LINGUIST List 5.799

Tue 12 Jul 1994

Disc: Linguist-bashing

Editor for this issue: <>

Directory

1. Paul Deane, Fromkin post;Linguist-Bashing
2. Jacques Guy, Linguist-bashing

I appreciate Vicki Fromkin's thoughtful comments on where linguistics has
been making progress in establishing its legitimacy, such as with respect
to neuroscience. There can be no doubt that the linguistics/neuroscience
connection provides one of the most exciting areas of progress in the field.
I also appreciate her tireless efforts to promote the field, including the
development of the linguistics section of the AAAS. (I've not yet gotten
around to joining--I will now, and I hope many other readers of the net
will too.)

However, there are many statements in Fromkin's post that are contentious
and require an answer.

> Re the Pyscholoquy book review -- do YOU believe language understanding
> is primarily about underlying conceptual structures? If you do then you
> would go along with this `linguist bashing'.

Why is disagreement with modularism necessarily an attack on linguistics?
That makes sense only if being a linguist automatically means agreeing
with the tenets of the Chomskyan school. But Fromkin later states:

> Re the fable -- this is more Chomsky bashing than linguist bashing, and
> if you change the name of the 'Linguesses' to 'chomskyites' or 'Mitniks',
> this could be a fable written by those 'anti-chomsky' linguists who put
> up straw men and then knock them down.

How can Fromkin suggest in one breath that disagreement with the
modularist position implies linguist bashing, and then proceed in the
above quote to distinguish 'Chomskyites' from 'anti-chomsky' linguists? If
both opinions exist in the field, an attack on one does not constitute an
attack on the field.

> I would expect this view from those who do not understand the nature of
> language but think they do.

As a scientist, I would say that none of us "understand the nature of
language"--but each of us has some working hypotheses about how
language works and how it fits into the bigger picture. The same goes for
neuroscientists: in a recent issue of _Cognitive Neuroscience_, Randy
Gallistel, a professor of psychology at UCLA states that, "We clearly do
not understand how the nervous system computes. We do not know what are the
foundations of its ability to compute." Any of us who think we REALLY have
an adequate understanding of the nature of language--much less its neural
instantiation--are quite possibly confusing working hypotheses
with absolute (hence unrevisable) knowledge.

> Obviously, understanding conceptual structures is necessary, but not
> sufficient, for understanding language, and, as much research of brain
> damaged patients, savants etc. shows, there can be a dissociation between
> understanding language structures and conceptual structures.

 [A list of examples follows]

One of the greatest banes to useful debate occurs when one insists in
framing the debate in one's own terms. Fromkin implicitly treats
linguistic structure as being a distinct ontological entity from
"conceptual structure", terminology which presupposes her own position. Of
course, linguistic structure is distinct from conceptual structure if by
'conceptual structure' one means concepts which exist independently of
language. But this is a fairly minor point. Much more important is the
fact that none of the cases Fromkin cites really provides a clean,
clear-cut proof of the point she is trying to make. There is a lot of
interpretation under the surface. For example, Fromkin cites cases of
people with "impaired language but preserved conceptual structure", or
"preserved language but impaired conceptual structure". Such formulations
presuppose that conceptual structure is some monolithic entity, either
preserved or damaged _en bloc_. Typical of the studies I have seen is a
methodology which assesses "preservation of conceptual structure" in the
clinical sense of still being able to function reasonably normally. But
typically what we get is a much more complicated picture in which some
conceptual abilities are impaired along with (or independently of)
language. The question then arises: what is a result of accidental
patterns (say of brain damage) and what is an effect of underlying
associations? It is very easy to discount data as irrelevant when
advancing one's own point of view. For instance, Fromkin has cited
Williams Syndrome a number of times on LINGUIST, claiming it provides a
clear example of her thesis. But as Bellugi has discovered, Williams Syndrome
children have deficits in spatial cognition too--more subtle than the
linguistic deficits, but present nonetheless. In the absence of a theory of
how the brain processes spatial concepts, it is highly misleading to
assume that such "minor" spatial deficits have nothing to do with
cooccuring language deficits and need not be discussed in linguistic debate
...

