Pun processing from a psycholinguistic perspective: Introducing the Model of Psycholinguistic Hemispheric Incongruity Laughter (M.PHIL).

Ambiguity processing was examined using a stimulus set consisting of homograph puns in which semantic salience, as measured by semantic co-occurrence, was manipulated. Two lexical decision tasks using puns as primes for ambiguous targets revealed that high co-occurrence meanings were processed faster than low co-occurrence meanings. A divided visual field protocol revealed involvement of both hemispheres, but with the pattern of priming from the right visual field more similar to that of the centrally presented condition than the left visual field pattern. In contrast to the lexical decision data that favoured high co-occurrence targets, data from a forced-choice relatedness task showed an advantage for the low co-occurrence associates. Results from this series of experiments are consistent with Bryden's [(1982). Laterality: Functional asymmetry in the intact brain. New York, NY: Academic Press] proposal that there are several different laterality effects when processing language and emotionally valent stimuli. The results are used to frame a working model of pun processing based on the Graded Salience Hypothesis [Giora, R. (1997). Understanding figurative and literal language: The graded salience hypothesis. Cognitive Linguistics, 8(3), 183-206].