3.1 The argument from interpretationThe most popular of such arguments has been the argument frominter-
pretation. This has several variants, depending upon what sort of inter-
pretation is claimed to require an imputation of rationality. It may be
maintained that taking people to be rational is a precondition of con-
sidering their doings to be genuine instances of agency, or of making sense
of their intentional states, or of taking them to be believers capable of
entertaining both true and false thoughts at all.
Something which should be conceded to this line of argument is that
genuine actions – as contrasted with bodily performances like snoring,
hiccoughing and blushing – are things which agents do for reasons. Deeds
of intentional agency all fall under the most general (and content-free) law
of folk psychology, that agents act in such a way as to satisfy their desires
in the light of their beliefs. So to take something a person does as an action
is to assume that it is done for a reason, whether or not we canWgure out
what that reason might be. In taking something to be an intentional action
we are assuming that it can be ‘rationalised’, in the sense that it can be
explained in terms of the agent’s reason for so acting. This point has been
repeated very frequently in the philosophical literature. Unfortunately,
many make the mistake of supposing that it establishes more than it really
does. If an agent acts intentionally, then the agent has a reason for so
acting. But that does not mean that the agent has agoodreason. The
action, or at any rate the attempt to act, needs to be appropriate in relation
to (some of) the agent’s own desires and beliefs. That gives no guarantee
that the beliefs themselves are rationally held.
In order to establish a deeper root for a guarantee of rationality it is
sometimes suggested that assuming subjects of interpretation to be
rational is a precondition of engaging in the practice of belief-ascription
(Davidson, 1973, 1974; Dennett, 1971, 1981). Taking people to be believers
just is taking them to be rational. This claim is regularly presented under
the guise of some such principle as theprinciple of charity– according to
which attributions of unwarranted and irrational beliefs to the interpretee
are to be avoided if at all possible. Our view is that any such principle has,
at best, a heuristic role and a defeasible status. That, of course, is the realist
view about intentional states. If anti-realism about beliefs were right, and
what a person believed wereWxed by the best available principles of
belief-ascription, then we would have to accept a guarantee of at least
limited rationality. But we have already found good grounds for rejecting
anti-realism about beliefs. So if you agree with the arguments advanced in
chapter 2, you will think that anti-realism has the relation of dependence
the wrong way round. Beliefs are not generated by practices of belief-
112 Reasoning and irrationality