be the position adopted by Sperberet al., 1995a.) Or alternatively, those
processes may all be connectionist ones, as Evans and Over actually
suppose, in which norms of reasoningWnd no direct application. Second,
as regards explicit processes, the extent to which they are rational 1 may be
highly variable, depending on what the subject’s goals actually are. Simi-
larly, one would expect that the extent to which they are rational 2
might depend upon the extent to which the subject has undergone appro-
priate training and instruction, since explicit normative criteria are usually
arrived at through co-operative social enquiry.
To ask about the extent of human irrationality is to ask how often
people reason as theyshould, or as theyought. But this question takes a
very diVerent colour when applied,Wrst to implicit, then to explicit, proces-
ses of reasoning. For on reXection it is clear that, since implicit cognitive
processes are outside of our control, it makes little sense to ask whether we
oughtto reason diVerently in respect of such processes. All we can really do
is investigate the extent to which it is a good thing, from our perspective as
agents, that those processes operate as they do. And as for explicit proces-
ses, it seems equally clear that there is unlikely to be any such thing as a
basic reasoningcompetence, in the sense of an innate body of reasoning
knowledge. Rather, explicit reasoning abilities will vary with subjects’
diVerent learning histories, and with subjects’ diVering theories of nor-
mative rationality. It does, however, continue to seem important to ask
what normsshouldgovern our explicit reasoning – but here it is vital to
bear in mind the limits on our time, and the limitations of our cognitive
powers.
5.2 Against the Standard Picture of rationalityIf we ask how weshouldreason, when reasoning explicitly, most people
assume that the answer is obvious – of course we should reason in accor-
dance with norms derivable from valid logical systems. Call this the
‘Standard Picture’ of rationality, which sees rationality as coinciding with
the set of norms derivable from more-or-less familiar principles of deduc-
tive logic, probability theory, and decision theory. We follow Stein (1996)
in thinking that the Standard Picture is false, and should be replaced by a
conception of rationality which is relativised to human cognitive powers.
The main point is that it fails to take seriously the humanWnitary predica-
ment(Cherniak, 1986). For example, the Standard Picture would seem to
warrant normative principles such as this: ‘If you believe some proposi-
tionP, then you should believe any valid consequence ofP’; or this:
‘Before you accept some new propositionP, you should check to see
whetherPis consistent with the other things which you already believe’.
128 Reasoning and irrationality