LINGUIST List 3.276

Sat 21 Mar 1992

Disc: Rules And The Brain

Editor for this issue: <>


  1. jj36, Re: 3.266 The Reality of Rules (Part 2)
  2. , reality of rules
  3. , Re: 3.265 The Reality of Rules
  4. Paul Saka, RE: rules [3.250]
  5. Rick Wojcik, Re: 3.265 The Reality of Rules
  6. Penni Sibun, in whose heads are grammars
  7. , Reality of Rules/Ambiguity of 'Grammar'
  8. AVERY D ANDREWS, The Reality of Rules (& Itkonen)

Message 1: Re: 3.266 The Reality of Rules (Part 2)

Date: Thu, 19 Mar 92 20:58 EST
From: jj36 <>
Subject: Re: 3.266 The Reality of Rules (Part 2)

A note on Martti Nyman's interpretation of Saussure's characterization of
"langue" as "fait social". In his very interesting posting on the reality
of rules, Nyman wrote: "Unlike mathematical entities, languages are subject
to constant (regional, social, stylistic) variation and constant change.
The variability and changeability of language results from human collective
activity, and so it's proper to characterize language as a social entity.
This was Ferdinand de Saussure's point ("langue" as "fait social")...
 This is precisely the opposite of my understanding. As I read Saussure,
the social nature of langue accounts for its stability, not its
instability. Ch.2, sec.2 of the _Cours_ ("Mutabilite' du signe"), states
with regard to langue: "situee a la fois dans la masse sociale et dans le
temps, personne ne peut rien y changer" ('situated simultaneously in the
social mass and in time, no one can change anything about it' -- dangling
participle intact). The passage goes on to state that it is the
arbitrariness of language that accounts for its variability and
changeability, and that this is always in tension with the stability
inherent to language by virtue of its being a social fact.
 --John E. Joseph
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: reality of rules

Date: Fri, 20 Mar 92 10:21:01 GMreality of rules
From: <>
Subject: reality of rules

It is rather agreeable to find linguists hung up on a preposition. Instead of
"in the brain" what about "of the brain". Linguistics, I point out trivially,
is no proper science if it is only, in some way which does not appeal to my
squeamish tastes, concerned with the material reality of the brain. Our
COGNITIVE science (and I cannot think of any linguist wishing to disagree with
that label) must MAP the collective action of the brain's componential
structure. If this is not so, and if we are not concerned with MAPPINGg of
the activity of the material content which is called the brain, then we are
nothing but dabblers and philologists. And we shall have no way of dealing with
1) the recovery of language after brain trauma
2a) the ability of "English" speakers to accept Ozarkian "for-to" as "possible
English" and not a footnote in KOL;
2b) the same non-physical restructuring of rules of English grammar to allow
for the ECM in finite clauses represented by the London Times'
 "at Evry he came out with guns blazing against the self-perpetuating elite of
France's grandes ecoles, whom one is assured run evrything worth running in
France." (12 June 1990)
or the London Evening Standard's
"Earlier came the murder of [...], whom police believe was shot as a result of
an underworld feud." (31 December 1990)
 The form of the rules which will allow for the acceptance
by American English speakers of the examples in (2) (without commas, which I
would find to be a different structure) is open (we know the difficulty of
"variable rules") to discussion, once we have acknowledged the subject-matter
of our science to be knowledge of language in the mind. Bill Bennett.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 3: Re: 3.265 The Reality of Rules

Date: Fri, 20 Mar 92 09:30 EST
From: <KINGSTONcs.umass.EDU>
Subject: Re: 3.265 The Reality of Rules

As a daily laborer in experimental vineyards (as a phonetician) I must confess
that abstract arguments of the sort that have been advanced in recent postings
on the "reality of rules" leave me unconverted to any of the positions argued
for (I can hear the howls about the literal-mindedness of phoneticians
already). To my mind (literal or not), it is fortunate that there is a
wealth of empirical evidence on the reality (or lack of it) of linguistic
rules. Some of this evidence is external (see Fromkin's posting), but
a great deal of it is internal, in the sense that has been obtained from
observations of what people do when speaking or listening (again, I can
hear the howls, that that's just performance, but I would respond that
even if none of us or our subjects are ever _ideal_ speaker/hearers, we are
also never utterly incompetent speaker/hearers either). Even more to the
point there have been a number of studies that directly address the
question of what in speakers' heads drives their behavior. The most
interesting of these is a paper by Derwing and Baker on the acquisition
of the regular allomorphy of the English past tense, plural, possessive,
and 3rd singular indicative morphemes which appeared in _Experimental
Linguistics_ (G. Prideaux ed.) and more recently the Pinker paper "Rules
of language" in _Science_ 253 (8/2/91) pp 530-535. D&B show that the
analysis most linguists favor for the allomorphy of these morphemes best
describes their acquisition and Pinker shows that both rule and associative
memory (connectionist) schemas are necessary to account comprehensively for
past tense formation (rules for regular formation and associative memory for
irregular ones). These papers as well as many papers in recent issues
of _J. of Memory and Language_ and other similar journals show that it's
not necessary to try to settle what is or isn't in people's heads _a priori_.
One can go find out.
John Kingston
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 4: RE: rules [3.250]

Date: Fri, 20 Mar 92 11:45:57 -0RE: rules [3.250]
From: Paul Saka <sakacogsci.Berkeley.EDU>
Subject: RE: rules [3.250]

>From: [LING 3.250]

>linguistic rules[...] could not be "real" in the sense of being open
>to representation in neurological terms. If that were so, then our
>science would be concerned with the brain and not with the mind.

