This subsection discusses the distribution of nominal objects in clauses containing a clause adverb, such as modal adverbs and adverbs of frequency. It will be shown that the position of the noun phrase is related to the information structure, especially the distinction between focus and presupposition (i.e. “new” and “given” information) of the clause, notions that will be explained in Subsection A. We begin with a discussion of definite noun phrases and personal pronouns, followed by a discussion of indefinite and quantified noun phrases.

Subsection I has shown that non-specific nominal objects are not easily scrambled across a clause adverb. However, we cannot conclude from this that non-specific nominal objects categorically resist scrambling. Consider the examples in (58), which show that non-specific indefinite nominal objects can either precede or follow °VP adverbials of time and place. The sentences differ in the assignment of the sentence accent. In the primeless examples, the sentence accent is preferably assigned to the nominal head of the indefinite object, whereas in the primed examples it is preferably assigned to the nominal head of the complement of the adverbial PP. This corresponds to the prominence within the focus field of the clause, i.e. within the part of the clause expressing “new” information, which can be roughly defined as the part of the middle field of the clause following the clause adverbs. In the primeless examples the object is the most prominent element in the focus field, while in the primed examples it is the adverbial phrase that is most prominent; cf. Broekhuis (2007/2008).

The hypothesis that the orders in (58) are related to prominence within the focus field predicts that the orders in the primed examples are only possible when the VP adverbial can be interpreted as part of the focus of the clause. Since indefinite nominal objects are more likely to be part of the focus of the clause than e.g. adverbial pro-forms such as daarthere and toenthen, it is not really surprising that the primed examples in (59) are unacceptable.

Example (60) shows that similar facts can be found with non-specific indefinite nominal objects with a quantifier or a numeral. Substitution of a pro-form for the adverbial phrase in the primed examples in (60) leads to unacceptability.

Definite nominal objects, on the other hand, do not easily follow VP adverbials. The primeless examples in (61) seem acceptable, but they are certainly marked compared to the primed ones. The primeless examples also show that one possible reason for their markedness may be that they require an emphatic or contrastive focus on the noun, indicated by underlining. In the primed examples, the adverbial PP can be replaced by the pro-forms daarthere and toenthen; this is most likely when these pro-forms are assigned emphatic or contrastive focus.

Note, however, that examples like (61a&b) are perfectly acceptable when we are dealing with more or less fixed NP + V collocations referring to e.g. common domestic activities like de afwas doento wash the dishes or het gras maaiento mow the grass/lawn in (62); cf. Van der Does & De Hoop (1998) for more examples.

The examples in this subsection have shown that we cannot claim that scrambling of non-specific indefinite nominal objects is categorically blocked, since scrambling of such noun phrases is possible across VP adverbials (here: of time and place). The effect of this kind of “short” scrambling seems to be related to prominence in the focus field (the field expressing the “new” information of the clause); cf. Broekhuis (2008:§2.4). This fact has received little attention in the literature so far, and we believe that more research is needed to gain a better understanding of the factors that influence the order of constituents in the focus field of the clause; cf. Slioussar (2007) for a relevant discussion pertaining to scrambling in Russian, which may be partly applicable to Dutch.

While VP adverbs must be placed to the right of clause adverbs, there are also adverbial phrases that can be placed to the left of typical clause adverbs, such as the modal adverb waarschijnlijkprobably. This is true, for example, of the time and place adverbs in (63). Example (63a) shows that such time adverbs can co-occur with time adverbs that function as VP adverbs; the former are used to restrict the relevant time interval in which the event can in principle take place (time interval j in the tense representations given in Section V1.5), while the latter pinpoint the time at which the event denoted by the main verb actually takes place (time interval k). In (63b) the two adverbial phrases of place behave similarly: the first restricts the place where the event can in principle take place, while the second pinpoints the actual place where it takes place.

The examples in (64) show that although it is possible to scramble a definite noun phrase across the clause adverb, scrambling across the higher locational or temporal adverbial phrase leads to a marked result; placing the noun phrase before the higher locational/temporal adverb is possible, but usually requires emphatic accent on that adverbial phrase.

