Abstract:
We investigated how colour and location discriminations control choice when the likely availability of food changed according to time since an event. Six pigeons responded on a two-key concurrent free-operant psychophysical procedure where trials were either short or long, and a pair of keylight colours signalled each trial duration. The physical location of keylight colours was either held constant, or varied unsystematically across trials. Choice came under stronger control by the contingencies, and stabilised more rapidly, when the availability of food in a trial changed consistently across colour and location than when it changed consistently across colour only. Control by, and adaptation to the time-based change in food delivery location was further enhanced when the duration of time to the reversal was short instead of long. Food deliveries tended to reduce control by the time-based reversal, rather than enhance it. When stimuli provided conflicting information about reversal time, and congruent information about the direction of the reversal in probe trials, stimuli that had been associated with the availability of food in the first half of a trial controlled choice to a greater degree than stimuli associated with food in the second half of the trial, regardless of whether those stimuli signalled a shorter or longer time to the reversal. When a colour stimulus had consistently appeared on a particular location during baseline training, control by this stimulus was stronger when it appeared in probe trials in the same location than on the other location. When stimuli provided conflicting information about reversal time, the stimulus associated with food at shorter times generally exerted more control than did the stimulus associated with food at longer times. When stimuli signalled conflicting information about the likely location of food in the second half of the trial, there was no effect of reversal time, or of congruence between colour and location of colour. These findings suggest a tendency to use multiple congruent stimuli to discriminate a time-based change in contingency. We suggest discrimination of changes in likely food location across time is influenced by the availability of congruent stimuli, and by discriminability of the time at which a change occurs.