Copy Link
Add to Bookmark
Report
NL-KR Digest Volume 05 No. 08
NL-KR Digest (8/09/88 22:25:02) Volume 5 Number 8
Today's Topics:
Generative grammar, competence/performance
Re: Chomsky reference
Submissions: NL-KR@CS.ROCHESTER.EDU
Requests, policy: NL-KR-REQUEST@CS.ROCHESTER.EDU
----------------------------------------------------------------------
Date: Thu, 4 Aug 88 19:00 EDT
From: Arnold M. Zwicky <amz@tut.cis.ohio-state.edu>
Subject: Generative grammar, competence/performance
The discussion on these topics has been remarkably confused, I think because
people are trying to talk about a set of complex, interrelated, and subtle
ideas as if they were really one simple, obvious idea. I can't sort out
much of this in a brief space, but I would at least like to try to clarify
some matters that surfaced in John Merrill's posting. The problem isn't
some idiosyncrasy on his part, however; in fact, the confusion here has
its roots in Chomsky's rhetorical strategies, which tend (rather wickedly,
to my mind) to juxtapose distinct ideas and to suggest much greater
continuity of thought - in the field in general, and in Chomsky's work
in particular - than there really is.
First, the competence/performance distinction. Whatever Chomsky might once
have meant by these words, or means now, he has caused most theoretical
linguists to distinguish KNOWLEDGE of language from linguistic BEHAVIOR
(speaking, comprehending, remembering). The analogy is to the information
embodied in a program versus what the machine actually does, in input, output,
and storage terms. (The machine I'm on
wasn't programmed to crash, but it has, several times already today.)
Now, linguistic knowledge is embodied in individual people, and linguistic
behavior is, at least in the first instance, the behavior of individual people.
There are then at least as many systems of linguistic knowledge as there
are people (more, because many people clearly operate with several such
systems) - at least as many LANGUAGES as there are people. [Try not to
worry about that stuff about languages being sets of sentences. That's
a different sense of `language'.] It is easy to demonstrate, in fact,
for any two people that they do not have quite the same knowledge of
language, by probing the way they associate phonology with semantics
in contexts; actually, there are surprisingly many differences between
people in this regard. Still, we sort of manage to get along with one
another, so that the systems must overlap to some considerable extent.
The task of a child, in this picture, is not somehow to figure out which
language is being spoken by the people around her - there IS no such thing -
but to develop some knowledge that will keep her behavior in reasonable
alignment with the behavior of the people around her. This will change
through time, given fresh experience - change dramatically at first,
then much more slowly later on.
The sense of `language' in expressions like `the English language' is,
of course, still another sense associated with this orthographic expression.
Language in this sense belongs, in some way that has to be clarified - this
is one of the things that the field of sociolinguistics is about - to a
SOCIAL GROUP rather than to an individual. [I've now hit at least three
importantly different senses of `language'. This is why we have technical
terms; ambiguities like these are quite intolerable in careful discourse.]
Oi. This barely scratches the surface, and some of what I've said is
awfully crude. Maybe we could spend less time scoring points (off one
another, off Chomsky, off whoever) and more trying to understand this
stuff?
- Arnold
------------------------------
Date: Fri, 5 Aug 88 01:23 EDT
From: Rob Bernardo <rob@pbhyf.PacBell.COM>
Subject: Re: Generative grammar, competence/performance
In article <19327@tut.cis.ohio-state.edu> amz@tut.cis.ohio-state.edu (Arnold M. Zwicky) writes:
+First, the competence/performance distinction. Whatever Chomsky might once
+have meant by these words, or means now, he has caused most theoretical
+linguists to distinguish KNOWLEDGE of language from linguistic BEHAVIOR
+(speaking, comprehending, remembering).
Okay. But we need to make another distinction. The self-conscious knowledge
that a person reports using for a skilled behavior may not really be the
knowledge they use in the behavior. Specifically with language, it is
not clear that the knowledge manifest in a native speaker's grammaticality
judgements of hypothetical instances of language behavior ("sentences made
up by the linguist") is really the knowledge that underlies that native
speaker's actual language behavior.
If a psychologist were to ask a person what cues s/he uses to judge
the distance of an object, the person may report some visual cues, but
not necessarily be aware of the ones actually at work. S/he may not
be aware of some, because they operate below the level of
consciousness, or because the person hasn't reflected on them; or s/he
may falsely report report or exaggerate the importance of others,
because s/he simply believes they are used.
