Taalportaal - the digital language portal

Dutch1
Frisian
Saterfrisian
Afrikaans

Show all

1.3.2.1.The set-denoting property

quickinfo

From a semantic point of view, set-denoting adjectives denote properties and in this respect they differ from verbs and nouns, which normally denote events and entities, respectively. From the perspective of set theory, on the other hand, verbs, nouns and adjectives are similar in the sense that they all denote sets of entities. An intransitive verb such as wandelen'to walk' denotes all entities in the domain of discourse that are walking, e.g., “Jan", “Peter" and “Marie". A noun such as jongen'boy' denotes all entities that have the property of being a boy, e.g., “Jan", “Peter" and “Henk". And, finally, an adjective such as groot'big' denotes all entities that have the property of being big, e.g., “Jan", “Henk" and “het gebouw" (the building). The domain of discourse that has been set up can be represented as in (36); the entities that are mentioned between the curly brackets are part of the sets that are denoted by the relevant word.

wandelenV: {Jan, Peter, Marie}

jongenN: {Jan, Peter, Henk}

grootA: {Jan, Henk, het gebouw}

Given that some entities are placed within more than one set, a more proper representation of our domain of discourse can be given as in Figure 1, where the relevant entities are placed in the intersections of the three sets.

Figure 1: Set-theoretical representation of domain of discourse

Note that we have simplified the discussion above by ignoring the fact that, e.g., transitive verbs or adjectives that take a PP-complement do not denote entities but ordered pairs of entities: a verb like slaan'to hit' denotes all ordered pairs <x,y> such that x hits y; similarly, an adjective like trots'proud' denotes all ordered pairs <x,y> such that x is proud of y. Some verbs and adjectives may even denote ordered triples or quadruples of entities. We will not discuss this any further here since this will lead us into a discussion of complementation, which is the topic of Chapter 2. This section will continue with a brief discussion of the syntactic uses of the set-denoting adjectives: Subsection I will begin with their attributive and predicative uses, which are the most common ones, followed by their adverbial use in Section II.

I. Attributive and predicative uses
II. Adverbial use

[+] I. Attributive and predicative uses

This subsection briefly discusses the attributive and predicative uses of adjectives, subsection A will start by discussing the interpretation that is typically associated with these uses: an attributive adjective normally enters into an intersection relation with the modified noun, giving rise to a so-called restrictive interpretation, whereas a predicative construction instead enters into an inclusion relation with the noun phrase it is predicated of, subsection B will show, however, that there are also cases in which attributive adjectives enter into an inclusion relation with the modified noun, giving rise to a so-called non-restrictive interpretation, subsection C discusses various types of predicative uses of adjectives, and show that the inclusion relation is also apt for describing these cases, subsection D concludes with a note on the use of attributive adjectives in complex proper nouns like de Stille Oceaan'the Pacific'.

[+] A. Intersection and inclusion

Set-denoting adjectives are typically used to specify the properties that are attributed to nouns or noun phrases. Two typical environments in which these adjectives occur can be distinguished: the prenominal attributive position shown in (37a) and the complementive position immediately preceding the verb(s) in clause-final position in (37b).

a.		grote	jongens
		big	boys

b.	dat	de jongens	groot	zijn.
	that	the boys	big	are
	'that the boys are big.'

However, the formulation that set-denoting adjectives specify the properties of nouns or noun phrases is still too general, since there is a difference in function between the adjectives in (37a) and (37b): due to the placement of the adjective in the prenominal attributive position in (37a), the noun phrase grote jongens denotes those entities that are both big and a boy, that is, it denotes the intersection of the two sets denoted by, respectively, jongen and groot. Thus, in the domain of discourse we have set up in Figure 1, this noun phrase refers to “Jan" and “Henk". The predicative use of the adjective in (37b), on the other hand, asserts that (all of) the boys are big, that is, that we are dealing with an inclusion relation between the set of entities referred to by the noun phrase the boys and the set denoted by big. Clearly, this inclusion relation does not hold in our domain of discourse in Figure 1, since “Peter" is not included in the set of big. This difference between the attributive and predicative use of the set-denoting adjectives can be schematically represented as in Figure 2.

