Reinforcement and Metacognition

Abstract

We return to the issue of reinforcement addressed by Smith, Beran, Couchman, Coutinho & Boomer. We argue that their concerns are unfounded because (a) reinforcing the ‘uncertain’ response need not undermine an otherwise convincing demonstration of animal metacognition, even for skeptics, (b) with or without metacognition, in the absence of the appropriate reinforcement contingencies, animals will not choose at all.