LINGUIST List 4.523

Wed 07 Jul 1993

Disc: GB

Editor for this issue: <>

Directory

1. Steven Schaufele, PPA and scientific methodology
2. Bill Bennett, T
3. Anjum Saleemi, GB THEORY
4. David Heap, Re: 4.461 French Subject Clitics

There's been some lively discussion recently on the net on the 'clout'
supposedly enjoyed by PPA aka GB aka Chomskyan theory; this discussion is,
indirectly and in part, due to a query of mine (4-405) about the
falsifiability of Pollock's hypothesis. I was interested to see, in
LINGUIST 4-507, back-to-back, two postings from David Pesetsky and Bruce
Samuelson that together seem to sum up a lot of the concerns behind my
original query.

I ask these two scholars' indulgence in my following summary. It seems to
me that Pesetsky is saying, effectively, 'C'mon, people, this is *science*
we're talking about here; our ideas should be judged on their *merits*, not
on our skills at marketing them, etc.'. Meanwhile, Samuelson is saying,
'C'mon, people, this is *science* we're talking about here; our ideas
should be judged on their *merits*, not philosophical and political
arguments.' In other words, both scholars seem to agree on an idealization
toward which our discipline should aspire, but while Pesetsky seems to feel
that PPA, at least, is doing a pretty good job of approaching that goal
Samuelson expresses great frustration at its shortcomings.

I confess to much sympathy with both points of view. While not precisely a
PPA-partisan i agree with Pesetsky that, with regard to the Dominical
criterion 'by their fruits you shall know them', PPA has an impressive
track record. Much exciting research has been fomented in and by the
Chomskyan school, and we have all learned a lot from it over the years. On
the other hand, i often wonder whether what we are learning really has
anything empirically to do with language and the human language faculty.
>From discussions public and private i often get the impression that, in the
minds of many linguists, the proper goal of linguistic research is not to
learn more about human language but to serve a particular theory and to
further its associated agenda, even at the risk of ignoring data that seems
to challenge the validity thereof. Especially when not all researchers are
convinced of the empirical validity of the theoretical school in question,
this approach strikes me as counterproductive.

There are a few questions that i believe should routinely be asked of any
linguistic hypothesis:

(1) Does a particular linguistic hypothesis H, couched in terms of
 a particular theory, make claims about (potentially) empirical
 data?

(2) If the answer to (1) is 'no', does H really tell us anything about
 language?

(3) If the answer to (1) is 'yes', are H's claims/predictions confirmed
 by relevant data?

(4) If the answer to (3) is 'no', is H nevertheless so valuable (e.g., in
 inspiring research) that it is worth revising?

(5) If the answer to (4) is 'yes', does the revision continue to make
 claims about empirical data?

I imagine that there is likely to be some disagreement over any of these
questions (for instance, what constitutes 'relevant data'?). My impression
of the kind of literature published in Science or Nature is that the
authors thereof strive to answer these questions explicitly, and that their
disciplines have very clear and generally-accepted standards in these
methodological issues. In my own work, i strive to formulate explicit
answers to these questions. But i do not sense that there is much
consensus amongst linguistic researchers on these issues, and i think if we
are going to aspire to scientific status there ought to be.
------
Dr. Steven Schaufele c/o Department of Linguistics
712 West Washington University of Illinois
Urbana, IL 61801 4088 Foreign Languages Building
 707 South Mathews Street
217-344-8240 Urbana, IL 61801
fcoswsux1.cso.uiuc.edu

*** Syntagmata linguarum liberamini humanorum!***
***** Nihil vestris privi nisi obicibus potestis! *****

I have once intervened in that discussion about GB. What I would like to ask of
the 3.5k professionals who subscribe: what is their primary theory ABOUT? The
discussion suggests that the reply would be that it was about a particular
(though perhaps fashionable) model. Should it not be, instead a theory about
the language-using organism? And a particular model would then not hold such a
sway over teachers and taught in linguistics. Does the model wag the theory or
the other way round?

RE 4.495
In all the discussion, "theory" seems to be used for "model". Structuralism and
rules produced results, of a sort. Nowhere in the discussion have I yet found
(sorry if I have missed it - exam time all over the place) mention of human
beings and chimps, but where training occurred and why GB should be the the
fashion. Linguistic theory is properly about the users of language (=langage
not langue) but if that was understood we might find ourselves needing more
than one narrow (say, GB or TG) model to account for the difference between
vernacular and careful).

I thought the advice to maintain silence on the sociological and political
aspects of linguistic theory, and also to return to the lab for more serious
work, was rather sound. However, the "silence of the lab" is being broken so
frequently and by so many people that I now feel it might perhaps be worthwhile
to make an attempt to steer the discussion towards a more productive set of
issues.

