Systematicity and Semantic Ambiguity

Abstract

The central issue in the study of semantic ambiguity has concerned whether all senses of a polysemous word are initially accessed as opposed to only the contextually relevant sense. As a simplifying assumption, many theorists assume implicitly that the content of each retrieved sense is static, containing the same information across retrievals. Other theorists make this assumption explicitly (e.g., Fodor & Pylyshyn, 1988, p. 45). But when investigators have assessed the stability of individual word senses, they have generally observed variability in the information retrieved. Reports of semantic variability in retention include Anderson and Ortony (1975), Anderson, Pichert, Goetz, Schallert, Stevens, and Trollip (1976), Barclay, Bransford, Franks, McCarrell, and Nitsch (1974), Geis and Winograd (1975), Greenspan (1986), Thompson and Tulving (1970), and Tulving and Thompson (1973). Reports of semantic variability in lexical access include Barsalou (1982), Conrad (1978), Johnson-Laird (1987), Tabossi (1988), Tabossi and Johnson-Laird (1980), and Whitney, McKay, and Kellas (1985). A review of variability in the representations that underlie decision making can be found in Kahneman and Miller (1986).