According to the proposed Austinian version of the Linguistic Underdeterminacy Thesis, it is claimed that the conventionally specified phatic meaning of an utterance underdetermines its content and force. This Austinian version plays a key role in a contextualist model of verbal communication. The model is eliminativist pertaining to rhetic content and illocutionary force: it considers contents and forces as one-off constructions whose function is to map individual utterances in terms of their respective representational and institutional effects . The aim of this paper is to reformulate the Linguistic Underdeterminacy Thesis by making use of Austin's theory of speech acts. Viewed from the post-Gricean perspective, linguistic underdeterminacy consists in there being a gap between the encoded meaning of a sentence uttered by a speaker and the proposition that she communicates. According to the Austinian model offered in this paper, linguistic underdeterminacy should be analysed in terms of semantic and force potentials conventionally associated with the lexical and syntactic properties of the pheme uttered by the speaker; in short, it is claimed that the conventionally specified phatic meaning of an utterance underdetermines its content and force. This Austinian version of the Linguistic Underdeterminacy Thesis plays a central role in a contextualist model of verbal communication. The model is eliminativist with respect to rhetic content and illocutionary force: it takes contents and forces to be one-off constructions whose function is to classify individual utterances in terms of their representational and institutional effects, respectively.