Differentiation versus unlearning of verbal associations.
Norman J. Slamecka · Journal of Experimental Psychology · 1966
Underwood and Postman (1960) theorized that learning of a list requires the unlearning of competing preexperimental associations to the list items. 3 experiments tested this assumption. Experiments I and II showed that obtained extralist and intralist preexperimental associations were not unlearned following the learning of one or two competing lists. Experiment III showed no weakening of natural associations after competing list learning, by latency measures. It was concluded that preexperimental associations were inhibited but not unlearned during list acquisition. The hypothesis was developed that strong associates are differentiated before they can be unlearned, and weaker associates are unlearned more rapidly than they can become differentiated. Underwood and Postman (1960) offered a theoretical account of verbal learning and recall which is based upon the traditional interference model, but which lays stress upon the role of existing language habits. Basic to their theory is the hypothesis that memorization of a list requires the unlearning of competing preexperimental associations to the list items. They recognize that S is no associative tabula rasa, but that he brings to the laboratory prepotent associative tendencies to the stimuli that will be encountered in the learning task, by virtue of long-standing language habits. If 5 must acquire new responses to items in the list, it is asserted that he must first break the older, preexperimental, associative bonds. To quote from Underwood and Postman (1960): As an illustration, assume that the word over is a unit in a serial list. Assume further that, due to long existing language habits,