Forgetting of verbatim information in discourse

Gregory L. Murphy, Amy M. Shapiro

Research output: Contribution to journalArticlepeer-review


In tree experiments, the causes of forgetting of verbatim information in discourse were investigated. In experiment 1, we tested the view that the surface form of a sentence decays when linguistic elements are integrated into larger discourse structures. Contrary to such a view, the results showed that when text could be integrated, both content and verbatim memory improved. In Experiment 2, we examined the possibility that task instructions could influence the level of memory for exact wording and content. The results showed that although task did influence the amount of verbatim memory, it did not affect memory for content. Experiment 3 was an investigation of the degree to which subjects would spontaneously encode surface information on the basis of the social interactiveness of the sentences. Past research has shown that verbatim memory for insults, jokes, and other personal utterances is quite good. In this experiment, identical sentences were tested in texts that had high versus low interactiveness. Verbatim memory was much higher for the same sentences in the highly interactive context. A pragmatic account of verbatim memory is given to explain how content and instructions influence the encoding of surface form.

Original languageEnglish (US)
Pages (from-to)85-94
Number of pages10
JournalMemory & Cognition
Issue number1
StatePublished - Jan 1994

ASJC Scopus subject areas

  • Neuropsychology and Physiological Psychology
  • Experimental and Cognitive Psychology
  • Arts and Humanities (miscellaneous)


Dive into the research topics of 'Forgetting of verbatim information in discourse'. Together they form a unique fingerprint.

Cite this