Since the copular verb zijnto be is homophonous with the passive auxiliary zijnto have been as well as with the auxiliary of time zijn selected by some unaccusative verbs, copular constructions with adjectival past/participle participles are sometimes difficult to distinguish from perfect-tense and passive constructions. The following subsections consider some tests that can be used to distinguish between them.

Section 31.2.2 has shown that past participles of nom-dat verbs can be used attributively to modify a head noun corresponding to the theme-subject, provided that the verb takes the auxiliary zijn in the perfect tense. This is shown again in (119).

Since the past participle ontbroken cannot be used attributively, it is not surprising that it cannot be used predicatively; cf. (120b). However, example (120a) shows that the past participle of the nom-dat verb opvallento strike/attract attention cannot be used predicatively either. This is consistent with the conclusion drawn in 31.2.2, sub II, that past participles of nom-dat verbs like opvallen do not have a truly adjectival interpretation; cf. (64).

For completeness’ sake, note that (119a) is not ambiguous between the perfect-tense and the copular construction, since truly adjectival participles generally do not allow nominal arguments (here the pronoun onsus). That (119a) cannot be a copular construction is also illustrated in (121): it only allows the addition of adverbs such as gisterenyesterday, which refer to a specific point on the time axis; cf. Table 6.

As also shown in Section 31.2.2, past participles of object experiencer psych-verbs can be used attributively if the modified noun corresponds to the [+human] object of the active verb. This is illustrated once more in (122).

The examples in (123) demonstrate that the past participles can also be used predicatively; in such cases there is no confusion with perfect-tense constructions, since these psych-verbs select the auxiliary hebbento have. Observe that the truly adjectival status of the participles is also evident from the fact that they can be modified by degree modifiers such as heel/zeervery.

This subsection has shown that the complementive use of past/passive participles is more restricted than their attributive use: it is only possible if the participle is truly adjectival, i.e. with a subset of transitive and monadic unaccusative verbs and object experiencer verbs; cf. Table 5.

Present participles of intransitive and transitive verbs cannot be used in copular constructions. This has already been seen in (83a&b), and some more examples are given in (124) and (125). The unacceptability of the predicative constructions in the primed examples contrasts sharply with the acceptability of the corresponding attributive constructions: cf. het vloekende/werkende meisjethe cursing/working girl and het zingende/etende meisjethe singing/eating girl.

However, the examples in (126) show that there are many metaphorically used present participles that can be used not only attributively but also predicatively. Since the meanings of these forms are highly specialized, we may be dealing with true adjectives. We added the (c) and (d)-examples to show that the corresponding non-metaphorically used present participles yield unacceptable results.

Example (83c) has already shown that present participles of unaccusative verbs cannot usually be used in copular constructions; further examples are given in (127). The unacceptability of the predicative constructions in the primed examples again contrasts sharply with the acceptability of the corresponding attributive constructions de vertrekkende gastenthe leaving guests and de vallende jongenthe falling boy.

The primed examples in (128) provide some possible counterexamples to the claim that the complementive use of present participles of unaccusative verbs is excluded.

However, it is not clear whether these are really copular constructions. First, the present participles in the singly-primed examples have an -e ending, which is normally not possible with predicatively used adjectives. Second, they seem to refer to ongoing processes, just like the attributively used verbal present participles in the doubly-primed examples: the subject of the clause is said to be undergoing a change of state. This is clear from the fact that the primed examples can be paraphrased by a durative aan het + infinitive construction; example (128b'), for instance, is virtually synonymous with Het schip is aan het zinkenThe ship is sinking.

Other potentially problematic cases arise with motion verbs like lopento walk, bussento travel by bus and liftento hitchhike, which can be used either as intransitive or as unaccusative verbs; cf. Section V2.1.2, sub I. The unaccusative forms of these verbs in (129b) require the presence of a directional PP and select the perfect auxiliary zijnto be. The forms of these verbs in (129a) behave simply as intransitive verbs; they occur without a predicative complement and select the perfect auxiliary hebbento have. The examples in (129c) show that the present participle of these motion verbs can be used as a predicate in copular constructions.

129

Motion verbs

a.		De jongen	heeft	gelopen/gebust/gelift.	intransitive verb
		the boy	has	walked/bused/hitchhiked

b.		De jongen	is naar Amsterdam	gelopen/gebust/gelift.	unaccusative verb
		the boy	is to Amsterdam	walked/bused/hitchhiked

c.		De jongen	is/bleek	(*naar Amsterdam)	lopend/bussend/liftend.
		the boy	is/ turned.out	to Amsterdam	walking/busing/hitchhiking
		'The boy has/appeared to have come on foot/by bus/hitchhiking.'

The examples in (129c) differ from the singly-primed examples in (128) in that they do not refer to an ongoing event. A sentence like Ik ben lopend does not imply that the speaker is walking at the moment of utterance, but merely expresses that he has come on foot. This suggests that the present participles in (129c) are truly adjectival, which seems to be supported by the fact that the predicative PP-complement naar Amsterdam cannot be used, in contrast to what is the case with the attributively used verbal participles in de naar Amsterdam lopende jongenthe boy who is walking to Amsterdam.

Nom-dat verbs usually resist the formation of truly adjectival present participles, regardless of whether they select the auxiliary zijn or hebben in the perfect tense; cf. (130). Section 31.2.2, sub IIB, has already shown that the present participle of the verb opvallento strike/attract attention is an exception in this respect.

This leaves us with the present participles of the object experiencer psych-verbs. The primed examples in (131) show that these can be used comfortably as predicates in copular constructions. The present participles in the primed examples can be modified by degree modifiers like zeer/heelvery, which confirms that we are dealing with truly adjectival participles in these examples.

The examples in (132) show that the result is occasionally unacceptable. When it is, the intended assertion can usually be expressed by a genuine adjective, suggesting that the adjectival use of the present participle is blocked by the availability of this alternative.

This morphological “blocking” approach to the unacceptable versions of the primed examples in (132) can be supported by the fact that the present participles cannot be used attributively on their adjectival/state reading either; this reading can only be expressed by the genuine adjectives. Note in passing that the primeless examples in (133) show that the attributive use of the present participles on their verbal/eventive reading leads to different degrees of acceptability.

The fact that the present participles in (132) cannot be used as complementives thus follows from the fact, illustrated in (133), that they are always verbal, if at all possible.

Table 7 summarizes the tendencies observed in the previous subsections. Occasionally, participles appear in copular constructions that are not expected on the basis of these tendencies, but these exceptions are mostly idiosyncratic in nature.

Apart from their attributive use, modal infinitives can also be used as predicates in copular and vinden-constructions. Predicatively used te-infinitives differ from attributively used ones, however, in that they only allow the ability reading; e.g. example (134a) does not easily allow an interpretation according to which the books must be read: the only readily available reading is that the books are easily accessible.

The adjective gemakkelijk acts as an adverb modifying the te-infinitive and not as a predicative complement of the verb zijn/lijken. Since adverbs are not morphologically distinguished from other adjectives in Dutch, the examples in (134) are easily confused with the easy-to-please construction, which does involve a predicatively used adjective. Fortunately, there are several criteria for distinguishing the two constructions: (i) the predicatively used adjective is obligatory in the easy-to-please construction, whereas the adverbially used adjective can be dropped in the case of modal infinitives; (ii) the infinitival clause has an obligatory complementizer om in the easy-to-please construction, whereas this complementizer cannot co-occur with modal infinitives; (iii) in attributive constructions, the infinitival clause follows the modified noun, whereas the modal infinitive precedes it. These tests are discussed in more detail in Section 28.5, sub IV.

Section 31.2.3 has shown that the noun modified by attributively used modal infinitives corresponds to the accusative object in the matching active sentence. Something similar holds for predicatively used modal infinitives: the noun phrase they are predicated of also corresponds to the accusative object in the matching active sentence. This is clear from the contrast between the examples in (134) and (135): the modal infinitives in (134) are transitive and the result is fine, while the modal infinitives in (135) are intransitive and unaccusative, respectively, and the result is unacceptable.

