coherence relations between large spans of text

Michael O'Donnell micko at WAGSOFT.COM
Mon Apr 13 10:35:57 UTC 2009


Dear Wallace,

 I stand corrected, yes, Labov and Waletzky  (but is is 22 years since
I actually read the reference). If anyone wants the referece:

  Labov, William and Joshua Waletzky. 1967. Narrative analysis. In J.
Helm (ed.),
  Essays on the Verbal and Visual Arts. Seattle: U. of Washington
Press. Pp. 12-44

But my point was not so much to provide the perfect schema for narratives,
but rather to point out that mono-nuclear relations (nuc-sat
dependency relations)
may not be the best way to model entire texts, and that the use of higher
level schemas (whose elements are realised by RST trees) can clarify
the structure.

Mick



On Mon, Apr 13, 2009 at 2:25 AM, Wallace Chafe
<chafe at linguistics.ucsb.edu> wrote:
> I've always felt disappointed, not to say annoyed, by the frequent
> references to the schema proposed by Labov and Waletzky (and why is the
> second author forgotten?). They very strangely left out the climax, which
> forms the nucleus of this very common pattern. It does indeed have a
> nucleus.
>
> At the risk of sounding paranoid, I wish people would occasionally notice my
> revision in Discourse, Consciousness, and Time (Chicago, 1994), especially
> pages 128-132. The schema set forth there consists of orientation -
> complication - climax - denouement - coda. I provided what I thought was a
> very nice extended example from a conversational narrative.
>
> I guess Labov and Waletzky entered the canon way back in 1967, and since
> then nobody has ever bothered to look further.
>
> Wallace Chafe
>
>> For stories, Labov proposed a schema consisting of:
>>
>>            Orientation^Complication^Resolution(^Coda)
>>
>> In this approach, there is no one nucleus to which other elements
>> are optional satellites. Rather, we have a multi-nuclear structure, where
>> each structural component has its part ot play in the whole.
>>
>> RST includes multi-nuclear relations, although they are generally
>> more generic (e.g., sequence, joint, etc.)
>
>
> __________________________________________________
> D O T E A S Y - "Join the web hosting revolution!"
>            http://www.doteasy.com
>



More information about the Rstlist mailing list