Interactive lexis: prominence and paradigms 201in (2). When the two issues are clearly separated in this way, it can be
seen that many of the covering remarks made by analysts attempting to
explain tonic segments are glosses for a variety of features connected with
prominence and tone choice which need clearly to be distinguished. Terms
such as ‘contrastive’ and ‘emphatic’ (Crystal 1975:8–10), or tonic placement
reflecting ‘the decision of the speaker’ on what is given and new (Halliday,
1967:27), Crystal’s reference to matter ‘assumed by the speaker to be
familiar or obvious’ (1975:26), Bolinger’s ‘every peak is semantic’ (1961),
his ‘semantic and emotional highlighting’ (1972a) or his reference to matter
‘worth getting excited about’ (1972b:24), Pike’s earlier allusions to intonation
as ‘an indication of the attitude with which the speaker expects the hearer
to react’ (1945:21), Newman’s (1946) remarks around the same time on
‘expressive accents’, and some of the statements earlier quoted in this
chapter, are all examples that can lead to a confusion of the significance
of two distinct intonational choices.
PART II: PROMINENCE AND THE LEXICON
Bolinger (1961) had hinted at what Brazil was later to formulate rigorously
when he said: ‘in a broad sense every peak is contrastive’, reiterated by
Schmerling’s (1976) remark that ‘there is a sense in which anything
meaningful is contrastive’. Both unconsciously echo Trier’s dictum (1931,
cited in Lyons 1977:270) that anything we utter implies its opposite, a
statement seen by Lyons as ‘more relevant to the construction of a theory
of language-behaviour than it is to the analysis of a language system’
(ibid.). These quotations serve not merely to underline difficulties associated
with notions such as contrastiveness in relation to stress (what is there
which is not contrastive?) but also provide a background to Brazil’s view
of the matter.
Brazil says that prominent syllables embody ‘speakers’ choices from
known alternatives’ (1985:53); the corollary is that non-prominent items
represent no selection and a situation where no alternative is relevant.
This generalization will have to be expanded later, but it does, for the
present, enable us to see the breadth of Brazil’s definition and the essentially
interactive nature of prominence. Questions such as ‘given’ versus ‘new’
in terms of old and new information (whether formally or propositionally
recoverable in a discourse), ‘emphasis’ and ‘contrast’ of items in a discourse,
focal or thematic centres, or ‘information centres’ (after Grimes 1975:280ff)
now become subordinate to a speaker’s perception or projection of a context
of interaction in which he/she may choose to realize an item as though it
were a ‘given’ in terms of negotiated, shared meaning, or to realize it as
‘new’ in the sense of discourse-forwarding, unnegotiated, matter-yet-to-
be-fixed (or indeed something in need of re-negotiation). By this definition,
since it is a social projection of a context (Brazil 1985:47), prominence
will not be explicated with reference to truth-conditions, factuality or