| Abstrakt: |
Rats received Pavlovian aversive (shock) conditioning in which white noise was established for different groups as a CS+, CSO, or CS−. Then, in an appetitive T-maze discrimination, the CSs were presented contingent upon a designated correct response for which food reinforcement was factorially varied at 0, 1, 2, or 4 pellets. Although the CS+ suppressed and the CS− facilitated speed of running in the correct arm at the start of discrimination training, these effects extinguished rapidly and did not interact with reward magnitude. Furthermore, choice learning was faciltated by the CS+ and retarded by the CS−, with these effects being comparable for the 1- to 4-pellet reinforcement conditions, but absent for the 0-pellet condition. These findings are difficult to reconcile with a transfer interpretation positing a general signaling property of the CS and are better interpreted as across-reinforcement blocking effects: By predicting a preferred outcome (safety) comparable to the preferred outcome of food reinforcement, the CS− blocks (retards) the association of reinforcement and the S; conversely, by predicting a nonpreferred (shock) outcome discrepant from the preferred food outcome, the CS+ 'counterblocks' (enhances) the association of reinforcement and the S. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink