ArticlePDF Available

Nouns and Countability

September 1980
Language 56(3):541

September 1980
56(3):541

Authors:

Monash University (Australia)

The customary disjunctive marking of lexical entries for English nouns as [± countable] does not match the fact that the majority can be used both countably and uncountably in different NP environments: this binary opposition is characteristic not of the nouns, but of the NP's which they head. Nevertheless, nouns do have countability preferences; some enter countable environments more readily than others. And not all nouns occur in all kinds of countability environments. A noun's countability preference can be computed by checking its potential for occurrence in a definitive set of countability environments. In the dialect examined here, well-formedness conditions on NP must consider eight levels of countability among English nouns-not, as custom has it, only two.

Content uploaded by Keith Allan

Content may be subject to copyright.

http://www.jstor.org

Nouns and Countability

Author(s): Keith Allan

Source:

Language,

Vol. 56, No. 3, (Sep., 1980), pp. 541-567

Published by: Linguistic Society of America

Stable URL: http://www.jstor.org/stable/414449

Accessed: 30/04/2008 17:02

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at

http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

http://www.jstor.org/action/showPublisher?publisherCode=lsa.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We enable the

scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that

promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

NOUNS AND COUNTABILITY

KEITH ALLAN

Monash University

Clayton, Victoria,

Australia

The customary disjunctive marking of lexical entries for English nouns as

[+ countable] does not match the fact that the majority can be used both countably

and uncountably in different NP environments: this binary opposition is character-

istic not of the nouns, but of the NP's which they head. Nevertheless, nouns do have

countability preferences; some enter countable environments more readily than others.

And not all nouns occur in all kinds of countability environments. A noun's count-

ability preference can be computed by checking its potential for occurrence in a

definitive set of countability environments. In the dialect examined here, well-

formedness conditions on NP must consider eight levels of countability among English

nouns-not, as custom has it, only two.*

NOUNS, COUNTABILITY,

AND NOUN PHRASES

1.1. Tradition has it that a noun like car is countable, whereas lightning is

uncountable: we talk about one car or many cars, but not normally about *one

lightning or *many lightnings. More generally, the following rule is observed in

English:

(I) If the head constituent of an NP falls within the scope of a denumerator,

it is countable.1

Here the phrase 'falls within the scope [or domain]' of a denumerator means 'is

denumerated' by it; i.e., the NP reference is quantified by the denumerator as a

number of discrete entities. All denumerators are quantifiers, and the paradigm

set of denumerators

consists of the natural (cardinal) numbers from one to infinity.

In addition, a denumerator is any quantifier which necessarily identifies one or

more discrete entities (but not necessarily a precise number of them), and which

can be substituted for a natural number within any-and perhaps every-NP,

ranging over the same domain and causing no change in the well-formedness or

ill-formedness of the NP. Thus the quantifiers

in 1 are all denumerators, but those

in 2 (other than the natural numbers one and two) are not:

(1) 'one ' one

a a

each >car]Np *[q

each lightning]Np

every every

,either, ,eitherd

* I am grateful to William Bright, Ray Cattell, James Hoard, Rodney Huddleston, and D. T.

Langendoen, whose comments on earlier versions of this paper led to a number of improve-

ments. The faults that remain are mine alone.

1 In order

not to prejudice

the issue, here and in Rule II, the word 'it' is ambiguous between

coreference with the NP head vs. the NP as a whole.

541

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

'three ' three

several several

many > cars]Np *[4 many lightning(s)]Np

both both

,(a) few, ,(a) few,

(2l [one carN *[one lightning]NP

(2)

[no; } ]NP [no lightning]NP

{two

{ *[two lightning(s)]NP

all J

some ca]NPll lightning]NP

In summary, the quantifiers no, all, and some do not necessarily denumerate; but

a(n), each, either, several, many, both and (a)few do. Only the latter are denumera-

tors within the terms of Rule I.

Predictably, the morphological category of number intersects with countability;

but there is no universally

applicable correlation between the two. Thus, in Sinhalese

(Indo-European) and Swahili (Bantu), uncountable nouns are formally plural; but

in English they are formally singular, and the following rule can be set up:

(II) If the head constituent of an NP is plural, it is countable.

The surface morphology of a head noun usually indicates whether it is plural by a

contrast of form; but when this is not the case (e.g. with such nouns as sheep,

means, scissors, measles, or physics), Rule II is opaque, and recourse must be made

to the syntactic context of the noun.

In many languages, not only is the NP reference named by the noun, but the

NP also contains a constituent that identifies its salient characteristic.

This supple-

mentary classification of the NP reference varies from language to language, both

in the type of characteristics

indicated and in the range of NP environments that

require supplementary classification. Thus Bantu noun class prefixes distinguish

material differences in the reference of different NP's headed by the same noun

stem: cf. Swahili m-tu 'person' vs. ki-tu 'artifact, thing'; Kikuyu mu-ti 'tree' vs.

gi-ti 'wooden artifact, seat' vs. ma-ti 'woody mass, undergrowth'-or the Luganda

series mu-ganda 'Ganda person', ba-ganda 'Ganda people', ki-ganda 'Ganda

culture', lu-ganda 'Ganda language', bu-ganda 'Ganda country'. In numeral or

demonstrative expressions, many languages require the presence of an NP con-

stituent to classify the NP reference

according to shape, configuration, consistency,

or material characteristics (cf. Allan 1977a for extensive discussion): three such

languages are Yucatec (Mayan), Kiriwina (Austronesian), and Thai (perhaps Sino-

Tibetan). The reference of locative expressions is classified in one of four ways in

the northeastern

Australian language Dyirbal. In the Wakashan

languages of north-

western America, the visibility and location of the NP reference must be encoded

within the NP. English requires

that the countability of the NP reference

be known,

and this affects NP constituency through application of the following rule:

(III) If his listeners do not already know the countability of the NP reference,

the speaker must make it known to them.

542

NOUNS AND COUNTABILITY

This rule states a condition for disambiguation, i.e. for marking contrasts; it is not

a rule for neutralizing contrasts. Thus the plural, which marks a contrast by

unambiguously indicating countability, will never be neutralized through the

operation of Rule III; but there are both countables and uncountables which are

singular, and so under the stated condition Rule III will operate to decide the

countability of singular NP's. Where the speaker judges that his listeners are

already aware of the NP reference because of either previous mention or general

knowledge, he uses a definite NP; otherwise the NP will be indefinite (cf. Allan

1977b:39 ff., Christophersen 1939, Grannis 1972, Hawkins 1978, Kramsky 1972).

Being aware of the NP reference entails knowing its countability; otherwise Rule

III would operate to mark the countability of definite singular NP's, which it

does not:

(3) THE LIGHTNING has frightened Caspar, and he's hiding under THE

CAR.

THIS CAR has front wheel drive.

THIS WATER'S

cold.

THAT FLOUR'S damp; but then, so's THAT CUPBOARD.

Here there is no visible formal distinction between the countables the car, this car,

that cupboard

and the uncountables the lightning, this water, that

flour. Indefinite

NP's indicate that the speaker judges his audience not to be aware of the NP

reference prior to this occasion of mentioning it; so he is required by Rule III to

indicate its countability. Typically, plural indefinites are morphologically marked

as countable; thus, by Rule II, all the emphasized NP's in 4 are marked for

countability:

(4) LIONS

can be dangerous.

ONE-ARMED

IRISH WASHER-UPPERS

exist only in TELEVISION SITUATION

COMEDIES.

WORKERS

came pouring out of the factory gate.

Within TEN MINUTES the sugar I had dropped was cleared away by ANTS.

But singular indefinites have no comparable countability-marking in noun mor-

phology; hence, to satisfy Rule III, countable indefinites are marked by a denu-

merator

which ranges over the NP head, as in 5. By contrast, uncountable

indefinites

cannot take a denumerator, and often lack any determiner

at all;2 cf. 6:

(5) A LION

can be dangerous.

There's A LION

escaped from the zoo.

ONE

CAR

is not enough for the average Australian family.

EVERY

RECORD HE

OWNS

is badly scratched.

(6) SEA WATER

is saline.

SEA WATER got into the petrol tank.

There is SOFT BROWN SUGAR

on the table.

Because every singular countable falls within the scope of one or another

determiner, we can usually assume that, if the head constituent of a singular NP

does not fall within the domain of a determiner, it is uncountable.

2 Strictly speaking, the Determiner is not a constituent of NP structure, but a portmanteau

term for Articles, Demonstratives, Possessives, and Quantifiers.

543

LANGUAGE,

VOLUME

56, NUMBER

3 (1980)

Exceptions to this rule of thumb are the closed class of pronouns and the

restricted class of proper names. Exceptions that may be more apparent than real

are the pronominal quantifiers

one, two,

fifty, many, several etc., as in

one.

(7) I would like to take jtwo.

several.

It seems likely that these quantifiers are the survivors of NP's in which they were

originally determiners, and which have been beheaded-i.e. have lost their head

nouns under certain specifiable conditions which will not be elaborated here (cf.

Allan 1977b:169 ff., Jackendoff 1971).

All such NP's, consisting only of pronominal

quantifiers,

are indefinite; they are also countable if the quantifier

is a denumerator.

Thus the principle is maintained that the countability of indefinite NP's is normally

indicated. All other undetermined NP's consisting of pronouns or proper names

are definite:

(8) Is this yours, or isn't it?

Hermione is beautiful; but she's Henry's, or was when we were in Paris.

And we have already seen that definite singular NP's are not marked syntag-

matically for countability. Indeed, for nouns like management, congress, or

government-which may function as either common or proper names-the proper-

name function is indicated by the lack of a determiner:

(9) Management reserves the right to dismiss staff who are persistently late.

Congress believes that shorter working hours are the right of every

worker.

Government will publish a law banning transvestism in public urinals.

When determined, these nouns cease to be proper names. The same is true for

regular undetermined proper names like John or Hermione, which are not used as

fully-defining proper names in the following:

(10) I don't believe there is a management in this country that would accept

such impossible conditions.

The John I spoke of is Egbert's brother.

Is there a Hermione in your class? I've always wanted to meet a

Hermione, but such has never been my good fortune.

