LAUSR

Abstract In this article we examine how Neo-Gricean Pragmatics, Relevance Theory, and Integral Linguistics account for the code-inference distinction in natural language. Although the distinction between encoded and inferred meaning figures prominently in each of these theoretical frameworks, the way the distinction is articulated varies considerably. We compare how the three frameworks establish the code-inference distinction with respect to seven issues: 1) the extension of the code (its degree of underspecification) vis-a-vis its intension (the content contained in the code), 2) the criteria for distinguishing code and inference, 3) the differentiation of the pragmatic domain into nonce realizations and default senses, 4) the linguistic phenomena to which the code-inference distinction is applied, 5) the role of truth-conditions, 6) the levels and heuristics involved in the process of getting from encoded meaning to speaker meaning, and 7) the scope of the distinction in the overall study of language. On the basis of the comparison the article explores the main strengths and limits of the three theoretical frameworks.