CUED AND UNCUED TERMINAL LINKS IN CONCURRENT-CHAINS SCHEDULES

Pigeons were trained on a concurrent-chains schedule. The initial link s were concurrent variable-interval schedules arranged on two side key s. Each terminal link was a fixed-interval schedule arranged on the ce nter key. In cued conditions, different center-key colors signaled the two terminal-link schedules. In uncued conditions, the same center-ke y color appeared for both terminal links. Experiment 1 arranged unequa l initial links and equal terminal links. Preference for the shorter i nitial-link schedule was greater when the terminal links were uncued. Experiment 2 arranged equal initial links and unequal terminal links. Preference for the shorter terminal-link schedule was greater when the terminal links were cued. Although the results of Experiment 2 succes sfully replicated previous research, the results of Experiment 1 are n ot easily reconciled with conditioned-reinforcement or discriminative- stimulus accounts of the role of terminal-link cues. Rather, terminal- link cues appear to decrease sensitivity to initial-link contingencies .