• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
2.2.4.The structure of complementive constructions
quickinfo

The question as to what structure should be assigned to examples containing a complementive has given rise to a long and still unresolved debate. According to some, the subject is part of a projection headed by the complementive, often called a small clause: it occupies a designated subject position in which it saturates the thematic role assigned by the predicate, as in (302a). According to others, the subject is generated in the regular object position of the verb, in (302b), the subject-predicate relation being established by other means, here indicated by the use of subscripts.

302
a. .... [VP ... V [SC DP [Pred]]]
b. .... [VP ... V DPi Predi]

The main difference between the two proposals is that in (302a) the noun phrase and the complementive form a constituent, whereas in (302b) they do not. An argument for structure (302a) is that the complementive and its subject do behave like a constituent when it comes to coordination, as shown in (303a). An argument for structure (302b) is that the noun phrase and the complementive need not be adjacent. Many proposals have been made to solve these problems. For example, proponents of the small-clause approach may explain an example such as (303b) by referring to the independently established fact that noun phrases can be scrambled in Dutch, while proponents of the alternative approach may claim that examples such as (303a) involve the coordination of a verbal projection smaller than VP.

303
a. Jan vindt [[SC Marie aardig] maar [SC Els een smeerlap]].
  Jan considers Marie nice but Els an asshole
b. Jan vindt Marie waarschijnlijk niet aardig.
  Jan consider Marie probably not nice
  'Probably, Jan will not consider Marie nice.'

A very compelling argument for the small-clause approach is that the noun phrase and complementive can be pronominalized together, as shown in example (304a), in which the pronoun dat replaces the whole small clause, i.e. both the predicate erg goed and its logical subject dat boek. For the sake of completeness, we have added the examples in (304b&c), which show that the noun phrase can also be pronominalized in isolation, and that the same holds for complementives in (at least) copular constructions.

304
a. Jan vindt [SC dat boek erg goed]i maar Peter vindt dat niet.
  Jan considers that book very good but Peter considers that not
  'Jan considers that movie very good, but Peter does not.'
b. Jan vindt het erg goed.
  Jan considers it very good
c. Jan is erg aardig maar Els is dat ook.
  Jan is very nice but Els is that too

Another possible argument for the small-clause approach is that in copular constructions such as (305) the nominative subject of the clause can follow the object pronoun hemhim. If we assume that the nominative phrase is base-generated within the small clause as the subject of the adjective bekend, this will follow from the fact that nominative subjects are only optionally moved into the subject position of the clause; cf. Section 13.2 and N21.1.2.

305
a. dat hem [SC die problemen bekend] zijn.
  that him those problems known are
  'that he is aware of those problems.'
b. dat die problemeni hem [SC ti bekend] zijn.
  that those problems him known are

A perhaps even more convincing argument is that the nominative subject may also follow the pronoun onsus in example (306a). The fact that the subject may follow this pronoun strongly suggests that it must be generated within the AP, given that the pronoun is selected by the modifier tetoo of the adjective; an example such as *dat ons die auto duur is shows that the pronoun cannot be present if the modifier is omitted; cf. section A24.2, sub I, for a detailed discussion.

306
a. dat ons die auto te duur is.
  that us that car too expensive is
  'that that car is too expensive for us.'
b. dat die autoi ons ti te duur is.
  that that car us too expensive is

We will adopt the small-clause approach here for the reasons discussed above, but remain somewhat agnostic about the internal structure of small clauses; this issue is still open, although the debate on the two structures in (302) has ultimately led to proposals that potentially reconcile and at least combine a number of advantages of the two competing ideas. These alternative proposals postulate some functional head between the DP and the predicate that expresses the predicative relation between the two; cf. Bowers (1993), Hale & Keyser (1993), and Den Dikken (2006) for different proposals. If we call this functional head pred, the structure of a small clause can be represented as follows: [PredP DP [Pred [XP ... X ...]]], where X stands for N, A, or P.

readmore
References:
    report errorprintcite