No abstract

No abstract

Two differential eyelid conditioning studies employed grammatically correct and incorrect adjective-noun phrases as conditioned stimuli. For different groups of subjects, the nouns were either high or low in imagery. The hypothesis that congruency between grammatical correctness and reinforcement consequences (Le., the aversive stimulus contingent upon presence of incorrect rather than correct grammar) would facilitate conditioned discrimination was not supported, but the hypothesis that high noun imagery would facilitate differential response to syntax received strong support. Cognitive awareness of the syntactic discriminandum was also related to effective differential responding, as well as being implicated as a mediating mechanism in the imagery effects. Finally, performance was also significantly related to conditioned-response topography, with better conditioned discrimination by voluntary-form (V) than by conditioned-form (C) responders, and also evidence of more effective utilization of contingency awareness by Vs than by Cs.The present research had four major aims. The first aim was to determine whether a congruency effect previously obtained with arithmetic conditioned discriminanda (Fleming, Cerekwicki, & Grant, 1968;Fleming, Grant, North, & Levy, 1968) would likewise apply to verbal stimuli involving syntactic correctness as the differential cue. The arithmetic studies employed correctly and incorrectly solved arithmetic problems as differential conditioned stimuli (CSs), and found that discrimination was best when incorrect rather than correct problems were followed by aversive reinforcement, or when correct rather than incorrect problems were followed by positive reinforcement. On the basis of these data, Grant's (1972) conditioning model proposed that the learning of CS-reinforcement contingencies is facilitated to the extent that these contingencies are congruent with contingency codes already present in the long-term memory store (LTMS), in that instance, pre experimental codes relating reward and punishment to right and wrong arithmetic, respectively. This model would likewise imply that Requests for reprints may be sent to

No abstract

No abstract