Brains in the Hands of Computers

Abstract

The article defends with the help of the simple dilemma the belief that people, i.e. our brains, are in some particular sense a computer. Copeland (1993) claims that empirical differences between a computer and a brain do not reject the classic approach to the identification of a brain with a computer, i.e. the strong symbol hypothesis grounded in symbolic representation of our knowledge of the world. He argues that these differences merely show that our brains are not the traditional von Neumann computer. The connectionist approach to the explanation of our cognition fails in this as well. Despite applying the non-symbolic way of representing information about the world, which is supposed to be alien to computers, it does not succeed in preventing their identification with brains. We may still be computers, but of some other sort, a sort which is not possible only in principle as it is demonstrated by the most developed hybrid model of our cognition ACT-R, which combines both approaches. And even if we, for the sake of argument, accept the belief that a computer as a symbol manipulator is not compatible with neural networks, we are still not safe to claim that we are not computers in some general way.