LINGUIST List 3.297

Sun 29 Mar 1992

Disc: Reality of rules

Editor for this issue: <>

Directory

1. Stavros Macrakis, 3.290 Reality of Rules
2. , Rules & langage
3. Vicki Fromkin, Re: 3.281 Rules
4. Swann Philip, 3.290 Reality of Rules
5. , Re: 3.290 Reality of Rules
6. Penni Sibun, grammar and generation
7. Lew Shapiro, Aphasia

Bill Bennett <WAB2phx.cam.ac.uk> says:

 ...we are all aware of the whole which is greater than the sum of
 the parts.... What are we to term the source of the creative output
 of the brain, if not "mind"?

 If such enlightened monism were right, why should we need two terms, both
 "mind" and "brain"? Either the terms are co-extensive, perfect synonyms, and
 there is a wasteful redundancy. Or they differ in reference, and it is this
 very difference which is of critical importance to such human sciences as
 linguistics. Monists are missing out on something.

They do differ in referent: `mind' is the name of a function; `brain'
is the thing that performs that function. A dead brain no longer
functions as a mind. Brains can have other functions, e.g. as food or
as displays in museums. Other things could perhaps function as minds,
although of course we have no examples of such today.

`Passenger-carrying road vehicle' (PCRV) is the name of a function;
`car' is the name of a thing. A broken car is no longer a PCRV.
Other things can function as PCRV, e.g. motorcycles with sidecars.

`Subject' is the name of a function; `noun' is the name of a part of
speech. What is the mysterious `something' that makes a noun into a
subject?

As it happens, English identifies function and mechanism in the names
of many things. When there is a name for function, it is often more
general than any particular mechanism (e.g. weapon vs. gun).

Of course, it is often useful to posit multiple levels of description
of the mechanism that allows the thing to perform a given function.

Now, please tell me what I am missing out on.

	-s

PS Part of the dispute seems to be in the word `is'. A car `is' a
passenger-carrying road vehicle if it works, just as a brain `is' a
mind if it is alive. This does not mean that `car' and `p-c r v' or
`mind' and `brain' are perfect synonyms.

As follow-up to: John_E_Josephumail.umd.edu (jj36) [3.276/1]

JJ36 argued that "langue" was for Saussure a stabilizing factor. This is to
ignore Saussure's emphasis on the antinomy between a "force du
clocher...agissant comme principe divisant" ("parish influence...acting as
divisive principle") and "intercourse...force unifiante" ("general
communication...unifying influence"). "Langue" was not a simple stabilizer. I
shall show that it was contrasted with "langage" as <Aspects><competence> is
contrasted with the capacity for knowledge of (all and any) language. This
latter, the proper study of linguistics, is not encompassed by "I-grammar",
which in <KOL, 22> is characterized as "some element of the mind of the person
who knows THE language" [my emphasis].

The 4 major terms used by Saussure were
 1) la parole, "speech performance"
 2) une langue, "a language"
 3) la langue, "tacit knowledge of a particular
 language" (cf. >Aspects<, 1965, 3 >competence< "the speaker-hearer's
 knowledge of his language")
 4) le langage, "tacit knowledge for language"

[Material henceforth in this posting is cited (accentless) from Godel (R., 1969,
Sources manuscrites du Cours de linguistique generale, Droz) translated, except
for technical terms, by me.]

3) was defined by Saussure as a "fait social" only as far as it was the trigger
of the universal potential for language (=4) Consider the 2 following
statements

1) > the faculty of "le langage" is a fact distinct from "la
 > langue", but it cannot be exercised without the latter.
 (p.266)
2) > "langue" and "langage" are but one single thing: one is the generalisation
 > of the other. "Le langage" is not an object which is immediately
 > categorisable.
 (p.266)

By the third course (1910-11), and close to the end of his life, Saussure had
begun to replace the notion of (capacity for knowledge of a specific language),
"la langue" by the wider notion of general linguistic capacity, "le langage".
This has been the intellectual route that Chomsky travelled between 1965 and
KOL, from the modelling of knowledge of A language to knowledge of language.
"Le langage" stands here in relation to "la langue" , as mind relates to the
knowledge of particular languages.

