Process adverbials restrict the denotation of the verbal predicate by adding specific information about the eventuality; they are characterized by the fact that they can be questioned by a wh-phrase. We will briefly discuss the semantic subclasses in (22).

Manner adverbials such as grondigthoroughly in (23a) are prototypical cases of process adverbials; they restrict the denotation of the verb phrase by specifying the manner in which the eventuality is performed. The (b)-examples show that manner adverbials pass the two VP-adverbial tests introduced in Section 8.1, sub III.

Instrumental adverbials such as met een schepwith a shovel in (24a) restrict the denotation of the verb phrase by specifying the instrument used to perform the activity; the (b)-examples show that instrumental adverbials pass the two VP-adverbial tests. Instrumental adverbials usually take the form of a met-PP, although there are incidental adjectival forms like handmatigby hand and machinaalmechanically.

Another set of process adverbials indicates the means used in performing the activity, as in (25a). Substantial subsets of these adverbials indicate means of transportation or communication; some typical examples are given in (25b&c). The primed examples again show that these adverbials pass the two VP-adverbial tests.

Adverbial phrases like juridischlegally, lichamelijkphysically, medischmedically, psychologischpsychologically, and wetenschappelijkscientifically are known as domain adverbials because they limit the process to a specific domain (e.g. legal, medical, or scientific).

Volitional adverbials like gedwongenforced, met opzeton purpose, met tegenzinreluctantly, met plezierwith pleasure, noodgedwongenby necessity, opzettelijkdeliberately, per ongelukby accident, and vrijwilligvoluntarily specify the relationship between the eventuality denoted by the verb (phrase) and the person performing/undergoing it. These adverbials are often considered subject-oriented, which is justified in the case of vrijwilligvoluntarily, since passivization of example (27a) shifts the orientation of this adverb from agent to theme.

However, adverbials such as opzettelijkdeliberately and per ongelukby accident are only oriented towards the (implied) agent, as can be seen from the fact that passivization of example (28a) does not affect the orientation of these adverbials.

The examples in (29) show for the adverbials vrijwillig in (27a) and opzettelijk in (28a) that volitional adverbials pass the two VP-adverbial tests.

That process adverbials are VP adverbials is also supported by the fact that, under neutral (i.e. non-contrastive) intonation, they follow modal adverbials such as waarschijnlijkprobably; this is illustrated in (30). However, we will see in Section 8.2.2, sub XI, that domain adverbials such as juridisch in (30d) can also be used as clause adverbials.

There are two kinds of agentive adverbials. The agentive door-PP in (31a) refers to the agent of the eventuality in passive constructions, while the comitative met-PP in (31b) introduces a co-agent. The singly-primed examples show that these adverbials pass the first VP-adverbial test, provided that we also passivize the conjoined pronoun doet dat clause in (31a'). The doubly-primed examples show that they also pass the second test.

That agentive adverbials are VP adverbials is also supported by the fact, illustrated in (32), that under neutral intonation they follow modal adverbials such as waarschijnlijkprobably. Note that comitative PPs can easily precede the modal adverbs, but only when the nominal complement of met can be accented, suggesting that this order is the result of focus movement; cf. Section 13.3.2.

Spatiotemporal VP adverbials in (33) restrict the denotation of the predicate by anchoring the eventuality in a particular location or time.

That the adverbials in de tuin and om drie uur function as VP adverbials in (33) is suggested not only by the fact that they follow the modal adverbial waarschijnlijkprobably but also by the fact, illustrated in (34), that they pass the VP-adverbial tests from Section 8.1, sub III.

The different subtypes of spatiotemporal VP adverbials are discussed in Subsections A and B. Note that we depart from traditional grammars by assuming that spatial phrases are used not only as adverbials but also as complementives. Semantically, adverbial and complementive phrases differ in that an adverbial phrase provides more information about the eventuality as a whole, while a complementive phrase provides more information about the subject or direct object of the clause (which originates as its logical subject). The difference is illustrated in (35): while (35a) expresses that the eventuality of Jan playing takes place in the garden, (35b) merely expresses that Jan’s location is in the garden.

