<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
Dear James,
<p>Knowing that you may not have time, I hope you will bear with my editing
your post.
<p>James Cornish wrote:
<blockquote TYPE=CITE>Zouhair Maalej
<p>Thank you for your reply to my "cohesion" question. I confess
to making loose references "off the top of my head" without checking the
exact dates of publication, therefore possibly misleading. I was
referring to Hoey's <i>Patterns of Lexis in Texts</i> (1991) wherein
he goes to great pains to lead to the argument that "Despite the fact that
lexical cohesion is covered in Halliday and Hasan's book in less than twenty
pages (compared with over fifty for substitution), it is the single most
imprtant form of cohesive tie, even in terms of Halliday and hasan's own
sample analyses at the end of the 1976 book." I took this claim by
Hoey as a minor challange and looked at the database of student writing
I've been building for a few years now and found that the actual occurances
of ellipsis and what Halliday and Hasan label as "substitution" simply
don't happen to any significant degree.</blockquote>
I think (it remains to be checked on actual data?) that substitution and
ellipsis are not equally distributed in spoken and written discourse (cf.
work done on orality and literacy by Tannen and others). I believe (I haven't
read anything to this effect) that they are more frequent in speech, which
may be the reason why you didn't find a significant occurrence of them
in your students' compositions.
<blockquote TYPE=CITE>I realize that the database is sckewed, but that
rather informal examination motivated me to begin to actually sit at the
computer screen and use Hoey's methodology (the flow charts in the <i>Patterns</i>
book) to examine a few essays in depth.
<p>> I take this to be an invalidation of the claim that cohesion is lexical,
unless Hoey
<br>>changed his mind in a different piece of writing I am not aware of.
Please note that _textual >relation_in this quote takes us to the conception
of text they offer, which is defined, quite >obviously, as _a SEMANTIC
unit_ (p.2).
<p>Yes, but I would argue that "text and textual relations" are semantic
units very different than a word. I have not read the article in
<i>Advances
in Written Discouse</i> (again thank you for the references) so am on shakey
ground here. But, at least in the <i>Patterns</i> book and the early
<i>On the Surface of Discourse</i>, Hoey is trying to leave the metaphor
of seeing internal text relationships like one sees internal sentence relationships
and using the same labelling system--to me an enourmously attractive departure.
<p>I take it that you are defining "lexial" as merely vocabulary
items and "sematic" as the meaning attached to a given stretch of language.
Then we are approaching the tautological problem facing any discussion
of this type: where do the lexical items leave off and the meanings
begin, or just as validly, vice versa.</blockquote>
I don't think that this is like the chicken-and-egg issue. Lexical matters
include relations (as you rightly mentioned it) such as co-classification
(similarity, contrast, hyponymy) and meronymy (this is when two lexical
items are related as whole to part or part to whole (e.g. "the legs going
up and down beyond the railings." One important case of meronymy
is synecdoche, which is the substitution of a part for the whole, or the
whole for a part.)). Hyponymy expresses relations of inclusion between
lexical items. "Co-hyponyms can be seen as kinds of SYNONYMS, since their
CONCEPTUAL MEANINGS partly overlap in respect of their superordinate. If
synonymy is symmetrical (a = b), hyponymy is asymmetrical: ... an oak is
a tree, but a tree is not necessarily an oak. But hyponymy, like synonymy,
often functions in discourse as a means of lexical COHESION by establishing
referential EQUIVALENCE to avoid repetition: Did you see the policeman
flag down that old car? I bet the vehicle wasn't taxed or insured properly"
(K. Wales, <i>A Dictionary of Stylistics</i>, 1989: 223). What is important
about lexical relations is that, they are, in De Saussure's words, paradigmatic,
i.e. selecting one excludes the other. One cannot use a lexical item and
its antonym simultaneously, although one can create a special effect by
using a lexical item and its near-synonym or a lexical item and its antonym
(which immediately takes beyond lexical considerations). A semantic relation
such as cohesion, however, is syntagmatic (i.e. a text-forming property),
and allows for co-occurrence (semantic stability), pro-forms (linguistic
economy), ellipsis (recoverability from a previous mention [the anaphora
issue], which contributes to economy, linguistic compactness, and difficulty
of language processing).
<blockquote TYPE=CITE>At this stage in my investigations into the text-forming
quality of lexical cohesion, I really center on patterns of vocabulary
items, their synonyms, antonyms, hypo- and hypernyms (is that a word?)
without, at least at this early stage, much concern for the ultimate MEANING
arising for the whole of each of these short texts. But I have to
respectfully disagree that lexical cohesion is not a text-forming quality;
it is simply one of the colors in the stream of the process called "text."</blockquote>
As I mentioned in my previous mail, H&H (1976) define a text not as
a structural but as semantic unit. Hasan (1989: 71) defines the texture
of a text as "manifested by certain kinds of semantic relations between
its individual messages." It so happens that it is cohesion, at least partly,
that provides a text with its texture. If you have patience, I will post
a short text for you with features of lexical cohesion underlined. Of course,
other kinds of cohesive devices exist in the text:
<p>What is IBM's Presence?
<br>Seldom does <u>an opportunity </u>come along that offers <u>such vast
riches</u> to business around the world and completely redefines <u>an
industry</u>. <u>The phenomenon of the Internet</u> has exploded <u>the
business world</u>. In 1991, roughly <u>a thousand businesses</u> were
connected to <u>the Internet</u>. Now, twenty-one thousand <u>businesses
</u>are attached! <u>New networks</u> are emerging at an astounding rate
of one every 10 minutes. And there are 37 million <u>surfers </u>out <u>there</u>
today.
<p>Kind regards.
<p>Zouhair Maalej,
<br>University of Tunis I,
<br>Department of English, Manouba,
<br>2010, Tunis, TUNISIA.
<br>Tel/fax: +216 1 362 871
<blockquote TYPE=CITE>I will leave this uncapped for now because of the
lack of time. But thank you so much for your response and references,
and I look forward to more of both.
<br>--
<br>James Warren Cornish - Texas A&M University
<br>English Department/ Discourse Studies
<br>213B Blocker Bldg. M/S 4227
<br>College Station
<br>TX 77840-4337
<br>409-845-3542 ex. 40
<br> </blockquote>
</html>