A Cognitive and Cross-Linguistic Approach to Polysemy

Abstract Polysemy refers to the phenomenon that a word possesses multiple different but related meanings. Past accounts have provided descriptions of the relations that an extended sense may have to the most central or fundamental ("core") sense of a word, but they do not provide an explanatory account of how such senses are generated. The current research investigates the cognitive basis for the generation of polysemous senses. I hypothesize that extended senses are built on the salient characteristics of referents of core senses. I also argue that to the extent that speakers of different languages find the same characteristics of default referents salient, different languages should tend to generate similar polysemous senses from the same core senses. Evidence is provided for this proposed language-independent mechanism using data from six psycholinguistic tasks administered to speakers of two historically unrelated languages: English and Chinese. I also propose that, assuming a bidirectional interaction between language and thought, more embodied words should generate a larger number of extended senses and a higher proportion of senses shared between the languages. Evidence is provided for these predictions through the same empirical tasks. Broadly, this project provides a novel avenue for the study of word senses by revisiting the cognitive link between thought and language.