We can distinguish a fully-defining proper name from a proper name used as an

appellative, viz. like a common noun. The proper name is fully defining when it is

the sole constituent of an NP and when, for any and every occasion, the NP

uniquely labels and guarantees the existence of the referent or set of referents (in

either the real or a fictional world). It is not necessarily the case that the reference

of a given proper name is identical on every occasion: not all Johns are the same

John. So the significant characteristic of a fully-defining proper name is the fact

that it always uniquely labels, and so fully defines, the referent or set of referents.

As a result of the fact that it uniquely labels the referent,

a fully-defining proper

name

is definite; because it guarantees the existence of the reference, it is specific (cf.

Fodor 1970, Karttunen 1976, Palacas 1977, Partee 1972, Rivero 1975, 1977); hence

the ungrammaticalness

of *Is there Hermione in your class ? Any proper name in a

544

NOUNS AND COUNTABILITY

non-specific NP, an indefinite NP, or an NP where it is not the sole constituent,

therefore is, not a fully-defining proper name. That is why the proper names in 10

are not fully defining; and also why fully-defining proper names do not permit

restrictive relative clauses, though appositional clauses are acceptable:

() The London? I am talking about is London, Ontario. (RESTRICTIVE)

London, which is for me the greatest city on earth, was my home for

many years. (APPOSITIONAL)

(12) The John who will be in your class is a minister's son. (RESTRICTIVE)

*John

John, who will be in your class, is a minister's son. (APPOSITIONAL)

(13) The congress taking place in Blackpool agreed on the following motion.

*Congress

(RESTRICTIVE)

Congress, taking place in Blackpool, agreed on the following motion.

(APPOSITIONAL)

(14) A management which denies the right of workers

to hold union meetings

*Managementf

during office hours will suffer the strike of an outraged work force.

(RESTRICTIVE)

Management, which mercilessly exploits the work force, must be made to

pay the price. (APPOSITIONAL)

Appellative proper names (ones that are not fully defining) are countable:

(15) I once knew a Euphemia.

My atlas lists six Lewistons but only one Paris.

(fNoam Chomsky }

There is only one Roam Chomsky known to me.

River Thames

Fully-defining proper names which are plural, e.g. the United States or the

Grampians,

are countable by Rule II-but nevertheless non-denumerable, because

denumeration changes the status of a fully-defining proper name to an appellative.

For the same reason, it is impossible (using NP-internal structural criteria) to see

any sense in which a singular fully-defining proper name can be countable.

1.2. In English, then, countability is marked syntagmatically in singular NP's;

in plural NP's it is marked morphologically, and on occasion syntagmatically as

well. But uncountability is never marked. That countability is the marked member

of the duo is 'natural': many languages mark countability morphologically or

syntagmatically (or both), but have no marking for uncountability. And although

languages exist which arguably make no such distinction between countables and

uncountables (e.g. Bantu languages), none systematically marks uncountables

while leaving countables unmarked.

1.3. For grammarians, the conventional wisdom is that countability is a

characteristic of nouns, and consequently should be accounted for by marking the

lexicon entry of a noun with a feature [

+ count]-or something equivalent (cf.

Jespersen 1933:206, Chomsky 1965:82, McCawley 1975, Quirk et al. 1972:127)

545

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

The descriptive inadequacy of this analysis has hitherto been overlooked except by

Weinreich-although McCawley had an inkling of it. Weinreich (1966:435) noted

that 'it fails to account for the ability of English words [sic] to be used as EITHER

count or mass nouns.' He suggested three alternative solutions. One was to create

separate countable and uncountable lexical entries for the same noun; but this

was rejected because it almost doubles the number of nouns in the lexicon, and

because it failed to show that some nouns are more countable than others. His

second suggestion was to make countability a feature on determiners, to be trans-

ferred to the (countability-neutral) noun 'by a concord-type rule'; but without

discussing the interesting possibilities of such an hypothesis (cf. Allan 1977b, ch. 5),

Weinreich rejected

it on the grounds that determiners like the, any, this, and my are

countability-neutral. The solution he adopted was to maintain the conventional

binary marking for countability on the lexical entries for nouns, and in addition to

mark NP nodes for countability. This bipartite proposal requires the use of a

semantic calculator to resolve the various effects of locating a [ count] noun in a

[? count] NP; but this is an unwelcome complication to the grammar of count-

ability, and my own analysis will show it to be unnecessary. Weinreich's proposal

contains two additional flaws. Although his model demanded it, he offered no

solution to the problem of determining whether a given noun is countable or

uncountable: e.g., on what grounds would he decide the countability of cake, and

what would be the grounds for agreeing or disagreeing with him ? His adherence to

the conventional view of countability as a binary feature on the lexical class of

nouns renders the problem insoluble. The second, and less serious flaw in Wein-

reich's proposal is the lack of explicit justification for his substantially

correct claim

that countability is a feature of NP's. I shall, below, justify a rather similar claim.

Weinreich's discussion of countability can be seen as an attempt to shore up an

indefensible traditional kind of analysis by introducing additional apparatus. In

this paper I present a radically different approach, based on the assumption that

countability is not in fact a characteristic of nouns per se, but of NP's; thus it

is associated with nouns in syntagmata, not with nouns as lexical entries. Clear

evidence for this lies in the fact that countability is indicated in singular indefinite

NP's not by any modification of the noun, but by its falling within the domain of a

denumerator. Thus a noun like cake can regularly be used both uncountably and

countably, occurring both in undetermined singular NP's and in the scope of

denumerators:

(16) Hetty likes to gorge herself on cake.

(17) Whenever Hetty gobbles down a cake, her diet 'starts tomorrow'.

There are many parallel examples involving nouns less frequently seen in both

environments:

(18) Oak is deciduous.

An oak is deciduous.

(19) Small farmers in Kenya grow corn rather than wheat.

Triticum

aestivum

ssp. vulgare

is a wheat suitable for high altitudes.

(20) Nick Frenzy plays guitar with Noise.

Carol has just bought a guitar.

546

NOUNS AND COUNTABILITY

(21) The scrapyard is full of smashed car awaiting recycling.

The driveway was blocked by a car with its front end stove in.

(22) We went to school by car.

Is there a school around here?

(23) In bed they were blissfully happy.

You could buy a bed cheaply down on the lane.

(24) There's not enough table for everyone to sit at.

We need a bigger table.

(25) Emmy finds squashed spider more nauseous than the thing alive.

We were worried that even a squashed spider would upset Graham

enough to make him suicidally depressive.

Innumerable examples may be found: there seems to be no absolute constraint to

stop any noun, other than pluralia tantum nouns and perhaps plural proper names,

from heading an undetermined singular NP (cf. Gleason 1965:137). It is true that

there may be constraints on the location of such NP's within sentence structure,

and restrictions on the inclusion of other NP constituents; but this will not alter

the fact that nouns commonly used within the scope of a denumerator-and

therefore countably-can elsewhere head undetermined singular NP's, and so be

uncountable. This makes it absurd to propose that countability is a function of

nouns per se. To maintain that lexicon entries for nouns are marked with a feature

such as [ +?

count] would, for the majority of nouns, be either (a) contradictory or

(b) vacuous. First, for the lexicon entry to be simultaneously countable and

uncountable is a contradiction. Second, disjunctive countability-marking would

entail that a pair of homophonous, homographous, yet different nouns exist-one

countable, the other uncountable (cf. Yotsukura 1970:60); but then we face the

fact (of much greater significance than the inadequacies noted by Weinreich)

that

the selection of one of the disjuncts rather than the other would necessarily be

determined by countability features in the intended host construction, viz. the NP.

This dependence proves that disjunctive marking in the lexicon is vacuous. Count-

ability must perforce be a feature of NP's and not nouns; only then can the

uncountable/countable pairs of identical nouns in 16-25 be properly accounted

for.3 Exactly how this should affect our view of NP grammar has been discussed

elsewhere (Allan 1977b:144 ff.), and will not be considered here; instead, I will

discuss the consequences for nouns.

3 I am grateful

to William Bright

for drawing

my attention to recent philosophical

literature

on mass terms, where ideas similar

to the ones presented

here have occasionally

been advanced.

For example, Pelletier 1975 imagines a 'universal grinder' into which can be fed any object

labeled by a countable; the grinder chops and grinds it into a homogeneous mass, which is

then appropriately

labeled by the same noun used uncountably. So he concludes that every

noun must have both a count and a mass sense (which, as we shall see, is not quite accurate).

Bunt (1976:81) writes: 'the count/mass distinction is not really a distinction between words,

but a distinction between

ways of using words ... The count/mass distinction

can be drawn not

only between bare nouns, but also between complex noun phrases. In what follows, when we

speak of mass nouns, this has to be understood as short for: noun phrases, used in a mass

sense' (cf. also Bunt 1979, Ware 1975). These philosophers

seek to elucidate the semantics of

count and mass terms; with the exception of Bunt (1979:250-51), they virtually ignore the

syntax of countability. By contrast, in this paper I shall virtually ignore the semantics of

countability (which I discussed in Allan 1977b), to concentrate on the syntax of nouns in

relation to countability environments.

547

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

COMPUTING THE COUNTABILITY

PREFERENCES OF NOUNS

2.1. Even though countability is characteristic of NP's and not nouns, it is still

true that some nouns are generally located in countable NP's, while others are

located in uncountable NP's. The former set includes car, boat, beetle, carpet,

closet, and table; the latter lightning, mankind,

equipment,

evidence,

furniture, and

physics. Between these polar countables and uncountables are words like cake,

coffee, and lamb-which are common in both countable and uncountable NP's in

everyday language. These particular characteristics of nouns I call their COUNT-

ABILITY PREFERENCES. The countability preferences of nouns can be computed by

comparing the relative frequencies

of their occurrence in countable and uncountable

NP's, respectively. One way of doing this would be to take a large sample of texts,

then add up and compare the actual countable and uncountable occurrences of the

same noun, to build up a picture of countability usage; but however extensive the

sample, this method would not provide a justifiable basis for a decisive statement

about the countability potential of nouns. To discover the extent to which a noun

is countable (as distinct from the extent to which it has been used in countable

NP's within a given sample of texts), we must set up a test battery of countability

environments consisting of every kind of unambiguously countable NP and every

kind of unambiguously uncountable NP: then, for all these, a given noun will be

tried as NP head, and the result judged for grammaticality. The set of such judg-

ments will give a countability scan for the noun, from which its countability

preference

can be computed.

