Copy Link
Add to Bookmark
Report
NL-KR Digest Volume 02 No. 61
NL-KR Digest (6/29/87 18:24:04) Volume 2 Number 61
Today's Topics:
S/O Raising
Re: S/O Raising
----------------------------------------------------------------------
Date: Sun, 21 Jun 87 19:25 EDT
From: Michael P. Smith <mps@duke.cs.duke.edu>
Subject: S/O Raising
[Forwarded from sci.lang, inappropriate meta discussion edited - BWM]
I'd like to see discussions that would inform a neophyte like
myself on trends in current lingustics. Anyone out there want to
explain LFG, GB, GPSG (or whatever is in fashion these days),
especially as they might relate to computational linguistics?
The best way to ensure that a newsgroup discusses issues that interest
you is to bring one up yourself. So here's a topic that undoubtedly
is not hot, but I just found about it myself: subject-object raising.
(I told you I was a neophyte.) If this a dead horse or otherwise
inappropriate, flame away. I am certainly leaving myself open for it.
A quick review: in S/O raising, the subject of a that-clause
complement is "raised" to the direct object of the matrix clause, and
the verb in the complement put into the infinitive. For example, S/O
raising applied to
Alice expected that Bob would be chosen.
yields
Alice expected Bob to be chosen.
The puzzle is that some verbs in the matrix permit S/O raising, and
others do not.
Alice feared that Bob would be chosen.
does not permit raising to
*Alice feared Bob to be chosen.
In a half-hour or so, I came up with about 50 verbs that do not permit
raising, and about 25 that do. I'll burden you only with a small
sample:
PERMIT RAISING PROHIBIT RAISING
know realize
suppose guess
declare state
deem decide
take infer
recognize learn
imagine hope
remember notice
say, protest, guarantee, etc.
The list is based on my own shaky intuitions, of course.
The problem is to find a principled distinction between the two
classes of verbs. I found out about this from SYNTACTIC ARGUMENTATION
AND THE STRUCTURE OF ENGLISH by Soames & Perlmutter (Cal 1979) which
BTW seems to be an excellent text for us neophytes. They say it is
still an open question. What do you think?
----
Michael P. Smith ARPA: mps@duke.cs.duke.edu
"Good sense is the best distributed thing in the world: for everyone
thinks himself so well-endowed with it that even those who are the
hardest to please in everything else do not usually desire more of it
than they possess." Rene Descartes
------------------------------
Date: Mon, 22 Jun 87 12:49 EDT
From: Jeffrey Goldberg <goldberg@su-russell.ARPA>
Subject: Re: S/O Raising
In article <9799@duke.cs.duke.edu> mps@duke.UUCP (Michael P. Smith) writes:
>What I expected were discussions that would inform a neophyte like
>myself on trends in current lingustics. Anyone out there want to
>explain LFG, GB, GPSG (or whatever is in fashion these days),
>especially as they might relate to computational linguistics?
May I recommand a book by Peter Sell: "Lectures on contemporary
syntactic theories: Government and Binding theory, Generalized
Phrase Structure Grammar, Lexical Function Grammar". It is
published in the Lecture Note series by the Center for the Study of
Language and Information (CSLI) and is distributed in conjunction
with University of Chicago Press, Chicago. It doesn't discuss any
computational issues.
Asking for an explanation of three linguistic theories over the net
is a tall order. Even explaining how they all take care of raising
would be a tough job. Could you ask a smaller and more managable
question.
>A quick review: in S/O raising, the subject of a that-clause
>complement is "raised" to the direct object of the matrix clause, and
>the verb in the complement put into the infinitive. For example, S/O
>raising applied to
(1)
> Alice expected that Bob would be chosen.
>yields
(2)
> Alice expected Bob to be chosen.
What I can tell you, is that in none of the modern theories is (2)
derived from (1).
In GPSG the verb "expect" has a dual subcategorization. It can
take either an S' (S-bar) as in (1) or an object NP followed by an
infinitival VP as in (2). (The relationship between the object and
the "logical subject" of the infinitival VP (VP[INF]) is gotten by
stuff that I won't go into right now.
