LINGUIST List 4.703

Tue 14 Sep 1993

Disc: The Linguistic Wars: Ross, Constraints

Editor for this issue: <>

Directory

1. Steven Schaufele, syntactic history and pedagogy
2. Vicki Fromkin, Re: 4.698 Ross and The Linguistic Wars
3. Michael Kac, Constraints

I considered sending a personal message to David Pesetsky and/or Helen Dry
on this subject after their recent postings, but perhaps it would be
worthwhile to raise this question for general discussion in the List. When
teaching courses in syntactic theory, how much of the history of the
discipline should we present to our students? I haven't yet had the
opportunity to do so (sigh!), but my personal view, for what it's worth, is
that students of syntactic theory should begin by being (made) aware, at a
bare minimum, of:

(1) the Aspects model
(2) Ross' constraints
(3) 'Remarks on Nominalization'

Part of my reasons for this belief is that these works, and the ideas in
them, are taken as (historically) basic premises (possibly to be argued
against) by all subsequent work in (at least formal) syntactic theory,
whether transformational or not. So before introducing students to the
variety of syntactic-theoretical frameworks available since the
mid-70's/early 80's, i think they should be presented with the historical
background that all such work assumes in common.

Any thoughts?
 ------
Dr. Steven Schaufele 217-344-8240
712 West Washington Ave. fcoswsux1.cso.uiuc.edu
Urbana, IL 61801

*** O syntagmata linguarum liberemini humanarum! ***
*** Nihil vestris privari nisi obicibus potestis! ***

Message 2: Re: 4.698 Ross and The Linguistic Wars

I too agree with Pesetsky but continue to be aggrieved
that our younger colleagues and graduate students know too little about
the historical roots of what we are doing now. Which is why I think courses
in the history of linguistic theory can be interesting and important --courses
which summarize and highlight major contributions to
our understanding of the nature of grammars and language going back
to pre-generative linguistics, and also from Syntactic Structures to
the present. No one course can do it all but even one term can whet the
appetite of our students to read stuff on their own and particularly in
areas where they are doing their own research. One doesn't want to quell
creativity but then one doesn't want to reinvent wheels or constraints
or ignore what has been said before. I don't think, however, that
reading work of the past is nearly as important as doing linguistics in
the present, so view such courses as sort of a luxury or an 'extra'.for
I think the students who have taken my course have been
interested in the kinds of recurring questions that arise and
the new approaches to finding solutions, as well as the paradigm shifts
that have taken place and the new questions that are being asked which
couldn't have been asked in the past because they arise out of the
work that has been accomplished to date.

In addition, we should all feel very good about the fact
that our field has grown so, in both linguists and linguistic theory,
that it is impossible to require graduate students to be
prepared to answer any question about any linguist living or dead or any
major paper or publication at their oral exams. I remember the
night before my orals working through Hjelmslev's Prolegomena
which noone questioned me on then or since.

But then I am one of the old-timers and will be happy to tell anyone
at the drop of a hat or the dangling of a participle what life was
REALLY like during the linguistic wars. I am glad I lived through that
but happier to see the work being produced today by those too young to
have read Ross when he wrote his truly important and impressive work, which
indeed remains important today.

Vicki Fromkin

Polly Jacobson's recent comment about conflation of distinct
senses of the term 'constraint' in the now much discussed book by
Harris (and elsewhere) prompts me to note that a related
phenomenon can be observed in regard to the term 'rule' --
except that what appears to be going on in some cases is the
opposite phenomenon.

For example: rules are commonly contrasted with constraints: in
1960's vintage transformational syntax, as exemplified by Ross's
*Constraints on Variables in Syntax* transformations are rules and
the Complex NP Constraint isn't. But if a rule, pretheoretically, is
something which tells you what you are or aren't allowed to do,
then the CNPC has to count as a rule too -- except that it's a rule
which applies to the conditions under which other rules are
allowed to apply. (The natural term to use to describe such a rule
would be 'metarule' but that's been pre-empted for other
purposes.)

To a certain extent, generative semanticists nonetheless did
eventually unify the idea of (transformational) rules and
constraints of the sort proposed by Ross by uniting them under
the rubric of 'derivational constraints'.

The multiplicity of senses in which the term 'rule' is used by
linguists raises a number of questions that I personally happen to
have an interest in and the temptation to engage in a combination
of tub thumping and self promotion in this regard is strong; rather
than indulge myself here, let me simply issue an invitation to
anyone who happens to be interested or curious to contact me
personally and I'd be happy to hawk my wares to them.

Michael Kac