We can look to Rules I-II above as the starting point for discovering appropriate

test environments for countability. Rule I states that an NP in which a denumerator

ranges over the head constituent is countable; and we found that the set of

denumerators is a(n), each, every, either, several, many, both, (a) few, and the

natural numbers from one to infinity. But not all nouns that fall within the scope of

one denumerator will fall within the scope of every other denumerator. Compare

these examples:

(26) John has bought his wife a car.

Each car cost more to produce than it can sell for.

Two cars were badly damaged by the falling wall.

Many cars are not properly maintained because of the exorbitant cost.

(27) Penelope's is an admiration that I treasure.

*Each admiration contains an element of awe.

*Both Charlotte and Emily admired Anna's poems, and it was these two

admirations which she most appreciated.

*Linda Ronstadt has received many admirations.

(28) *A cattle has died in the west paddock.

*Each cattle out in the hurricane was killed.

*Two cattle were severely injured by the falling wall.

Many cattle died in the cyclone.

Some nouns (like car) are countable with all denumerators;

others (like admiration)

are countable only with unit denumerators a(n) and one; and pluralia tantum

nouns (together with one or two others like cattle) are countable only with fuzzy

548

NOUNS AND COUNTABILITY

denumerators. Fuzzy denumerators are those like (a) few, several, many, a dozen

or so, about

fifty, and high round numbers as infive hundred

cattle, 70,000 cattle-

denumerators which do not state a precise number, unlike (say) two,

fourteen, or

twenty-one. The various co-occurrence conditions demonstrated in 26-28 are

tabulated in Figure 1.

TYPE OF DENUMERATOR HEAD NOUN IN THE DOMAIN OF

THE DENUMERATOR

car admiration cattle

A. unit, a(n), one + +

F. fuzzy

(plural)

e.g. several, about fifty + +

0. all others +

FIGURE

Type O denumerators are a motley collection; but as Fig. 1 shows, they can

be logically excluded from our test battery, which need only differentiate

environments containing unit denumerators from environments containing fuzzy

denumerators:

any noun which can fall within the domains of both unit and fuzzy

denumerators will also fall within the domain of all other denumerators. For

convenience, I shall refer to the test environment of NP's with a unit denumerator

as the A + N Test, and that of NP's with a fuzzy (plural)

denumerator

as the F + Ns

Test. With these two tests, four differences

in countability can be computed: nouns

like car occur in both environments; nouns like equipment

in neither; those like

admiration

pass only the A + N Test; and those like cattle pass only the F + Ns Test.

Rule II states that, if the head constituent of an NP is plural, the NP is countable.

This suggests that noun morphology, and in particular the acceptance of a plural

inflection, might be an appropriate test for countability; and on many occasions

it is. But the morphology of nouns like sheep, means, scissors, measles, and physics

does not indicate their grammatical number, which can be determined only from

the context of the noun. Thus we can do no better than decide the countability of

these nouns by using the two tests for countability that have just been instituted.

These reveal that sheep and means are fully denumerable, like car:

(29) Jim barbecued a sheep to celebrate.

A few sheep nibbled vacuously at the roadside herbiage.

(30) A phaeton is a most elegant means of transport.

Several means of transport are available from the station.

Scissors, like cattle, fails the A+N Test:

(31) *Don't you have even one scissors?

But, unlike cattle, it is not clearly grammatical

in the F+ Ns environment:

(32) ?How many scissors do we have?

?Quite

a few scissors have disappeared

that way.

Measles, like equipment, fails both tests:

(33) *Have you ever had a measles?

*Several measles broke out in the town.

549

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

And finally, physics is like admiration

in passing only the A + N Test:

(34) A physics in which energy is lost rather than transferred is quite

inconceivable; where would the energy go to?

*There are several physics: geophysics, astrophysics, nuclear physics-

and I don't know what else.

With the exception of certain proper names (to be discussed in due course), any

noun which can be plural can also head NP's within the domain of a fuzzy denu-

merator, and so will pass (if perhaps only dubiously) the F+Ns Test. Hence, if a

countability test were defined on whether or not a noun can take a plural inflection,

it would give results identical to those of the F + Ns Test, making one of these tests

redundant. And there is no doubt which test is the more satisfactory: whereas noun

morphology does not necessarily indicate countability, the F + Ns Test unambigu-

ously defines countables. The conclusion must therefore be that a countability test

based on noun morphology, and in particular

on the possibility of plural inflection,

would be redundant.

Consider the emphasized NP's of 35, which have adjectival head constituents of

immutable morphological form, and which refer to groups of human beings:

(35) THE MEEK

shall inherit the earth.

Politicians nowadays are forever claiming to succor THE POOR.

Labor governments get their kicks by soaking THE RICH.

In soap operas THE WICKED

are always punished.

None of these NP's is denumerable:

(36) *There isn't a poor now that we have a welfare state.

*We passed two poor begging under the bridge.

*Many poor cannot be blamed for their lot; but a few can.

However, we know that these NP's are countable-not only because of their

reference, but also because they govern plural NP-external number registration:

(37) THE MEEK

ARE going to inherit the earth, ARE THEY?

THE WEALTHY WERE being soaked by the wicked lefties, who hated THEM.

THE WICKED ARE always punished, ARE THEY?

Otherwise, only countable NP's govern plural external number registration:

(38) THESE

TWO

GIRLS

WERE stealing skirts, WERE THEY?

THREE

SHEEP WERE nibbling the carrot tops when farmer Giles noticed

THEM.

NOT ALL CATTLE

ARE bovine, ARE THEY?

There ARE THREE ELEPHANT

browsing in that thicket, and one of THEM

is a tusker.

It is not even necessary

for the internal number of a countable NP to be plural when

the NP head is collective (cf. Allan 1979):

(39) THE HERD WERE grazing peacefully when a lion disturbed THEM.

MY COLLECTION

ARE fetching higher prices than the valuers expected

THEM to.

But uncountable NP's can never govern plural number registration:

550

NOUNS AND COUNTABILITY

(40) *THE LIGHTNING ARE

frightening Caspar, who's always been terrified of

THEM.

*SUGAR ARE

cheap

this week,

AREN'T THEY!

*THE ADMIRATION

OF HIS FRIENDS ARE what he wants most of all, AREN'T

THEY?

*EQUIPMENT ARE getting ever more expensive, ARE THEY not?

*MEASLES ARE no fun to have, ARE THEY, doctor?

Therefore we can set up a countability test that will identify an NP as countable if it

governs plural external number registration. I will refer to this environment as the

EX-PL Test.

It was pointed out in ?1 that any undetermined singular NP which is neither a

pronoun nor a fully-defining proper name is uncountable. But any noun (except

pluralia tantum) can head such an NP; so this environment would be completely

unrevealing as a test of countability preference. The question arises whether any

restricted uncountable environment exists. To guard against the unusual use of the

noun in an uncountable NP, it would be best to look to one that makes universal

reference-either a generic NP, or an NP determined by a universal quantifier.

Generic NP's are difficult, if not impossible, to define. They are formally identical

with non-generic NP's; and the only intrinsic characteristic

peculiar to them is the

scope of their reference. It might be that formal definition of generic NP's could be

achieved through identifying the predication on them (cf. Allan 1977b:70, Chafe

1970:189, Dahl 1975)-which is to say that there are no generic NP's, only generic

statements couched in generic sentences. Compare the generic statements of 41

with the non-generic ones of 42:

(41) The lion is carnivorous.

Sea water is saline.

(42) The lion is hungry.

Sea water had got into the petrol tank.

It is a defining characteristic

of certain species to be carnivorous, but to be hungry

is not defining for any species; it is by recourse to this kind of knowledge, I think,

that we distinguish the generic NP the lion in 41 from the non-generic instance in

42: the projection of genericness on the subject NP of 41 is a result of the predica-

tion on it. It is often the case, as in 41, that a generic NP is the subject of a sentence

which describes some characteristic feature of the reference of that subject NP; but

this is not necessarily so:

(43) Caspar is afraid of the dark.

Emily likes eggs.

I greatly respect the lion.

In each sentence of 43, it is the object which is the generic NP, and this is so because

of the predication on it-or by the nature of the statement in which the NP is

located (if this is not the same thing). It is clear from Dahl, though he never

proclaims it, that generic predications do not take for their arguments only the

kind of NP's that have traditionally been regarded as generic; e.g., it seems to me

that the following sentences are generic in that they make law-like statements:

(44) Staring at some people makes them blush.

Tickling their feet makes many people writhe.

551

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

If this is correct, the question arises whether the NP's some

people and many

people,

despite the fact that they are explicitly NOT

universal, are generic in 43. Perhaps,

before making a serious attempt to define 'generic NP', it would be well to discover

whether the effort is necessary for the present purpose. It seems that generic NP's

can usually be recognized intuitively; on this assumption, let us consider the

possibilities that generic NP's define countables-and more importantly for us,

that there is a restrictive uncountable generic environment.

Although the plural generic defines countable NP's (by Rule II), all those nouns

which may head plural generics may also head other plural NP's, and they are all

accounted for already by either the F + Ns Test or the EX-PL Test. The the-generic

mostly occurs with countables:

(45) a. The car is 20th century man's horse.

b. *Car is 20th century man's horse.

(46) a. (*)The lightning frightens cowardly Caspar.

b. Lightning frightens cowardly Caspar.

The parenthesized asterisk of 46a indicates that the sentence cannot be interpreted

generically, though it may legitimately have a non-generic interpretation-viz., for

some particular occasion, there was lightning that frightened Caspar. The meaning

of 46b is quite different from this, of course. But in some environments, the the-

generic NP is headed by nouns that are typically uncountable:

(47) Milquetoast is afraid of the dark.

Caspar, our pet mole, hates the light.