As further confirmation that Saussure was thinking of mind is this statement
 > Faculty of "le langage": is only a power, faculty, the > organisation ready
 > for talking.
 (p.266)

Saussure gave the plan of his third and final course on 4 November 1910 as
 > 1. Les langues
 > 2. La langue
 > 3. La faculte et l'exercice du langage chez les individus.
 (p.77)
He justified this order by saying that
 > We must begin with what is given: "les langues"; then draw from it what is
 > universal:"la langue". Then only shall we be concerned with "le langage" in
 > individuals.
 (p.77)

It is a matter of academic regret that Saussure was not able to broach more
fully the third section of this last course. The result has been for linguists
to leave "le langage" and "mind" out of their discussions. The intellectual
route that Chomsky travelled between 1965 and KOL, had been traversed by
Saussure half a century before.

The triggering action of the speech-community as "fait social" cannot be
denied, but this should not be taken as the whole meaning of "langue", seat of
the potential for knowledge of a language. "Le langage" was used by Saussure to
denote specific competence for human language. It is beyond doubt that he meant
the term to apply to the mind and not to the brain.

 Bill Bennett.

Message 3: Re: 3.281 Rules

Another quote, re Alexis Manaster Ramer's reference to a quote from Chomsky.
Here it is:

 "...Suppose that someone were to discover a certain pattern of
electrical activity in the brain that correlated in clear cases with
the present of WH clauses, relative clauses (finite and infinital)and
WH questions (direct5 and indirect5)., Suppose that this pattern of
electrical activbity is observed when a person speaks or understands.
Would we now have evidence for the psychological reality of the postulated
mental representations?
 "We would now have a new kind of 'evidence', but I see no merit to
the contention that this new evidence bears on psychological reality
whereas the old evidence only relates to hypothetical constructions. The
new evidence might or might not be more persuasive than the old; that would
depend on its charactger and reliability, the degree to which the
principles dealing with this evidence are tenable, intelligible, compelling
and so. on." (Chomsky 1978. On the biological basis of language capacities.
in Psychology and biology of language and thought: Essays in honor of Eric
Lenneberg, ed. by G.A. Miller and E. Lenneberg. NY: Acad. Press. 199-220.)

 And further (for those who do not think Chomsky wants to consider any
 evidence except intuitions of the native speaker: "Suppose there is some
data from electrical activity of the brain that bears on, say, word
boundaries. Why should that be irrelevant to word boundaries? It just seems
absurd to restrict linguistics to the study of introspective judgments."
(N Chomsky 1982. On the generative enterprise: A discussion with Riny
Huybregts and henk van Riemsdijk. Dorderecht: Foris.

Not offerred because Chomsky's views are necessarily more interesting than
those expressed in LINGUIST but because he is so often misquoted. VAF

Regarding the reductionist claim that the "mind IS the brain". In
an obvious sense this is false: there are brain processes that are
inaccessible to consciousness, while the mind itself is aware of
the whole body not just the brain. It is equally false in the
context of the research agenda for which it acts as a slogan: the
mind is a process (not a thing) involving the continuous interaction
of an organism with its environment AND its own past history (both
individual and evolutionary). To equate the mind with brain
chemistry is a bit like equating the economy with the metallurgy
of coins.

Philip Swann
FPSE-TECFA
University of Geneva

Message 5: Re: 3.290 Reality of Rules

Avery Andrews writes:

"So, even granting that individual linguistic facts are about norms, we
are left with the question of what is causing the generalizations
that one discerns in the norms, and the structure of the brain remains
the only candidate answer that looks plausible today."

I partly agree: If true causes are what we are looking for, we probably
would have to look at brains. The question is, however, whether *grammars*
can be taken to describe (aspects of) such causes.

Grammars do not describe what *must* happen. It is possible to choose not
to conform to the norms for syntactic well-formedness described by a
grammar. Since it is possible to know (in the sense of being able to
follow) a grammatical rule without actually following it, the fact that
someone knows a rule cannot by itself explain the fact that he *does*
follow it.

Besides, even assuming that everybody always speaks grammatically, a
grammar does not explain facts occurring in time and space, such as the
fact that a given utterance is made. What it explains, is why utterances,
once made, have the institutional properties they do. I would hesitate to
regard such explanations as causal explanations.

I suppose the crucial question is: is it plausible to assume that whatever
causal mechanisms in there enable us to conform to the grammatical norms
must have the same structure as a grammarian's description of the norms
themselves?
To me, that looks like the question: When you know something, do you then
become internally like that which you know? (I believe the etymology of
'information' is related to such an idea: you are informed by something by
assuming its form.) I see no a priori reason to think that it is so.

Helge Dyvik

 From: ADA612csc1.anu.edu.au (AVERY D ANDREWS)
 Subject: the reality of rules

 Re Penni Sibun (Linguist 3.275):
 >...
 >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.