For further discussion, we refer the reader to Sections P32.1.2.2 and P35.2.1.1, where it is extensively argued that complementive PPs function as predicates denoting a (change of) location or a direction. Some representative examples discussed in these sections are given in (36).

Finally, note that spatiotemporal adverbials can also be used as clause adverbials; we will ignore this use here and provide the relevant data in Section 8.2.2, sub IX. The difference between the two cases will be explored in more detail in Section 8.2.3.

Contingency adverbials relate the eventuality expressed by the clause to some other concurrent circumstance. Prototypical examples are adverbial phrases indicating cause and reason; the primed examples show that these adverbials pass the two VP-adverbial tests. We will follow Quirk et al. (1979: §8.7) in assuming that cause can be determined more or less objectively, while reason involves a subjective and often personal evaluation. In Dutch, the distinction can be clarified by questioning: waardoorby what usually elicits an answer giving a cause, while waaromwhy usually elicits an answer giving a reason.

The concessive adverbial PPs headed by ondanksdespite in (50) refer to a potential cause of an effect that, contrary to the speaker's expectation, did not occur. The primed examples show that the concessive PPs pass both VP-adverbial tests.

Quirk et al. (1979) shows that reason is often difficult to distinguish from purpose. The actual interpretation depends on one’s point of view: in an example such as (51a), earning money or getting pleasure from it can be seen as indicating Els’ motivation or purpose for working here. Again, questioning can help clarify the two sides: while waaromwhy elicits an answer that provides a reason, waarvoorfor what elicits an answer that provides a goal. Similarly, purpose and result are difficult to distinguish, although the latter is often expressed by the preposition tot, as in (51b).

All the contingency adverbials discussed so far pass the two VP-adverbial tests. That they are indeed VP adverbials is further supported by the fact that they can follow the modal adverbs under neutral intonation. Note, however, that at least adverbials indicating cause and reason can also precede the modal adverb, which suggests that they can also be used as clause adverbials: we return to this in Section 8.2.2, sub X.

Quirk at al. (1979) also count conditionals as contingency adverbials. We leave the discussion of such cases to Section 8.2.2, sub X, because there is good reason to think that they can only be used as clause adverbials. We conclude by noting that Haeseryn et al. (1997:1212) provides adverbial phrases that do not seem to fall into any of the above semantic subclasses, but simply refer to a concomitant circumstance; some examples are given in (53).

Section A25.1 shows that there is a relatively large set of adjectival adverbials that are typically used as degree modifiers of adjectives: prototypical examples are ergvery and vrijrather in erg/vrij aardigvery/rather nice. A small subset of these adverbials can also be used as modifiers of verbal projections; the examples in (54) show, for example, that this is possible for the amplifying degree modifier ergvery, but not for the downtoner vrijrather.

The use of ergvery has more restrictions. Although it is not clear to us what exactly determines whether its use is possible or not, it seems that erg is common with verbs that denote involuntary bodily activities such as niezento sneeze, verbs that denote a psychological state such as zich vervelento be bored, verbs of sound emission such as gillento scream, and weather verbs such as vriezento freeze, while it is less felicitous with verbs denoting voluntary activities like werkento work, fietsento cycle and pratento talk.

Section A25.1 also shows that nominal degree adverbials modifying adjectives are always downtoners: cf. een beetje zieka little bit sick. This use of nominal modifiers as downtoners of verbal projections is quite common; it occurs not only with the verbs in (55), but also with the verbs in (56) denoting a voluntary activity.

The status of the degree adverbials differs from that of the VP adverbials discussed in the previous subsections in that they do not give very clear results when it comes to the pronoun doet dat + adverb paraphrase: the paraphrases of (55a)/(57a) in (58b) are perhaps not entirely impossible, but feel awkward. However, the degree adverbials have the property of restricting the denotation of the predicate expressed by the lexical domain of the clause, as shown by the fact that the entailment test in (58c) leads to a positive result.