The lightning delights Vulcan the most; he doesn't give a fig for thunder.

The admiration of one's peers gives a better trip than DMT.

So the the-generic NP's do not appear to provide a reliable test for countability.

That brings us to those generics in which we are most interested, namely the

uncountable generics-undetermined singular generic NP's like those emphasized

here:

(48) LIGHTNING

is caused by the discharge of ELECTRICITY

from the clouds.

SEA WATER

is saline.

EQUIPMENT gets more and more expensive as it becomes more and more

sophisticated.

These may be contrasted with

(49) *Car is 20th century man's horse.

*Spider has eight legs.

*Bed is for lying in, not sitting on, according to my landlady.

Is this the restricted uncountable environment we have been looking for? Alas, it

is not: against 49, we can set 50, in which the same head nouns which failed the

uncountable generic environment of 49 are successful with different predications:

(50) Car is the best mode of transport.

Spider is a shrike's favorite food.

Bed has so many happy associations for William and friend.

This finally wipes out the generic NP as a catalyst for testing countability, and so

makes it unnecessary to define a generic NP.

552

NOUNS AND COUNTABILITY

Of course, the generic makes characteristic reference rather than universal

reference; thus there is no contradiction in

(51) Lions can be dangerous; but not all lions are dangerous.

Dogs have four legs. But not every dog has four legs; and any dog that

does not is abnormal.

So let us turn our attention from generic to universal NP's, and consider universal

uncountable NP's with all. Compare 52 with 53:

(52) All lightning is caused by the discharge of electricity from the clouds.

All sea water is saline.

All equipment gets more and more expensive as it becomes more and

more sophisticated.

(53) *All car is 20th century man's horse.

*All car is the best mode of transport.

*All spider has eight legs.

*All spider is a shrike's favorite food.

*All bed is for lying in, not sitting on, according to my landlady.

*All bed has so many happy associations for William and friend.

This may seem a suitably restrictive uncountable environment for the test battery;

but, without qualification, it is flawed because of sentences like

(54) This may look like an old jalopy to you, but it's ALL

CAR.

That stripper is ALL WOMAN, drooled Dirty Dick.

My boyfriend is ALL

MAN,

boasted Anita Bussom to Penny Plane.

The emphasized NP's of 54 can surface only as predicates (cf. Allan 1973), and

therefore they are non-referring:

the noun in each has very much the quality of an

adjective, with all an adjectival qualifier in the sense of 'completely'. By contrast

the all+ N phrases of 52 are never predicates,4

and they are typically used to refer

to a genus. It is also significant

that, in 52, the N of the all+ N construction is not a

covert NP: if it were, the preposition of could be inserted between all and N (cf. all

the lightning = all of the lightning;

all Paris = all of Paris). But *all of lightning,

*all of sea water, *all of equipment

would be ungrammatical

in 52. Hence, we can

establish all+ N as the restrictive

uncountable environment we have been seeking,

provided only that the N is not a covert NP (such that all+ N = all+ of+ N), that

the NP is not predicative, and that it is-potentially-genus-referring. Note that

all is the sole determiner in both 52 and 53; to avoid controversy, I will make it

a condition of this test that all must be the sole determiner. The singularity of the

NP will have to be determined on the basis of NP-internal

and NP-external number

registration.

Thus we finish with a test battery containing only four NP environments to test

the countability preferences of nouns. Having described the battery, it would be

well to clarify exactly what is to be measured. Countability consists of the binary

opposition between countableness and uncountableness; and the countability prefe-

ence of a noun will be either the degree to which it potentially occurs in uncountable

4 In sentences like They are all animals, the word all is a floating quantifier appositive to

they, and so not part of the predicate.

553

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

NP's or that to which it potentially occurs in countable NP's. The degree to which a

noun is countable will be the negative complement of the degree to which it is

uncountable, and vice versa; so, for the sake of simplicity, we must decide either

to postulate that a noun is COUNTABLE to a certain extent (which implies that nouns

are basically uncountable), or to say that a noun is UNCOUNTABLE to a certain

extent (which implies that nouns are basically countable). All the evidence indicates

that nouns are basically uncountable, with many of them being countable to a

certain extent. Of the NP environments in the test battery, three out of four are

countable because these are more restrictive than uncountable environments:

so uncountable environments are the more general, and presumably more basic

than countable ones. Again, countableness is marked in both morphology and

syntax; uncountableness is unmarked, and so presumably is the basic form. Some

nouns, like lightning, head only uncountable NP's; I find nothing to prevent the

ungrammatical NP's *a lightning and *these lightnings from being interpreted as

'flashes of lightning', and their ill-formedness seems arbitrary. This opinion is

substantiated by comparing the grammar of such nouns as equipment

in standard

English and in certain dialects spoken where English is a second language. In

standard English, equipment

can head only uncountable NP's; but in many African

and Asian varieties of English, it is found in countable NP's, e.g. At great expense

we have bought

a new equipmentfor

the lab, or Many equipments

have been removed

from the laboratory by unauthorized

persons. Such instances of the countable use of

equipment are correctly interpreted in terms of the standard English classifier

construction 'pieces of equipment'-which is, of course, the denotation of the word

in the non-standard dialects. Because no semantic anomaly results from the count-

able use of this noun and others like it, it is difficult to convince the ESL speaker

that he should use it only in uncountable NP's. The point of interest here is that,

however arbitrary

it may seem, there are nouns that can never, in standard English,

head countable NP's. However, all English nouns (except pluralia tantum nouns)

can head uncountable NP's; and on those occasions, they occur in their unmarked

base form. Even the pluralia tantum cease to be plural when used adjectivally,5

cf. scissor movement,

pant(y)-hose, suspender

belt, Nutcracker Suite; and since the

adjectival form of the noun is identical with the uncountable and unmarked form

for all nouns, it follows that pluralia tantum have base forms which look just like

the uncountable forms of most other kinds of nouns-even though they cannot be

used to head uncountable NP's. Thus all nouns have what I have just called a base

form, which is the adjectival and uncountable form. In consequence, it is justifiable

to measure the countability preferences of nouns in terms of their degree of count-

ableness.

2.2. The four NP environments which will provide a test for the countability of

their head nouns are the A + N Test, the F +Ns Test, the EX-PL Test, and the

All+N Test: of these, the first three define countable environments, the last one

an uncountable environment. The procedure will be to try a noun as the head of

each of these test NP's, and judge the resultant grammaticality. The degree of

5 Viz. as a 'noun modifier' which holds the position in surface NP structure between the

rightmost adjective

and the head noun.

554

NOUNS AND COUNTABILITY

countability of a noun will be determined on the basis of its success (viz. gram-

maticality) in the countable frames, and of its failure (ungrammaticality) in the

uncountable frame. Where possible, I will conflate the F + Ns and the EX-PL Tests.

There is a set of nouns like car which are 100% countable, in the sense that they

occur in all the countable environments, but are ungrammatical in the all+N

uncountable NP's. Nouns like beetle, needle, banjo, ball, deed, wall, table, and cat

are in this set:

(55) The A+N Test

John has bought himself A NEW CAR.

Elspeth trod on A BEETLE.

Nick has A GUITAR

as well as A BANJO.

Lucky found ONE

BALL

after losing two the day before.

ONE GOOD DEED deserves another.

(56) The F+Ns and EX-PL Tests

SEVERAL CARS chased Idi round the parade ground.

The sight of A FEW BEETLES

gobbling up your roof timbers can make

your nights marish.

If you want THEM, there ARE ABOUT

FIFTY

NEEDLES in that drawer.

SEVERAL DEEDS ATTRIBUTED TO THE DYNAMIC DUO WERE actually

performed by Steed and Purdey-who carried THEM out without a

whisper.

There AREN'T TOO

MANY GUITARS that sound like a uke, ARE there?

(57) The All+N Test

*ALL CAR is 20th century man's horse.

*ALL BEETLE

tastes good to a woodpecker.

*ALL CAT is a feline quadruped.

*ALL GOOD

DEED is rewarded, isn't it?

2.3. The next set of nouns to consider are those which succeed in all four envi-

ronments, and which thus differ from nouns like car because they are grammat-

ical when heading all+N uncountable NP's. This set seems to comprehend four

kinds of noun: (a) those like oak, lamb, potato, jacaranda, and wahoo-in which,

characteristically, the countable use describes a discrete object that is the source

for the constituent substance described by the isomorphic uncountable; (b) those

like hair, stone, or cake, whose countable form is typically used to indicate an

instance of what is denoted by the same noun used uncountably; (c) those like wine,

wheat, or coffee, whose countable use typically connotes species, types, or kinds of

substances denoted by the same noun used uncountably; and (d) nationality nouns

like Greek or Russian, which are fully inflecting, and others ending in -ese, like

Chinese (but not English, Welsh, Irish, Dutch, or French, which form a separate

class). The uncountable use of these nationality nouns denotes the language-and,

except occasionally in A +N constructions, the countable refers to the people.

Illustrations of these four kinds of noun are:

(58) The A+N Test

There is ONE

OAK

standing alone in a beech copse.

'Waiter, there's A HAIR in my soup!'-'It's rabbit, sir.'

555

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

Myself I prefer A RED WINE

with white meat, but then I've never been

a gourmet.

Up in Nyeri you need A WHEAT

that likes high altitude.

I think the man I saw was AN ITALIAN

because his words didn't quite

keep up with his hands.

(59) The F +Ns and EX-PL Tests

Caspar poached A FEW OAKS

from the forest and planted THEM in his

garden, where THEY ARE a monument to his dishonesty.

Gerry tried to swap me SEVERAL WAHOOS

for one sailfish but I didn't

want THEM; THEY just DON'T have the cachet!

There ARE MANY CAKES with more egg in THEM, but none so tasty as

this.

We have ABOUT

FIFTY WINES on our list, sir. Would you wish to peruse

it?

SEVERAL

WHEATS

HAVE been developed for high altitudes.

MANY GREEKS BELIEVE that democracy is a thing of the past, DON'T

THEY?

(60) The All+ N Test

All oak is deciduous.

To me all lamb tastes better than pork.

