Here's my definition of a sentence: A sentence is a grammatical unit
consisting of at least one finite, independent clause. Finite is not
hard to define, but 'independent' is. The old 'can stand alone' is a
pretty good metric. When working with relatively simple, declarative
sentences, tests like 'I am convinced that _____' and tag-formation are
useful for determining this kind of independence. I think of it as a
unit which is not so constructed as to be a constituent of another unit.
A dependent clause, for example, is structured as a constituent of a
larger sentence, hence the term 'dependent'.

This definition conforms to the way traditional grammar has used the
term.  Questions, etc., are considered to be different types of
sentences. I don't like the notion of deriving one type from another,
but it _is_ one way of approaching the shared propositional content of
sentences of different types (e.g., active vs. passive). It is probably
also, for non-linguists, the most robust remnant of
transformational-generative grammar.

It is also a definition which will allow students to understand rules
for end punctuation as found in most grammar books and writing manuals.
I think it's important for students to realize that punctuating
non-sentences as if they were sentences is a stylistic deviation. It
certainly happens a lot, and is quite acceptable in numerous genres of
text, but it is not acceptable in all genres of text. (This should make
clear that I do not find 'a stretch of words between a capital letter
and a period' to be a good definition of a sentence. This definition
would certainly not help students avoid fragments and comma splices when
they need to.)

Maybe some people don't like the idea of defining one grammatical term
by referring to other grammatical terms. But grammar is a system of
interdependencies. No one would object to defining a fuel injector by
referring to other car parts and their functions; indeed, there is no
other way to understand the concept of a fuel injector. Grammar deals
with the parts of language and their functions. These all work in
concert to connect meanings to each other.

I wouldn't go near true/false judgments with a ten-foot pole. There is a
whole tradition in the philosophy of language, along with a
counter-tradition, which explores truth/falsity as a metric for defining
linguistic units. Keyword: truth-conditional semantics. There are
lengthy works that defend the notion that true/false judgments are
crucial to defining a sentence, and others that argue against that
claim. Have fun wading through all that!

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Johanna Rubba   Assistant Professor, Linguistics
English Department, California Polytechnic State University
One Grand Avenue  • San Luis Obispo, CA 93407
Tel. (805)-756-2184  •  Fax: (805)-756-6374 • Dept. Phone.  756-259
• E-mail: [log in to unmask] •  Home page: http://www.calpoly.edu/~jrubba
                                       **
"Understanding is a lot like sex; it's got a practical purpose,
but that's not why people do it normally"  -            Frank  Oppenheimer
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~