Bob, Bill, Herb,
   I have enjoyed the discussion very much, though I'm sure I'm not alone on the list in having the feeling I'm overhearing a conversation by insiders (rather than being included.)  But that's the nature of the beast (ATEG),  both our weakness and our strength.
     I'm certainly  glad that people feel free to explore the nature of language without having to find direct teaching application, and I'm glad other people are interested in these applications.  Like Bill (I hope I'm not misreading you), I can't help feeling that anyone who looks at language as if it is not interactive will simply misunderstand what they are seeing, much as classical biology would miss so much without an ecological  orientation.  We can look at life in the lab, but it helps to remember that life does not exist in a laboratory, and the forms of life are part of a dynamic world.
    Since my primary interest has been reading and writing, I am interested in grammar in context (and by that, I don't mean finding student error in their actual work, which is what context has come to mean in my field, but looking at how grammar participates in the making of a public meaning.) To me, a grammar that does not accommodate this is a bit like a human biology that has no connection to medicine or health.  (Grammar, of course, need not be reduced to a theory of disease.)  We need disinterested study, but we also give colossal grants and colossal resources to issues of great public weight, like finding a cure for cancer.  What we have in linguistics seems often to me to be the equivalent of saying that human health has nothing to do with biology.  And the fact that people are sick is not our concern.
    We do, in fact, have a present day crisis in my field (English) deeply exacerbated because so many specialists are at work asking questions that they themselves say are irrelevant to real world practice. The material I quoted from Diane Hacker on pronouns and antecedents is a case in point.  The deep disconnection between the work of linguistics and public understanding (Hacker's book is generally thought of as among the most current of handbooks) is absolutely staggering.
    I'm probably reducing the argument to the absurd, but it's a little like saying that much grammar has nothing to do with what people write, read, say, or hear in their everyday lives.  If you can't bring the knowledge back to that world, what use is it?
    Using a tag question, I see as a functional test.  It is made possible because a sentence (in traditional practice) requires a grammatical subject and finite verb (though they don't fully understand what they are asking for), and the tag question is a test for those elements.  (The finite verb auxiliary acts almost like a pronoun would, as a stand in for the predicate.)  SFG pays considerable attention to this.  This "mood element" is often what gets bandied about in conversation when we agree or disagree with a statement or even answer a question.  "Did your son ever graduate from college?" "He did."  "He" stands in for "your son" and "did" stands in for the predicate ("graduated from college.") The tag question also has the function within conversation of inviting a response from the listener, perhaps softening a statement from time to time. "You will be on time on Friday, won't you?" We wouldn't be able to use it as a trick for identifying the subject if it weren't already part of our functional repertoire.
    Subject and predicate are functional terms, not simply formal ones.  The sense I get of their meaning (again, traditional grammar is not all that reflective about what it is asking for) is that something has been isolated (the subject) and something has been predicated about it.  There are, indeed, formal ways in which this is realized, and I don't know of a functional grammar that would ever imply that function is not realized through forms.  SFG presents this as one of three subject functions, and it gives us the kind of framework that helps us understand why we produce sentences that do not conform to the prescriptive norms. Predicating statements are not the only kind of meaning language allows us, but traditional grammar tends to overemphasize their value by insisting on them in all formal contexts. It's hard to conform to (or amend) a practice without understanding what is being asked for.
    Rather than seeing the tag question trick as a great boon from linguistics, I would see it as  one more attempt to stave off the real teaching of grammar.  These minimalist approaches (avoiding error in context or what is sometimes called "grammar at the point of need") are simply stopgap solutions for writing teachers who are trying to help students who know little or no grammar to start with.  Since knowledge of grammar is considered either irrelevant (descriptive) or harmful (prescriptive), we use these tricks to get by with as little knowledge as we can.  It seems easier than saying that this isn't an independent clause, or whatever would be relevant, simply because we know that the student doesn't know what a clause is and somewhere along the line we have decided that it is not important or helpful to take the time to teach what a clause is or how subordination contributes to the creation of meaning.    
    We won't begin to solve this problem until we admit that a deep separation exists and admit that it is not necessarily everyone else's fault.

Craig  
       
Stahlke, Herbert F.W. wrote:
[log in to unmask]"> 
Bob,
 
Thanks for a thoughtful response.  As I wrote I wondered how you would respond.  And I certainly have forgotten more than I care to consider.  But, let's look at your comments.

        I am confident that Herb Stahlke has forgotten more linguistics than I
        will ever know.  So, at the risk of being shown
        to be wrong, I am still going to take issue in some of his observations
        about linguistic theories and pedagogy.
	
        Stahlke, Herbert F.W. wrote:
	
        >I strongly support Bill's concerns over the application of linguistic
        >theories to teaching.  We've seen this trend in motion from the days of
        >Charles Fries' structural grammar back in the 50s, and it flourished
        >like poison ivy with the Paul Roberts' reduction of early
        >transformational grammar in the 60s.
        >
        I have read some of the papers on the debate that took place in College
        English and College Composition and Communication
        in the 1950s and 1960s on the role of linguistics and pedagogy.
	
        First, linguistic theories were absolutely crucial in challenging the
        privileged position of Standard English.  I recommend a reading of
        Geneva Smitherman's 1999 paper in College Composition and Communication
        for a review of this debate.  The use of linguistic
        theory lead directly to the NCTE's resolution on a Students' Right to
        their Own Language.
	
