Structure and process in semantic memory: A featural model for semantic decisions

Proposes a model to account for recent findings on the time needed to decide that a test instance is a member of a target semantic category. It is assumed that the meaning of a lexical term can be represented by semantic features. Some of these features are essential or defining aspects of a word's meaning (defining features), while others are more accidental or characteristic aspects (characteristic features). This defining vs characteristic distinction is combined with a 2-stage processing mechanism in such a way that the 1st stage determines the similarity between the test instance and target category with respect to both defining and characteristic features, while the 2nd stage considers only agreement between defining features. This model is shown to be consistent with most semantic memory effects, and 2 experiments on category size and instance-category verification using undergraduates as Ss provide further detailed support for it. (11/2 p ref) (PsycINFO Database Record (c) 2006 APA, all rights reserved).