• Dutch
  • Frisian
  • Afrikaans
Show full table of contents

This chapter takes as its point of departure the discussion in Section 9.2, which has shown that finite verbs can be found in basically two positions: the clause-final position in embedded clauses and the verb-first/second position in main clauses; the latter position is normally occupied by a complementizer in embedded clauses.

Example 1
a. Marie zegt [dat Jan het boek leest].
  Marie says  that  Jan  the book  reads
  'Marie says that Jan is reading the book.'
b. Op dit moment leest Jan het boek.
  at this moment  reads  Jan the book
  'At this moment, Jan is reading the book.'

On the basis of these two positions, the clause is traditionally divided into various "topological" fields: the clause-initial position, the middle field and the postverbal field. This is illustrated in Figure (2), repeated from Section 9.2.

Example 2

This chapter will focus on the middle field of the clause. Section 9.1 has shown, however, that this notion has no independent theoretical status as it cuts across the more fundamental division between the lexical and the functional domain of the clause.

Example 3

The lexical domain consists of the main verb and its arguments as well as VP modifiers, which together form a proposition. In (4a), for example, the verb kopen'to buy' takes a direct object as its complement and is subsequently modified by the manner adverb snel'quickly', and the resulting complex predicate is finally predicated of the noun phrase Jan. The complex phrase thus formed expresses the proposition that can be represented by means of the logical formula in (4b).

Example 4
a. [Jan [snel [het boek kopen]]]
  Jan  quickly   the book  buy
b. buy quickly (Jan, the book)

If the proposition in (4b) is to correspond to the syntactic structure in (4a), we should assume that the VP in (3) must be replaced by a more finely articulated syntactic structure. In current generative research it is generally assumed that this structure is as given in (5). As the linking of semantic and syntactic structure is unlikely to vary across languages, it is often assumed that the structure in (5) is more or less invariant across languages, and that the surface differences between languages are due to movement. For example, the word order difference between Dutch and English with respect to the relative placement of the verb and the nominal direct object can be accounted for by assuming that English but not Dutch has obligatory V-to-v movement; see Section 9.4 for a more detailed discussion.

Example 5

The structure in (4a) can now be made more explicit as in (6): internal arguments such as the theme het boek'the book' are generated within VP, VP adverbials such as the manner adverb snel'quickly' are adjoined to VP, and external arguments such as the agent Jan are generated as the specifier of the "light" verb v.

Example 6
[vP Jan v [VP snel [VP het boek kopen]]]
       Jan  quickly  the book  buy

In what follows we will adopt the assumption that the lexical domain does have a more finely articulated structure, and we therefore replace the global representation of the clause in structure (3) by the one in (7). Observe that the lexical domain may be even more complex than indicated here, as we have ignored issues raised by structures with, e.g., indirect objects or complementives.

Example 7

The semantic information encoded in the lexical domain can be equated with the information expressed by traditional predicate calculus; the functional domain provides additional information. For instance, the functional head T in (7) adds the tense feature ±past and the functional head C indicates illocutionary force, as is clear from the fact that the complementizers dat'that' and of'if/whether' introduce embedded declarative and interrogative clauses, respectively. In addition to these functional heads there may be other functional heads, indicated by X in (7), which introduce other features. Section 10.1 has shown that in main clauses finite verbs are moved out of the lexical domain into the functional head C (or T), which accounts for the verb-first/second effect in Dutch.
      Although arguments, complementives and VP adverbials generally surface within the lexical domain, they can also be moved into the functional domain. Normally, this has a semantic motivation; Section 11.3.1 has shown, for instance, that wh-phrases are moved into clause-initial position in order to create structures such as (8a), which can be translated more or less directly into the logical formula in (8b): the interrogative pronoun wat in clause-initial position corresponds to the question operator ?x, while the trace of the wh-phrase corresponds to the variable x.

Example 8
a. Wati leest Peter ti?
  what  reads  Peter
  'What is Peter reading?'
b. ?x (Peter is reading x)

The effect of wh-movement is immediately clear in main clauses from the fact that the wh-phrase surfaces in the position preceding the finite verb. Movements targeting a clause-internal position are often less easy to observe. For instance, it is normally assumed that in passive constructions the internal theme argument moves from its original VP-internal position into the regular subject position, that is, the specifier position of TP in (7), but this can only be observed if other material is present between the two positions. This is illustrated by the passive example in (9b), which shows that the postulated movement is indeed possible in Dutch but optional if the derived subject is definite. Whether or not the movement applies is of course less easy to determine if the indirect object is left implicit as the effect of movement cannot be observed directly from the word order of the clause in that case.

