nl-kr-request@CS.ROCHESTER.EDU (NL-KR Moderator Brad Miller) (04/06/88)
NL-KR Digest (4/05/88 19:43:25) Volume 4 Number 38
Today's Topics:
Phonetic Spelling (was Linguistic Theories)
re: Spanish as a "phonetic language"
Pro-Drop one more...
Schmerling on truncation
What are grammars (for)?
Submissions: NL-KR@CS.ROCHESTER.EDU
Requests, policy: NL-KR-REQUEST@CS.ROCHESTER.EDU
----------------------------------------------------------------------
Date: Fri, 25 Mar 88 11:59 EST
From: Rick Wojcik <rwojcik@bcsaic.UUCP>
Subject: Phonetic Spelling (was Linguistic Theories)
In article <1041@cod.NOSC.MIL> rupp@cod.nosc.mil.UUCP (William L. Rupp) writes:
>
>... It is a misnomer that Spanish is a phonetic
>language. It is true that you do not run into such things a the 'gh'
>of 'tough' versus the 'gh' of 'ghost' versus the 'gh' of 'thought.'
>
I agree with most of what you say in your article. I would add the
technical point that spelling is 'phonemic', rather than 'phonetic'. It
is important to use the term 'phoneme' in describing alphabetic writing
because 'phonetic' connotes allophonic substitutes for phonemes, which are
almost never represented in alphabetic systems. For example, Spanish does
not distinguish orthographically between voiced stops /b d g/ and their
fricative counterparts, which occur between vowels and at the ends of
words. On the other hand, few real alphabetic systems have a one-to-one
correspondence between phonemes and symbols. Spanish uses two symbols--{b}
and {v}--for the phoneme /b/. The same symbol {c} may represent the /k/
or /s/ phoneme, depending on its orthographic environment. What makes
Spanish a more sensible writing system is that the phoneme-grapheme
correspondences are generally more predictable and less complex than
for English.
>What it amounts to is this; a language is phonetic as long as you do not
>consider different dialects to be legitimate pronunciations of your
>language. However much one may prefer his or her speech to that of
>
True. Consider the vast discrepancy between standard Spanish (even the
'seseo' dialect) and those dialects such as Cuban and Puerto Rican that
have re-phonemicized many words under the influence of 's-aspiration' and
other processes. For example, standard ?Como estas? /komo estas/ comes
out as [kom eta] in many dialects, which is probably close to the phonemic
representation. Those who are interested in American Spanish
pronunciations might be on the lookout for
_American_Spanish_Pronunciation:_Theoretical_and_Applied_Perspectives,
which is edited by P. Bjarkman and R. Hammond (Georgetown U. Press,
forthcoming in June 1988).
--
Rick Wojcik csnet: rwojcik@boeing.com
uucp: {uw-june uw-beaver!ssc-vax}!bcsaic!rwojcik
address: P.O. Box 24346, MS 7L-64, Seattle, WA 98124-0346
phone: 206-865-3844
------------------------------
Date: Fri, 25 Mar 88 13:12 EST
From: Linda G. Means <MEANS@gmr.com>
Subject: re: Spanish as a "phonetic language"
------------------------------
In NL-KR Digest Volume 4 No. 32, William L. Rupp <rupp@cod.NOSC.MIL> writes:
> It is a misnomer that Spanish is a phonetic
>language. It is true that you do not run into such things a the 'gh'
>of 'tough' versus the 'gh' of 'ghost' versus the 'gh' of 'thought.'
>That is all to the good, but does not change the fact that all languages
>have dialects whose distinctive pronunciations are not reflected in the
>written language. Consider the words 'greasy' and 'idea' in English.
>In the North and West, the 's' of 'greasy' is unvoiced. Head south, however,
>and you will hear a voiced 's' (greazy). The word 'idea' ends in an 'r'
>sound when spoken by many New Englanders, but not by speakers in most
>other areas of the country.
>And what about "Earl bought some oil for his car" when spoken by a New
>Yorker. Wouldn't that sound more like "Oil bought some earl for his
>car"? Well, the same thing happens in Spanish, and in all languages.
You're missing an important point, Bill. Certainly, in English,
the word 'idea' is pronounced differently by New Englanders than by
Texans. But a given New Englander also pronounces the word-final 'ea'
differently in 'idea' than he does in 'flea', and so does the Texan.