--
Rob Bernardo uucp: ...![backbone]!pacbell!rob internet: rob@PacBell.COM
Pacific Bell SRVAC Room 4E750 | R Bar JB (residence)
San Ramon, California (415) 823-2417 | Concord, California (415) 827-4301
------------------------------
Date: Fri, 5 Aug 88 13:02 EDT
From: Greg Lee <lee@uhccux.uhcc.hawaii.edu>
Subject: Re: Generative grammar, competence/performance
From article <19327@tut.cis.ohio-state.edu>, by amz@tut.cis.ohio-state.edu (Arnold M. Zwicky):
" ...
" much of this in a brief space, but I would at least like to try to clarify
" ...
" (speaking, comprehending, remembering). The analogy is to the information
" embodied in a program versus what the machine actually does, in input, output,
" and storage terms. (The machine I'm on
" wasn't programmed to crash, but it has, several times already today.)
If it was a software problem that caused the machine to crash, then
information embodied in a program was indeed involved. But we know that
it wasn't programmed to crash, because we know something of the intent
of programmers. Is this analogy which requires us to judge the intent
of a Designer a clarification? Maybe it's about as good as one can do.
For the case of language, if we think there was no Designer, or at least
that we have no way of judging His intent other than through
theory-making and experiment, the assumption of a performance/competence
distinction is of no use to us at all.
Why are we still spending time after all these years still trying to get out
of this muddle of Chomsky's? He's making up new ones all the time. Let's
get up to date. How about the one where we figure out that there are
only a very few settable Parameters in Universal Grammar? If there were
lots, children couldn't learn a language, since they have to set
all the Parameters to learn one. Sound plausible? Find the trick.
Greg, lee@uhccux.uhcc.hawaii.edu
------------------------------
Date: Sat, 6 Aug 88 17:23 EDT
From: morgan@clio.las.uiuc.edu
Subject: Re: Generative grammar, competence/
I'm all for it. Where is this new theory written up that avoids all these
problems? Where can I read about it?
------------------------------
Date: Sat, 6 Aug 88 23:21 EDT
From: Apoorva Muralidhara <apoorva@mind.UUCP>
Subject: Generative grammar
I tried to post this response to John Merrill's classic Goedel Numbers
and Generative Grammar: The Hidden Connection (whoops! pun
unintentional!) posting two days ago, but someone told me it hadn't
come through, so I am reposting it.
apoorva>The aim of the science of generative grammar is to fully and precisely
apoorva>characterize the state of the inborn, species-universal
apoorva>language-learning facility of the brain (for whose existence there is
apoorva>evidence from the poverty of stimulus argument--children learning a
apoorva>language as a native language do not receive enough data to do so if
apoorva>they were "tabula rasa"; hence they are not)
John Merrill:
*For whose existence there is evidence? The evidence consists of assertions
*of the form "All languages are infinite, and there is no way to produce an
*algorithm, which, given a finite subset of an infinite language, will
*compute the true G\"odel code for that language. Ergo, there must be
*innate facilities shared by all people which allow them to intuit the
*G\"odel number of the language from a finite sample."
What do Goedel numbers have to do with generative grammar?
*This argument is only valid if speakers learn the "true" language.
*But it depends essentially upon an unspoken assertion---speakers know
*the "true" language. It isn't even clear that there is a "true"
*language; even assuming that there is, I don't know anyone who speaks
*it.
There is no such thing as the "true" language, and generative grammar
cetainly does not claim this. It is not only an unspoken assertion,
it is an unbelieved assertion. I would go on to re-explain about
generative grammar, but since by this point in your article you have
assumed both that generative grammar has something to do with Goedel
numbering (now, really!) *and* that generative grammar claims that
there is some Platonic ideal for each language, it is difficult to
extract any sense from the rest of your article at all.
But I will try:
*The competence/performance distinction upon which that purported proof
*depends is a chimera. If one assumes that speakers do not, in fact,
*share a common language, but merely share languages which overlap in
*large part, then there is absolutely no reason to accept the Cartesian
*argument from the poverty of the stimulus. The fact that there is no
*way to compute the "true" G\"odel number of the language is
*irrelevant---all that is needed is an algorithm for computing an
*approximate G\"odel number.
First of all, why does it follow from anything you say in this article
that the competence/performance distinction is a "chimera"?
