be appropriate epistemic practice, if the realist’s approach is correct. For if
scientiWc theorising is driven, not just by a concern to discover true
generalisations, but to uncover real causal mechanisms, then it may be
reasonable, for example, to revise rather than abandon a theory in the face
of recalcitrant evidence, provided that some story about the diVerent
underlying mechanisms involved is available.
So our advice to the philosophers is: look to the psychological evidence
of human cognitive powers, seeking to devise suitablypracticalepistemic
norms; and our advice to the psychologists is: look to the best available
philosophical account of the norms which are valid in a given domain,
before devising tests of human rationality, and making pronouncements of
irrationality. This is an area which is ripe for further inter-disciplinary
collaboration.
6 Conclusion
In this chapter we have reviewed the philosophical arguments in support of
human rationality, and have found that they are very limited in their eVect.
And we have reviewed some of the psychological evidence of human
irrationality, from which a more complex picture emerges. In all likelihood
human reasoning is conducted on a number of diVerent levels in cognition,
and within a variety of diVerent modular sub-systems. What we do insist
on is that questions concerning human rationality should be relativised to
diVerent domains – although until a reliable map of the modular structure
of our cognition emerges, we can perhaps only guess at what the relevant
domains are. And we also insist that if questions concerning rationality are
to have normativebite, then they must becomepractical, being relativised
to humans’ limited powers and abilities.
selected reading
For a general overview of philosophical issues concerning reasoning and rational-
ity see: Stein, 1996.
For more details on reasoning tasks and experimental evidence consult:
Manktelow and Over, 1990; Evans and Over, 1996.
For the argument from reXective equilibrium see: Cohen, 1981.
For criticism of philosophical attempts to secure a guarantee for human rational-
ity: Stich, 1990; Stein, 1996.
On the cheater detection hypothesis: Cosmides and Tooby, 1992.
For the application of relevance theory to the selection task: Sperber,et al., 1995a.
On the signiWcance of humanWnitude: Cherniak, 1986; Stein, 1996.
130 Reasoning and irrationality