Lexical ambiguity and its role in models of word recognition.

Abstract

This article reviews research concerned with processing words that have more than one meaning. Three principal models of ambiguity processing are outlined, and the research on which they are based is discussed. The findings seem best served by a hybrid model that allows for activation to accrue for all meanings, with the degree of activation sensitive to the frequency of the meanings and context. The major results of this field are interpreted in terms of the literature in word recognition. Such interpretation is important because of the guidance that such models may give to ambiguity research and because of the importance that ambiguous words hold for more general models of lexical access. Ambiguity is extremely common in natural language. A single sequence of words may yield more than one interpretation because of different possible relations among deep or surface structure elements or simply because one of the words has more than one meaning. This latter form-lexical ambiguity-has generated more research in the past decade than have the other types, not surprising considering the general preponderance of research at the lexical level of language processing (e.g., Becker, 1980;Meyer & Schvaneveldt, 1971). More than simply a by-product of the interest in lexical processing, however, word-level ambiguity may be viewed as a critical, special case of more general word recognition processes, capable of constraining lexical access models in important ways. There has been little attempt in the past, however, to integrate the lexical ambiguity research into more global models of word recognition.In particular, lexical ambiguity is relevant to the study of the role of context in lexical access, the process by which a presented word makes contact with its representation in longterm memory. Ambiguity normally goes unnoticed by a listener or reader, and it is assumed that the context allows the person I am grateful to Mark Ashcraft and four anonymous reviewers for their criticisms of earlier versions of this article and to Bill Easton for his help in its preparation.Requests for reprints should be sent to

Hemispheric Contributions to Lexical Ambiguity Resolution: Evidence from Individuals with Complex Language Impairment Following Left-Hemisphere Lesions

“…This finding is most easily interpreted as a deficit in inhibitory mechanisms which normally serve to dampen the facilitation of inappropriate meanings on the basis of integrated contextual information. It is acknowledged, however, that the exact nature and time course of inhibitory processes involved in lexical ambiguity resolution remains unknown (Simpson, 1984). The particular mechanisms involved in dampening meanings for ambiguities presented in a single-word context and a short ISI are especially difficult to delineate, given that inappropriate meaning deactivation has previously been conceived as a slower attention-based process of inhibition or suppression which occurs on the basis of sentencelevel information or meaning dominance (e.g., Gernsbacher, 1990;Burgess & Simpson, 1988).…”

“…The structure of lexical ambiguities has provided a critical platform for investigating the functional architecture of the language processing system; specifically, speaking to the issue of whether lexical processing is modular (context-independent) or interactive (context-dependent) in nature (e.g., Simpson, 1984). Briefly, the exhaustive access theory of lexical ambiguity processing holds that all meanings of an ambiguity are automatically activated upon its presentation, independent of contextual constraints, but this multiple meaning activation may only be witnessed within a brief temporal window.…”

The basal ganglia circuits, dopamine, and ambiguous word processing: A neurobiological account of priming studies in Parkinson's disease

“…According to the exhaustive access model, the occurrence of an ambiguous word in contextually loaded discourse (e.g., sentences or paragraphs) automatically activates all possible meanings in memory for a brief initial period, until the appropriate meaning is selected on the basis of integrated contextual constraints and the inappropriate meaning is rendered inactive through attentional withdrawal, active suppression, or decay (Simpson, 1984). The postlexical stage of meaning selection and inhibition may occur through strategic0controlled processing whereby limitedcapacity attention is allocated solely to the contextually appropriate meaning (Simpson, 1984) or the inappropriate meaning is actively suppressed (Gernsbacher, 1990).…”

Sources of sentence constraint on lexical ambiguity resolution

“…It is not obvious as to what variables are contributing to the empirical differences, and, although a number of methodological issues have been raised (e.g., lexical decision task, stimulus onset asynchrony, intralexical priming), these explanations appear to be inadequate (see Simpson, 1994, for a review). Consequently, several researchers have proposed hybrid models that incorporate certain aspects ofboth views oflexical processing (see, e.g., Dopkins, Morris, & Rayner, 1992;Duffy, Morris, & Rayner, 1988;Garnham, 1985;Rayner, Pacht, & Duffy, 1994;Simpson, 1984;Simpson & Krueger, 1991;Tabossi & Zardon, 1993).…”