Whereas although a Spaniard pronounces 'ce' in 'centro' differently
than a Peruvian, the Spaniard pronounces 'ce' the same way in 'doce',
'hacer' and any other word, and the Peruvian also pronounces any given
sequence of letters in a consistent way. So for any given dialect of
Spanish, pronunciation is consistent (i.e. predictable from the orthographic
representation) within that speech community. Not so in English.
>By the way, my credentials for commenting in this area, such as they
>are; B.A (Spanish major, German minor), graduate work in Spanish at
>U.C.L.A.(two years), 17 years secondary teacher, Spanish, German, English, etc.
>Bill
If it is necessary to present one's credentials in order to participate
in this discussion, I have an M.A. and Ph.D. in Spanish Linguistics,
and taught Spanish in universities for five years.
Linda Means
means%gmr.com@relay.cs.net
------------------------------
Date: Wed, 30 Mar 88 05:51 EST
From: Celso Alvarez <sp202-ad@garnet.berkeley.edu>
Subject: Re: Phonetic Spelling (was Linguistic Theories)
For the record. The extended notion that Spanish spelling is
almost phonemic ought to be reconsidered. A list of correspondences
between graphemes and phonemes can be found in many introductory texts
to Spanish linguistics, but here is one. Accent marks,
tildes and diacritic marks go after the corresponding
letter. Some diacritic marks are bastardized (e.g.
up-down ^ stands for palatalization instead of down-up `')
-- computer limitations):
GRAPHEME PHONEME
a /a/
b /b/
c /k/, /0/ (zeta), /s/; /c^/ in combination "ch"
ch /c^/
d /d/
e /e/
f /f/
g /g/ /x/
gu /g/
h zero, except in "ch"
i /i/
j /x/; /y/ in loanwoards ("jazz" /ya's/)
k /k/; in loanwoards
l /l/
ll /ll/; /y/ in dialects with "yeismo"
m /m/; archiphoneme /M/ ("cambiar" /kaMbia'r/;
*/kanbia'r/ is impossible)
n /n/; idem /M/
n~ /n~/
o /o/
p /p/
qu /k/
r /r/ /rr/; archiphoneme /R/
rr /rr/
s /s/; becomes [h] or zero in dialects with s-aspiration
sh /s^/ in loanwoards
t /t/
u /u/; zero in "qu" and in "gu+e,i" combinations
u" /u/ (in context "g_+e,i")
v /b/
w /gu/, /u/ or /b/ in loanwords
x /ks/ (commonly pronounced [gs], [gz], [s])
y /i/ /y/
z /0/ (interdental zeta); /s/ in dialects with
"seseo"
However, when we consider dialectal variation, things
start to get more complicated. "s" and "j" merge in [h] in dialects
with s-aspiration (and deletion). In Argentinian Spanish "ll" and
"y" are realized [z^] or even unvoiced in [s^]. In Andalusian
Spanish final /-s/ deletion leads to a reconfiguration of the
vocalic system -- from /i : e : a : o : u/ to /i : e (medium-high)
: e (medium-low) : a : o (medium-low) : o (medium-high) : u/.
In certain areas with "ceceo" in Andalusia and southwestern Spain
[s] and [0] merge in /s/ or /0/ (both "s", "c+e,i" and "z" are
pronounced [s] or [0] indistinticly). In areas with "rotacismo",
"s+voiced consonant" is pronounced as a thrilled [r] ("mismo"
[mi'rmo]). In bilingual Galiza (northwestern Spain), native
Galician phonological patterns are transferred to Spanish;
final-word "-n" is velar (here represented as [n,], not alveolar.
The distinction [n : n,] acquires a phonological value:
"enojos" 'annoyances' /eno'xos/ vs. "en ojos" 'in eyes' /en,o'xos/;
"enagua" 'underskirt' /ena'gua/ vs. "en agua" 'in water' /en,a'gua/;
"sino'pticos" /sino'ptikos/ vs. "sin o'pticos" /sin,o'ptikos/;
"conexo" 'connected', [cone'so] in colloquial speech vs. "con eso"
'with that' [con,e'so].
In a similar sense,
in article <4598@bcsaic.UUCP> rwojcik@bcsaic.UUCP (Rick Wojcik) writes:
>Consider the vast discrepancy between standard Spanish (even the
>'seseo' dialect) and those dialects such as Cuban and Puerto Rican that
>have re-phonemicized many words under the influence of 's-aspiration' and
>other processes.