Example 9
a. dat de gemeente (de koning) het concert aanbood.
  that  the municipality   the king  the concert  prt-offered
  'that the municipality offered the king the concert.'
b. dat <het concert> (de koning) <het concert> aangeboden werd.
  that    the concert   the king  prt-offered  was
  'that the concert was offered to the king.'

In order to investigate whether some element has moved from the lexical into the functional domain we appear to need a demarcation of the boundary between the two domains; compare the notion of pivot location in Haeseryn et al. (1997:1328) and the notion of comment modifier in Verhagen (1986:ch.4). Clausal adverbs such as the modal waarschijnlijk'probably' can perform this function because they take scope over the full proposition expressed by the vP in (7). This fact is actually exploited by the standard adverb tests according to which clausal adverbs can be paraphrased by means of the construction Het is adverb zo dat ...'it is adverb so that ...', in which the adverb likewise has scope over the proposition expressed by the embedded clause.

Example 10
a. Jan werkt waarschijnlijk.
  Jan works  probably
  'Jan is probably working.'
b. Het is waarschijnlijk zo dat Jan werkt.
  it  is probably  the.case  that  Jan works
  'It is probably the case that Jan is working.'

Another argument in favor of assuming that modal adverbs demarcate the boundary between the lexical and the functional domain is the fact illustrated in (11) that they can precede an external argument, which is located at the left edge of the lexical domain, namely in the specifier of the light verb v in (5)/(7).

Example 11
dat [TP <de klant> waarschijnlijk [vP <de klant> v [VP het boek koopt]]].
  that    the customer  probably  the book  buys
'that the customer will probably buy the book.'

Example (11) shows that the movement of the subject into the regular subject position is not only optional in passive constructions such as (9b) but also in active constructions. We return to this fact in Section 13.2, where it will be shown that the movements indicated in (12), which we will refer to as subject shift because they affect a noun phrase that surfaces as the nominative subject, are restricted by the information structure of the clause; they apply only if the subject provides discourse-old information.

Example 12

That subjects raise into the regular subject position has been a standard claim in generative grammar for a very long time (especially for passive constructions). This chapter discusses a number of other movement operations that likewise move elements out of the lexical domain into the functional domain in as far as this results in reordering of the constituents in the middle field of the clause: the various forms of wh-movement, which place elements into clause-initial position, are not discussed here but in Section 11.3. Following Ross (1967), the reordering of the middle field is often referred to as scrambling but there are reasons not to follow this practice because it incorrectly suggests that we are dealing with a single, uniform phenomenon. We will show that scrambling is in fact a pre-theoretical cover term for a wider set of movement phenomena with diverging properties. Section 13.2 will discuss nominal argument shift, which was referred to as NP-preposing in earlier generative literature, for instance Van den Berg (1978) and De Haan (1979); this movement type affects nominal arguments only and plays an important role in distinguishing between the presupposition and the focus of the clause, that is, between discourse-old and discourse-new information; see the discussion of (9) and (11) above. Section 13.3 will show that negative and contrastive phrases can likewise be moved into a more leftward position; this movement is not restricted to nominal arguments but can also be applied to specific constituents of other categories. Section 13.4 concludes by showing that phonologically weak forms like the referential personal pronoun ʼm'him' and the locational proform er'there' are obligatorily moved into a position close to the regular subject position of the clause. Section 13.1 starts by introducing the notion unmarked word order, however.

  • Berg, Evert van den1978Fokus presuppositie en NP-preposingDe Nieuwe Taalgids71212-222
  • Haan, Ger de1979Conditions on rulesDordrechtForis Publications
  • Haeseryn, Walter, Romijn, Kirsten, Geerts, Guido, Rooij, Jaap de & Toorn, Maarten C. van den1997Algemene Nederlandse spraakkunstGroningenNijhoff
  • Ross, John1967Constraints on variables in syntaxBloomingtonIndiana university linguistics club
  • Verhagen, Arie1986Linguistic theory and the function of word order in Dutch. A study on interpretive aspects of the order of adverbials and noun phrasesDordrechtForis Publications