Salford Linguistics Reading Group

Posted on October 20, 2012 by

From Dr Akis Kechagias:

The second Salford Linguistics Reading Group meeting for this term will take place on Wednesday, 24 October, from 14.45 to (roughly) 16.00 in Maxwell 819. Yannis Kostopoulos will speak about ‘Relevance theoretic semantics for relevance theoretic pragmatics. The case of and-utterances’, an abstract of which can be found below the fold. The presentation will be followed by a discussion around the topic.

“Relevance theoretic semantics for relevance theoretic pragmatics. The case of and-utterances.”

There is a widespread assumption in the study of linguistic semantics that the meaning of some natural language connectors such as and, or and if can be captured by formal devices of the propositional calculus, viz. &, ˅, → respectively. Nevertheless, the meanings that can be communicated by utterances containing these connectors go far beyond what their formal counterparts denote. The case of and is probably the most puzzling one. On one hand, and-structures can communicate a wide and fine-grained variety of meanings (see (1)-(3) below); on the other, the presence of and in an utterance seems to restrict the range of available interpretations (see (4)-(5)):

(1) I took out my key and opened the door.

(2) He gave up semantics and felt much happier.

(3) Stephen is a linguist and he can’t spell.

(4) Max didn’t go to school; he got sick.

(5) Max didn’t go to school and he got sick.

For several authors, the explanation of these phenomena lies in the fact that and has some extended or non truth-functional semantics (see, Cohen 1971; Bar-Lev & Palacas 1980; Txurruka 2003). According to others, the interpretation of and-utterances involves some special principles or maxims that deliver the intended meanings (see Grice 1967/1989; Dowty 1986; Lascarides & Asher 1993).

Against these alternatives, relevance theorists have proposed a pragmatic explanation for the divergence between the logical meaning of and, and the interpretations grasped in and-utterances. Maintaining the position that natural language and is the counterpart of the formal device, Blakemore & Carston (1999, 2005) have argued that both restricted and possible interpretations can be successfully explained on the grounds of the pragmatic processing that the and-utterance impose.

In this paper, I scrutinize the relevance theoretic approach arguing that although Blakemore & Carston’s pragmatic explanation deals successfully with a great range of data, their assumption on the semantics of and is rather problematic. The meaning of the logical connector does not contribute the appropriate input for the inferential process that Blakemore & Carston describe. Moreover, the assumption that natural language and is the counterpart of the formal connector & restricts the application of the pragmatic explanation only to cases of syntactic coordination. Aiming to a treatment of these problems, I suggest a modification of the relevance theoretic account that is theoretically more plausible, and deals better with the full range of data (discourse initial uses of and, and-parentheticals, and-conditionals).