All brick is artificial; only stone is natural.

All wine is acidic.

All wheat is highly nutritious.

All Italian sounds mellifluous,

no matter

where the speaker

comes from.

This is the most interesting class of nouns, because of the well-recognized concrete

distinction between the countable and uncountable reference of the nouns, nicely

brought out in sentences like

(61) Mary won't touch lamb because she adores lambs.

I have shown elsewhere (Allan 1977b:270 ff.) how the very different denotation and

reference of the countable and uncountable use of the same noun derives from a

common semantic item, as well as a common lexical form; this justifies talking

about THE

NOUN

lamb, oak, hair, Greek etc.

The two sets of nouns considered hitherto have been successful in both the A + N

and the F+Ns Tests; in consequence, any of these nouns may occur within the

domain of all denumerators:

(62) Two cars collided head-on outside the house this morning.

Each beetle Graham sees he steps on.

Two oaks stood sentinel over the driveway entrance.

Each hair on her head is daubed with henna individually!

We shall have two wines: a rouge and a blanc.

Each hybrid wheat selects some advantages of its ancestors, but often

also has unpredicted weaknesses.

Two Armenians stood arguing in front of the three Chinese.

The rest of the nouns I shall discuss do not share this characteristic:

they can occur

only within the domain of a restricted

set of denumerators,

or of none.

556

NOUNS AND COUNTABILITY

2.4. Consider a number of pluralia tantum nouns with different countability

characteristics.

At first sight, there seem to be three classes of them; as an example

of each, I will take the nouns cattle and scissors and the NP head poor, which I will

temporarily treat as a noun. They differ from each other only in respect of the

F+Ns Test:

(63) The EX-PL Test

Those cattle are going to have to be shot, aren't they?

Those scissors you lent me are jolly blunt, aren't they, old girl?

The poor are a burden to us richer folk, so I don't know why you

should feel sorry for them.

(64) The All+N Test

*All cattle is bovine.

*All scissors was taken away, wasn't it?

*All poor needs the help of the better off, doesn't it?

(65) The A+ N Test

*One cattle was injured when the livestock carrier

crashed.

*Give me a scissors, will you?

*Whether we like it or not, we've got a poor.

(66) The F + Ns Test

a. We only run a few cattle.

Several cattle have been lost to rustlers.

About fifty cattle died in the cyclone.

b. ?How many scissors do you think we need to order?

?Only

a few scissors could never be used as weapons.

?We've

bought her several scissors since she started school, but she

loses them.

c. *A few poor stood or knelt by the road side begging for baksheesh.

*His charity has been extended to many poor.

*Each year about fifty poor starve to death in New Delhi.

Cattle is unique-the only noun in its set. Like other pluralia tantum nouns, it is

fully denumerable by using a classifier construction, and by ranging the denu-

merator over the classifier:6

(67) Not a single head of cattle was lost on the trail.

We flushed out two head of cattle from the brush by the creek.

We lost about fifty head of cattle in the cyclone.

All other pluralia tantum nouns (this excludes poor, of which more below) are like

scissors. As can be seen from 66b, their occurrence within the domain of a fuzzy

denumerator is not altogether impossible, though it is nearly always avoided by

using a classifier

construction instead-usually one containing the classifier

pair(s):

(68) How many pairs of scissors shall we need for the cut-out game ?

Bert's ruined several pairs of trousers that way.

6 What I am here calling a classifier might, in English grammar, be more appropriately

called a 'quantifying adjunct'; but I have already used the term 'classifier' for similar com-

ponents of other languages (cf. Allan 1977a), and it has a wide currency.

557

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

This group of pluralia tantum includes the /-s/ nationality names Cornish, Dutch,

English, French, Irish, Welsh, and one or two others for which the head noun

-man/men

(or alternatively

person/people)

is used in free denumeration,

instead of the

classifier construction employed with pluralia tantum NOUNS. These nationality

names are, of course, adjectives, and not subject to the morphological rules which

govern nouns-even when they head NP's. Nor, in themselves, do they have

any countability preference:

(69) He speaks an English that I can barely understand

at all; and I was born

in London.

(70) ?We encountered only two Russians, but several English-all wearing

long shorts and knee socks.

In 69 English is in the domain of a unit denumerator; in 70, it is in the domain of a

fuzzy denumerator. Because English can fall in both domains, it should be the case

that it can be denumerated by all denumerators

(cf. Fig. 1). But this entailment is

false: *two English, *each English etc. are all ungrammatical. However, we note

that, within the domain of a unit denumerator, English can only mean 'English

language':

(71) *We picked up an English who claimed to be hitchhiking round the world.

And, within the domain of a fuzzy denumerator, English can only mean 'English

people':

(72) *Traveling through Asia and Africa you come across several different

English(es).

It seems that, unless English

and other /-s/ nationality names constitute an exception

to the entailments on denumerability that hold for all other nouns, we are not

dealing with a single item English, but with two different ones. The nouns English,

Dutch etc., referring

to the languages, have a countability preference

identical with

that of admiration; I will discuss these nouns in due course. The pluralia tantum

English, Dutch etc. refer to 'English people', 'Dutch people' etc.; and in this use,

the nationality name is an adjective,

not a noun, and has no countability preference.

Elsewhere (Allan 1977b:177 ff.) I have accounted for the human reference of such

NP's as the police, the English, the poor etc. by, in effect, postulating a head noun

with the semantics of 'person/people'; this does not appear on the surface of these

NP's as a lexical item, but remains

understood. The same underlying

NP constituent

accounts for a number of other implicitly human-referring

NP's as well. A detailed

account of its distribution, and the conditions on its failure to surface, will not be

given here; but it is a subset of these conditions that accounts for the behavior of

the pluralia tantum English etc.-a behavior that has nothing to do with count-

ability preference. The underlying head of these NP's is fully countable; but under

certain conditions it may not surface, and under other conditions it cannot.

Essentially the same explanation holds for pluralia tantum like the poor, the

wealthy, the wicked, the pure in heart-in which the head is also an adjective not

governed by the rules of noun morphology. With these human-referring

NP's the

underlying head noun fails to surface only when the determiner is the and the NP

is plural: thus we never find *a poor, but only a poor man; never *these

poor, but

these poor people; not *some poor, but some poor man/men; not *all poor, but

558

NOUNS AND COUNTABILITY

all poor people. Thus poor etc. will be excluded from consideration here: they are

adjectives,

not nouns.

We are left with one set of nouns that consists only of cattle;7 and another

consisting of the regular pluralia tantum nouns scissors, braces/suspenders,

glasses/

spectacles,

pants, nutcrackers,

pliers, scales, tights, tongs, and tweezers,

all of which

have referents

perceived as two moveable leg-like members pinioned to a bridge at

one end, or so as to cross each other.

2.5. The next set of nouns to consider are those like admiration, heat, sincerity,

physics, thunder; the LANGUAGE nouns Cornish, Dutch, English, French, Irish,

Spanish and Welsh; and derived nominals with suffixed -ness, e.g. darkness or

redness. These are nouns which are very restrictedly countable-only by the unit

denumerator:

(73) The A+ N Test

Hermione's is an admiration that I value very greatly.

A dry heat is so much more bearable than a damp heat.

Einstein was responsible for the development of a new physics.

The cavalry arrived with a thunder of hooves.

Sean's is an English full of the lilt of the Western Isles.

An oppressive darkness hung all around us.

(74) The F +Ns and EX-PL Tests

*I am vying for several admirations at the same time and failing miser-

ably to win them.

*There are many heats and I have tried most of them: Lagos, the

Sahara, bed with Fluzy Suzy, and the inside of Granny's fish and

chip shop.

*You can detect several English(es) in the Caribbean.

*Few darknesses are as oppressive as this!

(75) The All+ N test (which shows these nouns to be predominantly uncount-

able)

All admiration does the ego good.

All heat is a form of energy.

All mathematics requires a logical mind.

All thunder is shattering to my nerves.

All Welsh has initial consonant mutation rules, though there are dialect

differences.

All darkness frightens Caspar Milquetoast, the booby.

The only countable NP's that such nouns can head are those with unit denumera-

tors; otherwise, they may only head uncountable NP's. As a result of the relatively

rare countable use of these nouns, countable occurrences

sometimes appear to refer

to instances or occasions of particular

note:

(76) We got up in darkness.

(77) *We got up in a darkness.

(78) We got up in a pitchy darkness.

7 For some Australian speakers of English, cattle is a fully countable noun just like sheep.

Both may be denumerated

using a classifier

construction and the classifier head.

559

LANGUAGE. VOLUME

56, NUMBER

3 (1980)

Clearly, the reference in 78 is to a particular kind of darkness, not just to ordinary

darkness.

2.6. The noun mankind is like admiration

in nearly all respects; it differs only in

passing the EX-PL Test because of its collective semantics (cf. Allan 1979):

(79) Anthropophiles believe in a mankind full of sweetness and light, but we

cynics see it rather differently.

(80) *Mr. Spock waxed eloquent on the differences

between the several man-

kinds he's met with; but as usual no one was listening.

(81) All mankind shall praise his name.

(82) If we are to believe Doctor Who, mankind are his favorite species: but

the Mekon fails to understand

what he sees in them!

But mankind seems to be very unusual in this; its only peer is womankind.

2.7. The last set of nouns to be subjected to the tests are the true uncountables,

which never head countable NP's: nouns like lightning, equipment, laryngitis,

measles, and ornithology,

and gerunds like thinking, knowing,

running, smiling, and

swimming. These are unacceptable in any of the countable environments, but are

grammatical in singular NP's within the domain of all:

(83) The A+N Test

*A lightning lit up the sky.

*He borrowed one equipment yesterday.

*I think she has caught a measles.

*A smiling will help you win friends.

(84) The F+Ns and EX-PL Tests

*Weren't they beautiful, those several lightnings that lit up the sky?

*We need many equipment(s),

but unfortunately

we cannot afford them.

*Few measles are as bad as these; I don't know what to do about them.

*I've done about fifty swimming(s) this year, and it's only January

third.

(85) The All+ N Test

All lightning is caused by the discharge of electricity from the clouds.

