I am sorry I am coming late to this discussion.  I agree with everything that has been said.  My colleague Jim Kenkel and I have been looking at a collection of essays written by first year native and non-native speaking college students to understand the non-standard punctuation.  ALL of the essays had sentences that were punctuated according to the standard rules.  Those that were non-standard appeared to be following principles to show the relationship between various ideas.

I am very reticent to question Herb, but there are reasons why the concept of the sentence is more that a "methodological choice."  It is a category that reflects English speakers knowledge of the language.

Herb writes: 

  [The S (for sentence)] represented a unit within which certain relationships, structures, and constraints could be discussed without the inconvenience of answering questions about discourse.  This usually got us into an argument about competence and performance, which I held, and hold, to be a corollary of the methodological choice of S as the domain of analysis and description.  In informal speech, in contrast to formal lectures, addresses, sermons, etc., sentences tend to correspond to the breath group, so that the spoken sentence tends to be what one can say in one breath.

***
Note the use of the word "tend."   I think Herb gives away too much with that word.

Do we need the category of "sentence" (or clause) to describe what people do with they speak?

A couple of thought experiences.

I. Try to describe well-formed tag questions (a structure that almost exclusively in the oral language) in English without the use of the category sentence.  You know what tag questions are, don't you?  Tag questions are easy to describe, aren't they?

II. Try to describe well-formed questions (again forms that are very frequent in the oral language) in English without the use of the category sentence.   Some sentences to consider in your description.

1) Is the woman from France?
2) Does the woman live next door?
3) Is the woman who lives next door from France?
4) Does the woman who is from France live next door?
5) Yesterday, did the woman leave?

III.  Try to describe the antecedents of her and herself in the following strings without reference to sentence.  (I recognize that these sentences may not be common in the oral language, but they can be easily understood in the oral language.) 

"herself" has to refer to Mary in the following.
6) Mary sees herself on television.
7) Mary wants to see herself on television.

"her" cannot refer to Mary.
8) Mary sees her on television.
9) Mary wants to see her on televison.

"herself" has to refer to Jane.
10) Mary wants Jane to see herself on television. (but remember 7)

"her" can refer to Mary
11) Mary wants Jane to see her on television. (but remember 9)

****
I recognize that I am suggesting the competence-performance distinction is crucial.  By the way, I am not alone in this regard.  The competence-performance distinction is the basis for the suggestions that De Beaugrande in his "Forward to the basics" paper and Noguchi in his NCTE book use for their suggestions in how to show students how to determine whether a string they have written is an appropriate sentence.

Finally, given what I have written above, I have no idea what the following means:

". . . what a sentence can be depends very much on medium, genre, discourse pragmatics, and social setting, among other things."

Do the principles of well-formed tag questions, yes-no questions, and the antecedents of personal pronouns and antecedents change depending on medium, genre, discourse pragmatics, and social setting?  

Obviously, the frequency of the use of various forms change and some forms are very rare in some kinds of discourse (just like certain lexical items), but I have no idea how medium, genre, social setting changes the principles for any grammatical structure in English.

Bob Yates, University of Central Missouri 

To join or leave this LISTSERV list, please visit the list's web interface at:
     http://listserv.muohio.edu/archives/ateg.html
and select "Join or leave the list"

Visit ATEG's web site at http://ateg.org/