• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
1.3.2.2.Semantic classification
quickinfo

Many semantic subclassifications have been proposed for the set-denoting adjectives, but most of them seem to have a rather arbitrary flavor. Nevertheless, some of these distinctions have been claimed to be syntactically relevant (especially in the realm of modification, which is extensively discussed in Chapter 3), which is why we will briefly discuss these distinctions in the following subsections. It should be kept in mind, however, that in principle many other distinctions can be made, for other purposes, and that the classes discussed below exhibit a considerable overlap; see Subsection III for discussion.

readmore
[+]  I.  Scales and scalar adjectives

Many set-denoting adjectives are scalar. The primeless examples in (58) express that both Jan and Marie are part of the set denoted by the adjective ziek'ill', which will be clear from the fact that they imply the primed examples. The function of the intensifiers vrij'rather' and zeer'very' is to indicate that Jan and Marie do not exhibit the property of being ill to the same degree. This means that the possibility of adding an intensifier indicates that some scale is implied; the function of intensifiers vrij and zeer is to situate the illness of Jan and the illness of Marie at different places on this scale. This can be schematized as in (58c).

58
a. Jan is vrij ziek.
  Jan is rather ill 
a'. Jan is ziek.
  Jan is ill
b. Marie is zeer ziek.
  Marie is very ill 
b'. Marie is ziek.
  Marie is ill
c. Scale of illness:

The schema in (58c) indicates that Jan is less ill than Marie. Further, it indicates that there is some point to the left of Jan where we start to talk about illness; the scale is bounded at its left side. However, as long as the person involved stays alive, there is no obvious point on the right side of the scale where we stop talking about illness; the scale is unbounded at the right side. This subsection will discuss several types of scalar adjectives on the basis of the properties of the scales that they imply.

[+]  A.  Antonymous adjectives

Many set-denoting adjectives come in antonym pairs, which can be situated on a single scale. Some examples are given in (59). The following subsections will show, however, that the scales implied by these antonym pairs may differ in various respects.

59
a. slecht 'evil/bad'
a'. goed 'good'
b. klein 'small'
b'. groot 'big'
c. vroeg 'early'
c'. laat 'late'
d. gezond 'healthy'
d'. ziek 'ill'
e. leeg 'empty'
e'. vol 'full'
[+]  1.  Scales that are unbounded on both sides

First consider the scale implied by the pair goed'good' and slecht'evil/bad', given in (60). The two adjectives each indicate a range on the scale, that is, they are both scalar. Further, the implied scale is unbounded on both sides. However, between the two ranges denoted by goed and slecht, there is a zone where neither of the two adjectives is applicable, and which we will call the neutral zone.

60
Scale of “goodness":

That there is a neutral zone is clear from the fact that slecht'evil/bad' and niet goed'not good' are not fully equivalent. The difference can be made clear by looking at the logical implications in (61a&b). The fact that slecht implies niet goed, but that niet goed does not imply slecht can be accounted for by making use of the scale of “goodness" in (60). As can be seen in (61c), niet goed covers a larger part of the scale than slecht: it includes the neutral zone.

61
a. Jan is slecht.
  Jan is evil 
a'. Jan is niet goed.
  Jan is not good
b. Jan is niet goed.
  Jan is not good 
b'. Jan is slecht.
  Jan is evil
c.

That we need to postulate a neutral zone is also clear from the fact that examples such as (62a) are not contradictory, but simply indicates that Janʼs goodness should be situated somewhere in the neutral zone. This is shown in (62b).

62
a. Jan is niet goed, maar ook niet slecht.
  Jan is not good  but  also  not  bad
  'Jan isnʼt good, but he isnʼt bad either.'
b.
[+]  2.  Scales that are bounded on one side

The scale of size in (63) implied by the measure adjectives klein'small' and groot'big' in (59b) is similar to the scale of “goodness" in most respects, but differs from it in that it is bounded on one side; the size of some entity cannot be smaller than zero. Observe that this implies that, unlike the scale of “goodness", the scale of size has a natural anchoring point. In this sense, adjectives like goed and slecht are more subjective than measure adjectives like klein and groot; see Subsection C below for more discussion.