 Is this program part of a theory that explains *why* we find the kinds
 of grammatical generalziations that we do? If so, it might be a
 genuine alternative to having mentally represented grammars (although
 people might disagree about whether there was or wasn't a grammar
 lurking in it somewhere). Otherwise, it isn't (at least, not yet),
 since it doesn't address the main reason for positing rules and grammars
 in the first place.

no, my theory doesn't explain grammatical generalizations, though i
would certainly like to be able to. what i'd first like to do is have
a reasonably coherent story of what's going on, and *then* work on
relating it to more traditional linguistic approaches. i think both
parts are hard, and i've really only started on the first one.

 From: ADA612csc1.anu.edu.au (AVERY D ANDREWS)
 Subject: Reality of Rules (vs. of Grammars)

 Here's a way in which I could be right about rules and Penni Sibun
 about grammars.
 The rules in a linguist's grammar could be `projections' of various
 heterogeneous aspects of the mental structure

yes, i would agree with you in principle.

 From: ADA612csc1.anu.edu.au (AVERY D ANDREWS)
 Subject: Reality of Rules (& social norms)

 between syntax and semantics here. In the case of semantics, one
 can explain a lot in terms of a division of labor, so that meanings
 are in a sense socially distributed: we have access to the intension
 of the term `plutonium' even though we don't grasp it ourselves because
 we are members of a speech community and a society in which there are
 people who do (we can vote to make bombs out of it, or not to).
 But the division of labor idea doesn't apply to syntax,
 since people mostly produce and understand sentences by themselves
 (parents finishing each other's commands to their children would be
 a minor exception in the case of production, but this sort of case
 is clearly derivative and parasitic on the normal one-speaker-per-sentence
 situation).

i must disagree with you here. i suggest looking at a careful
transcript (better yet, with a video too) of half an hour or so of
conversation, and you will see plenty of instances not only of people
continuing each other's sentences and other constructions, but many
sorts of carry over between what one person and the next one say.
examples include repeated intonational patterns, similar clause
structures, repeated or related (eg, ``opposite'') lexical items and
associated structures.

i would also think that many prosaic examples of ellipsis in
question-answering clearly suggest shared access to syntactic
constructions. eg,

 where did you put it?
 in the drawer.

it's certainly true that people who know each other well have more
shared context, but that's not particular to syntax. do you really
think that how we talk with people we know well is derivative? surely
we learned to speak with people we know well, and the bulk of our
speaking is with people we know well.

while people certainly produce lots of sentences by themselves, they
also produce a lot of non-sentences. i don't really have a sense of
what the proportions are: one of the first things one realizes when
faced with a transcript, esp. one including more than one speaker, is
that the stream of language simply cannot be partitioned into
sentences.

i'm enjoying this discussion; however, i will be offline for the next
ten days, so won't be able to rejoin it till then.

				--penni

Thoughts on Avery Andrew's comments about Grodzinsky's work
with move-alpha and aphasia:

1. The linear-order strategy doesn't explain the data.
The point is that these aphasic subjects do not perform below
chance, but only at chance (and not above chance).That is,
if the subject always used a linear-order strategy of
Agent first for a passive, for example, then the subject would
be consistently wrong (100% of the time) when chosing between
one of two pictures in response to a sentence (with one picture
showing a girl kicking a boy and the other showing a boy kicking
a girl, for example). This at-chance, but not below-chance
pattern of results is also found for object relatives (and
should be the case for any sentence that has a trace in direct
object position).

2. I don't really see what this has to do with the reality
of rules discussion, however. Grodzinsky's theory is
intended as a description of the deficit in agrammatic
Broca's aphasia, a description that cannot be made unless
move-alpha and traces are invoked. But just because one needs
to refer to these contructs to explain the neuropsychological
data does not necessarily mean that such things are "in the head."

3. There is an interesting processing explanation for
Grodzinsky's description of the agrammatic Broca's data. The
story goes like this: The brain damage underlying Broca's
aphasia has resulted in a slower-than-normal access
mechanism such that when the subject needs to "fill" a "gap"
in a sentence (i.e., reactivate an antecedent to the trace),
he/she doesn't do it quickly enough to conform to the rapid
on-line requirements of sentence processing. This theory
suggests that "Broca's area," an area traditionally
considered to be important for the production (and now,
comprehension) of aspects of syntax, might support rapid
on-line analyses of information, including - but perhaps
not limited to - lexical & syntactic information. I refer
you to the work of Edgar Zurif and Dave Swinney and colleagues.

ShapiroLFAUVAX.bitnet