However,
>For example, standard ?Como estas? /komo estas/ comes
>out as [kom eta] in many dialects, which is probably close to the
>phonemic representation. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
^^^^^^^^^^^^^^^^^^^^^^^
>Rick Wojcik
Why? Could you elaborate on that?
-----------------
C.Alvarez (sp202-ad@garnet.berkeley.edu.UUCP)
------------------------------
Date: Fri, 25 Mar 88 14:32 EST
From: John_M._Lawler@ub.cc.umich.edu
Subject: Pro-Drop one more...
Once again I come in late on a discussion that's had too many responses
already. However, into the breach:
--------------------------------
(1) the name "Pro-Drop"
Rob Bernardo <rob@pbhyf.UUCP> writes:
:We have been using words here like "drop" and "omit" wrt the subject
:of a clause. These imply that there is some canonical form of a clause
:in *those* languages and it has a subject.
:...
:Sounds like you are assuming that that the person and number would
:have to be deduced from the verb inflection and context and then from
:the person and number the actual referent is understood.
Quite right. There are some presuppositions lurking here which are es-
pecially dangerous in a mixed group like this one, where different un-
derstandings of terms and underlying assumptions wrt language regularly
cause miscommunication.
To begin with, there's a big difference between "PRO", "pronoun", and
"Subject". The first is a technical term (in effect, an arbitrary non-
terminal name) for a particular construct in a particular grammatical
theory. It doesn't exist in any empirical sense in real data, hence
can only be said to "drop" in the same very limited sense in which
(say) the start symbol "S" can be said to "drop" in a PS grammar.
That's not at all the same thing as deletion of an expected word in
(for instance) dropping "money" in "I gave to the Red Cross". "pronoun"
is a traditional term for a well-established word class that exists in
all languages (though the properties of pronouns vary widely).
Finally, "Subject" is a very controversial term. English, and I-E gen-
erally, make use of a Subject grammatical relation; lots of other lang-
uages can be analyzed to do so as well, and Relational Grammar makes
this an article of faith for all languages. However, it's not obvious-
ly a part of universal grammar, because many languages (e.g. Acehnese,
see Durie's and my articles in the new Language) can be fruitfully ana-
lyzed either with or without it. In any event, the three terms name
different *kinds* of things, whose extensions sometimes overlap. The
status of a rule of PRO-drop needs to be distinguished from one that
deletes pronouns, or one that deletes subjects.
Second, the "Drop" part presupposes a fully-specified underlying struc-
ture in which the "PRO" is "present". Note that this is a theoretical
presupposition; revising the theory to have a "PRO-Insertion" rule
(where "PRO" could be computed from context, say) is not a viable al-
ternative. Why not? That's what we're talking about, right? The idea
that the speaker starts with such a structure and proceeds to shave off
parts, which are then re-inserted by the listener, seems to me to be a
convenient but empirically irrelevant thesis, based on the "speaker-
centric" position that underlies generative-derived theories.
--------------------------------
(2) Another language example:
Malay (Indonesian) frequently uses sentences without expressed subjects
in speech, though the textbooks don't tell you this. In addition, there
is the usual uneasiness about which second-person pronoun to use,
complicated by the fact that in Malay pronouns are an open class (i.e,
one can freely use [human] nouns as anaphors). Therefore "you" is
usually not expressed, avoiding the touchy problem of which form of ad-
dress should be used in a given social context. Since third-person
subjects are also not expressed in many sentences, and since Malay has
no verbal inflection of any kind, this means that navigating through
spoken Malay is something of an exercise in anaphor resolution failure.
It's an open question whether this phenomenon (and that of English) is
"grammatical" or "pragmatic" in nature. A lot depends on what you
think a grammar should be capable of resolving, and that, as we know
all too well, varies a lot from grammarian to grammarian. There cer-
tainly seems to be a difference between the absence of a first-person
pronoun in Spanish "No se" (=[I] don't know), where the verb is unam-
biguous in the first person singular, and in English "Didn't see you at
the party last night," where the understanding that the speaker is the
subject is supplied by the speech situation and the conventions for
exploiting it.
--------------------------------
(3) Some references on various relevant topics:
a) re: Subject(?)-verb agreement in Acehnese:
J. Lawler (1977) "A agrees with B in Acehnese" in Cole & Sadock
(eds), _Syntax & Semantics 8_, Academic Press, NY.
M. Durie (1985) _A Grammar of Acehnese on the Basis of a Dialect
of North Aceh_, Foris Publications, Dordrecht.
