This section on finite prepositional object (PO-)clauses is relatively short because many of their properties, and of the anticipatory pronominal PPs that introduce them, are already discussed in Section 2.3. PO-clauses never have the form of main clauses and come in two types: declarative clauses introduced by the complementizer datthat or interrogative clauses introduced by the complementizer ofif/whether or some wh-phrase. Some examples are given in (246). Whether a declarative or an interrogative clause is used depends on the verb: the verb klagen (over)to complain (about) in (246a) selects a declarative clause, while the verb twijfelen (over) in the (b)-examples selects an interrogative clause.

The examples in (246) show that clause-final PO-clauses can be introduced by an anticipatory pronominal PP in the middle field of the clause. Depending on the verb, this PP can be optional or obligatory. The former holds for the verbs in (246) and the latter for the verbs in (247). An extensive sample of PO-verbs that may or may not omit the anticipatory pronominal PP can be found in Section 2.3.1, sub VI.

PO-clauses can be left-dislocated, in which case the anticipatory pronoun is replaced by a resumptive pronominal PP in the form daar + P. This is illustrated in (248) for the examples in (246); the (discontinuous) resumptive PP is italicized.

Although some speakers seem to allow the omission of the pronominal part of the resumptive PP, most do not. This is indicated by a percentage sign in (249), where pro stands for the empty/deleted resumptive pronominal part.

Note in passing that the fact that most speakers do not accept examples such as (249) might pose a problem for Koster’s proposal, discussed in Section 5.1.2.2, sub III, according to which sentence-initial object and subject clauses are left-dislocated and the first position of the sentence is filled by an empty pronominal element pro. If the resumptive pronoun can be phonetically empty in the case of subject and object clauses, why should this be excluded for most speakers in the case of PO-clauses? Note also that the examples in (249) do not improve when the prepositional part of the resumptive pronominal PP is omitted. Although some speakers might marginally accept examples such as (250), they are clearly degraded compared to the examples without an anticipatory PP in (246), which are perfectly grammatical.

PO-clauses cannot be placed in the middle field of the clause, regardless of whether an anticipatory PP is present or not. PO-clauses also do not normally occur as part of the PP-complement of the verb; examples such as (251) are marked compared to examples such as (246), which is indicated here by a question mark (although Section P33.4.1, sub I, discusses a number of exceptional circumstances that do seem to license PPs of the type in (251)).

Finally, we note that there are als-clauses which can easily be misinterpreted as PO-clauses; for a concrete case we refer the reader to Paardekooper (1986:1.18.9, B2), but we will use example (252a) here for the sake of presentation. Example (252b) shows that the two cases differ in that the als-clause, but not the dat-clause, can be followed by danthen when the als-clause occurs in sentence-initial position, suggesting that we are dealing with a conditional adverbial clause. This proposal is supported by the fact that there is a sharp difference between the two variants of example (252c) in which the clauses appear as part of the PP-complement: whereas the dat-clause gives rise to a marked but interpretable result, the als-clause gives rise to an unacceptable and uninterpretable result.

A final argument for assuming that the als-clause is a conditional adverbial clause is that it can occupy the main-clause initial position while the anticipatory pronominal PP erover is present; if the als-clause were a PO-clause, we would end up with two prepositional objects within a single clause. Analyzing the als-clause as a conditional adverbial phrase, on the other hand, is unproblematic, because we can then give (253a) an analysis similar to (253b).

For further discussion of the erroneous analysis of conditional als-clauses as argument clauses, we refer the reader to Sections 5.1.2.1, sub VI, and 5.1.3, sub I, for similar cases in the domain of object and subject clauses.