Interference Paradigms for Meaningful Propositional Memory

Abstract

Interference phenomena in the learning, transfer, and retention of meaningful, factual propositions were investigated. Subjects first learned 20 sentences asserting a fact about a character denoted by a profession name, e.g., The lawyer paid his insurance. Recall was cued by giving the sentence-subject (A-term) for recollection of the verb phrase (B term of A-B sentence). Different groups then received one of four different interpolated tasks: new predicates (A-D), reshuffled predicates (A-Br); new professions and predicates (C-D), or no learning (Rest control). Retention of the initial A-B sentences was tested by fast-paced forward recall, then slow-paced full recall of everything to the stimulus (MMFR test), then backward recall, then pair-recognition memory. Interference theory correctly predicted all results. A-Br showed greater negative transfer than did A-D relative to C-D. Forward retention was best for the Rest group, then C-D, then A-D, and A-Br worst. Backward recall and recognition memory were equally high for the Rest, C-D, and A-D conditions, and poor for the A-Br condition. It was concluded that interference theory applies as well to meaningful propositional learning with gist recall as it does to verbatim memorizing in standard rote verbal-learning tasks.