However, there are some problems that should be taken into account. First, it seems that in some cases the predicative use of transitive modal infinitives can be blocked by a competing construction, as illustrated in (136): we would expect the complementive modal infinitives te kopen/huren to be able to appear in these copular constructions with an ability reading, but this seems to be hampered by the possibility of using te koop/huurfor sale/rent.

Second, examples like (137) have been put forward as counterexamples to the claim that intransitive verbs cannot act as modal infinitives. The fact that these examples have an ability reading makes it plausible that we are indeed dealing with modal infinitives. The difference between the examples in (135) and (137) is not clear to us. For the moment, we can only observe that the examples in (137) are special in that there is a certain preference for using the verb vallen instead of zijn, and that some adverbial or quantified phrase like niet or veel must be present.

We have seen in Section 31.2.3 that arguments and predicative complements of the verb can only be expressed in the attributive construction when the modal infinitive has a verbal (i.e. obligation) reading; cf. (99). The fact that arguments and predicative complements cannot occur if the modal infinitives are used as complementives thus supports our claim that they are always truly adjectival in this syntactic function.

A potential problem for this claim, however, is that complementive modal infinitives can be combined with the predicative parts of collocations like schoon makento clean or kwaad/bang makento anger/frighten. This is shown in (139).

A similar argument can be construed for the claim that predicatively used modal infinitives are always truly adjectival: this is based on the fact that the addition of a (passive) door-phrase leads to a marginal result, whereas the addition of a voor-phrase is fully acceptable when the adverb gemakkelijk is present. The percentage sign in (140a) indicates that our judgment is controversial, since similar examples have been given as acceptable in the literature. However, speakers who accept the door-phrase in (140a) report that the sentence has the ability reading, suggesting that the modal infinitive here is adjectival, not verbal.

Another argument for assuming the nonverbal status of predicatively used modal infinitives is that for many (but not all) speakers they must precede the finite verb in clause-final position.

The fact that some speakers seem to allow the modal infinitives to follow the copular verbs in (141) is not conclusive for arguing that they are verbal, since postverbal placement of the modal infinitive is excluded for all speakers if there is more than one verb in clause-final position, as in (142).

On the other hand, the fact that modal infinitives can precede the verbs in clause-final position shows that they are not verbal; the examples in (143) show that unequivocally verbal te-infinitives never occupy this preverbal position.

We conclude, therefore, that complementive modal infinitives are truly adjectival, and hypothesize that their postverbal placement in the examples in (141) results from a performance error comparable to that found with complementive past participles; cf. the discussion of (112c) in Subsection I.

Before concluding the discussion, we would like to point out a potential problem for our earlier claim that modal infinitives cannot be used in predicative position with an obligation reading. Consider the examples in (144), which allow for an obligation reading.

The fact that the infinitives precede the clause-final finite verb indicates that they are not verbal. This raises the question as to whether we are dealing with predicatively used modal infinitives in this construction. An affirmative answer to this question is suggested by the fact that the verb hebben can also be used in other predicative constructions, such as (145); note that hebben can be replaced by krijgento get, an option that is also available for the examples in (144).

To our knowledge, the question of whether we are dealing with predicatively used modal infinitives in (144) has not yet been investigated. There are two possible arguments against the assumption that we are dealing with modal infinitives in these constructions. The first argument, illustrated in (146), is that the te-infinitive can be predicated of a subject of a transitive verb in the hebben construction, while we have seen in (134) that modal infinitives are usually predicated of direct objects of transitive verbs in the copular construction.

The second argument, illustrated in (147), is that the hebben construction occurs with intransitive and unaccusative verbs, whereas attributively or predicatively used modal infinitives of these verbs normally do not occur; cf. (93) and (135).

We leave it open here whether these two arguments are sufficient to refute the claim that we are dealing with modal infinitives in (144); the examples in (146) and (147) may be of a different nature than those in (144), given that hebben can only be replaced by the semi-copular krijgen in the first set of examples.