideal agent.


ev i de nc e aga i n st th e un i que ne s s as sum p t i on
So far it has been argued that there is theoretical space for the rejection
of the uniqueness assumption in both naturalistic and nonnaturalistic ideal
motive accounts. By itself the existence of this theoretical space should
be taken to place the burden of proof on those who wish to endorse the
uniqueness assumption. For the claim of uniqueness is a much stronger
claim than its denial. But at the very least the demonstration of the exis-
tence of this theoretical space should clear the path for arguments against
the uniqueness assumption. The current section is an attempt to provide
one such argument. The strategy is as follows. Ranges have two de¬ning
features: a minimum, and a maximum. If we can ¬nd a reason for which

23 For similar claims, see Ripstein (2001), p. 46.

124
The relation to other views

the minimum does not seem to be equal to the maximum, this will provide
some evidence against the uniqueness assumption.
Consider the minimum amount of motivation that the prospect of
avoiding a certain fairly substantial amount of pain (say, the pain of a
badly burnt hand) would cause in a perfectly rational agent.24 We would
expect a rational agent to go to fairly considerable lengths to avoid such
an amount of pain: paying a lot of money, for example, or spending all
day driving. Indeed, if it were clear that the agent could avoid the pain
by suffering a lesser amount of pain that was still quite substantial (say, the
pain of a badly burnt ¬nger) we would expect the agent to do so. The
minimum level of motivation that the prospect of avoiding the pain of a
badly burnt hand would cause in a rational agent is therefore, intuitively
speaking, fairly signi¬cant. Now consider the maximum motivation that
would be provided by the prospect of saving a stranger from the same pain.
It certainly seems rationally permissible, if somewhat saintly, for an agent
to be as strongly motivated to prevent someone else™s hand from being
burnt as he would be to prevent his own hand from being burnt. That is,
it would not be irrational to cause one™s own hand to be badly burnt, if
this were the only means of preventing a similar injury to someone else.
Indeed, it would not be irrational to sacri¬ce one™s life to save the life
of another, or to risk one™s freedom for a chance of gaining the freedom
of another. In fact, we can make the following general claim with some
con¬dence.
M The maximum rationally permissible degree of motivation
provided by an altruistic consideration is always at least as
great as the minimum rationally permissible degree of moti-
vation provided by an equivalent self-interested considera-
tion.25
M is the weakest claim that guarantees that it would always be rationally
permissible to suffer some harm oneself in order to prevent the same harm
to someone else.
By itself, of course, M is compatible with the uniqueness assump-
tion. After all, the maximum motivation acceptably provided by an
24 In the following, I will sometimes omit quali¬ers such as ˜in a perfectly rational agent,™
or ˜perfectly.™ But it is always to be understood that, as SIM speci¬es, the relevant degrees
of motivation are supposed to be those that a perfectly rational agent could feel.
25 ˜Equivalent™ here means ˜involving the same probabilities of the same substantive harms
or bene¬ts.™

125
Brute Rationality

altruistic consideration might always be as great as the minimum moti-
vation acceptably provided by an equivalent self-interested consideration,
simply because these reasons might always cause the same unique degree
of motivation in an ideally rational agent. Yet, while we would certainly
expect a rational agent (let us assume an agent with the ¬nancial position
of an average university professor) to spend quite a lot of money to avoid
a great deal of pain for himself, we would not regard it as irrational for
someone to decline to spend this money to spare someone else such pain.
We do not, after all, regard it as irrational for people to decline to give
money to Oxfam, or to similar famine-relief organizations, when they
have no alternate plans for the money, and will only leave it to swell a
bank account that they are in no danger of exhausting. If this is right, then
the following must be true.
N The minimum rationally permissible degree of motivation
provided by an altruistic consideration is sometimes less than
the minimum rationally permissible degree provided by an
equivalent self-interested consideration.
But it has already been argued that the maximum permissible degree of
motivation provided by an altruistic consideration is at least as great as the
minimum permissible degree provided by the equivalent self-interested
consideration. If M and N are both true, then an altruistic considera-
tion must be able to produce a nontrivial range of acceptable degrees of
motivation.
One might try to avoid the above conclusion by claiming that the
appearance of a range of rationally permissible degrees of motivation for a
unique altruistic reason is really only the result of con¬‚ating two distinct
altruistic reasons. True, one might admit, it would not be irrational to
save someone™s life by sacri¬cing one™s own life. Nor, one might further
grant, would it be irrational to decline to give money to famine relief,
even though this would also save lives, and even though one had no alter-
nate plans for the money. But this may be true because, when we ¬ll
out the descriptions of these actions, we ¬ll them out in such a way that
the altruistic reasons are essentially different. For example, in the ¬rst case,
we may be thinking of sacri¬cing one™s life to save someone in a burning
building, while in the second we may be considering the possibility of
saving the life of a faceless stranger in a distant land. And it may be that the
reason to save a person burning to death right in front of one is different
from the reason to save the life of a faceless stranger in a distant land.

126
The relation to other views

Perhaps this is all true. But even if the reasons in the two cases are different,
this does not argue against the plausibility of ranges of rationally permissi-
ble degrees of motivation. For the maximum degree of motivation would
be, for both of the altruistic reasons in these two cases, essentially the same.
This is because any sacri¬ce that it would be rational to make for the sake
of saving someone in a burning building would also be rational to make
if it were the only way of saving a faceless stranger in a distant land. Thus,
even if one takes the relevance of distance or facelessness seriously, or if
one favors some other explanation for why one is not rationally required
to donate to Oxfam, it is still hard to defend the idea that there is always
a unique rationally mandated degree of motivation for any given reason.
On the contrary, suggesting that distance is relevant only supports more
strongly the view that some reasons can produce a range of rationally
permissible degrees of motivation. For the suggestion focuses attention
more narrowly on the reason provided by the prospect of saving a distant
stranger. And this reason seems clearly to have a low minimum and a high
maximum.
In fact, however, distance does not seem relevant to the normative

<<

. 21
( 38 .)



>>