Two experiments tested a set of predictions regarding category definitions and categorization latencies. Neither prediction was supported by the experiment results, leading to the formulation of an alternative feature-based model of category definitions using the notion of a polymorphous concept. (Author/AM)

Research supported by a United States Public Health Grant to Edward E. Smith. (VM)

This study examines two views of lexical storage and word production, a derivational hypothesis and an independent unit hypothesis. Reaction times and errors were related to derivational complexity. A model of lexical retrieval process incorporating derivational processes is proposed. (Author/SW)

Examines, through an investigation of the effect of word frequency on lexical decision, alternative views of how the internal lexicon is structured and searched. A model is developed based on two internal lists: one a ready-access list, the other a complete list. (Author/AM)

Describes a process model for sentence verification, including sentences expressing the implications of their antecedents. The model's distinctive features are the assumption that verification is based upon a focal element in the test sentence, and the comparison of this element with information corresponding to it in the antecedent…

A variable time delay naming latency paradigm was used to investigate the processing of noun-verb lexical ambiguities (e.g., "watch") in syntactic contexts that biased either the noun or the verb reading. Results support a two-stage model in which all reading of ambiguous words are initially accessed, followed by suppression of…

Two experiments explored how memory for new information is affected by awareness of parallels to pre-experimental knowledge. Results suggest that the benefit of prior knowledge derives from the more elaborate encodings that analogy promotes. (Author/AM)

Examines the reasons for differences in conclusions about the way conditional sentences are comprehended. (AM)

The "property-set model" is proposed for concept learning and subsequent recognition and classification of old and new exemplars. In an experimental evaluation of alternative models, the property-set model was the best predictor of both recognition and classification performance. (Author/RM)

Describes an experiment to test two assumptions concerning what is stored in the memory with regard to sentence structure: (1) that the linear effect observed is based on a tally model rather than on an integrationist mechanism; (2) that this linear effect is not uniquely the result of a mechanism operating at comprehension. (CLK)

Research supported in part by a United States Public Health Service Grant from the National Institute of Mental Health. (VM)

Presents a computer simulation testing semantic networks and spreading activation models of human memory. Describes how a sentence is encoded by building a working memory structure from its words and from semantic memory concepts related to its meaning. Retrieval processes use cue words or sentences to locate working memory structures. (Author/MES)

Two experiments investigated the relationship of verification to the answering of yes-no questions. Subjects verified simple statements or answered simple questions. Various proposals concerning the relative difficulty of answering questions and verifying statements were considered, and a model was proposed. (SW)

This study deals with the patterns of word-by-word reading times over a sentence when the subject must code the linguistic information sufficiently for immediate verbatim recall. A class of quantitative models is considered that would account for reading times at phrase breaks. (Author/RM)