Figure 2: Attributive and predicative use of set-denoting adjectives

[+] B. Restrictive and non-restrictive uses of attributive adjectives

The intersection relation between the noun and the attributive adjective depicted in Figure 2 amounts to saying that the adjective restricts the denotation of the noun, and this reading is therefore generally referred to as the restrictive interpretation of the attributive adjective. Sometimes, however, attributive adjectives also allow a non-restrictive interpretation, and in that case their function is very similar to that of predicatively used adjectives. We will distinguish two cases that differ in whether the modified noun denotes a singleton or non-singleton set.

[+] 1. Nouns denoting singleton sets

Both examples in (38) are fully acceptable in case there is just a single crown prince in the domain of discourse. This means that the attributively used adjectives in (38a) are not needed to restrict the denotation of the noun kroonprins'crown prince', but added to express additional information about the crown prince: just like the predicative adjectives in (38b), the attributive adjectives are used to inform the addressee about the fact that the crown prince is tall and fair.

a.	De	lange,	blonde	kroonprins	trok	in China	veel aandacht.
	the	tall	fair	crown prince	drew	in China	much attention
	'The tall, fair crown prince got a lot of attention in China.'

b.		De kroonprins	is lang en blond.
		the crown.prince	is tall and fair

We need not construe the non-restrictive interpretation of (38a) as an exception to the general pattern given in Figure 2, but can simply consider it a special case of the left-hand representation. Given that the noun kroonprins denotes a singleton set, the intersection of the sets denoted by the noun and the two adjectives is either the singleton set denoted by the noun or empty. Since it would not be informative to attribute the properties of being long and fair to the empty set, it is clear that the speaker intends to refer to the singleton set; this gives the false impression that (38a) involves an inclusion relation, whereas we are actually dealing with a special case of the intersection relation.
Adjectives modifying proper nouns may also receive a non-restrictive interpretation, due to the fact that proper nouns are normally not used to denote a set, but an individual. The examples in (39a&b) therefore express that de Westerkerk can be characterized as a beautiful church, and that Schiphol is a crowded airport. Note that there is a certain tendency for non-restrictive attributive adjectives to be interpreted as epithets, example (39c) being a classical example of this.

a.		de	mooie	Westerkerk	a church in Amsterdam
		the	beautiful	Westerkerk

b.		het drukke	Schiphol	an airport near Amsterdam
		the crowded	Schiphol

c.		de	beeldschone	Helena
		the	beautiful	Helen

Note that proper nouns must always be preceded by a determiner if they are modified by an attributive adjective. The articles in (39b&c), for example, are obligatory despite the fact that the proper nouns in these examples are normally article-less: cf. (*het) Schiphol and (*de) Helena. For this, we refer the reader to the comprehensive discussion of modification of proper nouns in N3.

[+] 2. Nouns denoting non-singleton sets

There is no a priori reason for assuming that the non-restrictive reading is limited to proper nouns and nouns denoting singleton sets: we expect that, at least in some cases, we will find ambiguous examples when the modified noun denotes a non-singleton set. This is indeed borne out, as will be clear from the examples in (40). These examples can all be interpreted as referring to a subset of the denotation of the noun, but they also allow a non-restrictive interpretation, in which case the property denoted by the adjective is attributed to all members of the set denoted by the noun.

a.		De dappere soldaten	vochten	tot het einde.
		the brave soldiers	fought	to the end

b.	Het koude water	deed	Peter rillen.
	the cold water	did	Peter shiver
	'The cold water made Peter shiver.'

Example (40a), for example, can have two interpretations: under the restrictive reading of the adjective dappere'brave', it is asserted that only a subset of a larger set of soldiers fought till the end; the non-restrictive interpretation of the adjective, on the other hand, implies that all members of the set of soldiers fought to the end, for which reason they are called brave. Note that example (40a) is only truly ambiguous in written language and that, in speech, intonation will normally resolve the ambiguity: on the restrictive reading the adjective will usually be stressed, whereas it will normally be unstressed on the non-restrictive reading.