As for the rest, ignoring the irrelevant comments about Goedel numbers
(!!!), let me first remind you that the point is that the child
projects the data onto a *grammar* of the language consistent with
that of the language community and second of all: granted for the sake
of argument that even in this case, the child's grammar still differs
in some ways from that of the speakers who supplied the data, what is
there to explain the agreement between the acquired grammar of the
child and the grammar of the speaker, beyond that of the data? Her
grammar should be *anything at all consistent with the data*, but that
is not what in fact occurs. If you read research in generative
grammar, you will see (look at Luigi Burzio's book ITALIAN SYNTAX: A
Government Binding Approach, intended to be accessible to laymen, for
example) that grammars are much more complex than you think, and that
wellformedness judgements are firm and consistent on sentences which
could not possibly have been present in the data. By the way, how do
you explain judgements of illformedness? Children do not generally
receive negative data (i.e., information that a sentence is
illformed--e.g. *It seems John to be confused or *John seems that is
confused with the intended meaning of "John seems to be confused" or
"It seems that John is confused" (linguists use '*' to indicate
ungrammaticality).
Let me quote from Burzio's preface:
@In the course of our everyday lives, we generally take our knowledge
@of language for granted. Occasionally, we may become aware of its
@great practical importance, but we rarely pay any attention to the
@formal properties that language has. Yet these properties are
@remarkably complex. So complex that the question immediately arises
@as to how we could know so much. . . . [Goes on to discuss the difference
@between Italian transitive and ergative verbs and asks how we could
@know about these verb classes.--apoorva]
@Among the phenomena we will consider, only a small number have ever
@been noted by even the most thorough of traditional or pedagogical
@grammars. But, if the knowledge that speakers have thus exceeds--and
@by far--the contents of grammar books, then it could not be claimed
@that this knowledge is merely a reflex of what grammar books say,
@induced by some form of instruction. It would seem more plausible to
@hold the opposite view, that it is grammar books that represent some
@reflection of what we know, and a dim one at that. But even more
@significantly, it could not be claimed that the knowledge of language
@is, in any reasonable sense, derived by 'induction' from the
@linguistic facts that language learners are exposed to. For instance,
@many of the facts we will addresss, though conforming with the
@intuitions of all speakers, are very exotic from the point of view of
@normal life situations, as they can only be assembled artificially,
@through the painstaking work of the linguist. It is simply
@unimaginable that such facts could have occurred with sufficient
@frequency or consistency to provide an adequate inductive basis for
@what is a complex system of knowledge, so remarkably uniform among
@speakers. The much richer character of the knowledge attained,
@compared with the experience requried to attain it, compels us to
@postulate that the structure of the human mind is itself contributing
@significantly to that knowledge.
*And if that's all that is needed, then the claim that children cannot
*be tabulae rasae is nonsense.
The poverty of the stimulus argument, as I have tried, probably
unsuccessfully, to help you understand, shows (along with the many
other non-linguistic reasons to show babies' minds can't be tabula
rasae) that babies are not tabulae rasae.
What is nonsense is not generative grammar, but rather John Merrill's
posting, what with its talk of Goedel numbers and "true" languages.
--Apoorva
------------------------------
Date: Fri, 5 Aug 88 13:49 EDT
From: Greg Lee <lee@uhccux.uhcc.hawaii.edu>
Subject: Re: Chomsky reference
[Site mail problems has caused previous messages on this topic to be
deleted, sorry! -BWM]
From article <2713@mind.UUCP>, by apoorva@mind.UUCP (Apoorva Muralidhara):
"
" Generative grammar is a theory of linguistic *competence*, not
" linguistic *performance*. We are not trying to understand "language
" behavior," which is linguistic performance. We are not claiming
" anything about linguistic behavior.
Speakers' judgments are one kind of linguistic behavior.
" Linguistic competence, of course, includes grammaticality judgements,
" known by speakers, of sentences that have never been uttered. The
" *only way* to discover what this knowledge is is to ask the speakers.
I don't think that you completely appreciate the beauty and subtlety of
Chomsky's thought. We must also distinguish competence and performance
in the matter of speakers' judgments, since, after all, mistakes
due to inattention, plugged up ears, and so on have to be allowed for.
Those judgements that are *correct* are termed grammaticality judgments.
Others, or judgments without regard to their correctness, are termed
acceptability judgments. How do we know which judgments are correct? A
grammatical sentence is one that is generated by a given grammar (see
_Aspects_). Given the grammar, then, we can discard those acceptability
judgments which do not reflect the underlying competence of the naive
native speaker.* Thus we arrive at grammaticality judgments. How do we
tell whether the grammar is correct? We test the sentences it
generates, and those it does not, by asking native speakers' judgments.
