202 Advances in spoken discourse analysis
‘textual’ rules. The assumptions made by the speaker in the choice of
prominence may be challenged by the listener; the world projected by the
choices will be a half-way house in terms of the overall patterns inherent
in any discourse: the varying convergence and divergence of the speaker’s
and listener’s worlds. Any proper explication of such choices will be a
pragmatic one.
This last remark would seem, on the face of it, to leave Brazil’s theory
somewhat deflated: if all prominence choices can simply be attributed to
the speaker’s decision to project things as ‘choices from known alternatives’,
then, so what? Yet it is precisely here that the most powerful link between
the interactive and the linguistic is seen by Brazil. All language is indeed
choice from alternatives, but prominence choices represent not selections
within the general paradigms inherent within the fixed, stable, shared language
system as a global whole, but within a narrow, everchanging set of paradigms
whose limits are discourse-internal, fixed by the interaction of the participants
themselves and shifting in real time. These paradigms are existential paradigms,
they represent ‘that set of possibilities that a speaker can regard as actually
available in a given situation’ (Brazil 1985:41). They are sets of choices
imposed by the real world and the interaction of the moment. This view
can be seen to operate in Oakeshott-Taylor’s example at (1) above and
repeated here:
4 A: What d’you think of JOHN?
B: Oh, JOHN’s all right
John is projected by the speaker as a selection from a list of possible
candidates towards whom he/she has positive or negative feelings; that list
does not have to be realized in the discourse; it may be, but it is essentially
a projected world. One of Brazil’s own examples amply illustrates the
converse:
5 Q: who sent you away?
A: an insolent ofFICial
where non-prominent insolent is projected as no selection, from no putative
list of alternatives (i.e. what else do you expect officials to be?). In the
two cases illustrated, the paradigms are distinct: John is not a selection
from the system-paradigm of ‘all proper male adult names’ and insolent
is not one from the ‘open class of adjectives relating to manner or behaviour’
in any real sense. John is from a projected paradigm of real-world candidates;
insolent is from a paradigm with only one member. Yet clearly, in the last
example, insolent could be substituted by cheeky or bolshy, for example
by a speaker reiterating another speaker’s use of insolent, a feature that
occurs often in conversation. If the reiteration is non-prominent it will
pass with minimal disturbance to the message. Brazil’s explanation for
this represents another key concept in his overall theory: that the ‘slots’
amenable to prominence or non-prominence are not word slots (in the