All electronic equipment gets more and more sophisticated every year.

All measles is dangerous.

All running is said to be good for you-even running round the

corner for a packet of smokes.

2.8. All nouns discussed so far in ?2 have been common nouns; to complete the

picture of countability preferences

of English nouns, it is necessary only to consider

the behavior of proper names. It was pointed out in ?1 that a fully-defining proper

name is invariably the sole constituent of an NP-more precisely, a definite and

specific NP that cannot contain a restrictive relative clause. Thus a fully-defining

proper name is defined not as a particular kind of noun, but as a class of NP, and

conditions on its countability are not relevant to the present essay. However, an

appellative proper name is one used like a common noun (what distinguishes it is

that it is ordinarily used as a fully-defining

proper name, whereas a common noun

is not). An appellative proper name is, therefore, relevant to our analysis of the

countability preferences

of nouns. There may appear to be a problematic difference

between a proper name like Nepal and one like John: John has many potential

560

NOUNS AND COUNTABILITY

referents, but at first sight Nepal has only one. However, the latter name can be

used as an appellative; e.g., one can quite legitimately say things like

(86) I know there are two Oxfords, why not two Nepals ?

I shall therefore assume that, in principle, all proper names can be used as appel-

latives; even though for some, such use would be very unlikely.

Singular appellative proper names are 100% countable, like car; thus they pass

the A+ N Test (87), the F +Ns Test (88), and the EX-PL Test (89):

(87) There is a Richard Devantpeau on the phone, dear.

(88) There have been several Tarzans, but J. W. was the best.

(89) Those six Lewistons, are they all in America?

An appellative proper name will fail the All+ N Test:

(90) *All Noam is a linguist.

(91) (*)All Lewiston is agog at Rachel's exploits.

In 91, the parenthesized asterisk means that, although the sentence is grammatical,

all Lewiston fails the All+N Test on the grounds that Lewiston is a covert NP.

This is demonstrated by the synonymy between 91 and 91':

(91') All of Lewiston is agog at Rachel's exploits.

Like car, such proper names as these are genus-referring

when they head plural

NP's quantified

by all:

(92) All cars are instruments of pollution.

All Lewistons listed in my atlas are in the U.S.A.

It might be expected that plural proper names like the United States, the Hima-

layas, or the Grampians

would have the same countability preference as pluralia

tantum like scissors; but this turns out not to be the case, and they form a separate

preference set. Perhaps the most significant difference is in their denumerability.

The only denumerator that may range over a plural proper name is, rather sur-

prisingly, the unit denumerator; so they pass the A+N Test, but completely fail

the F + Ns Test. It seems that the plural morphology of these proper names is

misleading: when used to refer to a single entity, e.g. a country (The United States

is in dire financial straits), an institution (It is a great squanderer of money, the

United Nations), or a place (The Himalayas, that's where

Edmund

longed to climb),

they are singular. Thus they pass the A+N Test:

(93) To read Dickens you would never know there could exist a British Isles

that is not fogbound.

It isn't really a United Nations, but a Disunited Nations.

There is only one Himalayas-nowhere else like it.

Plural proper names really are plural when used in referring

to a country's people

(The United States are spying on Russia), members of an institution (The United

Nations are by no means united), or some other plurality

(The Himalayas [the mount-

ains] lie between China and Nepal). It is also the unmarked number for these names:

(94) There are some Grampians in Australia, as well as in Scotland; but I

don't know exactly where they are.

There is a Grampians in Australia, as well as in Scotland; but I don't

know exactly where it is.

561

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

From this we see that plural proper names will pass the EX-PL Test. However,

they fail the All+ N Test (95) and the F+Ns Test (96):

(95) *All United Nations is in New York.

*All Himalayas is in Asia.

*All Solomon Islands is to the west of PNG.

(96) *There are quite a few British Isles.

*We thought we could perceive several United Nations.

*Knowing how the Scots have fled their homeland, I wouldn't be sur-

prised to find several Grampians dotted about the globe.

It is impossible to determine to what degree the ungrammaticality of 96 is a

consequence of the very restricted class of real or imaginary potential referents for

plural proper names.

Plural proper names like Himalayas constitute a distinct preference set by

passing the EX-PL and A+N Tests, but failing the All+N and F+Ns Tests.

However, all the tests to which singular proper names were subjected have shown

them to fall in the same countability preference set as car.

THE COUNTABILITY PREFERENCES OF ENGLISH NOUNS

3. The four tests used to determine the countability preferences of English

nouns have been the EX-PL Test (NP-external plural registration), the F +Ns Test

(a fuzzy plural denumerator ranging over the noun), the A+ N Test (a unit denu-

merator ranging over the noun) and the All+ N Test (all ranging over the noun in a

genus-denoting or genus-referring singular NP). In addition, Fig. 1 showed that

any noun which passes both the A + N and F + Ns Tests is fully denumerable; i.e.,

it may occur within the domain of all denumerators. This actually gives us five

environments for deciding the relative countability of nouns: the fifth-that of

denumerators other than the unit, or fuzzy denumerators-will be referred to

below by the abbreviation O-DEN.

Eight sets of nouns were identified in ?2 as having distinct countability prefer-

ences; examples for each set are car, oak, cattle, scissors, mankind, admiration,

equipment,

and Himalayas. All these except the set represented by equipment

are,

to a greater or lesser extent, countable. None of the tests proves positive with all

seven countable sets, and thus none serves to define the class of countable NP's;

but that is not the intention of this paper, nor would it make any useful contribution

to the computation of the relative countability preferences of English nouns. The

results of subjecting nouns from each countability preference set to the five NP

environments that have been identified are tabulated in Figure 2,8 where the NP

NOUN

ENVIRONMENT car oak cattle Himalayas scissors mankind admiration equipment

EX-PL + + + + + +

A+N + + + + +

All+N + + + +

F+Ns + + + ?

O-DEN + + FIGURE

(+ indicates that the given NP environment defines the head noun as countable.)

8 In the All+N Test, it is FAILURE

that gets a plus, not success.

562

NOUNS AND COUNTABILITY 563

environments are ranked in order of choosiness: the least choosy, EX-PL, is at

the top, and the most choosy, O-DEN, at the bottom. The least choosy environ-

ment is the one with most plusses to its left, the most choosy that with fewest-the

one that excludes all but the most countable nouns; thus the choosiness of each

environment is in inverse proportion to the number of preference

sets that it defines

as countable. It follows that for a noun to be defined as countable by the EX-PL

environment is much less significant

than to be defined as countable by the O-DEN

environment: this can be represented by making each 'plus' a reciprocal of the

total number of plusses for a given NP environment. Thus each EX-PL plus is

worth 1/6, an A+N plus is 1/5, an all+N plus is 1/4, an F+Ns plus is 2/7

(because ' ?' is assigned the value of half a plus, see below), and an O-DEN plus

is 1/2. A countability score can be computed for each noun by summating the

values assigned to each plus in the column below it. There is, of course, a problem

with the value of' ?' in the scissors column on the F + Ns line; to give it the value

of a plus would imply that pluralia tantum like scissors have the same countability

profile as cattle, and that they are fully grammatical in the domain of a fuzzy

denumerator. If this were true, the following sentences would not be dubious:

(97) ?How many scissors do you need, for heaven's sake?

?Few nutcrackers are so constructed as to save the nuts from going the

same way as their shells.

?Joe bought a few tights from this bloke at the pub, and when he got

home he found the legs ended at the knee.

?The fire destroyed about fifty trousers in boxes at the back of the shop.

However, because these sentences are not ungrammatical, it would be wrong to

assign no value at all to the '?' in Fig. 2, thus classing true pluralia tantum with

plural proper names like Himalayas. A compromise is to assign half the value of

the usual F + Ns plus, viz. 1/7, to the countability score of scissors. Using the arith-

metically convenient denominator of 420, the countability scores for each noun

representing

a countability preference

set are given in Table 1, along with the same

scores expressed as percentages-on the basis of nouns like car being 100%7

countable. Also included is a digit indicating the level of countability of each set

of nouns in relation to the other sets, expressed on a scale from zero to seven

corresponding to the uncountable and the most countable, respectively. The rank

order of nouns according to levels of countability was incorporated into Fig. 2.

REPRESENTATIVE NOUN COUNTABILITY SCORE PERCENTAGE LEVEL OF

COUNTABLE COUNTABILITY

car 589 100 7

oak 484 82 6

cattle 295 50 5

Himalayas 259 44 4

scissors 235 40 3

mankind 154 26 2

admiration 84 14 1

equipment 0 0 0

TABLE 1.

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

I doubt that the countability scores or the concomitant percentages of count-

ability have any intrinsic value. To say that car is twice as countable as cattle is not,

in itself, an interesting

fact about English; a difference in countability is interesting

only when related to the different environments in which one noun is grammatical

and another is not. The significance of these scores is to rank the countability pref-

erence sets in respect of each other, i.e. to be the means of deciding the respective

levels of countability. Thus the significant part of Table 1 is the level-of-count-

ability scale from 0 to 7. Reference will have to be made to points on this scale when

stating well-formedness conditions on the locating of nouns in NP's. If the argu-

ments of this paper are accepted, the traditional matching of an a-countable noun

from the lexicon with an a-countable NP environment cannot be based on a binary

system, but instead will have to be based on an octal system correlated with the

eight levels of countability. For example, it will need to be specified that a unit

denumerator can range over a head noun only from levels 1, 2, 4, 6, and 7 if the NP

is to be well-formed. A fuzzy denumerator

may range over a plural head noun from

levels 5, 6, and 7 in an assuredly grammatical NP; over a noun from level 3 in a

dubious NP; or over a noun from level 0, 1, 2, or 4 only in an ill-formed NP.

Each preference set is defined by the behavior of a representative noun, as the

NP head, in just two or three of the five tests; cf. Table 2. So it would be possible

to guess the countability preference set of a noun, and to check the guess by

submitting the noun to the tests defining the predicted set. However, the gamble

would be inappropriate

in serious analysis of countability preference,

which would

entail submitting the noun to all the tests instituted in ?2.