        To give a very simple example, without a theory of language I think it
        is impossible to show that no dialect is superior to another.

(Herb writes)

I suspect that we are talking about two different kinds of theory.  What I addressed throughout my discussion was formal theoretical linguistics of the sort done, in the 60s, by Chomsky, Bach, Lees, Postal, Perlmutter, McCawley, Lakoff, Ross, Partee, etc.  Looking into the 70s I'd add people like Jackendoff, Lasnik, Culicover, Goldsmith and obviously I'm leaving out a number very important figures, but I'm not trying to build a catalog.  People I wouldn't include in this list include the later Lakoff work, Langacker, much of McCawley, Chafe, Fillmore, Shuy, Wolfram, Fasold, and Green, and many others, all of whom have produced high quality theoretical work that is also accessible and useful to teachers, although too often teachers haven't been trained in or directed to such work.  What distinguishes the two groups is that the former are interested in formal characterization of an idealized native speaker's knowledge, or, what amounts to much the same thing, Universal Grammar
.  This is important and interesting work, but not generally to teachers.  The latter deal with topics that teachers also deal with:  Pear stories, dialect variation, metaphor, the semantics and pragmatics of sentence types, etc.  The kinds of research that Smitherman and others depended on in their invaluable syntheses and arguments on the nature of dialect were largely not of the formal theoretical type but of the variationist, more broadly dialectological, and functionalist sorts.  That kind of theory has had more immediate application to pedagogy and teacher education, but it wasn't the kind I was unashamedly pillorying.

Second, if there has been any advance in how to teach certain
grammatical structures, it has been by work of DeBeaugrande and
Noguichi.  In noting that a yes-no question provides a way of
determining whether a string of works is an independent clause, both are
using a FORMAL property of English to identify an important structure
for students to understand what a "complete" sentence is.

Now, perhaps, there is a "functional" explanation for this fact about
English, but I don't know what it is.

(Herb writes)
What helps the yes/no question methodology is that it has a pragmatic function that students recognized readily.  That's what makes it work so well as a formal manipulation.  WH-movement out of bounded vs. unbounded structures would also do the job, but it's much more difficult to get students to understand what's going on in such cases.  The pragmatics is far less clear.  The y/n test has both functional and formal characteristics that are important.


The following statement is absolutely right.

  
But formal theories of language are about
language and formal theory, not about pedagogy and praxis.

    
One of the jobs of applied linguists is to figure out how such formal
theories applicable to pedagogy.

(Herb writes)

I have to agree with Bill on this one.  I helped found a doctoral program in applied linguistics, and we ended up calling it applied linguistics because the academic politics of higher education in the state of Indiana wouldn't allow us to call it linguistics.  But it's very difficult to distinguish between applied and theoretical linguistics.  Carol Chomsky's dissertation on child language draws on some pretty good theory but looked at an empirical, behavioral problem.  Applied or theoretical?  I don't really like to make the distinction because too often it's artificial and misleading.  Margaret Steffensen's dissertation critiquing the Bereiter and Engleman work on language deficit in African-American early elementary school children draws importantly on dialectology and ethnolinguistic methodology to discredit the claims of language deficit in ways that are readily accessible.  But, like Bill, I don't see figuring out how to apply formal theory to pedagogy as a major goal
 of what our doctoral students occupy themselves with.

  
I think the contemporary model theoretic concept that most influences
the non-linguistic world is modularity, the idea that different areas of
syntax, as well as phonology and different areas of semantics, represent
separate mental modules that communicate with each other rather like
functions in a computer program.  Some persuasive popular writing has
been done on this subject, Pinker's books especially.  But this is
simply another case of seeking psychological reality for a theoretical
construct, and that path doesn't help pedagogy.

    
Let me suggest there is value to these theoretical constructs.  If all
dialects of English share a lot of fundamental principles in common and
the differences we
notice are based on how some of these fundamental principles are
realized,  it is pedagogically useful for teachers to know how they can
describe those differences
to instruct their students.

(Herb writes)

Let's say there was dialect of English with PRO-drop.  It would be important to explain to teachers that the grammar of this dialect produced good sentences that lacked subject pronouns and perhaps had some related properties.  It would not be important to explain to them the nature of PRO-drop as a formal construct.  But this reflects a perhaps fundamental difference of approach.  I just finished revisions of a paper to appear in Word later this year on the phonetics and phonology of English obstruents.  It's a paper that draws heavily on the fundamental linguistic notions of contrast, variation, and distribution, the basic facts that we train linguists to observe, analyze, and describe.  I could have expressed my findings with Optimality Theory combined with some Autosegmental Phonology, but that would have turned the paper into an application of those theoretical frameworks, and perhaps a testing and extension of them, the latter being rather more interesting than the for
mer.  But that wasn't my goal.  There is a large body of basic fact that had to be drawn together to make it clear how the English obstruent system and things like voicing interact.  If someone wants to cast my arguments in some framework for theoretical phonology, I'd be interested in seeing the results, but I needed to be able to explain, for example, to MATESOL students, how the obstruent system and the accentual system are related to each other, and how voicing, vowel length, schwa devoicing, etc. are related to obstruent behavior.  The theoretical overhead for casting this, say, in OT, yields no significant benefit to MATESOL students and instead simply takes time away from covering material they really do need.  (I'm starting to sound scarily someone arguing for teaching grammar only in context, shudder.)

Herb Stahlke, Ball State University

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/