Dear all,
I have argued in a similar way in a 2005 paper in "Revue de sémantique
et pragmatique" and in a 2007 paper in Pragmatics & Cognition.
Diane Blakemore's work is also very important when it comes to
discourse and relevance. As for discourse as a process, Chafe,
Recanati, Carston have also argued in this direction. Reboul &
Moeschler in their 1998 book "Pragmatique du discours" even argued,
with quite a sound argumentation, that "discourse" if understood as a
preformatted structuration does not exist or can't be considered a
scientific object of study. They make in that book a particularly
insightful suggestion: discourses (undestood as non arbitrary sequences
of utterances) are understood as bearing layers of informative
intentions at various levels. What makes the unity of some particular
discourse or text lies in the fact that each utterance communicates an
informative intention which allows the addressee to enrich her/his
representation of other informative intentions that are of higher
textual (so to say) level. In other words: when processing an utterance
within some particular discourse, I am aware of global, discursive,
informative intentions, to be progressively uncovered, through the
accumulation of local informative intentions. Therefore, I conclude
that a particular text or discourse is particularly effect-productive
because of the convergence of local informative intentions toward a
global informative intention (in fact a hierarchy of more or less
global intentions), which has a correspondigly higher complexity. You
can get out of a novel some complex message, but you wouldn't get all
its weight through a short summary. I guess that when it comes to
production of discourses, notably literary works, such a convergence
shouldn't be missed. I would even add that what we tend to call
'coherence' in pragmatics, more or less intuitively, is probably
something like this convergence. Although 'coherence' as a notion is, I
think, an intuitive notion that applies before all to propositions,
concepts, thoughts, actions (as when talking of the coherence of
governmental policies or so) and only as a consequence of these applies
to language proper.
Anyway, it remains that a speaker shapes utterances in order for them
to be interpreted appropriately, and therefore anticipates the
inferences they are expected to lead to. She/he does so through the
recourse to her/his ability of exploiting the interpretive /
inferential properties of the human mind, an ability you may wish to
call 'mindreading module', 'folk psychology' or 'theory of mind'.
Therefore if she/he anticipates these inferences, she/he obviously
makes them (even unconsciously, which is another matter).
It remains that she/he may not anticipate all the inferences her/his
utterance may lead to in the context or in the particular place it
occupies in the considered discourse. This also leaves room for other
kinds of processes, which are due to semantic underdeterminacy. The
actual linguistic form provided by the speaker will only provide the
addressee with a symbolic (in the logical sense) form that resembles,
but is only rarely equivalent to, the original thought. Therefore,
there is always, I think, a possibility for a hearer to derive other
contents than the anticipated ones, not just because of the risk of
incorrectly disambiguate, attribute referents or exploit unintended
implicit premises, but also because of the approximate character of the
utterance itself with regard to the original thought 'corresponding' to
the literal meaning. It may end up with the writer herself/himself
derivating inferences she/he herself/himself didn't even expect, when
interpreting what she/he writes.
An important issue in this respect is that when writing a text, a
writer may well not only write but interpret her/his own utterances
more or less at the same time, which triggers possibly unexpected
inferences even for the writer. That sometimes happens when we write
scientific texts and feel that the simple act of putting our ideas on
the paper gives access to new representations which we didn't have
(full) access to before writing.
Well, food for thought.
Best to all,
Louis
Christoph Unger wrote:
Hi Mai,
a lot of early research in RT up to the present has been devoted to
pursue the idea that discourse structure is a consequence of
processing a text for relevance: cohesion and local coherence have
been shown to be consequences of relevance. I have looked at paragraph
structure and found it, too, follows from relevance rather than being
due to seperate discourse principles. In my 2006 book (Genre, Global
Coherence and Relevance; Palgrave) I have looked at global coherence in
particular (which has received relatively less attention in RT than
local coherence), including theories of discourse topic and
focus. My conclusion is that these phenomena can be best explained as
a by-product from the on-line processing of a complex ostensive
stimulus in search for relevance. I have also looked at claims about
differences between written and oral discourse, and have found no
convincing evidence that these different modes employ different
cognitive processes (sure, the different mediums do affect how
communicators form their ostensive stimuli, and there are statistical
differences, but that is a different matter). So I would firmly answer your first
question: we can and should use RT on its own to account for writer
strategies.
Christoph
On Tue, Jan 27, 2009 at 04:12:33PM +0000, Mai Zaki wrote:
Hi everyone,
Following from my question about the inferences speakers/writers make in the
encoding phase, I would like to ask about what relevance theory has to say
about discourse structure. In dealing with written texts, do we need to refer
to a theory of discourse structure to account for writers strategies in using
specific linguistic items or can we use RT on its own for this purpose? On the
other hand, what are the theories of discourse structure that could be said to
be compatible with RT? For example, Grosz and Sidner Theory of Global Focus?
Thanks a lot.
Mai Zaki
Received on Wed Jan 28 01:33:42 2009