Although case is not morphologically realized in Dutch, we can divide the set of adjectives that take a nominal complement into three groups on the basis of the case that would be assigned to the nominal complement in German. The German counterparts of the adjectives in A and B of Table 5 take genitive and dative complements, respectively, and the German counterpart of the adjective waard worth in C takes an accusative complement. The ability of the adjectives in Table 5 to take nominal complements should be seen as a relic of earlier stages, as evidenced by the fact that some of the adjectives in Table 5 are obsolete, e.g. deelachtig, indachtig and beschoren.

case	example	translation
A. Genitive	zich iets bewust zijn	to be conscious of something
	iets deelachtig worden	to acquire something
	iets gewend/gewoon zijn	to be accustomed to something
	iets indachtig zijn	to be mindful of something
	iets moe/zat/beu zijn	to be weary of something
	iets machtig zijn	to be in command of something
B. Dative	aangeboren	innate
	beschoren	given
	bespaard	spared
	duidelijk/helder	clear
	gehoorzaam	obedient
	goedgezind	well-disposed
	(on)bekend	(un)known
	toegewijd/toegedaan	devoted
	trouw	loyal
	vertrouwd/vreemd	familiar/foreign
C. Accusative	waard	worth

The case distinctions in Table 5 seem to correlate with the semantic type of the complement: dative complements are usually [+human], or at least [+animate], while the genitive and accusative complements of adjectives can also be [-animate]. This is illustrated in (26); the nominal complements are in italics.

The adjectives in Table 5A&B can easily be used in copular constructions, as shown in the primeless examples of (27) and (28). We have also given the German translations in the primed examples; the case of the nominal complement is indicated by a subscript.

The nominal complements differ from the PP-complements discussed in section 24.1 in that they cannot follow the adjective. The (a) and (b)-examples in (29) illustrate this for the Dutch examples in (27) and (28), respectively.

The examples in (30a&b) show that the adjectives in Table 5A&B are not common in resultative constructions. An apparent counterexample is (30c), where bekend maken (usually written as a single word) seems to function as a fixed collocation with the meaning “to announce”.

Adjectives with a genitive or dative complement differ in that only the former can be used marginally in the achten/vinden-construction; this is shown by the examples in (31a&b). However, if the dative noun phrase in (31b) is omitted, the construction becomes fully acceptable, although it is not immediately clear why omitting the complement should have this effect; cf. the discussion of (35) for a suggestion that may be helpful in this regard.

An example with the adjective waard from Table 5C is given in (32a). This adjective also appears in the syntactic frame in (32b), where the pronoun mij is in the dative and the noun phrase het geld is in the accusative in the German translation.

The genitive arguments of the adjectives in Table 5A are usually obligatory, although they can occasionally be replaced by a van-PP, as in (33). The use of the number sign in (33b) indicates that omitting the genitive argument is possible, but will have a dramatic effect on the meaning of the sentence: Jan is moe/zatJan is tired/drunk.

It seems that the adjectives in Table 5B can be divided into two subclasses in this respect. The adjectives beschoren, bespaard, goedgezind and toegewijd/toegedaan require the dative argument to be present, whereas aangeboren, duidelijk, gehoorzaam, helder, (on)bekend, trouw, vertrouwd and vreemd can be used without it. This contrast is illustrated in (34).

However, omitting the dative argument from the second subclass sometimes affects the meaning/denotation of the adjective. This is illustrated in (35). This raises the question whether our claim that the dative argument is optional with the second subclass of adjectives is factually correct; it may be that we are simply dealing with two homophonous forms.

In some cases, the dative noun phrases can be replaced by a voor or aan-PP. The former is possible, though not preferred, with the adjectives duidelijk and helder in the (a)-examples in (36), and the latter is an equally acceptable option with the adjectives gehoorzaam and trouw in the (b)-examples.

The genitive complement of the adjectives in Table 5A always follows the subject of the clause, as illustrated in (37) by the contrast between the primeless and primed examples.

However, the adjectives in Table 5B can be divided into two subclasses in this respect. The adjectives gehoorzaam and trouw in (38a) require the dative object to follow the subject of the clause; the order in example (38b) forces the reading that it is Jan who is loyal/obedient to the dog and is consequently degraded with the object pronoun hem.

The adjectives in (39), on the other hand, allow the dative object to precede the subject.

Because we find a similar kind of word-order variation with the subject of the nom-dat dyadic unaccusative (nom-dat) verbs discussed in Section V2.1.3, it has been suggested that the subject of the clause is actually an internal argument of the adjective; cf. Cinque (1990) and especially Bennis (2004); this claim will be discussed in detail in Section 28.5.1.

The genitive/dative noun phrase is a complement of the adjective (and not of the verb). One fact that supports this claim is that in German, the case assigned to the noun phrase is determined by the adjective. Furthermore, if the noun phrase can be replaced by a complement PP, then the combination A + PP can be placed in sentence-initial position and must therefore be a constituent; cf. the constituency test.

Given that the combination A + PP forms a constituent, it seems relatively safe to assume that the same is true for the combination NP + A, although it should be noted that there is a potential problem: the primeless examples in (41) show that at least for some speakers, topicalization of the combination NP + A leads to a degraded result, and all speakers clearly prefer the split patterns in the primed examples. We return to these “split” patterns in Section 24.3, while noting that German speakers seem to have much less trouble with the German counterparts of the primeless examples in (41).

Although the set of adjectives that take a nominal complement is quite small in Dutch, it should be noted that virtually any gradable adjective can be combined with a dative object if it is modified by the (evaluative) degree modifier tetoo; cf. Section 25.1.2, sub II. The examples in (42) show that in the resulting structure the dative object refers to the person whose evaluation is given. Note also that the subject of the clause can always follow the dative object; cf. the discussion of examples (38b&c) in Subsection C.

The examples in (43) show that similar facts can sometimes be observed in the case of the modifier genoegenough; cf. Section 25.1.3, sub IV.

The primeless examples in (44) show that the dative phrases in (42) and (43) alternate with voor-PPs. The primed examples show that the voor-PP can be easily pied-piped by topicalization of an AP modified by tetoo, but that this is more difficult in the case of the modifier genoeg. The fact that (44b') is also marked without the PP suggests that this is not due to problems with pied piping of the PP, but to problems with topicalization of the modified adjective.

Besides the examples in Table 5, there are some more or less fixed expressions in which the categorial status of the predicative element is not clear; examples such as (45) are often considered to be lexicalized phrasal verbs.