Citation:
Download PDF | 315 KB |
Abstract:
There is increasing consensus on the idea that certain sentences perceived as “ungrammatical” owe their status not to being syntactically ill-formed, but to their being L(ogically)-determinate and hence informationally trivial. Clearly, however, not every L-determinate sentence is perceived as ungrammatical, which raises the question of whether there is a principled way of sifting among the L-determinate sentences those that give rise to ungrammaticality from those that do not. Several interesting attempts have been made in this connection (Gajewski, Del Pinal), which, however, we argue fall short of the task. We propose a modification and generalization of such proposals based on the notion of ‘modulation’ of what are termed ‘the referential points’ of sentences (i.e. their non logical vocabulary and their variables). This approach has far reaching consequences for our understanding of the divide between logical and non logical vocabulary and for the very notion of semantic competence.