LINGUIST List 8.1759

Sat Dec 6 1997

Qs: Determiners and Quantifiers

Editor for this issue: Brett Churchill <brettlinguistlist.org>


We'd like to remind readers that the responses to queries are usually best posted to the individual asking the question. That individual is then strongly encouraged to post a summary to the list. This policy was instituted to help control the huge volume of mail on LINGUIST; so we would appreciate your cooperating with it whenever it seems appropriate.

Directory

  1. Piroska Csuri, Determiners and Quantifiers

Message 1: Determiners and Quantifiers

Date: Thu, 4 Dec 1997 16:21:20 -0500
From: Piroska Csuri <piroskaresearch.nj.nec.com>
Subject: Determiners and Quantifiers

DETERMINERS AND QUANTIFIERS


In building a lexicon for our syntactic parser, we started looking at 
the properties of quantifiers and determiners, or whatever element may 
appear at the "beginning" of a noun phrase in English (i.e., all the 
stuff left of adjectives), including 

	a/an, the, this/that, some, any, no, all, both, each, every, 
	many, most much, few, several, any, either, cardinals, which, 
	whose, unstressed some, unstressed this, 's (possessive), etc.

In trying to come up with relevant-and-appropriate features for these
elements, we came up against a lot of questions - some might be obvious, 
others might be trickier. We would like to know what is "out there" 
regarding these questions. If you can provide us with insights or maybe 
even answers to at least some of our questions, we would really 
appreciate it.

Since the parser works with English, Turkish, Hungarian, German, 
French, Spanish, Dutch, we are foremost interested in the properties 
of quantifiers/determiners in these languages. On the other hand, 
we would like to find a *principled* solution to the descriptive 
problem (as much as it might be possible), so data from other languages 
are also very important determining the relevant featural description
of the classes. 

While some of the questions below might sound too broad or naive, this 
might be due to the fact that I tried to be unassuming, and to avoid
anticipating any particular answer, so please bear with me.

Please send answers to 

	piroskaresearch.nj.nec.com

Finally, if there is enough interest, I will, of course, post a summary 
to the list.


Thank you for any suggestions, pointers, or insights, whether pro or contra.


	Piroska Csuri
	NEC Research Institute
	Princeton, NJ 

******************** BEGINNING OF QUERY ******************************

NOTE: To avoid presupposing a possibly biassed determiner/quantifier 
distinction, I will simply use the terms D/Q or D's/Q's. Once again, 
the elements I am interested in are: 

(0)	a/an, the, this/that, some, any, no, all, both, each, every, 
	many, most much, few, several, any, either, cardinals, which, 
	whose, unstressed some, unstressed this, 's (possessive), etc.


QUESTIONS IN BRIEF: 

1. What *principled* distinctions can one make between the class of 
determiners and the class of quantifiers, whether on morphological, 
syntactic, or semantic grounds? In other words, what makes a given 
element a determiner, and what makes it a quantifier?

2. What features are needed to adequately describe the behavior of
determiners and quantifiers? Are there seperate features for determiners
and quantifiers, or rather a common array of features describes them?

3. Under what conditions can a determiner/quantifier appear without an 
overt head noun, or with an "expletive noun" *one*?

4. What tests have been proposed to distinguish between strong and 
weak NP's/determiners/quantifiers? 

5. What tests can be used to distinguish the proportional v. cardinal
reading of quantifiers (and possibly of determiners)?


SOME QUESTIONS IN MORE DETAIL:

QUESTION 3. Under what conditions can a determiner/quantifier appear without an 
overt head noun, or with an "expletive noun" *one*?

Determiners/quantifiers seem to split into three groups, depending on
whether they can appear without an overt head noun or can take "one": 

(3.1)	Without overt N:	With "one":
	================	===========
	which (of them)		which one(s) (of them) 
	each (of them)		each one (of them)
	either (of them)	either one (of them)
	those (of them)		those ones (of them)

	few (of them)		*few ones (of them) 
	all (of them)		*all ones (of them)
	both (of them)		*both ones (of them)
	some (of them)		*some ones (of them)
	three (of them)		*three ones (of them)
	
	*every (of them)	every one (of them)
	*no (of them)		no one (of them)

Q. 3A: Does the possibility of appearing without an overt head noun or 
with "one" correlate with other known properties of those D's/Q's 
(other than the restriction that "one" must appear with count D's/Q's)

Q. 3B: Is it true that "one(s)" seems to not occur with weak D's/Q's 
(although it might not appear with every strong D/Q)?

(3.2)	*some ones
	*a one
	the one
	every one
	each one

Q. 3C: Is there any evidence, and if so what, that a covert head noun 
should be assumed for cases where no overt head noun is present? 



QUESTION 4. What tests have been proposed to distinguish between strong 
and weak NP's/D's/Q's? 

We are aware of existential sentences as contexts selecting for 
weak readings/determiners/quantifiers: 

(4.1)	There was a/*the/*every/*either cow on the pasture.
	 were few/many/*these/*both/no cows

We are also aware of another potential test, which involves restrictive
relative clauses: 

(4.2)	Some/few/many men appeared who were wearing a hat.
	*The/*Every man appeared who was wearing a hat.

Supposedly only weak NP's can appear in this context, while
strong NP's seem to be out.

Q. 4A: We are wondering, what other tests have been proposed to test 
for the strong v. weak distinction? 

Q. 4B: What evidence is there, if any, that strong v. weak D's/Q's 
would appear in different syntactic positions?

Q. 4C: What evidence is there that in a syntactic tree a restrictive 
relative clause would be attached lower than (i.e., is in the scope of) 
the determiner? 



QUESTION 5. What tests can be used to distinguish the proportional 
v. cardinal reading of D's/Q's?

Some D's/Q's allow for two readings: proportional/quantificational 
reading and a non-proportional/cardinal reading:

(5.1)	Many children came to the party.

	cardinal: 	the children who came to the party were many
	proportional:	of the children many came to the party 

We know that existential sentences seem to select/allow only the 
cardinal reading, 

(5.2)	There were many stars in the sky.
	i.e. the stars that were in the sky were many

while individual-level predicates supposedly prefer (only allow?) the
proportional/quantificational reading: 

(5.3)	Many children like spinach.
	i.e. of the children many like spinach

Q. 5A: Do individual-level predicates really only allow for a proportional 
reading? A cardinal reading is available in this context: 

(5.4)	Three children like spinach.

Q. 5B: How does the proportional v. cardinal/quantificational v. 
non-quantificational distinction relate to the strong/weak distinction?


************************** END OF QUERY *************************

Please send any resposes to:

	piroskaresearch.nj.nec.com

Thank you again for your help and insights.


	Piroska Csuri
	NEC Research Institute
	Princeton, NJ
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue