Listening to people reading 213Both of these demonstrate the reader’s strictly limited involvement in the
material being read. Both readers—though for different reasons—see their
task as to ‘say what the text says’ without being concerned with what its
potential communicative implications might be.
The lack of engagement that such reading represents can, in fact, be
fairly easily related to the categories of the Discourse Intonation model and
to the significances attaching to them. Among the decisions that an engaged
speaker must make there are two kinds that are evidently interdependent.
They both concern the way linguistic items realize, or do not realize, existentially
significant selections.
Firstly, there is the matter of how prominence is distributed. In a minimally
engaged reading, this kind of decision is made simply on the basis of whether
the words are perceived as ‘content words’ or not, a distinction we can
provisionally think of as being inherent in the language system. In any
engaged use of language, prominences occur as a result of decisions over
and above those that result in the use of particular words; they are, in effect,
decisions about the communicative status of the item in a particular conversational
nexus. The maximally disengaged reader avoids making such decisions by
regarding prominence as something which is already assigned in the citation
form of the word.
Then there is the matter of division into tone units. In interactive discourse,
a form of words like ‘It was very grand to know somebody who owned a
circus’ might be divided in various ways, for instance
// it was VERy GRAND // to know SOMEbody who owned a CIRcus //or
// it was VERy grand to KNOW somebody // who owned a CIRcus //The decisive questions for the interactant concern how the selections realized
by the various prominences combine, so as to articulate oppositions of a
larger kind, and therefore mesh with present communicative needs. Doubtless
an engaged reading of the context would provide a reader with grounds
for selecting one of these versions or the other. The minimally engaged
reading, however, is made without benefit of the kind of information which
would incline the reader towards any particular way of assigning tone unit
boundaries. The fairly common one-tone-unit-per-sentence arrangement
that we find in such a reading can be thought of as arising from a reliance,
once more, on the reader’s recognition of a linguistic item, this time the
‘sentence’; but it is worth noting that other considerations might tip the
balance: in the case of the disengaged reading of verse, for instance, one-
tone-unit-per-line is probably more common, something that reinforces
the impression that the end result is a rather arbitrarily chosen fall-back
position, arrived at in the absence of those situational factors on which
speakers usually rely.