Some of those judgments we discard as being failures of performance,
according to whether the sentences involved are generated by the given
grammar.
So, although in a sense generative grammar is a theory of linguistic
behavior, it needn't be, since it really doesn't matter what the
judgments are, you see. This has come to be more and more clearly realized
at MIT subsequent to the theory of _Aspects_.
*(footnote) If it occurs to you that there might be some other way
of detecting the incorrect acceptability judgments, bear in mind that
any such method would require a theory of performance, but
a theory of performance presupposes an adequate theory of competence
(as is well known).
Greg, lee@uhccux.uhcc.hawaii.edu
------------------------------
Date: Sat, 6 Aug 88 16:15 EDT
From: Rob Bernardo <rob@pbhyf.PacBell.COM>
Subject: Re: Chomsky reference
In article <2713@mind.UUCP> apoorva@mind.UUCP (Apoorva Muralidhara) writes:
+And you are quite correct that "generativists . . .[are] not
+describing language [I interpret this to mean lingustic
+performance--Apoorva] but rather accounting for the grammaticality of
+hypothetical sentences [I assume you mean that the sentences need not
+actually have been uttered--Apoorva] as judged by native speakers."
+They account for this by describing the initial state of the innate
+language-learning facility which, upon exposure to data from a
+language, produces a grammar, which is indeed a set of grammaticality
+judgements on sentential structures.
But how do you know if there is a separate thing that is the innate
language-learning facility? How do you know that it produces a grammar?
How do you know that a grammar is a set of grammaticality judgments?
How do you know that sentential structures are units of language, rather
than simply units linguists (and others biased by *written* language)
believe exist?
As I see it these are assumptions generativists make. And since they
aren't obviously true to me I think they need to be based on some lower
level assumptions that are more plausible to me.
--
Rob Bernardo uucp: ...![backbone]!pacbell!rob internet: rob@PacBell.COM
Pacific Bell SRVAC Room 4E750 | R Bar JB (residence)
San Ramon, California (415) 823-2417 | Concord, California (415) 827-4301
------------------------------
Date: Sat, 6 Aug 88 20:09 EDT
From: Apoorva Muralidhara <apoorva@mind.UUCP>
Subject: Re: Chomsky reference
Rob Bernardo writes:
*But how do you know if there is a separate thing that is the innate
*language-learning facility? How do you know that it produces a
grammar? How do you know that a grammar is a set of grammaticality
judgments? How do you know that sentential structures are units of
language, rather than simply units linguists (and others biased by
*written* language) believe exist?
As I've explained before and will explain again, we know that there is
an innate language-learning faculty because of the *poverty of the
stimulus*--children simply do *not* receive enough data to account for
their projection onto a grammar consistent with that of the language
community. Hence there is some innate constraint on generative
capacity involved, which interacts with the data in a structured way
in order to produce a grammar. We call this the "innate
language-learning faculty." See?
There is obviously empirical evidence that people have grammars, as
evidenced by their wellformedness judgements, which are independent of
semantics and so on. This point is clear.
Why do you think linguists are biased by written language?
It is true that we are assuming that sentential structures are units
of language, but this seems amply confirmed empirically. Anyway, the
point is, like in any other science, we construct theories to account
for the data, which include certain assumptions, and then test them
against the data.
Perhaps these assumptions aren't "obviously true" to you. But a
theory based on them *explains* the data. And what more could one
ask?
--Apoorva
------------------------------
Date: Sat, 6 Aug 88 22:18 EDT
From: Clay M Bond <bondc@iuvax.cs.indiana.edu>
Subject: Re: Chomsky reference
All right now. This is smug and condescending. We are all trained linguists,
and we know all the arguments as well as you do, Apoorva; some of us better,
I'm sure, since you can't discuss, but only be nasty.
What you say is true ONLY if you assume in the first place that there is a
difference between *brain* and *mind*, ergo *behavior* and *thought*, etc. We
do not make these assumptions; therefore you have said nothing, made no points,
accomplished no useful communication.
This is the last time I will listen to any kind of diplomatic approach which
places equal blame for mud slinging on both sides of the fence. That is not
true. It is not reflected in the literature (see Pinker and Prince--have you
read it, Apoorva? Have you read Rumelhart and McLellan's paper on learning
the past tense WITHOUT "rules" or any mentalist grammar? If not, you have
no business saying anything.)