63
Scale of size:
[+]  3.  Scales that are bounded on both sides

The implied scale can also be bounded on both sides. This is the case with the temporal scale implied by the adjectives vroeg'early' and laat'late' in (59c). When we assert that Jan is getting up early, that may be consistent with Jan getting up at 6:00 or 5:00 a.m., but presumably not with him getting up at 1:00 a.m. or at 11:00 p.m. Similarly, by asserting that Jan is getting up late, we may be saying that he is getting up at 11:00 a.m. or at 1:00 p.m., but presumably not that he is getting up at 11:00 p.m. or at 1:00 a.m. Beyond a certain point (which may be vaguely defined, and can perhaps be changed when the context provides information that favors that) the adjectives are simply no longer applicable (this is indicated by ### in (64)).

64
Temporal scale of vroeg and laat:
[+]  4.  Scales with one absolute and one gradable adjective

In the examples in the previous subsections, the two antonyms are both gradable. This need not be the case, however. The adjective gezond'healthy' in (59d), for instance, does not seem to be scalar itself; rather, it is absolute (see the discussion of (68)), and indicates one end of the scale. In other words, we may represent the scale of illness as in (65).

65
Scale of illness:

Many gradable adjectives that imply a scale that is bounded on one side are deverbal or pseudo-participles; cf. the primeless examples in (66) and (67). Their antonyms, which are situated at the boundary of the scale, are often morphologically derived by means of on- prefixation. In the case of the pseudo-participles occasionally no antonym exists, so that we can only express the negative counterpart by means of the negative adverb niet.

66
a. brandbaar 'combustible'
a'. onbrandbaar 'incombustible'
b. bereikbaar 'attainable'
b'. onbereikbaar 'unattainable'
c. begroeid 'overgrown'
c'. onbegroeid 'without plants'
d. toegankelijk 'accessible'
d'. ontoegankelijk 'inaccessible'
67
a. bekend met 'familiar with'
a'. onbekend met 'unfamiliar with'
b. bestand tegen 'resistant to'
b'. niet bestand tegen 'not resistant to'
c. gewond 'wounded'
c'. ongewond 'not wounded'
d. opgewassen tegen 'up to'
d'. niet opgewassen tegen 'not up to'
e. verwant aan 'related to'
e'. niet verwant aan 'not related to'

      That gezond and the adjectives in the primed examples in (66) and (67) are not scalar but absolute is clear from the fact that they can be modified by adverbial phrases like absoluut'absolutely', helemaal'completely' and vrijwel'almost', as in (68). We show these examples with topicalization of the AP in order to block the reading in which absoluut/vrijwel is interpreted as a sentence adverb. The examples are perhaps stylistically marked but at least the cases with absoluut become fully acceptable if we add the negative adverb niet'not' at the end of the clause.

68
a. Absoluut/vrijwel gezond is Jan.
  absolutely/almost  healthy  is Jan
b. Absoluut/vrijwel onbrandbaar is deze stof .
  absolutely/almost  incombustible  is this material
c. Helemaal/vrijwel onbekend met onze gewoontes is Jan.
  completely/almost  not.familiar  with our habits  is Jan

The examples in (69) show that these adverbial phrases cannot be combined with scalar adjectives; cf. Section 1.3.2.2, sub II.

69
a. * Absoluut/vrijwel goed/klein/ziek is Jan.
  absolutely/almost  good/small/ill  is Jan
b. * Absoluut/vrijwel brandbaar is deze stof.
  absolutely/almost  combustible  is this material
c. * Helemaal/vrijwel bekend met onze gewoontes is Jan.
  completely/almost  familiar with our habits  is Jan

      For completeness’ sake note that the adjective gezond'healthy' can also be used as a scalar adjective, provided that it is the antonym of ongezond'unhealthy'. In this use, gezond cannot be modified by the adverbial phrases absoluut and vrijwel. This is shown in (70).

