Control of choice in conditional discriminations by sample-specific behaviors.

When sample-specific responding is occasioned by sample stimuli in matching and oddity tasks with pigeons, such responding controls the choices between comparison stimuli. This control was investigated in four experiments. In Experiments 1 and 2, differential responding to the samples in line-line matching and line-line oddity was reversed following acquisition of these problems. The reversal interfered with reacquisition of the same conditional discrimination but facilitated acquisition of the opposite discrimination. In Experiment 3, pigeons initially trained on line-line matching were shifted to hue-line matching. Positive transfer occurred when correct choices in both tasks were paired with the same sample-response patterns. Conversely, negative transfer occurred when correct choices were paired with opposite patterns. In Experiment 4, two concurrent conditional discriminations were designed so that sample-response patterns were paired with specific sample stimuli, but not with correct choices. Faster acquisition occurred when response patterns differed within rather than between sample dimensions. Furthermore, sample-specific responding controlled choices when the dimensional stimuli were difficult to discriminate but not when they were easy to discriminate. The combined results are interpreted in terms of overshadowing. Visual stimuli that control choice can be overshadowed by sample-specific responding if the latter facilitates conditional discrimination acquisition.