Let me try to address the following question, which has sparked some of the
current controversy: How do GB researchers manage to agree on so many things
many of which most non-GB linguists find very hard to accept. I think some
insight can be obtained by trying to unearth the methodology that underlies
the GB research enterprise. One of the major (implicit or explicit) strategies
is to choose to disagree selectively, in a limited fashion; so if the field
at some point is working with a set of hypotheses H(i), it would be considered
desirable for someone to offer an alternative set H(j), or any set fairly
"close" to what was the point of departure, but something like H(t), for
instance, might be regarded less highly, or might even be misunderstood as it
would involve too great a leap in the process of collective reasoning. It is
conservative (and I think justifiable) research strategies such as this one
which, on the one hand, give the impression that there is perhaps a great deal
of unquestioning acceptance of many ideas, and, on the other, make it possible
for people like Sabine Iatridou to present some divergent proposals. It is
not a question of believing in one thing or the other: one just has to hold
certain factors constant so that the consequences of any suggested change can
be worked out maximally accurately. Clearly, any free-for-all methodology is
going to result in total theoretical chaos; I think this is probably what
Perlmutter (quoted in David Pesetsky's recent message) must have meant by what
he vaguely calls open-mindedness leading to one's brains falling out, etc.

The GB community, thus, shares a consistent method of inquiry. In addition, it
shares the broad philosophical/cognitive rationale (a la Chomsky, Fodor, etc.)
for this inquiry. Although Gazdar et al. (1985) reminded us that "virtually
all the work needed to redeem the promissory notes linguistics has issued to
psychology over the past 25 years remains to be done", things have progressed
consistently in the right direction in the past decade or so (see, for example,
much of the recent work in language acquisition theory).

Given the approach outlined above, it should not be too difficult to see why
GB is perceived to be dominant, prestigious and in some sense more successful
than other frameworks. So long it remains as productive and dynamic as it
is today, it'll continue to have the sort of influence it has. I think the
virtues of the framework under discussion far outweigh any drawbacks resulting
from factors other than the purely academic ones. As far as the "GB" linguistic
proposals are concerned, they are surely there to be amended by anyone who
finds the basic rules of the game agreeable. One can of course choose to play
a very different ball game, but then why complain about the other game - just
because it appears to be more successful? Do we need remind ourselves that
what's dominant today may cease to be so tomorrow, and that today's minority
science can become tomorrow's mainstream science? Those people who feel
strongly that GB is wrong, therefore, are free to choose to do their own stuff,
but the choice to work in a "non-dominant" framework has to be theirs!

Anjum Saleemi
National U of Singapore

Other than methodological and philosophical assumptions, the GB researchers
of course also agree on what is plausible in current linguistic theory, with
the set of plausible constructs varying to some degree from person to person
but there still being a sufficient amount of overlap to lend a sense of
collective responsibility to the research enterprise in question.

Message 4: Re: 4.461 French Subject Clitics

Steven Schaufele writes (June 14):

> He (Bernhard Rohrbacher) noted that it
> has been suggested that the obligatory use of nominative-case pronominal
> proclitics in some colloquial French usages is developing into a new,
> richer subject-agreement marking system. I granted that the hypothesis has
> been suggested (as long as as Tesniere, i believe), but that

> (1) 'it has been pooh-poohed by some Gallicists of my acquaintance' and
>
> (2) if it is true 'it is true only of certain colloquial registers,
> >while Pollock's arguments are based on formal standard French, for
> >which the near-obligatory use of subject pronominal clitics is not
> >characteristic. I question to what extent a fact about one dialect or
> >register can be related as a diagnostic to a fact about another, for
> >which the proposed diagnostic is not actually a fact.'

(1) is hardly an argument, in any framework: one could aduce
Gallicists (or other language-ists) of many people's acquaintance to pooh-pooh
just about any theoretical position, including the view that subject clitics
do NOT constitue agreement. If you are going to have recourse to
authorities, at least have the courtesy to NAME your authorities!

(2) while it may not be true of ALL varieties of French, the presence
of such clitics (plausibly analysable as agreement) in many varieties of
spoken French with rather a lot (i.e. millions) of speakers makes this
hypothesis worthy of attention in its own right, regardless of the facts in
"formal, standard French". Unless of course someone has come up with a
conclusive diagnostic as to which dialects/grammars are important, and which
are not, in which case they should let us know about it.

Finally (and this may be of interest to those following even part of
the long-running debate on GB & linguistic politics): this hypothesis
goes back considerably further than Tesniere. Roberge & Vinet
(1989:54-5, 63), writing on the 'subject-clitic-as-agreement' hypothesis from
a GB perspective, cite linguists from Meyer-Lu"bke (1895) to Schogt
(1968). It seems obvious that parallel conclusions from linguists working in
other frameworks (traditional philology and martinetiste functional-
structuralism, respectively) can only strengthen a theoretical position
(although it may deflate those who are convinced that THEIR formalism really
is the first invention of a given wheel). Contra two Davids (Pesetsky &
Perlmutter), neither Roberge nor Vinet seems to be "so open-minded that their
brains fall out"; a healthy dose of open-mindedness never hurt anyone's
research (whatever it may do to job and publication prospects).

David Heap
University of Toronto
heapepas.utoronto.ca