Scalar Implicatures: a Gricean vs. a Relevance Theory Approach

Abstract

Griceans have always supported the idea that scalar implicaturesare Quantity-based generalized conversational implicatures (GCI). With the purpose of explaining this phenomenon, they derived their own principles inspired in Grice’s Quantity maxims and concluded that whenever a speaker uses a weak linguistic item of a given scale, no matter the context she is in, she will be implying (a) that she is not in a position to be more informative and that therefore (b) a stronger item of the scale will not hold. For instance, if there is a scale such as <all, some>and the speaker chooses the weaker element some, she will automatically be implying not all. However, Relevance theorists believe that scalar implicatures are context-based. Thus, they claim that theseimplicaturesdo not depend on any Quantity maxim to arise but on a notion of relevance that considersthe context in which sentences are uttered. So although the speaker chooses the linguistic term somefrom the previously mentioned scale,the context should be taken into account to infer that she is implicating not all. Thepresent paper aims to compare both pragmatic theories’approach to the scalar implicature case. Consequently, I will first explain the main features of each theory: how human verbal communication works, how implicatures are generated and how the information contained in such implicaturesis recovered by hearers according to them. Then, I will introduce scalar implicatures and contrast the Gricean and Relevance Theory approaches (based on the previously explained features of each pragmatic tendency) to such phenomenon. To conclude, I will discuss what experimental pragmatics says about each proposal.The results from the application of each theory’s approach to the case of scalar implicatures indicate that Relevance Theory predicts their generation more accurately. In the same way, experimental pragmatics denies that the Griceansystem for generating scalarimplicatures isthe correct and most efficient one and verifies Relevance Theory’s predictions about it as well as its analysis of scalar implicatures.