Condescension is the refuge of the defenseless, is it not?
--
<<<<<<<<<<<<***<<<<<<<<<<<<***<<<<<<***>>>>>>***>>>>>>>>>>>>***>>>>>>>>>>>>
<< Clay Bond, IU Department of Linguistics >>
<< ARPA: bondc@iuvax.cs.indiana.edu >>
<<<<<<<<<<<<***<<<<<<<<<<<<***<<<<<<***>>>>>>***>>>>>>>>>>>>***>>>>>>>>>>>>
------------------------------
Date: Sun, 7 Aug 88 00:16 EDT
From: Apoorva Muralidhara <apoorva@mind.UUCP>
Subject: Re: Chomsky reference
Clay Bond writes:
*What you say is true ONLY if you assume in the first place that there
*is a difference between *brain* and *mind*, ergo *behavior* and
**thought*, etc. We do not make these assumptions; therefore you have
*said nothing, made no points, accomplished no useful communication.
"Mind" is simply an abstract description of what information is stored
in the brain and how it is changing. I find it difficult to believe
that you don't believe in the existence of mind. (In that case, where
would I be posting from? :-) ). I have certainly said plenty (a habit
of mine!), made several points, but I agree with you that I have
accomplished no useful communication. It takes two to communicate,
and if you are intent on your smug and condescending refusals to
understand, there is no way my explanations can help. And, by the
way, I am not making a distinction between behavior and thought, OR
between brain and mind (where did you get this from? I just posted
saying that brain and mind were essentially the same thing, didn't I?
Wasn't it you who said that they were different, that generative
grammar was a theory of the *mind*, not the *brain*?), but between the
knowledge of grammar in someone's mind/brain and the activity of
speaking. This distinction should be clear.
Clay, I understand that you are a connectionist, and I have nothing
against using neural networks to understand the mind--especially
since, as you guess quite correctly, I have not read Rumelhart and
McLellan, apparently the founders (? correct me if I'm wrong) or some
of the founders of connectionism. "If not, you have no business
saying anything?" And why not? Just having read Chomsky and
completely misunderstood him does not qualify you to talk about
generative grammar either, you know. But this isn't about
"qualifications." What I don't understand is why, if the problem of
linguistics can be explained by neural networks, why this means that
generative grammar is wrong. Surely if the neural networks and the
symbolic, rule-based theory are isomorphic in the relevant way (i.e.,
agree in terms of generative capacity), why is either theory wrong?
Specifically, why does "mentalism" have anything to do with "rules"
anyway? Yes, linguists use both, but while I believe mentalism to be
*true*, *rules* are merely one way of constructing a theory. You
connectionists think you have a better way. This is fine. It'll take
some time to see which, or both, make progress.
*Condescension is the refuge of the defenseless, is it not?
Precisely.
--Apoorva
------------------------------
Date: Sun, 7 Aug 88 11:17 EDT
From: Greg Lee <lee@uhccux.uhcc.hawaii.edu>
Subject: Re: Chomsky reference
From article <3739@pbhyf.PacBell.COM>, by rob@pbhyf.PacBell.COM (Rob Bernardo):
" ...
" How do you know that sentential structures are units of language, rather
" than simply units linguists (and others biased by *written* language)
" believe exist?
Since there are no constant phonetic correlates for grammarians Ss, it
seems clear they have the latter status. To conclude because
linguists simply believe them to exist that they therefore do not exist
requires a level of scepticism I'm not sure I can attain. I'll work
on it.
Greg, lee@uhccux.uhcc.hawaii.edu
------------------------------
Date: Sun, 7 Aug 88 14:35 EDT
From: Clay M Bond <bondc@iuvax.cs.indiana.edu>
Subject: Re: Chomsky reference
In article <2208@uhccux.uhcc.hawaii.edu>, Greg Lee writes:
> From article <3739@pbhyf.PacBell.COM> ... (Rob Bernardo):
>> How do you know that sentential structures are units
>> of language, rather than simply units lintuists (and
>> others biased by *written* language) believe exist?
> Since there are no constant phonetic correlates ...
> seems clear that they have the latter status. To conclude
> ... that they therefore do not exist requires a level
> of scepticism I'm not sure I can attain.
All good points, and certainly nothing here I would necessarily disagree
with. For the sake of discussion, though, let me throw in another, pos-
sibly alternate, question.