The behavior of definite pronouns differs markedly from that of definite noun phrases; pronouns must cross not only the clause adverb, but usually also the higher place/time adverb.

The contrast between the examples in (64) and (65) suggests that definite pronouns must be moved further to the left than definite noun phrases. If true, this would seem to cast new light on the word order pattern in example (56) from Subsection IE, repeated here in a slightly different form as (66a): the fact that (66a) is marked can now be attributed to the fact that the pronoun has not been moved into the designated position of the definite pronouns, but has remained in the position of the scrambled non-pronominal object.

For completeness’ sake, note that we can now also account for the obligatory scrambling of the indirect object in (66) by claiming that the leftward shift of the pronoun must be preceded by scrambling, i.e. the pronoun is first scrambled into the position in (66a) before it is moved into its position in (66b): scrambling of the indirect object is then forced by the first movement, just as in the analysis proposed for the wh-examples in (57) from Subsection IE.

In negative clauses with a neutral intonation pattern, the negative adverb nietnot cannot normally be followed by a direct object; cf. Kraak (1966:§25). Since niet can be followed by other types of constituents, such as the PP-complement op zijn vader in (67b), we cannot account for the fact that the noun phrase het boek must precede the negation by assuming that the negative adverb niet and the verb are somehow merged; the fact that the PP-complement follows the negative adverb suggests that the latter is external to the VP. If so, we should conclude that nominal objects must be scrambled to a position preceding the negative adverb.

Example (68a) shows that indefinite nominal objects in negative clauses normally neither precede nor follow the negative adverb; rather, they require that the negation be expressed by the negative article/quantifier geenno, as in (68b). Note that the examples in (68a) are acceptable if een is stressed, in which case we are probably dealing with the numeral éénone; cf. Subsection IC for similar data with clause adverbs. Example (68a) with the nominal object preceding the negative adverb then receives a specific interpretation, and the one with the nominal object following the negative adverb receives a not-a-single reading. For a more detailed discussion of geenno and data of this kind, see Section 19.1.5.

The examples in (69) show that indefinite noun phrases containing an attributive adjective like volgendenext or nieuwenew or an ordinal numeral are exceptions to the rule that indefinite nominal objects cannot precede the negative adverb niet. Indefinite noun phrases of this type are also exceptional in that they can precede modal adverbs such as waarschijnlijkprobably; cf. example (41). Note that the indefinite nominal object cannot follow the negative adverb niet, in accordance with the general rule that noun phrases cannot follow a negative adverb. However, the doubly-primed examples show that the negative article/quantifier geenno cannot be used in such examples either.

From the above discussion we can conclude that the adverb niet cannot be followed by a noun phrase. However, the primeless examples in (70), in which the nominal objects are assigned contrastive focus, are acceptable. This is due to the fact that we are not dealing with sentential negation here, but with constituent negation; this can be seen from the fact that in the primed examples the adverb niet is preferably pied-piped by topicalization of the noun phrases; stranding of the negation is at best marginally possible. The primeless examples therefore do not contradict the general rule that noun phrases cannot follow sentential negation.

This may be different in what we see in (71), where not just the object het/een boek receives contrastive focus but the whole VP het/een boek gelezen. In this case, the negative adverb is preferably stranded by topicalization of the VP, although pied piping is at least marginally possible. If this indicates that we are dealing with sentential negation, then example (70b) must be seen as an exception to the general rule that noun phrases cannot follow sentential negation. We leave this issue for future research.

It is not easy to decide whether indefinite nominal objects preceding sentential negation allow a non-specific interpretation or not. Although the noun phrase vier boeken can be placed either before or after the clause adverb waarschijnlijk, a non-specific interpretation is not easily obtained. Rather, the nominal object following waarschijnlijk seems to be interpreted specifically (while the one preceding it seems to prefer a partitive reading). It seems that the quantifier iemandsomeone in (72b) also receives a specific interpretation in both positions, although in this case the judgments seem a bit murky.

