In the komen + adjective constructions in (90), the adjective natwet is predicated of the noun phrases jijyou and de badkamerthe bathroom, respectively.

The komen + adjective construction is very limited: it only occurs in interrogative clauses such as (90), in which a cause is questioned, and in yes-no questions such as (91), which contain a causative door-phrase.

The examples in (92) show that the interrogative komen + adjective construction is also possible in embedded clauses and in yes-no questions with a verb-second order. Note that the interrogative sense of (92b) is triggered by the adverb immers/toch; the negative element nietnot must be present.

Example (93a) shows that the komen + adjective construction is excluded in declaratives, although somehow it becomes perfectly acceptable in syntactic frame (93a'). The (b)-examples show that the construction is also excluded in interrogatives that question something other than a cause.

The deictic element zo is usually required in the interrogative komen + adjective construction. It often modifies the complementive adjective, as in (94), but it can also modify some other element, as in (95), where zo modifies the adverbial modifier plotselingsuddenly.

It seems possible to save the examples in (94a&b) without zo by adding a discourse particle such as nou, which is used to express astonishment. This is illustrated in (96); to our knowledge, examples like these deserve careful attention.

The fact that a causative phrase (e.g. a door-PP) must be present in the interrogative komen + adjective construction suggests that this construction is related to the declarative construction in (97); these also have an obligatory causative phrase, but involve a finite clause rather than a complementive adjective.

Further evidence for the claim that the two uses of komen are related is found in the fact that, in both constructions, komen can only be combined with epistemic modal verbs: deontic modals, such as willen with the meaning “to want”, are excluded.

That the two uses of komen are related is also clear from the fact that the examples in (90) and (91) are near-synonymous with those in (99) and (100). Note that in (97) to (100) hetit is an anticipatory pronoun introducing the embedded finite clause. This is clear from the fact that het must be dropped when the finite clause is preposed: compare (97a) with Dat jij zo nat bent, komt door de regen.

The (a)-examples in (101) show, however, that the komen + adjective and the komen + clause constructions differ in that only the latter is compatible with sentence negation. The (b)-examples show that, if the negation has a more limited scope, as in the case of adjectives prefixed with on–, both constructions are equally good.

Paardekooper (1986) suggested that the examples in (90) and (91) are “derived from” the infinitival counterparts of the examples in (99) and (100) by replacing the infinitival copula te zijnto be with a “null sign”. Although this suggestion may be on the right track, it should be noted that, contrary to what Paardekooper claims, the overt realization of te zijn does not lead to a very felicitous result according to most speakers of Dutch.

The modification possibilities of the adjective in the interrogative komen + adjective construction depend on what the element zo modifies. If it modifies a constituent unrelated to the complementive adjective, as in (103), the possibilities are rather limited: (103a) shows that amplifiers like ergvery or verschrikkelijkterribly and downtoners such as vrijrather are then impossible. The examples in (103b&c) show that comparative/superlative forms and adjectives preceded by tetoo are also degraded.

The constructions in (104) with a declarative complement clause, on the other hand, are less deviant, or even perfectly acceptable.

If the element zo modifies the modifier of the adjective, as in (105a), amplifiers like ergvery or verschrikkelijkterribly are possible in the interrogative komen + adjective construction; downtoners such as vrijrather, on the other hand, remain unacceptable. The examples in (105b&c) show that comparatives and adjectives modified by tetoo can be used; this does not hold for superlatives, which is to be expected because they never combine with degree modifiers.

The corresponding constructions with a declarative complement clause are also acceptable, as shown in (106). The fact that zo vrij klein is again unacceptable in (106a) shows that its unacceptability is independent of the interrogative komen + adjective construction.

Finally, note that the complex modifier zo A mogelijkas A as possible in (107), in which zo is used non-deictically, is excluded. Note that speakers who use komen as a copular verb comparable to wordento be accept (107a) with an inchoative meaning: “how can you become as small as possible”; this is irrelevant for our present discussion.

Adjectives that take a non-dative nominal complement, such as zatweary in (109), do not occur in attributive position; cf. Section 27.3, sub IA.

Adjectives such as gek in the sense of "fond" must be followed by their PP-complement; cf. the contrast between (110a) and (110a'). Such adjectives cannot occur in attributive position due to the head-final filter on attributive adjectives; cf. Section 27.3, sub IB.

Example (111) provides some examples of adjectives that can only occur in combination with a (pseudo-)copular verb.

The primed examples in (112) show that such adjectives cannot be used attributively. The doubly-primed examples show that these examples become fully acceptable when the verb appears as an attributively used present participle.

The examples in (113) also involve copular constructions, but are special in that the adjective seems to take a nominal complement.

The fact that these adjectives can only be used predicatively may therefore follow from the general restriction discussed in Subsection A that adjectives with a non-dative complement cannot appear in attributive position. Note that the doubly-primed examples, involving the present participles of the verbs in (113), are again fully acceptable.

For completeness’ sake, example (115a) shows that the adjective kwijt can also be used without a nominal complement; here the nominal complement of kwijt in (114c) appears (surprisingly) as the logical subject of kwijt. Example (115b) further shows that kwijt cannot be used in attributive position in this case either.

The examples in (116) are comparable to resultative and vinden-constructions. Note that the adjectives in (116b&d) can also occur in a copular construction; cf. (111a&d).

That the adjectives in these fixed combinations cannot be used attributively is demonstrated in the primed examples in (117). The acceptable cases in the doubly-primed examples again involve an attributively used present participle.

Some adjectives can be modified by a van-PP containing a bare noun. There are at least two types, which are illustrated in (118). The van-PP in (118a) expresses a restriction on the adjective “big as far as stature is concerned”, and the van-PP in (118b) indicates the cause of the occurrence of the property denoted by the adjective “red caused by excitement”. The primed examples show that these two constructions cannot be used attributively.

However, the two constructions differ in that the former denotes an individual-level (i.e. permanent) property, whereas the latter denotes a (i.e. transitory) property. Therefore, the latter, but not the former, can also be used as a supplementive. The two A + van + N sequences in (118) are discussed in more detail in Section 25.5, sub I.

There are a number of isolated cases of adjectives that can only be used in complementive position: alleenalone, andersdifferent, bekaf/doodopdone in, klaarready, onwelill, and wegaway.

Finally, note the remarkable contrast between (121a) and (121b), which differ only in that in (121b) klaarready is part of the compound kant-en-klaarinstant.