Logic in Cognitive Science: Bridging the Gap between Symbolic and Connectionist Paradigms

What can logic contribute to cognitive science? In the early days of cognitive science, logic was taken to play both a descriptive and a normative role in theories of intelligent behavior. Descriptively, human beings were taken to be fundamentally logical, or rational. Normatively, logic was taken to define rational behavior and thus to provide a starting point for the artificial reproduction of intelligence. Both positions were soon challenged. As it turns out, however, logic continues to be at the forefront of conceptual tools in cognitive science. Rather than defeating the relevance of logic, the challenges posed by cognitive science have inspired logicians to enrich the repertoire of logical tools for analyzing reason, computation, and communication. After a brief survey of the wide array of roles played by logic in cognitive science, we will examine the role of non-monotonic logics and complexity theory in more detail. Classical logic provides the groundwork for the abstract theory of computation, which in turn was used by Alan Turing (1950) to define the challenge of human level artificial intelligence. Turing proposed that the true test of machine intelligence is indistinguishability from human intelligence and suggested a concrete method for determining if this criterion is satisfied. A human judge sits in front of two terminals, one allows him to communicate with a computer and the other one with a human being. The judge can type whatever text he chooses into the terminals. His task is to use the answers he receives to decide which terminal is connected to a human and which to a machine. If the judge cannot tell the difference, the computer has passed the Turing test. The Loebner Prize for artificial intelligence is an annual competition which awards cash prizes to the computers which come closest to passing the Turing test.