Reinforcing the occurrence or nonoccurrence of interim drinking

Four pairs of rats were studied in a yoked control design intended to determine if an interim activity (schedule-induced drinking) was sensitive to operant contingencies. Food was always presented on a fixed-time 30-sec schedule. Additionally, a positive or negative operant contingency was in effect during the first 6 sec of each interval. The positive (drink/food) contingency produced an extra food presentation at the 6th second of an interval if the lead rat drank at least once in the first 6 sec. The negative (no-drink/food) contingency produced an extra food presentation only if the lead rat did not drink in the first 6 sec. Two pairs of rats were first exposed to the positive contingency and then to the negative contingency. Two pairs received training in the reverse order. In drink/food training, all lead rats developed patterns of drinking that produced extra food presentations in most intervals. There were some indications that the positive contingency facilitated early acquisition of drinking, but the yoked rats eventually developed temporal distributions and asymptotic levels of drinking comparable to those that occurred in lead rats. In no-drink/food training, the two lead rats initially exposed to the positive contingency showed high levels of drinking inappropriate to the negative contingency, but the two lead rats initially exposed to the negative contingency showed appropriately low levels of drinking. The latter effects seem attributable to the no-drink/food contingency.