De dappere soldaten vochten De dappere soldaten vochten

tot het einde.

restrictive

De dappere soldaten vochten

tot het einde.

non-restrictive

Something similar holds for (40b): under the restrictive interpretation there are several contextually determined amounts of water with different properties, and it is asserted that the water that was cold made Peter shiver: under the (more likely) non-restrictive reading, there is just one amount of water, and we will infer that the water made Peter shiver because it was cold. The ambiguity can again be dissolved by means of the intonation pattern.

Het koude water deed Peter rillen.Het koude water deed Peter rillen.

restrictive

Het koude water deed Peter rillen.

non-restrictive

Similar ambiguities may also arise in constructions with a demonstrative determiner or a possessive pronoun. Example (43a) can be used in order to ask for a certain pen from a contextually determined set of pens, in which case the adjective is used restrictively to identify the intended object. The adjective may, however, also be used if there is only one entity that answers to the description of the noun pen, in which case the adjective has a non-restrictive, purely descriptive function, which may help the addressee to recognize the object referred to, but does not serve to identify the intended object. Similarly, example (43b) can be used in a context where the speaker had an old computer that did not function well, in which case the adjective is used restrictively and will be emphasized. A non-restrictive interpretation is, however, also possible; in that case no comparison with other computers is involved, and the intended computer may in fact be the very first one the speaker ever had.

a.	Kun	je	mij	die rode pen	aangeven?
	can	you	me	that red pen	prt.-hand
	'Can you hand me that red pen?'

b.		Mijn nieuwe computer	werkt	prima.
		my new computer	works	fine

Sometimes, the context also favors the non-restrictive reading. In (44), for example, it is clearly the non-restrictive reading of the adjective that is intended: the intended message is not that only the well-informed advisors will be glad to be of service, but that all advisors are well informed and will be glad to be of service.

Onze welingelichte adviseurs	zijn	u	graag	van dienst.
our well.informed advisors	are	you	gladly	of service
'Our well-informed advisors are glad to be of service to you.'

Example (45) shows that certain construction types may also favor the non-restrictive reading. The pseudo-partitive construction van die heerlijke truffels,which is more extensively discussed in N4.1.1.6, looks like a PP but in fact functions as a noun phrase that refers not to a particular subset of truffles but to a certain type of truffles, which is claimed to be delicious. The attributive adjective does not serve to restrict the denotation of the noun, because this is assumed to be known to the addressee, but is used purely descriptively; we may infer from (45) that the speaker wants to have more truffles of the intended type, because he considers them delicious.

Mag	ik	een half pond	van die heerlijke truffels?
may	I	a half pound	of those lovely truffles
'Can I have half a pound of those lovely truffles?'

Indefinite noun phrases do not readily allow a non-restrictive interpretation of the attributive adjective; in all examples in (46), the adjective restricts the set denoted by the noun, thus (implicitly) contrasting this subset with the remaining members of the set.

a.	Kun	je	mij	een rode pen	aangeven?
	can	you	me	a red pen	prt.-give
	'Can you hand me a red pen?'

b.		Nieuwe computers	werken	prima.
		new computers	work	excellent

c.	Wij	hebben	alleen	welingelichte adviseurs	in dienst.
	we	have	only	well.informed advisors	in employ
	'We only employ well-informed advisors.'

[+] C. Supplementives and appositions

The semantic representation of the predicative use of the adjective depicted in Figure 2 was based on a discussion of complementive adjectives. The set-theoretic interpretation of supplementive adjectives like dronken in (47) is, however, very similar.

a.	De gasten	gingen	dronken	naar huis.
	the guests	went	drunk	to home
	'The guests went home drunk.'

b.	Ik	bracht	de gasten	gisteren	dronken	naar huis.
	I	took	the guests	yesterday	drunk	to home
	'Yesterday, I took the guests home, drunk.'

