In their work McCulloch and Pitts describe an idea of representing all of
nervous activity in terms of propositional logic. This idea was quickly
challenged. One of reasons for this challenge was rising believe that logic is
unable to describe most of human cognitive processes. In this paper we will
analyse premises of original McCulloch and Pitts proposition. Following that,
we will ask about ability of symbolic (logical) systems to represent human
cognition. We will finish by analysing relation between symbolic and
subsymbolic computing, in hope of bridging the gap between the two.