11 Listening to people reading
David Brazil
READING AND TALKING
Since the so-called ‘Discourse Model’ for the analysis of intonation was
proposed (Brazil 1975), a considerable number of people have sought to
apply its categories to what happens when we read aloud. Areas of interest
have varied widely, from the intonation of native learners when they read
out to adults in the very early stages of literacy to the performance of
actors, poets and others when they read verse aloud. Among the matters
investigated have been the different styles of reading that go with different
kinds of public communication on radio, television and elsewhere; the possibility
of using material read aloud as a means of identifying the special characteristics
of the intonation that foreign learners of English use is currently being
explored.
Personal contact with some of this work convinces me of the need to
look systematically at the great variety of activities that common usage
includes under the collective label ‘reading aloud’. Evidently, we cannot get
very far on the basis of a straight comparison between the intonation of
speech and the intonation of reading. We need, first of all, some way of
clarifying, in a way which is relevant to the task in hand, what readers are
doing on a particular occasion when they read aloud. By relevant I mean
that any typology that emerges should be capable of being mapped onto the
intonation in some way. There are many ways in which one might classify
the multifarious acts of reading aloud of which literate people are capable.
In trying to contribute to the essential task of systematizing the diversity, I
am concerned here with distinctions that have, or seem likely to have, regular
intonational consequences.
The discourse-based description of intonation which is set out in some
detail in Brazil (1985) was based upon the examination of a great deal of
spontaneous, naturally produced, data of various kinds. Its categories are
defined, moreover, in terms which make direct reference to the here-and-
now state of speaker/hearer convergence that is assumed to provide the
matrix for each successive increment in an interactive speech event. Its
central organizing principle could be said to be the notion of a step-by-step