Something out of nothing? Rethinking unarticulated constituents

In this paper we look into the notion of unarticulated constituent (UC) as characterised by Relevance theorists and the philosopher Francois Recanati. We show that postulating that constituents are added to the proposition expressed subject only to pragmatic restrictions, as they do, makes empirically wrong predictions and leaves a number of interpretation phenomena unexplained. We argue that the need for UCs arises inevitably for pragmatic approaches that subscribe to the thesis of the underdeterminacy of linguistic content – as we do –, but which view the syntax of the surface sequence as the only source of structural complexity. We defend an alternative view of how language articulates content and how it interacts with context which combines insights from both Conceptual Semantics and Dynamic Syntax which can accommodate the recovery of unpronounced constituents of the proposition expressed in terms of constraints on the construction of logical/conceptual representations along the same lines as the recovery of any content derived from overt linguistic input. We focus our discussion on three alleged major sources of UCs: implicit arguments, subsentential utterances and optional modification.

► Evidence of theoretical and empirical problems with unarticulated constituents.
► The notion of meaning articulation underlying free enrichment proposals challenged.
► An alternative model of natural language processing proposed.
► Alternative analyses of implicit arguments, sub-sententials and optional modification.