LEVEL OF COUNTABILITY GRAMMATICAL IN THESE UNGRAMMATICAL IN THESE

DEFINED ENVIRONMENTS ENVIRONMENTS

7 O-DEN All+N

6 O-DEN, all +N

5 F+Ns A+N

4 F+Ns, all+N

3 ?F

+ Ns all+N, A+N

2 all+ N, EX-PL F + Ns

1 A+N EX-PL

0 A+ N, EX-PL

TABLE

In Fig. 2, the five countability environments were ranked in order of choosiness;

but more general relations also exist between them. For instance, there is the plural

hierarchy:

a noun grammatical

in the O-DEN environment will also be grammatical

in the F + Ns environment; and any noun that heads a well-formed F + Ns phrase

will also be grammatical in an NP which has plural external number registration.

This can be stated more formally:

(98) O-DEN -+ F + Ns -+ EX-PL

(where Vx,y E {O-DEN, F+Ns, EX-PL} when a head noun is wf in x

then it is wf in y iff x -* y)

The relations between members of the plural hierarchy

are (predictably)

maintained

within another hierarchy that encompasses all five countability environments; it is

564

NOUNS AND COUNTABILITY

defined on the lowest level of countability exhibited by any noun which may in that

environment head a well-formed NP. Plural number registration of any kind is a

mark of countability, but NP-internal plural is associated with more exclusive

countable environments than is NP-external plural registration; and, of course,

among internal plurals, nouns that enter F + Ns are of lower countability than those

in O-DEN environments. Nouns of levels 0 and 1 can only enter singular environ-

ments, those with zero countability only the all+N environment. The hierarchy

between the five countability environments is shown in Table 3.

NP environment O-DEN > F+Ns > EX-PL > A+N > all+N

Lowest level of

countability

of NP head 6 5(?3) 2 1 0

TABLE 3.

CONCLUSION

4. This paper began with a sketch of what is meant by countability. In the

paradigm cases, that which is countable is denumerable. Although countability is

a linguistic category, it typically has perceptual correlations: the reference of what

is linguistically countable is ordinarily perceived in terms of one or more discrete

entities. What is uncountable is typically, though not necessarily, perceived as an

undifferentiated unity. The significance of the referential distinction between

countable and uncountable is quite clear from 99, where it is just this distinction

which gives the sentence its bite:

(99) It is because I like lambs that I don't like lamb.

The rule for English is that the countability of an NP must be known or made

known to the audience. In a definite NP, it can be assumed to be known, because

such an NP is used only where the speaker believes his audience to be aware of the

reference, which will ordinarily mean knowing the countability. In indefinite NP's,

however, this knowledge cannot be assumed; therefore countability must be

indicated. An NP is either countable or uncountable, so only one of these need be

marked; and in English, as in all other languages which mark only one, the

countable NP is the marked opposition. In plural indefinite NP's, the only obliga-

tory marking is the plural suffix on the head noun; in singular indefinite NP's,

where the head-noun morphology is identical for countables and uncountables,

there is obligatory inclusion of a denumerator. No problem arises from the fact

that fully-defining proper names and pronouns-other than quantifiers

and one(s)-

are non-denumerable, because they are invariably definite NP's.

Although countability has been traditionally thought a subcategory of the

lexical class of nouns, this view does not accommodate itself to the facts. Most

nouns can be used either countably or uncountably, making it impossible to

propose seriously that either plus or minus countable is the intrinsic property of

each of them-although countability can be associated with each particular use of

a noun in an NP. But then the fact that marking for countability is syntagmatic, in

the singular, shows that countability is not intrinsic to the particular instance of a

noun, but is a feature of its environment. So countability can be properly accounted

for only as a subcategory of the NP.

565

LANGUAGE, VOLUME 56, NUMBER 3 (1980)

Even though countability is characteristic

of NP's, not of nouns, it is nonetheless

a fact that nouns do show countability preference-insofar as some nouns more

often occur in countable NP's, others in uncountable NP's, and still others seem

to occur quite freely in both. In ??2-3, it was shown that these preferences can be

computed in a non-arbitrary way by setting up a test battery of countability

environments, defining the head noun as either countable or uncountable. A given

noun is tried as the NP head, and the result is judged for grammaticality.

The set

of judgments gives a countability scan for the noun, from which its countability

preference is computed. Countability is a binary opposition between the countable

and the uncountable; and it must be decided whether to assume nouns are basically

uncountable, and then environmentally induced to be countable to a degree-or

vice versa. All the evidence points to nouns being basically uncountable, though

most of them exhibit a degree of countability.

One unambiguously uncountable NP environment has been discovered, and four

unambiguously countable ones.9 A representative sample of all English nouns has

been tested, and eight levels of countability found. The results are summarized in

Figs. 2-3. This analysis of the countability preferences of English nouns is exhaus-

tive for my dialect, in which the eight levels of countability exist. However, it is

certain that other dialects of English will differ, because of different responses to

the various countability tests.

What I hope to have shown in this paper is the true relationship between nouns

and countability: that countability is a subcategory of the NP, not of nouns, and

that nouns have computable countability preferences which can be incorporated

into well-formedness conditions on English grammar.

REFERENCES

ALLAN, KEITH. 1973. Complement noun phrases and prepositional phrases, adjectives

and verbs. Foundations of Language 10.377-97.

. 1977a.

Classifiers.

Lg. 53.285-311.

. 1977b.

Singularity

and plurality

in English

noun phrases:

A study

in grammar

and

pragmatics.

Doctoral thesis, Edinburgh University.

- . 1979. Number

registration

in English: Concord

and discord.

Talanya

6.1-13.

BUNT, HARRY

C. 1976. The formal semantics of mass terms. Papers from the 3rd

Scandinavian

Conference

on Linguistics, ed. by Fred Karlsson, 81-94. Turku:

Academy

of Finland.

-- . 1979. Ensembles

and the formal semantic

properties

of mass terms. In Pelletier

1979:249-77.

9 Bunt (1979:270) writes: 'Certain adjectives,

like size adjectives

(large, small, big, tiny, huge,

...), are always count; if we meet a noun combined with such an adjective,

it must be classified

as a count noun.' But since these adjectives will combine with nouns from the three lowest

levels of countability (0, 1, 2), as well as from the five higher levels, he is surely wrong. Cf. these

data:

The equipment

will be too large to fit inside this room.

The tiny equipment for Empress Eugenie's doll's house was made by the outfitters for a

flea circus.

I have a huge admiration

for Bette.

Mankind is small compared with the blue whale.

Restrictions do seem to exist on the co-occurrence of size adjectives

and nouns of low count-

ability, and I do not understand

what they are. But (pace Bunt) size adjectives

cannot define an

unambiguous

countability

environment.

566

NOUNS AND COUNTABILITY

CHAFE,

WALLACE L. 1970. Meaning and the structure of language. Chicago: University

Press.

CHOMSKY,

NOAM. 1965. Aspects of the theory of syntax. Cambridge, MA: MIT Press.

CHRISTOPHERSEN,

PAUL.

1939. The articles: A study of their history and use in English.

Oxford: University Press.

DAHL, OSTEN.

1975. On generics. Cambridge Colloquium on Formal Semantics of

Natural Language, ed. by Edward L. Keenan, 99-111. Cambridge: University

Press.

FODOR,

JANET DEAN. 1970. The linguistic description of opaque contexts. MIT dis-

sertation.

GLEASON,

HENRY A. 1965. Linguistics and English grammar. New York: Holt, Rinehart

& Winston.

GRANNIS,

OLIVER

C. 1972. The definite article conspiracy in English. Language Learning

22.275-89.

HAWKINS,

JOHN A. 1978. Definiteness and indefiniteness. London: Croom Helm.

JACKENDOFF,

RAY. 1971. Gapping and related rules. LI 2.21-35.

JESPERSEN,

OTTO. 1933. Essentials of English grammar. London: Allen & Unwin.

KARTTUNEN,

LAURI. 1976. Discourse referents. Syntax and semantics 7: Notes from the

linguistic underground, ed. by James D. McCawley, 363-85. New York: Academic

Press.

KRAMSKY,

JIRi. 1972. The article and the concept of definiteness in language. The

Hague: Mouton.

MCCAWLEY,

JAMES

D. 1975. Lexicography and the count-mass distinction. BLS 1.314-

21. [Reprinted in his Adverbs, vowels, and other objects of wonder, 1'65-73.

Chicago: University Press, 1979.]

PALACAS,

ARTHUR L. 1977. Specificness in generative grammar. Studies in descriptive

and historical linguistics: Festschrift for Winfred P. Lehman, ed. by Paul J.

Hopper, 187-208. Amsterdam: Benjamins.

PARTEE,

BARBARA H. 1972. Opacity, coreference and pronouns. Semantics of natural

language, ed. by Donald Davidson & Gilbert Harman, 415-41. Dordrecht: Reidel.

PELLETIER,

FRANCIS J. 1975. Non-singular reference: Some preliminaries. Philosophia 5.

[Reprinted in Pelletier 1979:1-14.]

(ed.) 1979. Mass terms: Some philosophical problems. Dordrecht: Reidel.

QUIRK, RANDOLPH; SIDNEY GREENBAUM;

GEOFFREY

LEECH; and JAN SVARTVIK. 1972.

A grammar of contemporary English. London: Longman.

RIVERO,

MARiA-LUISA.

1975. Referential properties of Spanish noun phrases. Lg.

51.32-48.

-. 1977. Specificity and existence: A reply. Lg. 53.70-85.

WARE, ROBERT X. 1975. Some bits and pieces. Synthese 31.379-93. [Reprinted in

Pelletier 1979:15-29.]

WEINREICH,

URIEL. 1966. Explorations in semantic theory. Current trends in linguistics

III: Theoretical foundations, ed. by Thomas A. Sebeok, 395-477. The Hague:

Mouton.

YOTSUKURA,

SAYO. 1970. The articles in English. The Hague: Mouton.

[Received 5 July 1979.]