A. G. Asyik (1987) _A Contextual Grammar of Acehnese Sentences_
Univ of Michigan PhD dissertation.
b) re: Japanese "subject" omission:
M. Tokunaga (1986) _Affective Deixis in Japanese_, Univ of Mich.
PhD dissertation.
c) re: English "subject" (etc.) omission:
R. Thrasher (1974) _Shouldn't ignore these strings: A study of
conversational deletion_, Univ of Mich. PhD dissertation.
Thrasher's work is especially interesting in the present context
because he overtly compares and contrasts the English phenomenon with
the situation in Japanese; Thrasher is a Japanese specialist who's
lived there (Japan is an anaphoric island) for around 20 years. If
memory serves, his findings about English were roughly:
(a) "Deletion" (I'll continue to call it that, purely for terminologi-
cal convenience, pace the remarks above) takes place only at the
beginning of sentences. A strip of words can be deleted, termina-
ting at the first non-predictable word (normally the main verb).
Hence: (((If) I) would have) seen him, I would have told him.
[normally delivered in contracted form, of course]
(b) First person subjects are deletable in statements, second person
in questions.
(c) Third person subjects are deletable by discourse reference.
(d) There are *many* idiosyncratic constraints.
--------------------------------
(4) Non-expression of non-subjects:
Chuck Fillmore at Berkeley has been doing some interesting work on the
cases where expression of direct objects is optional in sentences with
transitive verbs, like "I gave (money) to the Red Cross", versus "Give
*(it) here". He's identified a number of different situations. I know
he's written an article on the topic, but I don't know where he's pub-
lished it.
Like the "Subject" cases, "Object-Drop" might be thought of as either
grammatical or pragmatic; the nature of the missing object is
recoverable, but variously from grammatical, lexical, or conventional
pragmatic facts. So the object of "eat" in "I've already eaten" has to
be some kind of food, which one might predict based on subcategoriza-
tion, while the object of "drink" in "She doesn't drink" refers to al-
coholic beverages, which might be hard to build into a grammar. Again,
where does the grammar stop?
And why does this sound like a familiar question?
John Lawler (jlawler@ub.cc.umich.edu)
Linguistics
University of Michigan
"Adde parvum parvo, magnum acervum erit"
-Add [a] little to [a] little, [and there] will be [a] big pile-
---Ovid (cited by Frederick Brooks in _The Mythical Man-Month_)
------------------------------
Date: Mon, 28 Mar 88 14:42 EST
From: Rick Wojcik <rwojcik@bcsaic.UUCP>
Subject: Pro-drop (naturalism/conventionalism)
Date: Sun, 20 Mar 88 19:41 EST
>From: kathryn henniss <henniss@csli.STANFORD.EDU>
Subject: pro-drop
Kathryn Henniss writes:
KH> Just to hammer one more nail in the coffin of the recent
KH> conjecture that pro-drop is necessarily linked to to verbal
KH> agreement morphology...
KH> Malayalam (a Dravidian language) has subject-drop, object-drop,
KH> indirect-object drop and even preposition-phrase drop; basically,
Wait! I call for an autopsy! The conjecture was only "linked", not
"necessarily linked". This entire discussion suffers from the lack of a
clear definition of just what pro-drop means. It is not the same as
pragmatic ellipsis, and it is not the same as "null-anaphor". We now have
"pro-drop" and "null-anaphor" as classifications of languages. It reminds
me of the ultimate classification of people into two types: those who
classify people into two types and those who don't. In any case,
Malayalam doesn't get you a hammer and nail until you cite some data or a
reference or something to show us how it is an example of a language with
"non-subject pro-drop". And why do you think that English doesn't have
non-subject pro-drop? You can certainly omit non-subject arguments to
many verbs--eg. John sliced the bread (with a knife).
I will try to define two distinct approaches on the question of pro-drop
(and grammatical theory in general). One position, which I call
CONVENTIONALIST (aka "formalist"), has it that pro-drop is an arbitrary
feature of language, which some languages have and some do not. Pro-drop is not
necessarily tied to any strategy of language use, nor is it ruled out as
interacting with performance strategies. The other position, which I call
NATURALIST (aka "functionalist"), has it that all linguistic rules are
grounded or motivated in behavioral strategies. (This is not to claim
that such rules are only triggered by the phenomena that motivate them.
So a behavioral strategy that requires an overt subject may be obligatory
even for environments where the subject is pragmatically predictable.)
