Differences and similarities between scalar inferences and scalar modifiers: The case of quantifiers
Despite the rich theoretical and experimental work on scalar implicature, many of the studies on this topic were limited to some vs. all, neglecting the cross-categorial pervasiveness of the phenomena. The few experimental studies involving a more diverse group of scalar implicatures have found variation among expressions in the likelihood they give rise to scalar implicature, thereby challenging the assumption that scalar implicature (and generalized conversational implicature) is a uniform phenomenon (Doran, Baker, McNabb, Larson & Ward 2009; Doran, Ward, Larson, McNabb & Baker 2012; Van Tiel, Van Miltenburg, Zevakhina & Geurts 2014). This paper presents a first, systematic investigation of the degree to which a large group of quantifiers give rise to the implicature 'not all' using an utterance compatibility task with a modified Likert scale. Two accounts for the variation among quantifiers are proposed: (i) Shared semantic properties among three coherent groups of quantifiers account for the degree they give rise to upper- bound implicature; or (ii) the likelihood of an implicature is a function of the scalar distance between the various quantifiers and 'all'. The predictions these two accounts make are discussed, charting the way to a future investigation of the heterogeneity of scalar implicature.