Interactions between Pavlovian first- and second-order conditioning.
Joan C. Bombace · Scholarworks (University of Massachusetts Amherst) · 1982
Four experiments investigated the interactions between Pavlovian first-and second-order conditioned stimuli in the rat CER preparation.The first three experiments examined the interactions of these kinds of stimuli with the Kamin two stage blocking procedure.The fourth experiment examined these interactions with Pavlov's summa- tion technique.Experiment 1 showed that second-order conditioning to an added cue was attenuated if such condi- tioning took place in the presence of a pretrained second- order conditioned stimulus (CS) .Experiment 2 failed to produce evidence for "super" second-order conditioning, i.e., rapid conditioning when a first-order conditioned inhibitor was compounded with a new cue during second-order condi- tioning.Experiment 3 showed that first-order conditioning to an added cue was attenuated if such conditioning took place in the presence of a pretrained second-order CS .vThis experiment also replicated the results of Experiment 1.Experiment 4 showed that 1) compounding a stimulus trained as a first-order inhibitor with a stimulus trained as a second- order excitor resulted in less suppression to the compound than to the second-order excitor alone and less suppression than that controlled by the excitor and a neutral cue that were compounded in a control group, 2) compounding two stimuli trained as second-order excitors resulted in greater suppression to the compound than to either stimulus alone and greater suppression than a compound of one of the excitors and a neutral cue in a control group.Experiment 4 failed to show summation when a second-order excitor was compounded with a first-order excitor.This failure was probably due to a conditioning floor effect.Results were discussed in terms of what assumptions would have to be made by the Rescorla and Wagner (1972) model in order to account for across-order interactions.An alternative interpretation suggested that across-order interactions did not necessarily account for the results.