Appetitive-to-aversive CS-counterconditioning, positive transfer of excitatory conditioning
Johns, Kenneth Wade
Theories of Reciprocal Inhibition (Dickinson & Pearce, 1977; Dickinson & Dearing, 1979) state that a stimulus (CS) that predicts the occurrence of an appetitive unconditioned stimulus (US), that is an appetitive excitor, is the functional equivalent of a CS that predicts the absence of an aversive US (i.e., an aversive inhibitor). This functional equivalency should result in poorer performance (negative transfer) in appetitive-to-aversive transfer of training (i.e., CS-counterconditioning) procedures in which appetitive conditioning to a CS precedes aversive training to it. The magnitude of the deficit on the transfer is assumed to be directly related to the magnitude of the appetitive conditioning acquired. Experiments designed to test this theory have had mixed results, revealing positive, negative, and null transfer effect . To further explore the theory, the present experiments attempted to vary the strength of appetitive conditioning by introducing delays between the CS and the appetitive US. In Experiment 1, rats were given Pavlovian appetitive conditioning in which three auditory stimuli were either unreinforced (UN), or reinforced by a food pellet immediately (0 s delay) or at a delay of 10 s. Head poking in the foodhopper during the CS-US interval was recorded as the conditioned response (CR). Thereafter, the subjects were divided into three groups, each tested with one of the three appetitive CSs by giving CS-footshock pairings at a delay of 10 s. Suppression of operant VI-60 responding was measured as the CR for aversive conditioning. In Experiment 2, all delay were lengthened to 20 s and in Experiment 3, to 30 s. In spite of excellent temporal conditioning in the appetitive phases, if anything, positive, not negative, transfer effects were observed in the CS-counterconditioning phases. The results failed to support the Theory of Reciprocal Inhibition, but were more consistent with an affective-behavioral-cognitive model (DeVito & Fowler, 1994).