Excitation and inhibition in unblocking.

In four experiments the nature of learning established with unblocking procedures in the appetitive conditioning of rats was examined. Measures of learning included response topography, effects of selective satiation, and summation and retardation tests of conditioned inhibition. One cue (A) was first paired with either a single unconditioned stimulus event, US1, or a sequence of two events, US1----US2. US2 was either qualitatively similar to (US2-Same) or different from (US2-Diff) US1. Then, a compound of A and a novel cue (X) was reinforced with US1 or US1----US2. Conditioning to X was blocked if either the single US1 or the US1----US2 sequence was used in both phases. If X accompanied an upshift in the reinforcer, from US1 to US1----US2, it acquired conditioned responding, especially when US2-Diff was used. Responding in the latter case was the consequence of both X-US1 and X-US2 associations. In Experiments 1-3, if X accompanied a downshift from US1----US2-Same to US1, it acquired conditioned responding that was based on X-US1 associations, but if it accompanied a downshift from US1----US2-Diff to US1, it acquired conditioned inhibition based on X-US2 associations. In Experiment 4, X acquired net inhibition at short US1-US2 intervals and net excitation at longer intervals, with downshifts from either US1----US2-Same or US1----US2-Diff to US1. However, the interval gradient was broader with downshifts from US1----US2-Diff. These data, and several other contradictory findings in the unblocking literature, are consistent with the views that (a) the surprising addition or deletion of US2 in unblocking experiments both facilitates the acquisition of excitatory X-US1 associations and establishes either excitatory or inhibitory, respectively, X-US2 associations, and (b) that the gradient of X-US1 facilitation is broader than that of X-US2 association.