Behavior systems approach: A better explanation for sexual conditioning studies.

Abstract

General process learning theory has accounted for many instances of both instrumental and Pavlovian conditioning. The theory suggests that general laws of learning should apply across species, regardless of what stimuli are used or what response is measured. The literature on sexual conditioning provides a wide array of findings that indicate the importance of stimulus features, and the importance of careful consideration of the topography of the conditioned response to be measured. The present review addresses these issues in a sexual conditioning paradigm in which temporal contiguity (CS-US interval) and stimulus features were manipulated. The methodology also involved measuring numerous response topographies to maximize the likelihood of detecting learning. The findings of these experiments are discussed with regard to general process learning theory and the behavior systems approach. ********** General process learning theory assumes that the principles of learning apply across behavior systems and to many different stimuli and responses. This approach to the study of learning has had a long and distinguished history in the field of learning dating back to Thorndike, Pavlov, Skinner, Hull, and others (e.g., Bower & Hilgard, 1981). In particular, Skinner who was a general process theorist developed the operant chamber, presumably to encourage the study of general laws of learning by allowing scientists to study arbitrary stimuli and responses (Skinner, 1938). Despite the challenges proposed by many findings (Breland & Breland, 1961; Garcia & Koelling, 1966; Bolles, 1970; Seligman, 1970; Shettleworth, 1972), the general process learning approach has maintained its acceptance in the field of learning. When I first joined the laboratory of Dr. Michael Domjan, it was my impression that I would be adopting a general-process approach and that the research I would conduct would likely provide support for general laws of learning. Although I knew about findings such as the misbehavior of Pliny (Breland & Breland, 1961), selective associations in aversion learning (Garcia & Koelling, 1966) and others, I learned that these were exceptions to general rules of learning due to biological constraints on learning. That is, they were species-specific adaptations assumed to influence the manifestations of learning, not its mechanism. Thus, as a graduate student, I was excited that my research might demonstrate that general laws of learning also applied to a somewhat unconventional system, the sexual behavior system. Ironically, findings published from my dissertation (Akins, Domjan, & Gutierrez, 1994) would later provide strong evidence for alternative thinking about general process theory. This paper is a review of those findings, as well as more recent ones, and a discussion of how the findings have contributed to our understanding of theories that have been proposed to explain response systems. The CS-US Interval and Sexual Conditioning Temporal contiguity, a recurrent theme in the field of learning, suggests that two events have to occur temporally close together to become associated. In Pavlovian conditioning, temporal contiguity may be altered by increasing the time between the onset of the conditioned stimulus (CS) and the onset of the unconditioned stimulus (US), the CS-US interval. Although there are a handful of experiments that have provided evidence for learning at relatively long CS-US intervals (e.g., Garcia, Ervin, & Koelling, 1966; Kamin, 1965; Millenson, Kehoe, & Gormezano, 1977; Holland, 1980), the common finding has been that acquisition of responding is inversely related to the CS-US interval (e.g., Schneiderman & Gormezano, 1964). Long intervals result in poorer learning or the absence of learning. Only one investigation on sexual conditioning and temporal contiguity had been reported at the time I began my dissertation research in 1991. Zamble, Mitchell, & Findlay (1986) studied the Pavlovian conditioning of sexual arousal in Long-Evans rats. …