A joint in nature between cognition and perception without modularity of mind
Is there a fundamental difference in kind between perception and cognition? In the 1940s, the "New Look" approach to perception argued that expectations can determine the contents of perception, giving rise to a crisis in epistemology and philosophy of science concerning the "theory-loadedness" of observation. At the height of this crisis, Thomas Kuhn's The Structure of Scientific Revolutions captured the ethos of this generation with its anti-realism about science. According to this book, the shift between Newtonian physics and Einsteinian physics is more like a religious conversion than a rational process. The "Modularity of Mind" movement later argued for the opposite view, that perception is an "informationally encapsulated" module that operates autonomously, putting epistemology on more secure ground. But now the consensus around modularity is breaking down, raising the issues of Kuhn's book all over again. This talk will argue that the debate of the last 75 years has had a crucial false presupposition, that a joint in nature between perception and cognition requires modularity of perception. The talk will focus on what perception is and how it differs from cognition. It will be argued that perception differs from cognition in the format of representations, the computations involving those representations and in the architectures of these different systems. More specifically, perception—unlike cognition—is iconic, non-conceptual and computationally and architecturally restricted.