The examples in (47) express that the guests were drunk when they went/were taken home, that is, that the set denoted by de gasten'the guests' is a subset of the set denoted by dronken. Therefore, as far as the noun phrase and the adjective are concerned, the same set-theoretical implication is expressed as in the case of the predicative use of the adjective in de gasten zijn dronken'the guests are drunk'. For simplicity, we have ignored that (47b) is actually ambiguous and that the adjective may also be predicated of the subject ik'I'; we will return to this fact in Section 6.3.
The set-theoretic interpretation of appositively used adjectives is also very similar to that of the complementive adjectives: example (48a), for instance, implies that all of the men are angry about the rejection. The appositive phrase in this example can be paraphrased by means of the nonrestrictive relative clause in (48b), which contains a copular construction; see Section 6.4, sub II, for restrictive appositives.

a.		De mannen,	kwaad over de afwijzing,	schreven	een gepeperde brief.
		the men	angry about the rejection	wrote	a spicy letter

b.	De mannen,	die kwaad	waren	over de afwijzing,	schreven	een gepeperde brief.
	the men	who angry	were	about the rejection	wrote	a spicy letter
	'The men, who were angry about the rejection, wrote a spicy letter.'

[+] D. A note on complex proper nouns and classifying adjectives

The primeless examples in (49) show that adjectives are sometimes an inherent part of proper nouns. We are not dealing with attributively used adjectives in cases like these, which is clear from the fact, illustrated in (49b), that complex proper nouns like Magere Hein cannot be preceded by a definite determiner; the discussion of the examples in (39) has shown that a determiner normally must be present if a proper noun is preceded by an attributive adjective. For completeness’ sake, example (49b') shows that the complex proper noun Magere Hein does not behave differently in this respect.

a.		de	Middellandse	zee
		the	Mediterranean	Sea

b.		(*de)	Magere Hein
		the	Grim Reaper

b'.	*	(de)	schrikaanjagende	Magere Hein
		the	terrifying	Grim Reaper

The examples in (50) show that an attributive adjective may sometimes also form a fixed collocation with a common noun. It would not be proper to say that the attributive adjective wit'white' has a restrictive function in (50a): rather the adjective and the noun function as a lexical unit that refers to a certain type of wine. The adjectives in (50) are sometimes referred to as classifying adjectives; cf. Alexiadou et al. (2007:part III, §3.3).

witte wijn

'white wine'

magere melk

'skim(med) milk'

Chinese thee

'Chinese/China tea'

Franse kaas

'French cheese'

The collocations in (50) come close to compounds. In this respect we may refer to the difference between Dutch witte wijn and its German counterpart Weißwein, which does not exhibit attributive inflection and therefore must be the result of compounding; cf. Booij (2002:12). A similar contrast between Dutch and German is more or less consistently found with color adjectives in the names of animal species; this is illustrated in (51) by means of a number of bird names. taken from an extensive list found at mezen.madelen.nl/VogelnamenWereld.xls.

blauwe reiger

a'.

Graureiher

a''.

Grey Heron

bruine vliegenvanger

b'.

Braunschnäpper

b''.

Brown Flycatcher

gele ral

c'.

Gelbralle

c''.

Yellow Rail

grijze spotlijster

d'.

Grau-Spottdrossel

d''.

Grey Thrasher

rode wouw

e'.

Rotmilan

e''.

Red Kite

witte specht

f'.

Weißspecht

f''.

White Woodpecker

Note. however, that if the bird name is based on some characteristic body part, as in (52a), Dutch invariantly uses a compound form. This may be to avoid confusion with attributive constructions such as (52a'), although this leaves open the question why compounding is also preferred in examples such as (52b), where such confusion is not likely to arise, given that the form borstbijeneter does not exist. We will leave the contrast between the primeless examples in (51) and (52) for future research.

blauwborst

'Bluethroat'

a'.

blauwe borst

'blue throat'

blauwborstbijeneter

'Blue-breasted Bee Eater'

b'.

blauwe borstbijeneter

[+] II. Adverbial use

Adverbially used adjectives are not morphologically distinguished from attributively or predicatively used adjectives in Dutch; that is, Dutch has no equivalent of the English suffix-ly. This means that the forms in (53) can be translated in English either as adjectives or as adverbs. Note that, although it is not clear whether there is a categorial distinction between the English adjectives and adverbs in (53), we will treat the adverbs in question as (inflected) adjectives here.

