In the late 1950s, the hypothesis was advanced that human thinking is information processing, alias symbol manipulation. Like most new ideas, this one had many harbingers, especially those collected under the general rubric of cybernetics. Allen Newell and I gave some account of these precursors in the historical addendum to our Human Problem Solving (1972). What was new, beginning about 1956, was the translation of these ideas into symbolic (nonnumerical) computer programs that simulated human mental activity at the symbolic level. The traces of these programs could be compared in some detail with data that tracked the actual paths of human thought (especially verbal protocols) in a variety of intellectual tasks, and the programs’ veracity as theories of ...
Computational Theories of Cognition*
Computational theories of cognition