be said about assessments of predictive power. It is diYcult to assess the
‘predictive power’ of folk psychology, since many of the reliable expec-
tations we form about the conduct of others (including the example which
Fodor cites, of arranging over the phone to meet someone at an airport)
might appear to owe more to social rules and cultural order than they do to
the application of folk psychology. (In this case the rule may just be: ‘If
someone utters the words ‘‘I will do A’’, then they generally do A’ –
nothing mentalistic is required.) Thus, when you hand over the money for
your fare to the bus driver you expect to get something like the correct
change back. But that expectation has little to do with any beliefs or desires
you might attribute to the driver. Myriads of mundane personal interac-
tions of this kind involve unthinking expectations to which social custom
has habituated us. When we have to treat other people less superWcially
and make predictions on the basis of their attitudes and thoughts, our
success rate may not be so spectacularly high.
To get a proper perspective on this matter one needs to appreciate that
in applying any body of general theoretical knowledge, predictive success
depends upon the quantity and quality of information available. Predictive
ability is not likely to be good when information is inadequate or when
reliance is placed on incorrect data. Living as modern humans do in vast
(urbanised and industrialised) societies, individuals are repeatedly being
brought into contact with strangers. There is not enough in the way of
background psychological knowledge to allow folk psychology to work
very well for many of these interactions. However, to a considerable extent
settled social roles and practices enable us to cope with these situations, at
least for a range of transactions which can be turned into social routines.
(So we could add thesocial-role stanceto Dennett’s list of ‘stances’, and
such a social role stance is very important to the way in which a large-scale
society works.) But this does not reveal a defect in folk psychology; much
less does it show that folk psychology is a false theory. All it shows is that
folk psychology has its limitations, particularly in regard to the infor-
mational demands it imposes. These informational demands are, as noted
above, very much less than those for the physical stance, but may still be
overstretched inXeeting contacts with strangers. This does not give us any
reason to think that we could not account for these strangers’ actions and
reactions in folk-psychological terms, if we only knew more about them
and more about what they wanted, valued, and believed.
The home terrain for folk psychology consists in the purposes which it
has evolved to serve, we believe (see chapter 4). Initially this would have
been for purposes of both co-operation and competition between humans
in the sorts of sizes of groups to be found, not nowadays, but hundreds
of thousands of years ago (at least). The mind-reading basics of folk
Realism and eliminativism 45