Analysing everyday conversation 149As the ‘third’ utterance is one tone unit, which means there can be no
exchange boundary within it, the problem of indeterminacy remains unsolved
and a decision must be taken by the analyst. Similar problems are discussed
in the next section.
DOUBLE LABELLING
One criticism often levelled at the Sinclair—Coulthard system is that it
assumes that each utterance or part of an utterance has one and only one
function (e.g. Open University 1981:23). So each act must be either, say, a
qualify or an informative, a move must be either eliciting or acknowledging,
and so on. Yet, the critics claim, in practice a single act or move can perform
two functions at once.
The system as amended in Coulthard and Brazil (1981) does allow for an
element of structure which is both predicted and predicting: R/I. In a sense
this represents a dual function. At the ranks of move and act, however, such
double labelling is not used. The utterance is coded according to its ‘dominant’
function. In the following made-up examples, we may code the third utterance
as R/I but must decide between eliciting and informing at the rank of move.
If the third utterance is not structurally or intonationally an interrogative,
the moves are re-coded as follows (although ‘yes’ is both a statement and
a question):
(Note that the various eliciting and informing moves in these exchanges are
realized by different acts. The exchanges therefore satisfy the criteria laid
down on p. 144.)
The above problem is relatively simple as it involves labelling within the
exchange. A more severe difficulty arises when an utterance which appears
to have a dual function occurs at the boundary between one exchange and
another. This frequently happens in casual conversation where participants
may produce a string of utterances, each one simultaneously responding to