82 Advances in spoken discourse analysis
An alternative point of view, which I prefer, is to expect the text to
supply everything necessary for its own interpretation; what we need is not
an external knowledge base but a better understanding of text structure. If
we do not rely on the text to indicate its own interpretation, then we invoke
mysterious processes for which it is difficult to find evidence.
It is not, however, to be expected that texts will be fully explicit; the text
will organize its meaning up to a point regarded as appropriate by the
speaker; in interaction it is open to other participants to press for greater
explicitness where they feel the need for it. Interpretations are always provisional
during the course of a conversation, and a large number of points remain
obscure because it is not regarded as important to clarify them.
STATUS OF MOVES
There were several problems associated with the number of moves in an
exchange, and the status of the moves. A response to an elicitation is very
often a statement. If this statement is not ellipted, then it looks suspiciously
like an inform. But an inform is an initiation not a response.
The response to an inform is very often a simple acknowledgement, and
in such cases there is often no follow-up element present. The realizations
of responses and follow-ups overlap when they are minimal, such as yes,
mhm.
Some responses take the form of questions, for example:
I It’s red
R Dark red?
F yes
(data from Francis and Hunston, this volume, Chapter 7)These responses then have some of the character of elicitations, in that they
strongly prospect an answer to their questions. But again, elicitations are
initiations, that is they must come first in exchanges.
Various attempts were made to solve these problems, notably those of
Coulthard and Brazil (this volume, Chapter 3), which are summarized and
built into a revised model by Francis and Hunston (1987 and this volume,
Chapter 7).
There were also further enhancements and elaborations of the model
proposed from time to time, in particular Burton (1980) and Berry (1981).
Burton emphasized the importance of moves which challenge presuppositions
and contrasted them with all the others, which she called supporting moves.
Berry concentrated on the transmission of information, and established a
version of exchange structure which has been adopted by systemic linguists,
and increases steadily in complication (O’Donnell, forthcoming).
I would like to put forward, in response to all this activity, a position
which is not very far removed from the original one (Sinclair and Coulthard