General Process Learning Theory: Challenges from Response and Stimulus Factors

Abstract

Traditionally, general theories of learning have focused on associative and other mechanisms that are responsible for conditioned behavior without seriously considering how those mechanisms might vary depending on the stimulus being learned about and the response that provides evidence of learning. Recent studies of sexual conditioning in male domesticated quail have revealed both quantitative and qualitative variations in the functional properties of conditioned behavior depending on the response that is measured and the events or objects that serve as conditioned stimuli. For example, sexually conditioned sign tracking behavior is directly related to the ratio between context exposure (C) and trial duration (T) in a conditioning procedure, but sexually conditioned goal tracking is inversely related to the C/T ratio. Other studies have shown that conditioned stimuli that include limited cues from a female quail support different forms of sexually conditioned behavior than conditioned stimuli that lack female features. Furthermore, these various conditioned responses are differentially sensitive to extinction and reinforcer devaluation. The implications of these findings for general process learning theory are discussed. Investigators have found the allure of general process theory to be irresistible since the inception of the study of learning. The first modern learning theorists, Thorndike and Pavlov, both examined learning in a variety of situations. Thorndike, for example, used 15 different puzzle boxes in his research, and Pavlov examined both appetitive and aversive conditioning. However, Thorndike and Pavlov were not interested in how the unique features of a particular situation produced unique forms of learning. Rather, they were interested in the commonalities of learning across situations. Thorndike's famous Law of Effect, for example, states that the presentation of a satisfying event (S*) after the performance of a response (R) results in an S-R association that connects the response to the stimuli (S) present when the response was made. Thorndike knew that some responses were resistant to instrumental reinforcement and introduced the concept of "belongingness" to describe these exceptions. However, qualifications to the Law of Effect were ignored by learning theorists for more than 50 years after the original statement of the principle. All of the great learning theorists of the twentieth century focused on general mechanisms of learning that were stated in terms of abstract stimulus and response units. Hull, Tolman, Guthrie, and Estes all followed