ABSTRACT
The intellectual roots of traditional learning theories can be found in the principles of associationism advanced by the British empiricist philosophers of the 17th, 18th, and 19th centuries (e.g., John Locke, David Hume, & John Stuart Mill). They viewed the problem of how experience causes specific changes in subsequent behavior as essentially answerable in terms of mental links, called associations, between internal representations of contiguous events, that is, events that occur in temporal and spatial proximity to one another. Although Pavlov (1927) was primarily interested in the physiological underpinnings of learning, at the psychological level of analysis he too subscribed to contiguity as the central tenet of associationism despite his concern with the relative weakness of simultaneous conditioning compared to forward conditioning. Today, most theories of learning are couched in the language of associationism with the various theories being differentiated by their choices of qualifiers and additions to contiguity, e.g., whether reinforcement (Hull, 1943) or surprise (Rescorla & Wagner, 1972) are necessary in addition to contiguity and whether reactivation of the mental representation of a past event can act as a surrogate for the event itself (Gordon, 1983; Riccio & Ebner, 1981).
