The Predication Semantics Model: The Role of Predicate: Class in Text Comprehension and Recall
Althca Turner, Paul B. Andrcassen, Bruce K. Britton, Deborah McCutchen · RePEc: Research Papers in Economics · 1992
Previous models of text comprehension generally do not maintain a coherent set of propositions in working memory. Readers produce coherent summaries of texts, however, even when they lack extensive background knowledge. We present a computational model of text comprehension and recall that makes extensive use of predicate class information, which explains how readers extract the gist of a text. Predicate class determines the sequence in which propositions are reduced, and the rules that are used for gist extraction. A computer simulation of the model significantly predicted subjects ’ immediate recall and maintained coherence in working memory on most processing cycles. Predication Semantics Model 3 The primary assertion of this paper is that predicate class contains semantic information that readers use to make generally accurate predictions about the relative importance of a given proposition, and then which arguments to hold, pass to another proposition, or eliminate as they construct a summary representation of the text. We will show how different classes of predicates do more than define the relationship among each proposition’s arguments.