Extended exposure to a discriminated, limited-hold temporal schedule does not produce stimulus control

Four rats were exposed to two different stimuli (either lights or tones), each stimulus being correlated with independent probabilities of water delivery in a temporally defined schedule. The schedule consisted of a 60 s T cycle with 30 s t(D) and t(-) successive subcycles; t(D) was correlated with a probability of water delivery of 1.0 and t(-) was correlated with a probability of water delivery of 0.0. The schedule was maintained during 180 sessions and extended for 25 extra sessions omitting the stimulus in t(-). The four rats showed low frequencies of responding, response frequency being slightly higher in t(-) than in t(D). The percentage of lost reinforcers was independent of response frequency. The rats which lost less reinforcers were those which obtained more water deliveries during the first 15 cycles of each session. These results show that stimulus control does not develop in limited-hold temporal schedules, and that response-reinforcer effectiveness may depend on the initial contact with reinforcers in the first cycles of the session.