This section discusses clauses in which the clause-initial position in representation (113) is filled by some constituent. This results in verb-second clauses (i.e. main clauses with the finite verb in second position) or embedded clauses with some constituent preceding the complementizer (which is then usually phonetically empty). For a more detailed discussion of representation (113), see Section 9.1.

In main clauses, the unmarked filler of the clause-initial position is the subject of the clause, as in (114a), but there are several semantically marked constructions in which some other constituent precedes the finite verb: examples are wh-questions such as (114b), topicalization constructions such as (114c), and exclamative constructions such as (114d). The use of traces in (114b-d) reflects the traditional hypothesis in generative grammar that these examples are derived by movement; traces indicate the base position of the moved phrases (here: the object position in the middle field of the clause). Alternative hypotheses exist, but are not discussed here.

The examples in (115) show that the clause-initial position of declarative embedded clauses usually remains empty; the complementizer precedes the subject and topicalization is excluded. Note that the complementizer in (115b&c) is given in brackets because this element is normally not phonetically realized in standard Dutch when the first position is filled by phonetically realized material; when the first position is (phonetically) empty, as in (115a), the phonetic content of the complementizer cannot be omitted.

The examples in (116) show that the clause-initial position is filled in embedded wh-questions, relative clauses and embedded exclamative constructions. It has just been noted that the phonetic content of the complementizer of/dat is usually omitted in written and formal standard Dutch when the clause-initial position is phonetically filled; cf. taaladvies.net/taal/advies/vraag/592 for further discussion. The complementizers in (116) are given in brackets, however, because it is often possible to overtly express the complementizer in such cases in colloquial speech as well as in certain dialects. Note also that in some southern dialects the complementizer datthat is used instead of of in embedded questions, and that for some speakers it is even possible to realize both simultaneously: Jan vroeg [wie of dat hij bezocht had] Jan asked who he had visited. We will not discuss this issue here, but refer the reader to De Rooij (1965), Hoekstra & Zwart (1994) and Barbiers et al. (2008: §1.3.1.6) for more details.

In examples like (114) and (116), the result of movement is immediately visible, but there are also constructions for which it has been argued that wh-movement affects a phonetically empty element, as a result of which the movement can only be detected by the presence of an interpretive gap. An example of such a construction is the so-called comparative deletion construction illustrated in (117), in which the gap is indicated by e. Note in passing that in such constructions the verb in the als-clause is often omitted under identity with the verb in the main clause. We have indicated this in (117) by strikethrough for examples with and without extraposition of the als-clause; in the former case elision seems to be preferred, and in the latter case elision is even obligatory (probably to avoid a sequence of two occurrences of is).

Recall Section 11.2 above that certain main clauses with verb-first orders have also been analyzed as involving a phonetically element in clause-initial position.

One of the most important findings in generative grammar is that the syntactic relation between clause-initial constituents and their traces is subject to a set of general conditions; cf. Ross (1967), Chomsky (1973/1977), and Van Riemsdijk & Williams (1986: Part II) and Cheng & Corver (2006b) for two brief historical overviews. For this reason, the movements found in the constructions in (114), (116), and (117) are often referred to under a single umbrella term, wh-movement. This name derives from the fact that in English the moved constituents often contain a word beginning with wh-, like the interrogative and relative pronouns who and which in (118), but one should keep in mind that the term wh-movement covers all cases in (114), (116), and (117), not just those involving an interrogative or a relative pronoun.

The following sections will examine wh-movement in wh-questions (Section 11.3.1), in relative clauses (Section 11.3.2), in various types of topicalization constructions (Section 11.3.3), in exclamative constructions (Section 11.3.4), and in comparative deletion constructions (Section 11.3.5), and in comparative correlative constructions (Section 11.3.6). Section 11.3.7 then deals with the phenomenon of reconstruction, i.e. the fact that wh-moved phrases behave in certain respects as if they still occupied the position of their trace; reconstruction is therefore often taken as strong evidence for a movement analysis of wh-questions. Section 11.3.8 concludes this section on wh-movement with an appendix discussing so-called parasitic gaps; these are interpretive gaps in the structure which (under certain additional conditions) only occur if another gap is present, including those resulting from wh-movement.