Pro-drop is not necessarily a single rule under naturalism, but a pattern
that results from the interaction of one or more performance strategies.
A naturalist account of English might posit speaker-based rules that
delete subjects under certain pragmatic conditions. A naturalist account
of Russian (a pro-drop language) might sanction the lack of subjects on
morphological grounds but posit the existence of hearer-based rules that
insert pronoun subjects under certain pragmatic conditions. Where the
conventionalist might posit a "null-subject" parameter for languages,
such that the language learner need only decide whether it is "on" or
"off" for the language in question, the naturalist seeks structural and
pragmatic strategies that sanction the use of subjects.
My point in going into all this is that naturalists need not take the
simpleminded approach that morphological redundancy is the only factor
motivating pronoun omissions. It is a good place to start, and it may in
fact be the most important factor for many languages. But the omission of
subjects in languages like Chinese and Japanese has to be grounded in
other factors. The question is whether or not "pro-drop" in those
languages has the same behavior as "pro-drop" in Russian and Spanish.
If not, then the conventionalist viewpoint leads us to miss important
observations. If so, then the naturalist viewpoint leads us to miss an
important generalization.
Modern linguistic theory is strongly dominated by the conventionalist
viewpoint, and functional explanations are seldom explored adequately in
the literature. Although naturalism is not as much in vogue as it once
was, at least outside of California, it ought not to be rejected out of
hand. Proponents of "use-less" grammar still have to explain how it gets
used--someday maybe. (My terms "conventionalism" and "naturalism" are
based on David Stampe's, although he is not responsible for anything that
I have said above.)
--
Rick Wojcik csnet: rwojcik@boeing.com
uucp: {uw-june uw-beaver!ssc-vax}!bcsaic!rwojcik
address: P.O. Box 24346, MS 7L-64, Seattle, WA 98124-0346
phone: 206-865-3844
------------------------------
Date: Fri, 25 Mar 88 15:09 EST
From: morgan@uxe.cso.uiuc.edu
Subject: Schmerling on truncation
Some people have asked for the Schmerling reference on English subject
dropping (which aint subject dropping in the same sense as pro-drop, if
Schmerling is right). Here it is:
Susan F. Schmerling, Subjectless sentences and the notion of surface
structure. CLS 9 (i.e. Proceedings of the Ninth Regional Meeting,
Chicago Linguistic Society. (1973)
You might also be interested in
Jerrold M. Sadock, Read at your own risk: syntactic and semantic horrors
you can find in your medicine chest. CLS 10 (1974)
Sadock's paper discusses ellipsis in labels, etc (e.g. Shake before using,
keep away from children,...) and is great fun to read.
------------------------------
Date: Mon, 28 Mar 88 12:26 EST
From: Rick Wojcik <rwojcik@bcsaic.UUCP>
Subject: What are grammars (for)?
John Nerbonne has written:
JN> procedures. The point, however, is that natural language
JN> expressions are ambiguous, and that the information in
JN> grammars may be used to reduce the degree of ambiguity one
JN> would otherwise postulate. This is useful in building
JN> natural language understanding systems.
We can certainly both agree that a linguistic approach is necessary to
language processing. If you mean to say that spurious ambiguity can
be avoided with a syntactic parser, then we have no quarrel. That still
leaves a whole lot of territory where generative grammars are useless or
even a hindrance in language understanding: the resolution of legal
ambiguity, the processing of ill-formed structure, etc. What we need is a
clarification from orthodox generative grammarians on how grammars get
used to process language. Since the language listener is less concerned
with grammaticality than with comprehension, grammatical theory needs to
explain how grammars aid comprehension. As a computational linguist, you
know very well that two of the greatest problems in the field are the
resolution of ambiguity and the interpretation of ungrammatical language.
Arild Hestvik writes:
AH> You have misunderstood. As has been pointed out repeatedly by Chomsky, the
AH> notion "grammar" is ambiguous between the two meanings: (i) the theory about
AH> the grammar, and (ii) the grammar itself. The THEORY of the grammar will
AH> tell us WHY something is illformed or wellformed, just as much as a theory
Gaaa! You quoted Chomsky, and I left my crucifix in my briefcase! :-) I
don't think that I misunderstood your use of the term "grammar" so much as
your wording about how a generative grammar is to treat ill-formedness. I
certainly took a more pragmatic view of ill-formedness. I was thinking of
it as the type of ill-formedness that we encounter every day--foreign
accents, slips of the tongue, etc. You were thinking of it in terms of
linguistic argumentation. Your example clarifies this:
AH> ... But the sentence 'How did you wonder whether Bill
AH> fixed the car', with the intended reading that 'how' is a question about
AH> the manner of fixing...
