Max,

I appreciate your detailed response to my question.  I am wondering if I
could draw you out a little bit more on the issues that you raise.  First,
forgive my ignorance, but I am not sure what a T-unit is, how it is
defined.  Could you enlighten me?  In the same vein, you say that
Kelly Hunt discovered that "no one could define a sentence successfully."
Did he make this argument in a book or article that I could consult?

More generally, I was left wondering whether you felt that one could
produce counterexamples to the Aristotelian definition that I want to
propose, i.e., non-sentences that lend themselves to true or false
judgments or sentences whose format could not be used to produce true or
false propositions.  You include a number of what I would call elliptical
sentences as afterthoughts in your message.  "Though good writers do that
(punctuate stretches of discourses without subjects and finite verbs as
sentences) all the time.  All the time.  And not just to answer
questions."  In all of these cases, it seems clear to me that a subject
and main finite verb may be readily inferred from your context to complete
the thought of what might otherwise seem to be "sentence fragments."  No
one would consider "all the time" a correctly formed sentence in the
absence of such a context, would they?  Anyway, are these the kinds of
stretches of discourse that have persuaded you that the sentence cannot be
defined?

This all seems important to me because, frankly, I don't see the point of
grammar that is not anchored in some "notion."  Otherwise, don't
grammatical rules just constitute a huge circular argument?  T-units
consist of "subjects" and "finite verbs," whose importance is derived from
the fact that they are the indispensable elements of T-units.  My students
would like to know what it all adds up to in non-grammatical terms.

Regards,
David











On Sun, 10 Sep 2000, Max Morenberg wrote:

> David, I've never been much of a notionalist (am I making up this word?).
> I prefer to see myself as a pragmatist: if it allows students to
> understand, use it, whether it comes from Chomsky, Pike, Lamb, Halliday, or
> Curme.  Or even Dionysius Thrax.
>
> At any rate, I'd define a sentence as a stretch of written discourse that
> begins with a capital letter and ends with a period.  Sentence has always
> been a questionable term.  It was originally a period, a stretch of
> discourse that ends with a punctuation mark that came to take its name.  It
> was always a rhetorical stretch (No pun intended.  Well, maybe a little pun
> intended).  I guess this is all to say that I don't think you'll ever
> successfully define a sentence.
>
> But a clause. Now,  that's got grammatical limits.  Or, if you want a
> greater possiblity, go for a T-unit.  When you've got subjects and
> predicates to discern, you can come close to putting definable limits on
> stretches of discourse.  Kelly Hunt discovered that no one had defined or
> could define a sentence successfully.  That's why he measured clauses and
> T-units.  As a graduate student, I worked for Kelly counting T-units.  And
> I can tell you from experience that even T-units are blasted hard to
> discern sometimes.  Though not nearly so difficult as sentences.
>
> Ed is clearly right about what you might be able to teach to students.
> Notional ideas are tough to get across.  And why would you want to anyway?
> The idea of a sentence is only an issue in writing.  It's in part a visual
> issue.  It offends some people if you start with a capital and end with a
> period a stretch of discourse that doesn't have a finite verb and/or a
> subject.  Though good writers do that all the time. All the time. And not
> just to answer questions.
>
> I think what you want to teach students is how to craft written texts.  And
> how to see craft in the writing of others.  It seems to me that learning
> about certain kinds of constituents helps.  It gives you some vocabulary
> and a conscious knowledge to discern that the last three "sentences" of the
> previous paragraph aren't. Sentences, that is.
>
> I'm not sure I helped you, David.  But I'm not sure anyone can help you
> define a sentence. Not with a definition that will hold up very long.  Not
> evern Aristotle.  Max
>
> *************************
> Max Morenberg, Professor
> English Department
> Miami University
> Oxford, OH 45056
>
>
>
>
> >---------------------- Information from the mail header
> >-----------------------
> >Sender:       Assembly for the Teaching of English Grammar
> >              <[log in to unmask]>
> >Poster:       David D Mulroy <[log in to unmask]>
> >Subject:      What is a sentence?
> >-------------------------------------------------------------------------------
> >
> >Hello, everyone.  The list has been quiet lately.  I had an exchange with
> >Ed Vavra that I thought was interesting.  It concerned defining the
> >sentence.  I would really appreciate a sampling of your views.  My idea is
> >based on Aristotle, who connects sentences with the notions of truth and
> >falsity in his essay, On Interpretation, without offering a complete
> >definition of a sentence.  What I proposed to Ed was that a declarative
> >sentence is an utterance that has the formal qualities of utterances that
> >can be classified as true or false.  I don't mean that every sentence can
> >itself be classified as true or false, since it is easy to come up with
> >counterexamples, e.g., "This sentence is false."  What I am
> >saying is that every declarative sentence has formal qualities that can be
> >realized in such a way as to produce true or false statements.  Formally,
> ><"this sentence is false"> consists of a noun phrase, a linking verb, and
> >an adjective.  Obviously, that format can make any number of true or false
> >statements, e.g., "This patient is dead." Is it true that every
> >correctly formed declarative sentence has such a format, one that is
> >at least potentially productive of true or false statements,
> >while no non-sentence does?
> >
> >I would think that questions or commands could be explained as variations
> >or transformations of declarative sentences.  I don't think that
> >elliptical sentences like <"In the refrigerator."> as an answer to a
> >question like <"Where is the cheese?"> are really a problem, since no one
> >would suggest that <in the refrigerator> is a correctly formed sentence
> >all by itself.
> >
> >Ed seemed to accept this definition but had misgivings about its use in
> >teaching grammar to younger students.
> >
> >I would be most interested in your thoughts.  I hope that everyone's
> >semester is off to a good start!
> >
> >David
>