Another, basic, point has to do with the distinction between EXPLAINING
THE CAUSAL BASIS FOR LINGUISTIC COMPETENCE and EXPLAINING THE PROCESS OF
LINGUSTIC PERFORMANCE. I think there's plenty of evidence that our
linguistic performance is an ability relatively independent of higher
mental processes. But so is our ability to drive a car, which not only is
independent of higher mental processes but is clearly functionally
segregated from other sensory and motor activities such as smoking a
cigarette or watching the scenery. It does not follow that car-driving is
an innate ability independent of other mental processes. In the same way,
I would argue, we can admit the existence of abilities (even learned
abilities) that are functionally distinct and even differentially affected
by brain damage. Why? Because an ability can depend on a specific TYPE of
neuronal structure, without using the same SPECIFIC nerve cells as
another ability of the same type.
 In other words, I am making the following argument: 1. an ability may be
important enough to be processed automatically--meaning that it gets
neurons dedicated to it--whether it is learned or innate. 2. An ability
may be performed by specific parts of the brain for reasons of
computational efficiency, even if that part of the brain is not
specifically designed for that function. 3. It is therefore possible for
a learned (that is, not specifically innnate) ability to have its
own, dedicated neural substrate. Its distribution in the brain would
depend on its functional requirements; its independence from other
abilities (even those using the same underlying computational abilities)
would derive from its functional need for speed of computation. And of course,
something as complex as language is going to be all over the map--no other
ability is likely to use the same exact configuration of brain regions,
even if language were 100% a learned skill (which I doubt). So Fromkin's
use of the word "Obviously" in the quote given above is rather
contentious. To accept it we have to assume a lot of things other than the
clinical facts she cites.

> Whatever one thinks of Chomsky's work, linguists should be grateful
> that he has probably done more to gain the respect of the non-linguist
> world --both scientific, despite the AI view above, and general public.

There is no question Chomsky has had a huge impact. That impact includes
forcing a hearing for views that were then (and today probably still are)
unpopular outside linguistics. But with that hearing comes a high profile
and a lot of fire. That's fine. But I don't like being shot at because
opponents outside linguistics equate "Linguist" with "Chomskyan".

I think Fromkin is being a bit sanguine, though, when she states that
linguistics is making headway _in general_ as a legitimate field of
enquiry. At the linguistics/neuroscience junction where her research is
focused, that is probably correct. (Though I should note that her list of
people making an impact on neuroscience--Pinker, Jackendoff,
Bellugi--rather noticeably does not include anybody doing current GB
syntax. I find Pinker, Jackendoff and Bellugi's work highly useful, too,
even where I disagree with underlying assumptions. I think their impact
has some connection with the nature of their research, which is highly
interdisciplinary as well as being informed by linguistic theory.

However, I am a lot less sanguine about linguistic's reception in the
humanities and social sciences (where most of us are administratively
housed). A lot of English and English education scholars appear to view
transformational syntax with less than great enthusiasm. (In the early '70s
they tried to use transformational grammar instead of traditional grammar
in the secondary schools. That fiasco may have influenced a lot of attitudes
negatively.) And much of the social sciences and humanities are imbued
with deconstructionist views based ultimately on rather flawed theories of
language--a point of view which may make their reception of theoretical
linguistics less than enthusiastic. None of this may be our FAULT (in the
sense of reflecting on our scientific credibility) BUT if we aspire to
"make it" as a hard science, while the majority of linguists depend for
jobs upon academic units outside of the hard sciences, we have a lot of
fence-mending to do to avoid getting targeted by those who think that our
claims to scientific rigor are unjustified.

That's all I have to say at this point. I would like to encourage those
who posted to me privately to present their positions publicly if they
can--I'd rather not post them as a summary unless I must.

I have only one question: I figured out who "Chumpsky" was, of
course, and what country the kingdom of Pork was (hint: its
capital is Barrel), but rack my brain as I may, I failed
at guessing who "Borne" was. Does anyone know?