70
* Absoluut/vrijwel gezond/ongezond is spinazie.
  absolutely/almost  healthy/unhealthy  is spinach
[+]  5.  Scales with two absolute adjectives

The fact that gezond (i.e., the antonym of ziek'ill') is not scalar shows that the placement of an antonym pair of adjectives on a scale is not sufficient to conclude that the adjectives are both scalar. In fact, they can both be absolute. This is the case with the adjectives leeg/vol'empty/full' in (59e); they both typically denote the boundaries of the implied scale. That leeg and vol are not scalar but absolute is clear from the fact that they can be modified by adverbial phrases like helemaal'totally', vrijwel'almost', etc.

71
a. Scale of “fullness"
b. Het glas is helemaal/vrijwel leeg/vol.
  the glass  is totally/almost  empty/full
[+]  B.  Context dependent adjectives—the neutral zone

In the scales in (60), (63) and (64), we have indicated a neutral zone to which neither of the two adjectives is applicable. This zone is often more or less fixed for the speaker in question. With some adjectives, however, the neutral zone is more flexible and may be determined by the entity the adjectives are predicated of, or the context in which the adjectives are used. This holds in particular for the measure adjectives, of which some examples are given in (72).

72
a. dik 'thick'
a'. dun 'thin'
b. oud 'old'
b'. jong 'young'
c. groot 'big'
c'. klein 'small'
d. lang 'tall/long'
d'. kort 'short/brief'
e. hoog 'high'
e'. laag 'low'
f. zwaar 'heavy'
f'. licht 'light'
g. breed 'wide'
g'. smal 'narrow'

That the placement of the neutral zone, that is, that the interpretation of the measure adjectives depends on the argument the adjective is predicated of can be demonstrated by means of the examples in (73a) and (73b). Below, we will discuss the examples with the adjective groot, but the discussion is also applicable to klein.

73
a. Deze muis is klein/groot.
  this mouse  is small/big
b. Deze olifant is klein/groot.
  the elephant  is small/big

      Although groot can be predicated of both the noun phrase deze muis'this mouse' and the noun phrase deze olifant'this elephant', it is clear that the two entities these noun phrases refer to cannot be assumed to be of a similar size: the mouse is considerably smaller than the elephant. This is due to the fact that the placement of the neutral zone on the implied scales of size differs. In the case of mice the scale will be expressed in term of centimeters, as in (74a), while in the case of elephants the scale will instead be expressed in meters, as in (74b).

74
a. Scale of size for mice in centimeters:
b. Scale of size for elephants in meters:

This shows that the placement of the neutral zone is at least partly determined by the argument the adjective is predicated of; it indicates the “normal" or “average" size of mice/elephants. In other words, examples such as (73) implicitly introduce a comparison class, namely the class of mice/elephants, which determines the precise position of the neutral zone on the implied scale. Often, a voor-PP can be used to make the comparison class explicit, and clarify the intended neutral zone, as in (75).

75
Jan is groot voor een jongen van zijn leeftijd.
  Jan is big  for a boy  of his age

The comparison class and, hence, the neutral zone are not fully determined by the argument the adjective is predicated of; the context may also play a role. If we are discussing mammals in general, the statement in (76a) is true while the statement in (76b) is false: the comparison class is constituted by mammals, and therefore the neutral zone is determined by the average size of mammals, and Indian Elephants are certainly bigger than that. However, if we discuss the different subspecies of elephants, the statement in (76a) is false while the statement in (76b) is true: the comparison class is constituted by elephants, and the Indian Elephant is small compared to the African Elephant.

76
a. De Indische Olifant is groot.
  the Indian Elephant  is big
b. De Indische Olifant is klein.
  the Indian Elephant  is small
[+]  C.  Subjective/objective adjectives

Although the placement of the neutral zone on the scale implied by the measure adjective depends on extra-linguistic information, the scale itself can be considered objective in the sense that once speakers have established the neutral zone, they can objectively establish whether a certain statement is true or false. The fact that the scale implied by the measure adjectives is objective is also supported by the fact that (in some cases) the precise position on the scale can be indicated by means of nominal measure phrases like twee dagen and twintig meter in (77).

