Explanation versus accommodation: Reply to Bowers (2010)

Abstract

Bowers (2010) presented a critique of the account of masked priming presented by Norris and Kinoshita (2008) based on the Bayesian Reader model of visual word recognition (Norris, 2006). In this reply, we point out that: (1) Bowers' criticisms are based on his misunderstanding of both the Bayesian or ideal-observer approach, and the theoretical claims made in Norris and Kinoshita, and (2) Bowers' alternative account of the key data presented in Norris and Kinoshita relies on a series of ad hoc assumptions that are combined in different proportions to postdict any pattern of data.