Is it so much whether they exist or not (an incredibly sticky topic, rather
reminiscent of the "does the tree make a sound with no\"one to hear it"
problem), as _do sentences exist as units with some special status_ (as
opposed to "non-sentential phrases")? Discourse would seem to indicate
that if they do, then the criteria for categorizing "sentences" vs. "phrases"
must be much more complex than the traditional criteria; deixis, for example,
is clearly not restricted to sentential structures. And this question
stands whether or not we take constituents, phrasal or sentential, as having
linguistic reality or not.
Comments from other linguists out there?
--
<<<<<<<<<<<<***<<<<<<<<<<<<***<<<<<<***>>>>>>***>>>>>>>>>>>>***>>>>>>>>>>>>
<< Clay Bond, IU Department of Linguistics >>
<< ARPA: bondc@iuvax.cs.indiana.edu >>
<<<<<<<<<<<<***<<<<<<<<<<<<***<<<<<<***>>>>>>***>>>>>>>>>>>>***>>>>>>>>>>>>
------------------------------
Date: Mon, 8 Aug 88 15:17 EDT
From: Rob Bernardo <rob@pbhyf.PacBell.COM>
Subject: Re: Chomsky reference
In article <2723@mind.UUCP> apoorva@mind.UUCP (Apoorva Muralidhara) writes:
+
+Rob Bernardo writes:
+
+*But how do you know if there is a separate thing that is the innate
+*language-learning facility? How do you know that it produces a
+grammar? How do you know that a grammar is a set of grammaticality
+judgments? How do you know that sentential structures are units of
+language, rather than simply units linguists (and others biased by
+*written* language) believe exist?
+
+As I've explained before and will explain again, we know that there is
+an innate language-learning faculty because of the *poverty of the
+stimulus*--children simply do *not* receive enough data to account for
+their projection onto a grammar consistent with that of the language
+community. Hence there is some innate constraint on generative
+capacity involved, which interacts with the data in a structured way
+in order to produce a grammar. We call this the "innate
+language-learning faculty." See?
Hm. You missed a point. I asked how you know there is a **separate**
thing that is the innate **language** learning facility. What I'm
getting at is that perhaps there is a facility (in the mass noun
sense), but you have been using "facility" in the count noun sense
to mean a *thing/object* that has that ability. I don't think you've
begun to such that there is such a facility (in the count noun sense),
i.e. one particular to language. After all, your reasoning could be
applied to any learned human behavior. I.e. "People learn how to drive.
.... There is an innate driving-learning facility." :-)
What I'm more underlyingly getting at is this: Just because we humans
have the capability to learn something (language, driving), doesn't mean
this capability is "housed" in some "facility". Nor does it mean that
any set of rules were learned (cf. grammar). The behavior may appear
rule-governed (i.e. it is patterned in such a way as to be describable
by rules we scientists come up with), but that doesn't mean that whatever
it is I have in my brain that enables me to engage in the behavior
includes something in the form of rules. Rules are a convenient way
for us to talk about the regularities we observe in the behavior.
Let's not confuse the map with the territory. (I think this overlaps
with what Clay was getting at with the mind and brain stuff. The mind
is convenient way of talking about what we observe through introspection
and provides a convenient way for talking about what we suppose
goes on in the brain.)
+There is obviously empirical evidence that people have grammars, as
+evidenced by their wellformedness judgements, which are independent of
+semantics and so on. This point is clear.
I don't see how this follows. Just because people can make wellformedness
judgements (and not even consistently from time to time for *one*
person), you suppose people have grammars. Well, even supposing we
simply *define* grammar as that knowledge which people bring to bear
in making wellformedness judgements, how do you conclude that this
grammar is involve in un-selfconscious language behavior?
+It is true that we are assuming that sentential structures are units
+of language, but this seems amply confirmed empirically.
Well, it seems clear to me that utterances have an intended (and perceived)
structure, but the structure is not very much like that of
hypothetical sentences that linguists use. Have you ever transcribed
actual speech of someone talking for a few minutes? It's gruesome work
because it is so terribly different from written language and from
the generativist's hypothetical sentences.
--
Rob Bernardo uucp: ...![backbone]!pacbell!rob internet: rob@PacBell.COM
Pacific Bell SRVAC Room 4E750 | R Bar JB (residence)
San Ramon, California (415) 823-2417 | Concord, California (415) 827-4301
------------------------------
End of NL-KR Digest
*******************