77
a. Dit poesje is twee dagen oud.
  this kitten  is two days  old
b. De weg is twintig meter lang.
  the road  is twenty meters  long

      In the case of adjectives like lelijk/mooi'ugly/beautiful' and saai/boeiend'boring/exciting', on the other hand, establishing the precise position of the relevant entities on the implied scale is a more subjective matter; in fact, it can depend entirely on the language user, which can be emphasized by embedding the adjective under the verb vinden'consider', as in the (a)-examples in (78). Occasionally, the entity whose evaluation is assumed can be syntactically expressed by means of a voor-PP; some examples are given in the (b)-examples.

78
a. Ik vind De Nachtwacht lelijk/mooi.
  consider  The Night Watch  ugly/beautiful
a'. Ik vind Shakespeares dramaʼs saai/boeiend.
  consider  Shakespeareʼs tragedies  boring/exciting
b. Dit gereedschap is handig voor een timmerman.
  this tool  is handy  for a carpenter
  'These tools are handy for a carpenter.'
b'. Dit boek is interessant voor elke taalkundige.
  this book  is of.interest  to every linguist

      The pairs of measure adjectives in (72) can be considered true antonyms. This is clear from the fact that the two (a)-examples in (79) are fully equivalent. However, this equivalence does not seem to hold for the subjective adjectives in the (b)-examples, which suggests that the comparative forms mooier and lelijker are not true but quasi-antonyms.

79
a. Jan is groter dan Marie.
  Jan is bigger than Marie
a'. Marie is kleiner dan Jan.
  Marie is smaller than Jan
b. De Nachtwacht is mooier dan De anatomieles. ⇎
  The Night Watch  is more beautiful than  The Anatomy Lesson
b'. De anatomieles is lelijker dan De Nachtwacht.
  The Anatomy Lesson  is uglier than  The Night Watch

This difference may be related to the following observation. The use of the comparative form of objective adjectives like klein'small' and groot'big' in the (a)-examples of (79) does not necessarily imply that the argument the adjective is predicated of is actually small or big. The use of the comparative form of the subjective adjectives mooi'beautiful' and lelijk'ugly', on the other hand, at least strongly suggest that the argument the adjective is predicated of is indeed beautiful or ugly. This difference between objective and subjective adjectives may be lexically encoded; reasons for assuming this will be given in Subsection F below.
      For completeness' sake, it can be observed that the true antonym of mooier is the comparative form minder mooi'less beautiful', as is clear from the fact that the equivalency does hold between (80a) and (80b). The true antonymy relation of course also holds for groter'bigger' and minder groot'less big'.

80
a. De Nachtwacht is mooier dan De anatomieles. ⇔
  The Night Watch  is more beautiful than  The Anatomy Lesson 
b. De anatomieles is minder mooi dan De Nachtwacht.
  The Anatomy Lesson  is less beautiful  than  The Night Watch
[+]  D.  Measure adjectives—the (non)neutral form of antonymous adjectives

The examples in (77) have already shown that the measure adjectives can be modified by means of a nominal measure phrase. However, for each antonym pair in (72), only the adjective in the primeless example can be used. Some examples are given in (81). Observe that the acceptable example in (81a) does not express the fact that the kitten is old; on the contrary, it is quite young, which can be emphasized by using the evaluative particle pas'only'. Therefore, it is clear that the adjective oud has lost the antonymous part of its meaning. The same thing holds for the adjective lang in (81b). Since these adjectives have lost this part of their meaning, oud and lang can be considered as neutral forms of the relevant pairs; the adjectives jong and kort cannot be used in this neutral way.

81
a. Het poesje is (pas) twee dagen oud/%jong.
  the kitten  is only  two days  old/young
b. De weg is (maar) twintig meter lang/%kort.
  the road  is only  twenty meters  long/short

Similar conclusions can be drawn from the interrogative sentences in (82): the neutral form oud/lang gives rise to a perfectly natural question and does not presuppose that the subject of the clause should be characterized as being old/long, whereas the non-neutral form jong/kort gives rise to a marked result and seems to express the presupposition that the kitten is young/the road is short.

