Redundant Scalar Implicatures

Suppose a listener hears ‘Some of the students passed the exam’ in a context where everyone already knows that not all of them passed. Should the mind still engage in the process of implicature calculation where encountering the word ‘some’ initiates a process of enrichment to some but not all, even though it is already part of the shared knowledge? The example above where the enriched meaning is already in the common ground highlights an ontological issue about the nature of scalar implicatures (SI): the term can refer either to the process of pragmatic reasoning or to the product, the enriched proposition itself, what Grice called the implicatum. In the common ground example, the product is effectively null because the enriched meaning is already part of the shared knowledge, yet the central question is whether the enrichment process can still be discerned to take place. SIs have been the focus of a long-standing debate over whether they are ‘default’ or ‘contextual’. Defaultism covers both the possibility that SIs are automatic enrichments of lexical meaning (Levinson 2000) and that they are achieved by the activation of a grammatical exhaustivity operator (Landman 2000; Chierchia 2004; Fox 2007). In the experimental literature, contextualism has predominantly been identified with a Relevance Theoretic (RT) approach (Sperber and Wilson 1995) in which SIs arise only if they are germane to the needs of the current discourse context (Noveck 2001, Geurts 2009). Our work focuses on an understudied prediction which potentially distinguishes default from RT accounts. As noted by Magri (2009), default accounts predict that SIs arise even when their content is informationally redundant. Using a priming paradigm, we explore whether scalar implicatures are actively computed even when not all is shared knowledge or whether the implicature is simply part of the interpreted meaning without requiring computation. The findings will contribute to open questions in semantics and pragmatics, specifically, the processing patterns and interpretive mechanisms of scalar implicatures.

Link to session