200 Advances in spoken discourse analysis
‘part of the total pragmatic import of the message, arrived at by extracting
the “new” information from the set of “focal” items’. For the listener, nuclear
accents signal ‘new’ information, while non-nuclear accents attach to ‘potentially
new’ items whose interpretation depends on the listener’s ‘assessment of
the total context’.
Taglicht acknowledges that Brazil (1978a) perceives newness as relating
to ‘something like accent rather than to tonicity alone or to tonicity plus
syntax’ and Taglicht’s separation of intonation structure from the pragmatics
of information assessment to a certain extent moves the given/new issue
(as an aspect of information-status) to its rightful place as a feature of
context of situation, by virtue of including non-nuclear accentuation in
the debate. This partially echoes Brazil’s unified approach to accent placement
which deals with notions such as given and new under more generalized
headings. These headings account for choices of accent placement
independently of the choice of nuclear tone and so no nuclear placement
need be anomalous since the two features represent choices made within
different systems; the system of tone choice will be one thing, the system
of accent placement (within the tone group, not word accent), what Brazil
calls prominence, will be another. It is prominence which is at the higher
level: prominence is the domain within which tonicity operates; the rules
of prominence subsume all rules of givenness and newness. Brazil puts
it succinctly: the last prominent syllable in the tone unit is where tone
choice (i.e. choice of nuclear tone) operates and these ‘tonic syllables are
to be understood as constituting a sub-set of prominent syllables’ (1985:21).
This clearly and unambiguously separates the question of prominence
from syntactic or lexical concerns. In short, the difficulties inherent in
the tonicity/given—new debate arise because explanations of one set of
phenomena (nuclear tones) are being sought in features which belong to
an independent level of choice (prominence) and vice-versa, and anomalies
are created.
This last point is best illustrated by returning to examples similar to
those highlighted by Bolinger (1972a). The apparently anomalous:
2 I’ve got some BOOKS to sell
(uttered, for example, at a bring-and-buy sale) is not problematic if we
tackle sell as a choice of non-prominence within the prominence system. If
we can explain this non-prominence without recourse to syntactic or lexical
structure (as Brazil’s work does) then the tonic placement is not anomalous
to any norm and is dealt with in terms of its internal choice of, say, a falling
tone rather than a fall—rise, and not in terms of tonic placed on book rather
than sell. The converse utterance (by a customer in a bookshop):
3 I’ve got some books to SELL
now need only be considered in relation to a positive choice of prominence
on sell rather than book; the question of tone choice is identical with that