LINGUIST List 5.297

Wed 16 Mar 1994

Disc: Mainstream Linguistics

Editor for this issue: <>


  1. "Karen S. Chung", Re: 5.294 Mainstream Linguistics
  2. Benjamin Macias, Re: 5.294 Mainstream Linguistics
  3. Paul Deane, 5.294 Mainstream Linguistics

Message 1: Re: 5.294 Mainstream Linguistics

Date: Wed, 16 Mar 1994 07:23:33 Re: 5.294 Mainstream Linguistics
From: "Karen S. Chung" <>
Subject: Re: 5.294 Mainstream Linguistics

 One comment in the 'mainstream linguistics' discussion
has stuck with me and bothers me. That is that teaching 'mere'
language somehow puts you on the sidelines. Linguistics is and
will always be subordinate to language and individual languages.
For me, the real meat (as a former vegetarian, I would have
preferred a different metaphor, but somehow 'tofu' didn't fit)
of the subject I like best is to be found in language and
languages. And there's hardly a better way to dig in and discover
the treasures a language has to offer than by teaching it to
people with a different native tongue, who don't take those
'treasures' for granted and thus fail to recognize them. Teaching
English to Chinese in Taiwan, for example, has taught me more
about English intonation (and thus intonation in general) than I
think I could have learned from books or theoretically-oriented
research--though I do read those, too.

 Karen S. Chung
 Dept. of Foreign Languages
 and Literatures
 National Taiwan University
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: Re: 5.294 Mainstream Linguistics

Date: Tue, 15 Mar 94 18:39:52 +0Re: 5.294 Mainstream Linguistics
From: Benjamin Macias <>
Subject: Re: 5.294 Mainstream Linguistics

As a Mexican who's been both under the influence of the country's
political system, as well as various syntactic theories, I'm
amused to see a comparison between the PRI and mainstream
Linguistics. Can I just point out that the Mexican system is,
among other things, not democratic because most people in
the country are not 'democratic' (as the term is normally
understood in the US, for example)? In other words, most
do not behave as if they believe that disputes can by settled
by rational argumentation, or that what counts is _what_ is
said, and not _who_ says it, or that one should think in terms
of individuals responsible for their own positions, and not in
terms of groups that vie for power.

The question is then: are then linguists, specially syntacticians,

Ben Macias
U Cambridge Computer Lab
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 3: 5.294 Mainstream Linguistics

Date: Wed, 16 Mar 1994 07:57:37 5.294 Mainstream Linguistics
From: Paul Deane <>
Subject: 5.294 Mainstream Linguistics

The most recent postings on the subject "Mainstream Linguistics" illustrate
precisely what is wrong with the way the debate is tending to go. On the one
hand, postings like Mark Llewellyn's may help to relieve feelings of frustration
but they do little to improve communication across the the theoretical divide..
On the other hand, Pesetsky's posting illustrates exactly why such complaints
are so tempting to the "case-theoretically challenged" among us.

What good does it do anybody to compare mainstream linguists to the PRI of
Mexico or to the Communist Party of China? Such comparisons, no matter how
apt they may seem to the one flinging them around, are simply: name calling.
The Linguist List has, so far, been mercifully free of flaming. Please let's
keep it that way.

Pesetsky's response may be summarized as follows:

1. We have the right to prefer to hire people whose views we think are
 *correct* and *likely to lead in interesting directions*.

2. So stop complaining.

A response of this nature misses what the complaints are really about. Of
course, people are going to choose to hire people whose work they consider
worthwhile. That's not the issue. To explain the issue, consider the following
hierarchy of "research worthiness":

 I. People whose research is both "correct" and theoretically
 II. People whose research I believe to be on the wrong track,
 but which is dealing with critical issues and is likely to
 yield interesting results.
 III. People whose research I believe to be based on fundamentally
 flawed assumptions, but who are still doing valuable work which
 may prove to be correct (though I doubt it).
 IV. People whose research is based on assumptions so fundamentally
 in error that it cannot possibly yield any useful information or
 lead to worthwhile lines of research.
 V. People whose research is better characterized as "pseudoscience"
 or ideology than as valid scientific inquiry.

All of us categorize other people's research along these lines. But we differ as
about where we rank things. Obviously, we want our universities to have lots
of people doing (i), a smaller proportion of people doing II, and enough people
doing III to protect us against getting stuck in blind alleys if it turns out
that our assumptions were mistaken. We mostly read work by (i) and (ii), and
try to stay conversant with the most important people doing (iii). If we judge
someone falls into categories IV and V, we don't read their work and we don't
even consider hiring someone of their ilk at our university.
 Now, the complaint against the "mainstream" is not that it exercises such
preferential treatment, but that it operates with an extremely cramped
definition of what constitutes "worthwhile linguistic research". The hierarchy
that seems to be in operation is:

 I. Whatever is current at MIT.
 II. Other versions of the G.B./P.P.A. approach.
 III. Other generative theories, such as HPSG.
 IV. Other formal linguistic theories outside the generative
 V. Functional and cognitive theories of language.

That is, the complaint that is really being made is that the "mainstream"
is writing off dissident work as intellectually valueless, to the point of
not needing to be answered or refuted or included (via university hiring
decisions) in the ongoing disciplinary discourse.

To the extent that this is true (and I know perfectly well that it is not
true of all mainstream linguists), it has a fundamentally perverse effect
on the way people structure their research. I have had people tell me in
effect "the interesting part of my paper is in the data. Don't worry about
the analysis; I put it in there to make the paper palatable to the audience
I was presenting it to." If one must present one's work within very narrow
parameters even to win it a hearing, the field is taking on an ideological
tinge which is not conducive to genuine scientific discourse.

Of course, if the research I am doing is NO GOOD for reasons other than its
failure to conform to the current theoretical "mainstream", so be it. But
I don't like being ignored when I am making points directly relevant to
critical issues in the field. Writing off the dissidents guarantees that
facts unsupportive of the current orthodoxy just won't be noticed.

Some examples: If language is a distinct module, or faculty, we would not
expect to find any direct connection in the brain between language and
other fundamental cognitive abilities. And yet, the posterior language
areas (Wernicke's area and the adjacent parts of the inferior parietal
lobe) seem to be fundamentally involved in the visual perception of part
whole structure. (see a 1991 article by Robertson & Lamb in Cognitive
Psychology) Since constituency is fundamental to language, and may
plausibly be viewed as a type of part/whole structure, the coincidence
hardly seems coincidental: but it's the sort of thing one would not notice
if one started with modular assumptions. Similarly, a syndrome like Williams
syndrome is often cited as a pure case of preserved language abilities in
the face of general cognitive deficit. And yet I have found references which
indicate that Williams syndrome children are better at some types of spatial
cognition (those involving details of spatial structure) than at others
(those involving overall shapes and patterns). Since linguistic competence,
too, is a matter of perceiving detailed structure the coincidence may not be

These are neurolinguistic examples. The purely linguistic examples
These are neurolinguistic examples. The purely linguistic examples are
familiar enough (at least in kind) not to belabor here. But the main point
is simple: writing off the dissidents is to eliminate a presence which will
notice facts that may prove critical to the future development of the field.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue