Discrimination training with positive and negative reinforcement surmounts reciprocal inhibition in appetitive-aversive interactions

In a stimulus-compounding test, separately trained discriminative stimuli ($\rm S\sp{D}s$) are presented singly and simultaneously. If more total test responses occur to compound than to single presentations, the result is called additive summation. Weiss and Schindler (1989) compounded $\rm S\sp{D}s$ after training one in a food schedule and the other in a shock-postponement schedule. Such crosspolar training resulted in the compound and single presentations controlling similar response quantities. This nonadditive outcome supports an extension of Konorski's (1967) characterization of heterogeneous (appetitive and defensive) reflexes as reciprocally inhibiting to discriminative operant behavior. The present experiment tested whether reciprocal inhibition in discriminative operant behavior could be attenuated by operating food and shock avoidance schedules in the same $\rm S\sp{D}s$. This bipolar training was conducted in conjoint or mixed schedules of both food and shock avoidance that operated in tone and in light. In the absence of tone and light, extinction was in effect. A subsequent stimulus-compounding test revealed additive summation in the five bipolar-trained rats tested. When other rats received crosspolar training in separate tone and light $\rm S\sp{D}s$, with their absence controlling response cessation through suspension of reinforcement schedules, additive summation was not produced in the two rats tested. In Phase 2, the effectiveness of food $\rm\underline{or}$ shock postponement (unipolar) schedules in tone and in light was assayed for the bipolar-trained rats. Where unipolar control was weak, remediation consisted of prolonged unipolar training, increased frequency of scheduled reinforcers, increased deprivation, and/or increased shock intensity. A subsequent stimulus compounding test reproduced additive summation. In Phase 3, bipolar and crosspolar training conditions were reversed. Subsequent stimulus-compounding tests revealed little or no additive summation for rats switched from crosspolar to bipolar and a small additive summation for a single rat switched from bipolar to crosspolar. In Phase 4, the original baseline schedules were restored and a subsequent stimulus-compounding test reproduced additive summation for bipolar-trained rats. These results demonstrated that discriminative operant reciprocal inhibition can be surmounted through bipolar training and that initial reinforcement class affected training duration and total test output in opposite ways during bipolar and crosspolar training.