Abstract

It is indeed important to identify the rich variety of systems for the adaptive control of behaviour, rather than squeezing this richness into a few boxes. We need to recognise both the variety of systems for the cognitive control of adaptive behaviour and to chart the relationships between such systems. But I shall argue that these projects are not best pursued by asking about the extent of animal rationality. The argument develops in three stages. The first outlines a picture of the selective regimes that drive the evolution of the sophisticated use of information by animal agents. The second argues that hominid cognition has evolved in response to a somewhat different set of " 3 " challenges and that the transmission of social information and skill has come to be both a critical and an unusual feature of hominid selective and developmental environments. The third draws upon the ideas of Dan Sperber and others in arguing that the social transmission of information introduces a vetting problem. I shall suggest that we see rationality as an evolved response to this vetting problem