84 A Grammar of Spoken English Discourse
lexical item, concluded that where the ‘context’ was biased in favour of the
dominant meaning of the lexical item their results favoured the ‘context’
sensitive theory, but where the ‘context’ was biased in favour of a subordin-
ate meaning their results supported the exhaustive theory.
Scholars such as Halliday (1994: 15), Hasan (1996: 100), Hunston and
Francis (2000), Matthiessen (1995: 5) Sinclair (1991: 104) and, from a
different tradition, Jackendoff (1997: 89) argue that lexis and grammar
are not distinct. Lexical items are not bricks which are joined together by
the mortar of grammar; they are instead an integral part of a unifi ed lexico-
grammar. Some support for this view is found in Hasan’s (1996: 74–9)
exploration of the semantic differences between nine lexical verbs (gather,
collect, accumulate, scatter, divide, distribute, strew, spill and share). She demon-
strates quite clearly that it is possible to establish each verb as an independent
lexical item which can be distinguished by a sequence of paradigmatic choices
in a systems network with major clause as input.
She argues that system networks exploring only the ideational metafunc-
tion^38 may be unable to distinguish all lexical items and suggests that lexical
items such as ask/enquire, buy/purchase, smile/grin, cry/bawl realize identical
sets of paradigmatic choices in the ideational metafunction but realize
different sets of choices in systems networks describing the interpersonal
metafunction. She further speculates that lexical items such as day/today
and two/both can only be distinguished by systems networks exploring the
textual metafunction (ibid. 99).
This section has demonstrated strong theoretical support for Brazil’s the-
oretical assumption that speakers select lexical items with communicative
values which are negotiated by the participants as the discourse unfolds.
The value of the lexical items depends both on oppositions existing within
the lexical system and the physical and verbal context.
3.7 Conclusion
This chapter has shown that there is theoretical support in the literature
for all of Brazil’s premises. Such theoretical support is vital because had
Brazil’s premises proved to have been unreliable there would be little
point in attempting to undertake the outward exploration of the grammar
proposed in Chapter 1. The division of speech into telling and asking incre-
ments was shown to be sound though its reliance on theories predicted
on the mutuality of knowledge was shown to be problematic. The key