snel: quick (A), quickly (ADV)

langzaam: slow (A), slowly (ADV)

behoedzaam: cautious (A), cautiously (ADV)

Due to the lack of morphological marking, predicative and adverbial uses of adjectives are sometimes hard to distinguish in Dutch; cf. Sections 5.2, sub IV, and 8.2.2. Often, we can only appeal to the meaning of the example to determine whether we are dealing with the former or the latter. The crucial difference between attributively and predicatively used adjectives, on the one hand, and adverbs, on the other, is that whereas the former only modify nouns and noun phrases, the latter specify VPs, APs (including adverbial phrases), and PPs. Consider the examples in (54a&b). Although the syntactic frames in which boos'angry' and snel'quick' are used seem identical, we are dealing with a supplementive adjective in (54a) and with an adverbially used adjective in (54b). This can be made clear by using the paraphrases in the primed examples, which show that boos modifies the noun phrase Jan, whereas snel modifies the VP weg lopen'walk away'. The paraphrase in (54b'') may be clumsy, but the contrast with (54a'') is pretty sharp.

a.		Jan liep	boos	weg.
		Jan walked	angry	away

a'.	Jan liep	weg,	terwijl	hij boos	was.
	Jan walked	away	while	he angry	was
	'Jan walked away while being angry.'

a''.	*	Zijn weggaan	was boos.
		his going away	was angry

b.		Jan liep	snel	weg.
		Jan walked	quickly	away

b'.	*	Jan liep weg,	terwijl	hij snel	was.
		Jan walked away	while	he quick	was

b''.		Zijn weglopen	was snel.
		his walking away	was quick

In example (55a), too, we are dealing with an adverbially used adjective, since what is expressed is not that Jan is cautious (he may be reckless in several respects), but that the activity of investigating the meal was undertaken cautiously. Example (55b), on the other hand, is ambiguous: it can be interpreted either as meaning that Jan was greedy or as meaning that the investigation was undertaken eagerly.

a.		Jan onderzocht	de maaltijd	behoedzaam.
		Jan investigated	the meal	cautiously

b.		Jan onderzocht	de maaltijd	gretig.
		Jan investigated	the meal	greedy/eagerly

In (56a), the adverbially used adjective goed'well' modifies an adjective. There are reasons for assuming that intensifiers like erg'very' and afdoende'sufficiently' in (56b&c) also belong to the adjectival class; cf. the primed examples and the discussion in Section 3.1.2.

a.		een	goed	leesbaar	handschrift
		a	well	readable	handwriting

a'.		een	goed	handschrift
		a	good	handwriting

b.		een	erg	mooi	boek
		a	very	beautiful	book

b'.		een	erg	ongeluk
		a	bad	accident

c.		een	afdoende	gemotiveerd	antwoord
		a	sufficiently	motivated	answer

c'.		een	afdoende	antwoord
		a	conclusive	answer

In (57), we are dealing with adverbially used adjectives modifying, respectively, a locational and a temporal PP.

a.	De kerk	stond	ver	buiten het dorp.
	the church	stood	far	outside the village
	'The church was far from the village.'

b.		Jan voltooide	zijn artikel	lang	voor de deadline.
		Jan finished	his paper	long	before the deadline

References:

Alexiadou, Artemis, Haegeman, Liliane & Stavrou, Melita2007Noun phrases in the generative perspectiveBerlin/New YorkMouton de Gruyter
Booij, Geert2002The morphology of DutchOxfordOxford University Press

⇑ Back to top

About this article

Author(s): Hans Broekhuis

Category: Dutch Syntax

Keywords: AdjectiveAdverbialAppositiveAttributiveClassifying AdjectiveComplementiveEpithetNon-restrictive InterpretationNounPredicativeProper NounRestrictive InterpretationSet TheorySupplementive

Version: 1.0, December 2015

Version history:

version	editor(s)	date	remarks
1.0	Hans Broekhuis	December 2015

report errorprintcite

Jump to:

introduction
references
related articles