The picture seems clearer in the case of the frequency adverbs. In (73a) the nominal object precedes the adverbial phrase al drie keer, and we are clearly dealing with a specific indefinite noun phrase. As we have seen in Section 22.1.3, sub IC, indefinite nominal objects following adverbs of frequency must be given a non-specific interpretation. The fact that example (73b) is marked therefore suggests that indefinite nominal objects preceding the negative adverb niet cannot easily receive a non-specific interpretation. For completeness, (73c) shows that, in accordance with our earlier observation, the indefinite nominal object cannot follow the negative adverb niet either.

Let us assume that on the basis of (73a-b) we can conclude that indefinite nominal objects preceding the negative adverb niet cannot be construed as non-specific. The question then arises whether there is a syntactic reason for the fact that indefinite nominal objects preceding the negative adverb niet cannot easily be construed as non-specific: since the negative adverb normally follows the clause adverb (cf. (67a)), there is no a priori reason to assume that scrambling of a non-specific indefinite nominal object across it is blocked. Indeed, it seems plausible that there are pragmatic reasons for assuming that non-specific indefinite nominal objects cannot precede negation and thus fall outside its scope: it simply does not seem very informative to claim about some unidentified entity that a certain proposition does not apply to it. Of course it does make sense to have a non-specific nominal object within the scope of negation, since that would amount to having a negative existential quantifier, as in the English example I didn’t see a thing, but Dutch uses special negative forms in such cases: the negative element geenno, and the negative quantifiers nietsnothing and niemandnobody.

This subsection has discussed scrambling of nominal objects and has shown that different types of noun phrases have different scrambling possibilities: in general, we can say that definite pronouns scramble more often than definite noun phrases, which in turn scramble more often than indefinite noun phrases. We have also seen that the domains in which scrambling occurs are different for the different types of noun phrases. Indefinite noun phrases can cross certain VP-adverbs but not clause adverbs; definite noun phrases can cross clause adverbs if they are part of the presupposition of the clause, but cannot easily cross adverbs preceding these clause adverbs; finally, definite pronouns must precede clause adverbs as well as the adverbs preceding them.

The existing literature on scrambling generally does not take into account that the different types of nominal object may involve different domains of application, with the result that the occurrence of a presuppositional definite noun phrase after any adverb is sometimes taken as counterevidence for the claim that such noun phrases must undergo canonical A-scrambling; cf. De Hoop (2000/2003/2016), Van Bergen & De Swart (2010), Schoenmakers & De Swart (2019), and Schoenmakers (2022). The primeless examples in (74) show again that this view is too simple: scrambling of the definite noun phrase can only be observed when a lower clause adverb such as waarschijnlijk is present (or certain discourse particles such as maar; cf. Zwart 2011). It is therefore not surprising either that the primed examples in (74) allow two readings: one in which the definite noun phrase is part of the focus and one in which it is part of the presupposition of the clause.

The studies cited above found a high proportion of cases like the primed examples, and concluded from this that presuppositional definite noun phrases scramble only optionally (or not at all). We do not think this conclusion is justified for the reason given, but of course we cannot exclude the possibility that presuppositional definite noun phrases may fail to scramble under certain conditions (or in certain varieties of Dutch). Indeed, one case was already identified in Subsection E; presuppositional direct objects cannot A-scramble across VP-internal indirect objects. That this is a real possibility is also shown by the restrictions on Scandinavian object shift, which shares many properties with A-scrambling of nominal objects; cf. Broekhuis (2020/2022) and the references cited there. However, we believe that a deeper insight into such restrictions on A-scrambling can only be gained by first examining more closely the adverbs that can be crossed by the different types of A-scrambling. Again, a comparison with the Scandinavian languages seems to be quite useful; cf. Thráinsson (2007:ch.2) for a very appropriate overview in this respect. For a further discussion of A-scrambling from the perspective of the internal structure of the middle field of the clause, see Section 13.3.