Introduction

In the last chapter we looked at the interface between lexis and grammar. Building on this, we consider here the ways in which using corpora can promote a better understanding of the relationship between grammatical patterns and their contexts of use. We will set out to show that grammatical choices are rarely arbitrary and that pragmatic factors often account for particular ways of using grammar. As with so much of this book, we shall base our evidence on spoken corpora, largely because research into spoken grammar is still in many ways relatively young and overshadowed by research into the grammar of written language. To illustrate our points, we take three common structures and look at how they are used in everyday conversation, with occasional reference to their use in writing, for comparative purposes. The three structures are non-restrictive (or non-defining) which-clauses, if-clauses and wh-cleft clauses.

Non-restrictive which-clauses

This section is very much based on research by Tao and McCarthy (2001). They looked at the distribution and functions of non-restrictive which-clauses in two spoken corpora. They used a one-million-word sub-corpus of CANCODE and a 160,000-word sample of the Corpus of Spoken American English (CSAE). The CSAE project was undertaken at the University of California, Santa Barbara (see Chafe et al. 1991; see also appendix 1), and is composed of recordings made in a variety of settings, with a focus on casual conversation.