Blocking and the Prevention of Blocking from the Unconditioned Stimulus in Pavlovian Conditioning

Abstract

<p>Seven experiments are reported which demonstrate and analyze interference with autoshaping from unsignaled USs. In Experiment 1, eight groups of 10 homer pigeons each received conditioning sessions consisting of an 880 s wait followed by a train of USs spaced 10.5 s apart, followed by a second 880 s wait. For all groups except one, one of the USs in the train was signaled by a 10 s red keylight CS. The position of this single trial in the train was varied across these groups; for the remaining group, all the USs were signaled by the CS. Conditioning of the CS was shown to be a function of the position of the CS in the train. Specifically, conditioning was achieved only when the single trial appeared at or near the beginning of the train (referred to as the trial location effect) or, when all USs were signaled by the CS.</p> <p>The trial location effect was analyzed in Experiments 2 through 4 which showed that the interference from prior USs cannot be explained by any of the currently accepted explanations. Further analysis showed that the source of the interference is blocking of conditioning of the CS by the preceding USs which become established as signals of the next US.</p> <p>In Experiments 5 - 7 it is shown that blocking from unsignaled USs can be prevented through prior conditioning of the CS. It is argued that this result requires a view of conditioning as producing a long term change in the associability of the CS. A recent model of conditioning which encompasses such a change in associability is reviewed. A preliminary examination of this model shows that it might provide an account for the present results.</p>