Abstract

In two recent papers, Dan Sperber and Peter Carruthers have addressed the issue of cognitive flexibility, giving us different but somehow complementary accounts of it. Here I intend to focus on another cognitive mechanism which plays some role in allowing flexibility, and has been given little emphasis in their accounts. This mechanism is sensory imagination. In so doing, I have to confront with the assumption, which is widespread in the philosophical domain, that perceptual representations cannot convey any thought process. In the first place, I argue that this assumption rests more on the notion of accessibility than on that of systematicity. In the second place, I argue that – as an argument from the frame problem seems to show – accessibility of propositional representation has been largely overestimated; besides, there are reasons to think that systematicity and accessibility of perceptual representations have been largely underestimated. Those arguments are held to support the conclusion that people, and nonhuman animals too, can make use of sensory imagination as a cognitive strategy to confront with novel state of affairs. This conclusion fits well with a large amount of research in comparative psychology, and converges with current models of controlled thought processes which are based on the notion of a mental global workspace. The differences with the accounts of Carruthers and Sperber are briefly investigated