It is difficult to provide a watertight semantic characterization of the category of verbs, because they fall into two main groups with quite different semantic properties: main and non-main verbs. Main verbs can be characterized as verbs that denote specific states of affairs, which usually involve one or more participants; they can be semantically characterized as n-place predicates in the sense of predicate calculus. Main verbs thus function as semantic heads of their clauses and form propositions by combining with one or more arguments. This is illustrated for intransitive, transitive and ditransitive main verbs in (1).

Non-main verbs do not function as predicates in the sense of predicate calculus: the perfect auxiliaries hebbento have and zijnto be, aspectual verbs like gaanto go and epistemic modal verbs like willenwill and kunnenmay are not (or at least not primarily) argument-taking predicates, but add additional information to the proposition expressed by the main verb and its arguments. For example, the perfect auxiliary hebben in (2a) expresses that the event of Jan reading the book was completed before the time of speech, the aspectual verb gaan in (2b) focuses on the starting point of this event, and the epistemic modal zullen expresses that the speaker is confident that the proposition will come true. Note in passing that we have ignored here that zullen is often analyzed as a future auxiliary; cf. Section 1.5.4 for discussion.

Since it is difficult, if not impossible, to find a semantic characterization of verbs that can be applied equally well to both main and non-main verbs, it seems advisable to look elsewhere for a proper characterization of the category of verbs; Subsections II and III will show that morphology and syntax provide better means of characterizing this set. We return to the semantic properties of verbs and the distinction between main and non-main verbs in Section 1.2.

Verbs are characterized by the fact that they can be inflected in certain specific ways. We will restrict ourselves here to the inflection of finite verbs, which can be either main or non-main verbs; cf. Section 1.3 for more discussion of verbal inflection. Finite verbs are characterized by the fact that they agree in person and number with the subject of their clause and can be marked for [±past] tense. Table 1 gives the finite inflection of the so-called regular (or weak) verbs. A note on the translations given in this table may be in order, as the Dutch present and past tenses have different conditions for their use than the English present and past tenses. Here we provide translations that correspond to the (default) progressive reading of the simple present/past forms; cf. Section 1.5 for a detailed discussion of the actual use of the Dutch tenses.

	present		past
	singular	plural	singular	plural
1p	Ik huil-Ø ‘I am crying’	Wij huil-en ‘We are crying’	Ik huil-de ‘I was crying’	Wij huil-de-n ‘We were crying’
2p	Jij huil-t ‘You are crying’	Jullie huil-en ‘You are crying’	Jij huil-de ‘You were crying’	Jullie huil-de-n ‘You were crying’
3p	Hij huil-t ‘He is crying’	Zij huil-en ‘They are crying’	Hij huil-de ‘He was crying’	Zij huil-de-n ‘They were crying’

Table 1 shows that the past tense is expressed by the affix -de, which must be directly adjacent to the verb stem. This marker has the allomorph –te if the verb stem ends in a voiceless consonant: Ik vis-teI was fishing, ik pak-te een koekjeI took a cookie, etc. Table 1 also shows that there are two agreement markers in Dutch. First, we find the invariant plural marker -en, which is phonologically reduced to -n after the past suffix -te/-de. Second, we find the singular marker -t for second and third-person subjects in the present; there is no morphologically realized affix for singular agreement in the past or for first-person singular agreement in the present. In addition to the regular pattern in Table 1, there are a number of irregular patterns, which will be discussed in Section 1.3; here we just want to point out that the presence of finite inflection is sufficient to conclude that we are dealing with a verb.

Verbs are also characterized by their position in the clause; main verbs always occur in the right periphery of embedded clauses and typically follow their nominal arguments. However, they are typically followed by their clausal complements and can optionally be followed by their PP-complements; this shows that although the literature usually refers to the main verbs in (3) as clause-final verbs (i.e. verbs in the clause-final position), this should not be interpreted in such a way that main verbs in the right periphery of the clause are the rightmost elements in the clause.

The examples in (4) show that non-main verbs like auxiliaries and aspectual verbs are also clause-final in embedded clauses. Note in passing that these examples show that the position of non-main verbs relative to the main verb is not fixed, since they sometimes follow and sometimes precede it; this is the topic of Section 7.3.

In the northern varieties of standard Dutch, clause-final non-main verbs behave like main verbs in that they also usually follow the nominal arguments of the clause, but this is not the case in the southern varieties; in particular, the varieties spoken in Belgium allow nominal arguments between modal/aspectual verbs and main verbs; cf. Section 7.4. Note in passing that there are other elements, like verbal particles and complementives, that also tend to occur in the right periphery of the clause.

Non-finite verbs also occupy a clause-final position in main clauses, as shown in (6a) for the past participle gelezenread and in (6b) for the infinitive lezenread.

Finite verbs, on the other hand, do not. In yes/no questions, for example, they are in the first position of the sentence, as shown in example (7); such main clauses are often referred to as verb-first sentences.

In wh-questions the finite verb occupies the so-called second position of the sentence, i.e. the position immediately following the preposed wh-phrase. This is illustrated by the main clauses in (8), which are often referred to as verb-second sentences.

In declarative clauses the finite verb is also in the second position of the sentence, i.e. the position immediately after a clause-initial subject or a topicalized phrase. This is illustrated by the verb-second sentences in (9); the (a)-examples are subject-initial sentences and the (b)-examples involve topicalization of the time adverbial morgentomorrow.

Note that the technical terms verb-first and verb-second are used in strict opposition to the technical term verb-final; this has the somewhat odd result that certain verbs that find themselves in the final position of a clause do not count as verb-final, but as verb-first or verb-second. For example, the main clause in (10a), which consists of nothing more than an intransitive verb and its subject, does not count as a verb-final clause in the technical sense, because the verb must appear in second position when more material is added, as is shown in (10b).

If a verb occupies the first or second position in main clauses, this is usually sufficient to conclude that this element is a finite verb. Since it is normally easy to construct such examples for all main and non-main verbs, this would seem to provide a foolproof syntactic test for determining the verbal status of a lexical element. In short, a verb must be able to be used in the bold V[+finite] position in the structure of main clauses sketched in (11), where XP stands for the sentence-initial constituent in declarative clauses and wh-questions, and NP, PP and Clause refer to complements selected by the verb (if any), while the dots stand for an unspecified number of other constituents; cf. Section 9.1 for a more precise discussion of clause structure.