The evolution of decoupled representation

Sterelny (2003) has developed an influential account of the origin of human-like beliefs, which complements Godfrey-Smith's (1996, 2002a) account of the evolution of cognition. In this thesis, I focus on the first part of Sterelny's account, which addresses the evolution of cognition in non-human animals. In this thesis, I defend Sterelny’s account against two criticisms. Christensen (2010) argues that Sterelny’s most complex behavioural control system evolves as early as the nervous system. Akins and Pollon (in preparation) argue that Sterelny’s simplest control system cannot explain a great deal of the most basic behaviour. According to both these arguments, the findings of empirical research demonstrate that Sterelny’s account requires major revision or should simply be rejected. I counter these arguments, and show that Sterelny’s account does not require major revision. Although Sterelny’s account does not require major revision, Christensen’s and Akins and Pollon’s criticisms show that there are few details of the continuity between simple and complex behavioural control in Sterelny’s account. Akins and Pollon’s criticism also shows that behavioural control often involves on-going coordination and that the boundaries between input and output are sometimes more indistinct than Sterelny’s account indicates. In this thesis, I incorporate on-going coordination into Sterelny’s account of behavioural control, and expand upon Sterelny’s account of continuity.