82
a. Hoe oud/%jong is het poesje?
  how old/young  is the kitten
b. Hoe lang/%kort is deze weg?
  how long/short  is this road

In this context it is also relevant to observe that only the neutral forms of the measure adjectives can be the input of the morphological rule that derives nouns from adjectives by suffixation with -te. The formation *oudte in (83c) is probably blocked by the existing noun leeftijd'age'. See Section 3.1.2, sub II, for more discussion of measure adjectives.

83
a. breedte 'width'
a'. *smalte
b. dikte 'thickness'
b'. *dunte
c. *oudte 'age'
c'. *jongte
d. lengte 'length'
d'. *kortte
e. hoogte 'height'
e'. #laagte
f. zwaarte 'weight'
f'. *lichtte

      The adjectives in (84) exhibit a behavior similar to the measure adjectives in (82): the primeless examples are unmarked, and do not presuppose that the property denoted by the adjective is applicable; the primed examples, on the other hand, are marked, and strongly suggest that the property denoted by the adjective is applicable.

84
a. Hoe schoon is de keuken?
  how clean  is the kitchen
a'. % Hoe vies is de keuken?
  how dirty  is the kitchen
b. Hoe veilig is die draaimolen?
  how safe  is that merry-go-round
b'. % Hoe onveilig is die draaimolen?
  how unsafe  is that merry-go-round
[+]  E.  Positively/negatively valued adjectives

Often, the subjective adjectives not only imply a subjective scale, but also express a negative or positive evaluation. Of the pair slecht'bad/evil' and goed'good', the first adjective clearly denotes a negatively valued property, whereas the latter denotes a positively valued property. The examples in (85) show that this distinction is also reflected in their modification possibilities: the primeless examples contain negatively valued adjectives, and modification by the elements knap'quite' and flink'quite' is possible; the primed examples, on the other hand, contain positively valued adjectives and modification by knap and flink is excluded.

85
a. knap brutaal/moeilijk/lastig/ongehoorzaam
  quite  cheeky/difficult/troublesome/disobedient
a'. * knap beleefd/makkelijk/eenvoudig/gehoorzaam
  pretty  polite/easy/simple/obedient
b. flink moeilijk/lastig/ongehoorzaam
  quite  difficult/troublesome/disobedient
b'. * flink makkelijk/eenvoudig/gehoorzaam
  quite  easy/simple/obedient

      The examples in (86) show that litotes (the trope in literary and formal language by which one emphasizes a property by means of the negation of its antonym) also requires an adjective denoting a negatively valued property; if the adjective denotes a positively valued property, as in the primed examples, the desired interpretation does not normally arise; a notable exception is colloquial Daʼs niet goed!, in which the deictic force of the demonstrative has bleached; cf. English Thatʼs not good! (Carole Boster, p.c.).

86
a. Dat boek is niet slecht.
  that book is not bad
  'That is very good.'
a'. # Dat boek is niet goed.
   that book  is not good
  Not: 'That is very bad.'
b. Hij is niet lelijk.
  he  is not ugly
  'Heʼs quite handsome.'
b'. # Hij is niet knap.
   he  is not handsome
  Not: 'Heʼs quite ugly.'

The modifier wel'rather', on the other hand, requires an adjective that denotes a positively valued property. This is illustrated in (87). Note that the primed examples are fully acceptable if wel is interpreted as the affirmative marker wel; the two forms differ in that the affirmative marker receives accent, whereas the modifier does not. These uses of niet and wel in (86) and (87) are more extensively discussed in Section 3.3, sub II.

87
a. Jan is wel aardig.
  Jan is wel  kind
  'Jan is rather kind.'
a'. * Jan is wel onaardig.
   Jan is wel  unkind
b. Hij is wel knap.
  he  is wel  handsome
  'Heʼs rather handsome.'
b'. * Hij is wel lelijk.
   he  is wel  ugly

      Occasionally, the modifier is sensitive both to the positive/negative value of the adjective and the syntactic environment. The modifier een beetje'a bit', for example, requires a negatively valued adjective in declarative clauses (or an adjective that does not have an antonym such as verliefd'in love'). In questions and imperatives, on the other hand, this modifier prefers an adjective that denotes a positively valued property.