This argument needs to be rejected. First, it begs the question to
claim that Lx is concerned with the mind rather than the brain.

Second, according to the identity theory -- and most
scientists ARE monists, by the way -- the mind IS the brain. Even
for dualists, mental rules might very well correspond in one way
or another to neurological structure. Therefore just because something
is concerned with the mind does not preclude it from being concerned
with the brain.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 5: Re: 3.265 The Reality of Rules

Date: Fri, 20 Mar 92 09:59:39 PSRe: 3.265 The Reality of Rules
From: Rick Wojcik <>
Subject: Re: 3.265 The Reality of Rules

I do not want to get into a lengthy rebuttal of Rob Stainton's remarks on the
relevance of psychology to linguistic grammars, but he does express a strong
opinion against psychologistic linguistic theory. Part of the problem seems
to be that he construes proponents of psychologism as claiming that linguistic
generalizations are valid only if they represent descriptions of psychological
function. To me, the following remark represented a key element of his

> So linguistics might be like mechanics: no one attributes *knowledge*
> of the gravitational constant to rocks. But their behaviour conforms
> to that law.

First of all, I don't think it is fair to claim that Stanton is arguing that
linguistics *must* be like mechanics. Nor do I claim that linguistics *must*
be like psychology. My claim is that it is valid to construct grammars that
describe mental rules. The plain fact is that people differ from rocks in
a very interesting way: people have brains. I suspect that no one attributes
*knowledge* of the gravitational constant to rocks because they don't have
brains. And, if you are interested in describing the behavior of people, as
opposed to rocks, then you certainly do want to figure psychological function
somewhere in the causal chain of events.

			-Rick Wojcik (
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 6: in whose heads are grammars

Date: Fri, 20 Mar 1992 15:16:35in whose heads are grammars
From: Penni Sibun <>
Subject: in whose heads are grammars

i previously sent a message to this list stating that grammars are in
linguists' heads but there's no reason to suppose that they are in
speakers'. despite a poster's assumption that i was being sarcastic,
i mean this very seriously, though perhaps i expressed myself too

grammars may well be a useful tool for linguists to describe and
theorize about language and how people use it. however, just because
grammars are a good *descriptive* tool, it doesn't follow that
grammars play any role in language *use*. rob stainton's waltz
analogy illustrates this point very well.

one way to think about the locus of grammars is to try to model
language use and see if grammars (syntactic, discourse, etc.) are
necessary or sufficient for producing (or understanding) language.
within computational linguistics, there is a small ``natural language
generation'' community which builds or designs computer programs that
produce (presumably intelligible) text. of course, a computational
model doesn't prove anything about the insides of people's heads;
however, if a program can generate text without representing and using
grammars, this suggests that maybe people can too. further, if we can
agree that we can't find a grammar anywhere in the program, yet we can
discuss the output language in terms of a grammar (eg, decide that
the output is a grammatical sentence), then we have an example of the
grammar being in *our* heads, as linguists talking about grammatical
sentences, but not in the ``head'' of the producer of the grammatical
sentence under discussion. it is reasonable and appropriate for us to
talk about the grammaticality of the sentence; it would not be
reasonable or appropriate to assume that it follows that there is a
grammar in the program.

for my just-completed phd thesis, i wrote a generation program that
produces coherent texts that are up to a page long, without
representing or using either a discourse or a sentence grammar. the
program's job is to talk about something, and it is continually
deciding what to say next. it makes this decision based on things
like what it's already said, what it hasn't talked about that is
closely related to what it's just mentioned, and what words and
linguistic patterns it knows about that might express what it chooses
to say. when it has decided what to say it says it, and goes on to
the next choice. while the program does not concern itself with
syntactic or discourse-level grammaticality in making its choices, its
output can be judged as to whether the discourse structure is coherent
and whether the syntax is grammatical (the program usually does ok on
both counts).

this work and the arguments i present here are described in more
detail in an article in the next issue of _computational
intelligence_; i'd be happy to send preprints on request.

				--penni sibun
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 7: Reality of Rules/Ambiguity of 'Grammar'

Date: Fri, 20 Mar 92 18:41:10 ESReality of Rules/Ambiguity of 'Grammar'
From: <>
Subject: Reality of Rules/Ambiguity of 'Grammar'

Since the issue of linguists' grammars vs. speakers' grammars
has reared its ugly head (and with it the old business of
Chomsky's "systematically ambiguous" use of the word 'grammar'),
I thought I would point out that, as I read Chomsky's KNOWLEDGE
OF LANGUAGE, he has now given up this ambiguity. In the
new terminology,

 grammar = linguist's grammar
 I-language = speaker's grammar

And, of course,

 E-language = language
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 8: The Reality of Rules (& Itkonen)

Date: Sat, 21 Mar 1992 17:56:05 The Reality of Rules (& Itkonen)
Subject: The Reality of Rules (& Itkonen)

Re: Nyman on Andrews on Itkonen

My interpretation of what Itkonen meant (maybe I got it wrong) was that
he thought that linguists had some kind of a priori intuitive knowledge of what
generalizational statements about their language were true, such as that
articles are initial in NPs in English, and that what their task was was to
express that knowledge in the best possible way, similar to looking for nice
axiomatizations of mathematical systems

What I would take as a knockdown refutation of what I said is a passage
where Itkonen clearly says that linguist's generalizations are in the
first instance arrived at by induction from the properties of particular
strings of words (contexts being involved in the properties, of course).
But this would leave me puzzled as to in what sense linguistics was supposed
to be nonempirical, or like mathematics.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue