For much of the 20th century, thinking about signification, the production of meaning in language, was strongly influenced by the structuralist tradition, deriving from the semiotic theories of thinkers such as Peirce and Saussure (Buchler, 1955; de Saussure, 1974). Structuralist theory tends to view signification as the product of contrasts between the formal properties of signifiers, which can be modelled in terms of binary oppositions, e.g. ± voiced for the contrast between the sounds /d/ and /t/ in English. Signifiers are then arrayed in sequences according to rules governing permitted structural combinations (the language's syntax) to generate sentences with a determinate semantic content (Deacon, 1997). Others, however, have criticised the structuralist approach for ignoring the way in which language use is bound up with action in a social setting, arguing that to model it as a logical system for the exchange of neutral information misses out the fact that, in any episode of verbal interaction, some people's voices carry more weight than others'; that to understand language, in other words, we need to recognise how it is connected with questions of power, authority and inequality from which it cannot be divorced (Fairclough, 1989, 1992; Hodge & Kress, 1988, 1993; Milroy & Milroy, 1991). However, relatively little attention has been paid to the role of the dynamic aspects of speech communication, to intonation, pace and accental stress, to prosody in short as a constitutive feature of the utterance (Bolinger, 1986, 1989).

In this chapter, I will review three chief theoretical sources which help to amplify our understanding of language as a material activity practised by people as social beings. I go on to explore the significance for the relationship between language and society of the gestalt properties of speaking (such as intonation), which cannot be adequately modelled in terms of binary contrasts or encapsulated in a matrix of formal distinctive features.