Abstract
Dissociating the Three-Term Contingency: The Role of the Response.
IVER H. IVERSEN (University of North Florida)

In spite of its popularity, the three-term contingency is not well researched. Does the subject learn about the stimulus or does the subject learn about the response? In other words, does a subject learn that the stimulus signals reinforcement and that the absence of the stimulus signals absence of reinforcement, or does the subject learn to emit a certain response to the stimulus and withhold that response when the stimulus is absent? Rats were trained with food reinforcement to press a lever when a light turned on; pressing was not reinforced in the absence of the light (inter-trial intervals were 60 s). When the discrimination was acquired with prompt pressing when the light turned on and absence of pressing when the light was absent, the contingency was shifted to another response while the original stimulus was the same. Thus, when the light turned on pressing another lever now produced the food while pressing the original lever merely turned the stimulus off. Pressing the new lever was acquired quickly during the light. However, the rats also pressed the new lever at a high rate when the light was absent. Thus, the rats had to acquire the discrimination again when a new response was introduced; the rats had not leaned that absence of the light "signaled" no reinforcement. When the contingency was switched back to the original response the rats quickly switched the response. Interestingly, the rats immediately switched back to absence of responding during the absence of the light. This shows that the S-delta function of a stimulus is response specific. Additional responses were introduced in subsequent phases of the experiment while keeping the stimulus the same. In each case, the rats went through extinction of the new response in the absence of the light. The rats apparently never learned that the absence of the light signaled absence of reinforcement; instead the rats had to lean that the absence of the light was an S-delta for each response. Apparently the rats had to acquire the discrimination anew each time a new response is introduced even though the stimulus was kept the same. The results illustrate some of many complexities involved in the "simple" three-term contingency and suggest that rats do in fact no learn that stimuli "signal" reinforcement or absence of reinforcement.