567

Proper Names and the “Noun”/“Name” Categories: Pseudo-Nouns, Real Names

Chapter

Mar 2024

Manon Philippe

This chapter summarises a number of arguments regarding the syntactic and semantic nouniness of proper names, which are not nouns (or noun phrases) but nominals that imitate the syntactic and semantic behaviour of (common) nouns. As such, they really are pseudo-nouns. However, the semantic role they play is not devoid of categorising potential—thanks to their parallel with nouns—and that makes them perfectly good names, on a cline from identification to categorisation.

Countability in Sorani Kurdish

Article

Jan 2024

This paper has two primary goals: to contribute to the discussion of typology of countability systems, and to document the countability system in Sorani Kurdish. While grammars show that Sorani Kurdish has a system of numeral classifiers system, the data reviewed in this paper includes counting constructions without classifiers and a number of morphosyntactic properties that are only used with a subset of nouns depending on their countability. We show that Sorani Kurdish has a mass/count distinction despite the fact that all nouns can be counted with classifiers, and that it has at least five countability classes. Knowing these characteristics of Sorani Kurdish helps to create a clearer picture of the linguistic variation regarding countability and suggests a relationship between the number of countability properties and countability classes in a given language.

Bare nouns, indefinite articles and partitivity in an Early New High German cookbook

Article

Full-text available

Jan 2024

Elvira Glaser

The starting point of the present article is the usage of mass nouns with indefinite articles, known from modern Bavarian and neighbouring dialects. Our analysis is dedicated to the use of the indefinite article varying with bare nouns in a historical perspective, based on a cookbook handwritten in 1556 in the East Swabian variety of Augsburg, containing about 900 instances of mass nouns with and without articles. Like in modern Bavarian, the readings OBJECT and QUALITY can be distinguished. A comparison with the de- nominals in Old Spanish recipes shows that the indefinite articles appear in equivalent positions with mass nouns mostly denoting non-specific regular objects as instantiations of the kind. The discussion of quantifiers and measuring expressions shows a special syntactic and semantic behaviour of ain wenig ‘a little’. The final discussion leads to the assumption that the indefinite article does not formally express a partitive relation, but, at most, produces partitive effects.

The syntax of individuating and measuring pseudo-partitives in Alasha Mongolian

Article

Full-text available

Dec 2023
J EAST ASIAN LINGUIS

Luis Miguel Toquero

Pseudo-partitive constructions give rise to multiple interpretive ambiguities including a container interpretation (i.e. individuating) and a contents (i.e. measuring) one. There are two competing analyses: one based on structural ambiguities (Landman in Indefinites and the types of sets, Oxford University Press, Oxford, 2004; Rothstein in Brill’s J Afroasiat Lang Ling 1:106–145, 2009. https://doi.org/10.1163/187666309X12491131130783, a.o.) and one based on a uniform syntax (Lehrer in Lingua 68:109–148, 1986; Matushansky and Zwarts in Lamont and Tetzloff (eds) North East Linguistic Society (NELS) 47, Volume 2, pp 261–274, GLSA, Amherst, 2016, a.o.). I contribute to this debate with data from Alasha Mongolian (Mongolic), which differentiates each interpretation via case marking on the quantizing noun: glass-comitative = individuating vs. glass-genitive/Ø = measuring. I argue that there is no large-scale structural ambiguity: the numeral and the quantizing noun always form a constituent introduced in the specifier position of a null functional head (Schwarzschild in Syntax 9(1):67–110, 2006. https://doi.org/10.1111/j.1467-9612.2006.00083.x; Svenonius in McNally and Kennedy (eds) Adjectives and adverbs: syntax, semantics and discourse, Oxford Studies in Theoretical Linguistics, pp 16–42, Oxford University Press, Oxford, 2008; Ott in J Comp Ger Ling 4:1–46, 2011. https://doi.org/10.1007/s10828-010-9040-x). I propose that (i) case differences on the quantizing constituent boil down to the presence or absence of a case probe on a higher Agr head; (ii) and, the interpretive differences between the individuating and measuring pseudo-partitives are the result of a more subtle syntactic distinction in the feature content of the quantizing noun, i.e. an interpretable [±Container] feature.

Morpho-syntaxe sémantique des nominaux propres

Article

Nov 2023

Manon Philippe

Teaching Countable, Uncountable, Plural Nouns

Chapter

Jan 2018

Shigenori Tanaka

In discussing English articles, we refer to the distinction between countable nouns and uncountable nouns. A countable noun is expressed in the “a + noun” form (e.g., an apple) or the plural noun form (e.g., apples), whereas an uncountable noun in the “Ø + noun” form (e.g., apple). Distinguishing what is countable from what is not seems fairly straightforward, and the use of countable and uncountable nouns in English should be easy. L2 learners, however, consider that English articles are among the most difficult grammatical items. A cognitive linguistics, which is an alternative to the traditional way of presenting English articles, claims that the countable / uncountable distinction applies not to the word, but to the referent the word denotes. The perception‐based principle of countability and a more general principle of individuation will be discussed here to draw pedagogical implications.

Portions and countability: A crosslinguistic investigation

Article

Full-text available

May 2023
NAT LANG LINGUIST TH

We examine three constructions across several languages in which a mass noun is embedded in what appears to be a count environment, but the construction as a whole remains mass. We argue that the discussed phenomena—“Q-noun” constructions like lots of water, bare measure constructions like kilos of sugar, and pluralised mass nouns in languages like Greek and Persian—all involve portioning-out of the embedded mass denotation. We provide a structural account of portioning out and propose structures that derive both mass and count portioning out. Adopting an overlap-based approach to the mass/count distinction (e.g. Landman 2011; Rothstein 2011; Khrizman et al. 2015; Landman 2016) we provide a compositional semantics for the proposed structures. The examined phenomena all share an inference of large quantity or abundance that, we argue, cannot be reduced to the lexical meaning of the portioning-out expression, nor to a multiplicity inference contributed by plural morphology. We show that our cases of mass portioning-out involve a total order ≤ on portion size and propose to analyse the abundance inference in terms of an uninformativity-based Quantity implicature, following the analysis of the positive form (Mary is tall) in Rett’s (2015) approach to adjectival gradability.

Procesamiento de la contabilidad nominal en cuatro pacientes anglófonos con afasia: El déficit en el uso de los sustantivos incontables

Article

Full-text available

Mar 2023

Esta investigación indaga sobre el procesamiento de la contabilidad nominal. De hecho, añade más evidencia a favor de una posible influencia del número semántico en el procesamiento de los sustantivos incontables en cuatro pacientes diagnosticadas con afasia de Broca, siendo su lengua materna es el inglés. Esto sugiere que el procesamiento de estos sustantivos sería un híbrido tanto sintáctico como semántico. Además, corrobora al igual que estudios previos, un déficit específico en la gramática de los sustantivos incontables.

Probing the role of bounding, definiteness and other factors: bare noun and determiner use in Guianese French Creole

Article

Full-text available

Nov 2022

Evelyn Wiesinger

From a cognitive‑semantic perspective, two important conceptual schemas underlie determiner use and the count/mass distinction in languages such as English and French, namely bounding and definiteness. This paper investigates to what extent these concepts are related to noun phrase marking in Guianese French Creole (GFC). The corpus‑based analysis shows that GFC uses bare nouns not only with unbounded entities and in non‑referential and generic contexts, but also for representing singular and plural indefinite and definite referents. At the same time, GFC shows a clear count/mass distinction and also has definite and indefinite determiners. However, in contrast to article languages such as English and French, bounding and definiteness need to be examined alongside semantic‑pragmatic and discourse‑informational criteria in order to explain the use of determiners and bare nouns in GFC. Thus, determiner use as a grammatical individuation strategy clearly has a different weight in GFC, which does not fit into cross‑linguistic typologies of articleless and article languages (e.g. Chierchia 1998). The comparative results on GFC, French and English also serve to evaluate two recent typological claims: while they confirm Le Bruyn et al.’s (2017) hypothesis on a correlation between the richness of the article system and the freedom of bare noun use, Abraham et al.’s (2007) claim of a universally complementary distribution of articles and grammatical aspect distinctions in the verb phrase is not supported by the data.

The Acquisition of English Grammar and Phonology by Cantonese ESL Learners: Challenges, Causes and Pedagogical Insights

Book

Dec 2022

Alice Yin Wa Chan

Specificity and Existence: A Reply

Article

Full-text available

Mar 1977

María-Luisa Rivero

This paper, a reply to Rojas 1977, re-affirms my conclusion (Rivero 1975) that specificity is marked in Spanish NP's by the mood of restrictive modifiers, and that it is independent of the degree of definiteness and the existential claims associated with those NP's. Specific and non-specific NP's alike may be referring expressions depending on their position in sentences. However, contrary to what I previously assumed, the notion of specificity is not connected with the pragmatic distinction between the referential and attributive uses of definite descriptions.

Complement Noun Phrases and Prepositional Phrases, Adjectives and Verbs

Article

Full-text available

Jan 1973

Keith Allan

Non-Singular Reference: Some Preliminaries

Article

Full-text available

Jan 1975

Francis Jeffry Pelletier

One of the goals of a certain brand of philosopher has been to give an account of language and linguistic phenomena by means of showing how sentences are to be translated into a “logically perspicuous notation” (or an “ideal language” — to use passé terminology). The usual reason given by such philosophers for this activity is that such a notational system will somehow illustrate the “logical form” of these sentences. There are many candidates for this notational system: (almost) ordinary first-order predicate logic (see Quine [1960]), higher-order predicate logic (see Parsons [1968, 1970]), intensional logic (see Montague [1969, 1970a, 1970b, 1971]), and transformational grammar (see Harman [1971]), to mention some of the more popular ones. I do not propose to discuss the general question of the correctness of this approach to the philosophy of language, nor do I wish to adjudicate among the notational systems mentioned here. Rather, I want to focus on one problem which must be faced by all such systems — a problem that must be discussed before one decides upon a notational system and tries to demonstrate that it in fact can account for all linguistic phenomena. The general problem is to determine what we shall allow as linguistic data; in this paper I shall restrict my attention to this general problem as it appears when we try to account for certain words with non-singular reference, in particular, the words that are classified by the count/mass and sortal/non-sortal distinctions.

The Article and the Concept of Definiteness in Language

Book