The inductive category formation framework, an influential set of theories of learning in psychology and artificial intelligence, is deeply flawed. In this framework a set of necessary and sufficient features is taken to define a category. Such definitions are not functionally justified, are not used by people, and are not inducible by a learning system. Inductive theories depend on having access to all and only relevant features, which is not only impossible but begs a key question in learning. The crucial (...) roles of other cognitive processes (such as explanation and credit assignment) are ignored or oversimplified. Learning necessarily involves pragmatic considerations that can only be handled by complex cognitive processes.We provide an alternative framework for learning according to which category definitions must be based on category function. The learning system invokes other cognitive processes to accomplish difficult tasks, makes inferences, analyses and decides among potential features, and specifies how and when categories are to be generated and modified. We also examine the methodological underpinnings of the two approaches and compare their motivations. (shrink)

Harry Harlow is credited with the discovery of learning set, a process whereby problem solving becomes essentially complete in a single trial of training. Harlow described that process as one that freed his primates from arduous trial-and-error learning. The capacity of the learner to acquire learning sets was in positive association with the complexity and maturation of their brains. It is here argued that Harlow's successful conveyance of learning-set phenomena is of historic significance to the philosophy of psychology. Learning set (...) is said to reflect the affirmation or rejection of hypotheses. Hypotheses are generated by the learner's brain, not its muscles. Thus, learning-set research served to advance the perspective that even nonhuman primates think and that their thinking reflects the active processing of information accrued from efforts to solve problems. Their learning processes are not simply the strengthening of some motor responses over others. Hence, learning-set research served to advance studies of animals as rational agents. This trend is serving to supplant the radical-behavioristic models, formulated earlier this century, with models predicated on rational processes for animals' complex learning and behavior. (shrink)

It has become customary in modern behavioristics to speak of stimuli as though they elicit responses from organisms. But logically this is absurd, for analysis of the grammatical roles of stimulus and response concepts shows that stimuli and responses differ in logical type from causes and effects. The "S elicits R" formula thus stands revealed as elliptical for a more complicated form of assertion. The trouble with this ellipsis, however, is that by suppressing vital components of formal structure in behavioral (...) principles, it has led to gratuitous assumptions about the environmental antecedents of behavior and seriously undermined the ability of behavior theory to assimilate the "higher mental processes.". (shrink)