88
a. Hij is een beetje onaardig/*?aardig.
  he  is  a bit  unkind/kind
b. Is hij een beetje aardig/?onaardig?
  is  he  a bit  kind/unkind
c. Wees een beetje aardig/#onaardig!
  be  a bit  kind/unkind
[+]  F.  Truly antonymous adjectives and the licensing of negative polarity items

This subsection discusses the fact that negative polarity items can be licensed by the subset of antonymous adjectives that were called true antonyms in subsection C above. In order to be able to do that we should first discuss certain logical properties of these adjectives. True antonyms have the defining property that they allow the inference in (89a), in which A and A' represent antonymous adjectives; cf, subsection C. In (89b), we repeat example (79a): if we say that Jan is bigger than Marie, we may conclude that Marie is smaller than Jan, and, similarly, if we claim that Marie is smaller than Jan, we may conclude that Jan is bigger than Marie. This equivalency does not hold for quasi-antonymous adjectives like mooi'beautiful' and lelijk'ugly'; see example (79b) for discussion.

89
True antonyms
a. x is more A than y ⇔ y is more A' than x
b. Jan is groter dan Marie. ⇔ Marie is kleiner dan Jan.
  Jan is bigger than Marie Marie is smaller than Jan

True and quasi-antonymous adjectives are similar in that the implications in (90a) do not hold for either, which is due to the fact that in both cases the implied scale may have a neutral zone. This was already discussed for the quasi-antonymous adjectives slecht'bad' and goed'good' in Subsection A, so we confine ourselves here to giving similar examples for the true antonymous adjectives groot'big' and klein'small'.

90
a. not A ⇏A'; not A' ⇏A
b. Jan is niet groot. ⇏ Jan is klein.
  Jan is not big  Jan is small
b'. Jan is niet klein. ⇏ Jan is groot.
  Jan is not good  Jan is evil

Despite the fact that the implications in (90a) do not hold, we will show in this subsection that for the true antonyms above, the pair not A and A'as well as the pair not A' and A do exhibit certain similarities in semantic behavior, which may be relevant when it comes to the licensing of negative polarity items like ook maar iets'anything'. In order to demonstrate this, we have selected the adjectives in (91). These adjectives were chosen because they may take a clausal complement, which is crucial for our purposes below because negative polarity items like ook maar iets are normally only possible in embedded clauses.

91
a. gemakkelijk 'easy'
a'. moeilijk 'difficult'
b. verstandig 'clever'
b'. onverstandig 'foolish'
c. veilig 'safe'
c'. gevaarlijk 'dangerous'
[+]  1.  Upward and downward entailments

Consider the examples in (92). In the primeless examples, the complement clause refers to a wider set of events than the complement in the primed examples; the addition of an adverb in the latter cases makes the event the complement clause refers to more specific, and hence applicable to a smaller number of situations. For example, there are many occasions in which a problem is solved, but only in a subset of those occasions is the problem solved fast.

92
a. Het is gemakkelijk om dat probleem op te lossen.
  it  is easy  comp  that problem  prt.  to solve
  'Itʼs easy to solve that problem.'
a'. Het is gemakkelijk om dat probleem snel op te lossen.
  it  is easy  comp  that problem  quickly  prt.  to solve
  'Itʼs easy to solve that problem fast.'
b. Het is verstandig om een boek voor Peter te kopen.
  it  is clever  comp  a book  for Peter  to buy
  'Itʼs clever to buy a book for Peter.'
b'. Het is verstandig om hier een boek voor Peter te kopen.
  it  is clever  comp  here  a book  for Peter  to buy
  'Itʼs clever to buy a book for Peter here.'
c. Het is veilig om hier over te steken.
  it  is safe  comp  here  prt.  to cross
  'Itʼs safe to cross the road here.'
c'. Het is veilig om hier met je ogen dicht over te steken.