S
______|_______
| |
NP VP
| ________|________
Alice | | |
V NP VP[INF]
| | ____|_____
expected Bob | |
V[INF] VP[BSE]
| ____|_____
to | |
V[BSE] VP[PSP]
| |
be V[PSP]
|
chosen
(Don't worry about all the features such as BSE (VerbFORM BaSE), or
PSP (VFORM PaStParticiple), etc)
LFG is very similar to GPSG in this respect. (Historically, LFG is
prior to GPSG.) Except for the mechanism for what I didn't discuss
and that the VP[INF] is probably in S with no subject.
GB is the most divergent from what you know. Again the verb
"expect" has two different subcategorization but the structure of
(2) involves no Raising. "Bob" in in the subject position of the S
"Bob to be chosen".
S
_______|_______
| | |
NP INFL VP
| | _____|______
Alice ed | |
V S
| _____|______
expect | | |
NP INFL VP
| | |
Bob to be V
|
chosen
(I am not really sure where the "be" goes. To my knowledge, it is
not discussed in the GB literature, but maybe someone will correct
me about this.)
The object like properties of "Bob" is gotten by stuff that I won't
explain here, but what makes "expect" a Raising to Object verb is
that the S that it subcategorizes for is an S and not an S' (so
"Bob" works out to be in a position with respect to "expect" so as
to have certain object properties.
In a message I posted a while back I suggested that the split
between Chomsky and his current generative grammar rivals really
started when rejected Raising to Object. (LFG and GPSG aren't the
only rivals to the GB tradition, and all the rivals that I know of
put "Bob" in (2) in object position at their respective equivalents
to surface structure.)
>The puzzle is that some verbs in the matrix permit S/O raising, and
>others do not.
> PERMIT RAISING PROHIBIT RAISING
> know realize
> suppose guess
> declare state
> deem decide
> take infer
> recognize learn
> imagine hope
> remember notice
> say, protest, guarantee, etc.
A longer list can be fond in Rosenbaum's (1969?) dissertation.
>The problem is to find a principled distinction between the two
>classes of verbs. I found out about this from SYNTACTIC ARGUMENTATION
>AND THE STRUCTURE OF ENGLISH by Soames & Perlmutter (Cal 1979) which
>BTW seems to be an excellent text for us neophytes. They say it is
>still an open question. What do you think?
Please note that there is a semantic difference between (1) and
(2). Your use of "would" in (1) tries to cover this. But, subtle
differences in subcategorization ("give" and "donate", "eat" and
"devour", etc) are sometimes just arbitrary. All of the current
theories treat these as differences in subcategorization. Granted,
this is no explanantion.
Rosenbaums dissertation is probably cited in P&S (I am at home; so
I can't give complete references to things.) Joan Bresnan's
dissertation discusses the semantic differences of different types
of complementation.
>Michael P. Smith ARPA: mps@duke.cs.duke.edu
--
Jeff Goldberg
ARPA goldberg@russell.stanford.edu
UUCP ...!ucbvax!russell.stanford.edu!goldberg
------------------------------
Date: Mon, 22 Jun 87 18:22 EDT
From: M.BRILLIANT <marty1@houdi.UUCP>
Subject: Re: S/O Raising
In article <9799@duke.cs.duke.edu>, mps@duke.cs.duke.edu (Michael P. Smith) writes:
> A quick review: in S/O raising, the subject of a that-clause
> complement is "raised" to the direct object of the matrix clause, and
> the verb in the complement put into the infinitive. For example, S/O
> raising applied to
> Alice expected that Bob would be chosen.
> yields
> Alice expected Bob to be chosen.
which in turn can be turned into a passive with the original subject,
now an object, as subject again:
Bob was expected to be chosen.
> The puzzle is that some verbs in the matrix permit S/O raising, and
> others do not.
A further puzzle is that some that do not permit S/O raising do permit
the passive transformation. Here's the original sample list with Y
after those that allow the passive, and N for those that don't:
> PERMIT RAISING PROHIBIT RAISING
> know Y realize ?
> suppose Y guess ?
> declare Y state Y
> deem Y decide N
> take Y infer ?
> recognize Y learn ?
> imagine Y hope ?
> remember N notice Y
> say Y
protest N
guarantee Y
>
> The list is based on my own shaky intuitions, of course.
Mr. Smith, the list is firmly based on your idiolect: your personal
dialect of the language. No apology necessary. It happens that my
idiolect pretty much agrees with yours here ("remember" is one
exception). I'm not so sure about my own usage in the passive
versions, though.
