Descriptive versus Explanatory

[Gisela Redeker]

Dear Andy,

One of the ways in which RST-analyses can support explanatory conclusions is
to compare subgenres which you'd expect to differ in respects that should be
reflected in differences in the analyses (e.g., the depth of the trees, the
frequency with which relations are used, maybe the direction of N-S
relations). Examples would be good vs bad writers, planned vs unplanned
texts, simple vs complex/difficult expository texts, written vs. oral texts
(where oral texts might i.a. prefer N-S over S-N order), and of course
differences in overt persuasiveness (see e.g. Abelen, Redeker & Thompson
1993*, where we found that U.S. fundraising letters were much more openly
persuasive than Dutch ones -- reflected i.a. in much more frequent use of
presentational relations, esp. in the highest levels of the text structure).


	* Abelen, E., Redeker, G., & Thompson, S.A. (1993). 
	The rhetorical structure of US-American and Dutch 
	fund-raising letters. _Text_ 13: 323-350. 

While corpus studies (comparing subgenres of naturally produced texts) can
obviously not support strong causality claims, they can still yield pretty
good evidence for any theory that can provide a plausible account of
systematic differences found between subgenres. An experimental variant
would be to have writers produce texts under controlled experimental
conditions, but the unnaturalness of a lab setting can severely limit the
external validity and generalizability of the results of such studies.

Another way to use RST in explanation-oriented research is to manipulate
texts experimentally and then test how readers process, understand, and/or
appreciate them. E.g., in Abelen et al (1993), we report a small study on
the acceptability of manipulated texts (with relations changed from
subject-matter to presentational and vice versa) at the end of the paper. 

Maybe this can give you some ideas and/or other list members can react.

Concerning you intriguing remark about the document type you've been looking
at: What kinds of texts are they and what are the problems in analysing them
with RST? 

Cheers,
Gisela

-- 
Gisela Redeker, Professor of Communication
Department of Communication and Information Sciences
University of Groningen
P.O. Box 716, NL-9700 AS Groningen, The Netherlands
g.redeker at rug.nl tel: +31-50-3635973 fax: +31-50-3636855
http://www.let.rug.nl/~redeker
 


-----Original Message-----
From: RST Discussion List [mailto:rstlist at LISTSERV.LINGUISTLIST.ORG] On
Behalf Of Andy Potter
Sent: woensdag 10 augustus 2005 5:31
To: RSTLIST at LISTSERV.LINGUISTLIST.ORG
Subject: [RST-LIST] Descriptive versus Explanatory


Several years ago in this list Bill Mann wrote:

"RST is defined in a way that makes it a descriptive approach rather than an
explanatory approach. While RST helps in identifying things that go on in
coherent monologue text, it does not say how or why they occur."

(http://listserv.linguistlist.org/cgi-bin/wa?A2=ind0011&L=RSTLIST&P=R36&I=-3
)

I am using RST as a means for exploring the nature and extent of coherence
of a type of document (a type of document problematic for RST, but that is a
separate discussion), and I find I can perform the analyses and these
invariably lead to useful insights.  However, reaching these insights is,
thus far, ad hoc, and thus, for dissertation research, a scary thing.

It might be fun to think about what an explanatory approach would look like.

Short of that, if you have any wisdom, suggestions, citations, or
methodological hints you would care to share, I would love to hear from you.

 Andy