Memory limitations of stimulus-response models.
Michael A. Arbib · Psychological Review · 1969
Patrick Suppes has suggested that just as higher mathematics may be deduced from set theory, so may automata theory be deduced from stimulus-response theory. In other words, he seeks to destroy criticisms of stimulus-response theory based on the need for an internal state to mediate between stimulus and response by showing that memory can be conditioned by a suitable schedule of reinforcement. The fundamental axiom of stimulus-response theory is that for each stimulus-response table there is a stimulus-response model that asymptotically becomes isomorphic to it in that there is a reinforcement schedule which will ensure with repeated trials. In fact, finite-memory systems are well studied by some mathematicians and computer scientists under the label finite automata. Suppes notes that Miller, Galanter, and Pribram objected to the conditioned reflex as the building block for a scientific psychology. They formulated a theory of plans in terms of test-operate-test-exit (TOTE) units, a plan then being defined as a TOTE hierarchy.