> The problem is to find a principled distinction between the two
> classes of verbs. I found out about this from SYNTACTIC ARGUMENTATION
> AND THE STRUCTURE OF ENGLISH by Soames & Perlmutter (Cal 1979) which
> BTW seems to be an excellent text for us neophytes. They say it is
> still an open question. What do you think?
four classes of verbs?
What about just saying that there are these classes of verbs? After
all, what is it that defines a verb, anyway, other than that it is
allowed in certain constructions? That is, can you find a "principled
distinction" between verbs and non-verbs? Don't just fall back on the
grade-school definition that "a verb is a word that expresses an action
or state of being": it's not very accurate.
M. B. Brilliant Marty
AT&T-BL HO 3D-520 (201)-949-1858
Holmdel, NJ 07733 ihnp4!houdi!marty1
------------------------------
Date: Tue, 23 Jun 87 23:05 EDT
From: Frank Adams <franka@mmintl.UUCP>
Subject: Re: S/O Raising
In article <9799@duke.cs.duke.edu> mps@duke.UUCP (Michael P. Smith) writes:
|A quick review: in S/O raising, the subject of a that-clause
|complement is "raised" ...
Just a quick thought: describing this as "raising" the subject of the
that-clause may be backwards. Instead, it may be that the "base" form is
the direct object form where that is legal, and that the that-clause form is
secondary for such verbs (perhaps by analogy with those verbs where it is
expected).
--
Frank Adams ihnp4!philabs!pwa-b!mmintl!franka
Ashton-Tate 52 Oakland Ave North E. Hartford, CT 06108
------------------------------
Date: Thu, 25 Jun 87 22:26 EDT
From: Michael P. Smith <mps@duke.cs.duke.edu>
Subject: Re: S/O Raising
In article <307@su-russell.ARPA> goldberg@su-russell.UUCP (Jeffrey Goldberg) writes:
>In article <9799@duke.cs.duke.edu> mps@duke.UUCP (Michael P. Smith) writes:
>
>>anyone out there want to explain LFG, GB, GPSG (or whatever is in fashion
>>these days), especially as they might relate to computational linguistics?
>
>May I recommand a book by Peter Sell: "Lectures on contemporary
>syntactic theories: Government and Binding theory, Generalized
>Phrase Structure Grammar, Lexical Function Grammar". [...]
I went right out and bought it, and although I'm hardly in a position
to recommend it yet, it certainly looks just like what the doctor
ordered. Sell claims in the preface that he is trying to be
accessible to the non-linguist, and it covers the three major
syntactic theories of the 80s in roughly 200 pages. It is reasonably
priced in paper.
>>A quick review: in S/O raising, the subject of a that-clause
>>complement is "raised" [...]
>>the verb in the complement put into the infinitive. For example, S/O
>>raising applied to
>
>(1)
>> Alice expected that Bob would be chosen.
>>yields
>
>(2)
>> Alice expected Bob to be chosen.
>
>What I can tell you, is that in none of the modern theories is (2)
>derived from (1).
I don't know if you meant to include the (extended) standard theory,
but in that theory as well (2) is not derived from (1), but both from
a common deep structure. I knew I was being loose, and in retrospect
I think too loose.
> [lots of good stuff worth looking up]
>Please note that there is a semantic difference between (1) and
>(2). Your use of "would" in (1) tries to cover this. But, subtle
>differences in subcategorization ("give" and "donate", "eat" and
>"devour", etc) are sometimes just arbitrary. All of the current
>theories treat these as differences in subcategorization. Granted,
>this is no explanation.
I don't see the semantic difference between (1) & (2); further, the
subjunctive seems to be required in (1) by the sequence of tenses
rather than any subcategorical peculiarities of 'expect'. Maybe I
should have said
(1') Alice expects that Bob will be chosen.
and
(2') Alice expects Bob to be chosen.
It's rude to push after Jeff admits that an appeal to
subcategorization by itself is no answer, but -- how does the
subcategorization story go? Is it that each failure to raise has its
own reason, or is it that verbs otherwise nearly synonymous sometimes
differ along some single dimension that explains why one allows raises
and the other does not? See my reply to Marty on why I don't think
every failure to allow raising can have its own story.
------
Michael P. Smith "It could also be said that a man thinks
ARPA mps@duke.cs.duke.edu when he learns in a particular way."
Wittgenstein
------------------------------
End of NL-KR Digest
*******************