(Please continue to use the expression 'with the intended reading that'.
It serves to remind us all that grammaticality judgments don't exist
outside of pragmatic contexts.) Your claim was that grammatical theory
"should give a structural description (and analysis) of ... ill-formed
expressions." I'm still totally unclear as to what the analysis of the
above sentence looks like in your theory. My understanding of generative
grammar is that no structural analysis is possible for the intended
reading. That is how it gets recognized as ungrammatical. (By the way,
the ill-formedness of the above sentence might be considered outside the
purview of generative syntactic theory if it could be shown to reside in
presupposition failure with respect to WH question formation rather than
the contents of a putative COMP constituent. It is not a foregone
conclusion that your example is syntactically ill-formed.)
AH> It appears that Rick Wojcik thinks that the main interest of linguists
AH> is empirical coverage (i.e. to account for any possible string of words you
AH> might care to put together). However, that would be very misleading (at
I don't recall accusing anyone of being interested in empirical coverage.
Please don't go spreading malicious rumors about me :-).
AH> least for part of the field). Rather, the main interest is to try to
AH> understand the very nature of grammars ... namely the
AH> psychologically represented mechanism that underlies e.g. language acquisition
AH> and language processing.
Here we agree totally. This is why I believe that generative theory needs
a coherent position on the way in which grammars interact with linguistic
performance.
"When a linguist uses the word 'theory', put your hand on your wallet."
--J.D. McCawley
--
Rick Wojcik csnet: rwojcik@boeing.com
uucp: {uw-june uw-beaver!ssc-vax}!bcsaic!rwojcik
address: P.O. Box 24346, MS 7L-64, Seattle, WA 98124-0346
phone: 206-865-3844
------------------------------
Date: Sun, 3 Apr 88 18:24 EDT
From: Stan Friesen <sarima@gryphon.CTS.COM>
Subject: Re: What are grammars (for)?
In article <4630@bcsaic.UUCP> rwojcik@bcsaic.UUCP (Rick Wojcik) writes:
>
>We can certainly both agree that a linguistic approach is necessary to
>language processing. If you mean to say that spurious ambiguity can
>be avoided with a syntactic parser, then we have no quarrel. That still
>leaves a whole lot of territory where generative grammars are useless or
>even a hindrance in language understanding: the resolution of legal
>ambiguity, the processing of ill-formed structure, etc. What we need is a
>clarification from orthodox generative grammarians on how grammars get
>used to process language. Since the language listener is less concerned
>with grammaticality than with comprehension, grammatical theory needs to
>explain how grammars aid comprehension.
>
Indeed, I agree totally. In fact I maintain that it is impossible
to produce a performative interpretation of generative grammar that is
adequate for proper semantic processing. That is why I have esentially
abandoned generative models of grammar in favor of semantically based
grammatical models. My current favorite is functional grammar as proposed
by Simon Dik. There was also an interesting book written a few years back
called something like "How to Generate Sentences from a Semantic Base". It
presented a very interesting model of linguistic performance that has largely
been ignored as far as I can tell. It had some very interesting observations
about regularities in the order of adjectives that were quite surprising.
>(Please continue to use the expression 'with the intended reading that'.
>It serves to remind us all that grammaticality judgments don't exist
>outside of pragmatic contexts.) Your claim was that grammatical theory
>"should give a structural description (and analysis) of ... ill-formed
>expressions." I'm still totally unclear as to what the analysis of the
>above sentence looks like in your theory. My understanding of generative
>grammar is that no structural analysis is possible for the intended
>reading.
In a semantically based grammar this type of thing is handled easily,
since the "internal" structure of an utterance is a representation of its
meaning. Thus an ill-formed utterance will be parsed onto a 'reasonable'
internal structure allowing at least partial comprehension, as is seen in
real life.
>AH> least for part of the field). Rather, the main interest is to try to
>AH> understand the very nature of grammars ... namely the
>AH> psychologically represented mechanism that underlies e.g. language acquisition
>AH> and language processing.
>Here we agree totally. This is why I believe that generative theory needs
>a coherent position on the way in which grammars interact with linguistic
>performance.
Or why generative grammars must be thrown out, since they do not seem
to correspond to any real psychological process!
------------------------------
End of NL-KR Digest
*******************