The generalized matching law as a predictor of choice between cocaine and food in rhesus monkeys

Abstract.

Rationale: The generalized matching law predicts that the relative rate of behavior maintained by different reinforcers will match the relative rate of reinforcement. It has previously been shown that responding maintained by either food deliveries or cocaine injections under concurrent variable-interval (conc VI) schedules is well described by the generalized matching law. However, the generality of this conclusion to the choice between a drug and a non-drug reinforcer has not been well established. Objective: The objective of the present study was to determine the extent to which the generalized matching law could account for choice between cocaine and food. Methods: Four male rhesus monkeys (Macaca mulatta) lever pressed under various pairs of conc VI schedules with food and/or cocaine injection as the maintaining events. Two doses of cocaine (0.025 and 0.05 mg/kg per injection) were selected to provide information about reinforcer magnitude. Results: As has been found in a context of choice between identical reinforcers, the generalized matching law accounted for most behavior. As in earlier studies with identical reinforcers, there was less responding apportioned to the alternative with the greater reinforcement frequency than predicted by the generalized matching law, i.e., undermatching was observed frequently. There was a tendency for more responding to be emitted toward the food alternative when the lower dose of cocaine was available and toward the drug alternative when the higher dose of cocaine was available. Conclusion: These results suggest that, as proposed by the generalized matching law, relative reinforcement rate is an important determinant of choice between a drug and a non-drug reinforcer.