Abstract
If A asks B “Do you like berries?”, and B replies “I like some berries,” B would infer that A does not like all kinds of berries. Such inference derived by negating the stronger alternative (all) is known as the scalar implicature (SI). Earlier experimental studies showed that computation of SI requires additional processing time compared to literal interpretation, and hence they argued that derivation of implicature is cost-demanding. Some recent experiments, however, found that derivation of implicature does not require any additional processing cost. The present study re-examines the comprehension of implicature using a Truth Value Judgement task. The hypothesis of this study is that the computation of implicature is as immediate as the computation of literal meaning if the sentences are preceded by prior context and communicative intent as in real conversation. The study uses a two-between-subject design where 32 native English speakers were required to read a ‘context’, followed by a ‘question’ and an ‘answer’. The context followed by the question either demanded the lower-bounded ‘literal’ meaning or the upper-bounded ‘pragmatic’ meaning of the under-informative answers which is the implicature. The result indicates that when a prior context and a clear communicative intent guide the hearer toward the intended meaning, both literal and pragmatic meaning comprehension is immediate. The result certainly indicates against the Default Inference accounts, but it also opposes the Literal-first hypothesis of the Contextualist school. The result strongly supports the Constraint-Based account of implicature derivation and brings additional support to the studies which argue for immediate implicature computation.
Publisher
Asian Research Association