Semi-aspectual verbs usually correspond to main verbs like zittento sit, liggento lie, and staanto stand, which indicate a certain posture or position of the subject of the clause. The examples in (153) show that these semi-aspectual verbs are usually interchangeable, but that the meaning of the corresponding main verb can sometimes affect the preferred option; for example, activities that are normally performed while standing, such as afwassento wash dishes, will usually take the semi-aspectual staanto stand.

That semantic properties of the main verb carry over tot the corresponding semi-aspectual verb is perhaps even clearer with inanimate subjects. Example (154a) shows that the verb staan refers to situations in which an entity such as a lamp is resting on its base; the verb liggen cannot be used in such a case but refers to a case in which the lamp has another position, e.g. when it has toppled. Objects without a clear base, such as sinaasappelorange in (154b) are rather combined with the posture verb liggento lie. The primed examples show that the same preference is found in the corresponding semi-aspectual constructions.

That the lexical meaning of the main verbs corresponding to the semi-aspectual non-main verbs may or may not be present is also supported by example (155a) can be used without problems when the speaker cannot observe the referent of the subject of the clause and thus cannot tell whether this referent is actually sitting at the moment of speech. Furthermore, example (155b) shows that the semi-aspectual verb zitten can co-occur with the posture main verb zitten, which would be very surprising if the former had preserved the lexical meaning of the latter.

The verb of movement lopento walk can also be used as a semi-aspectual verb; according to Van Pottelberge (2002) this is a more recent innovation, which may be supported by the finding in Lemmens (2005:§4) that the semi-aspectual use of this verb is more restricted in that it is less frequent, and less productive in that only occurs with animate agents, as (156). Furthermore, he suggests that the use of lopen is regionally restricted in that it is perceived as more common in the Netherlands than in Belgium. This is further supported by comparing the use of semi-aspectual lopen in a Belgian and a Dutch quality newspaper of similar size over a period of two months: 3 cases were found in the Belgian newspaper against 16 cases in the Dutch one. This might indicate that lopen is a more recent innovation of Netherlandic Dutch.

Lemmens (2005:§4) takes the fact that the subjects of the semi-aspectual constructions with lopen in his corpus are always animate agents as an indication that the semantic bleaching of the motion verb is less advanced than that of the posture verbs (although this fact may also be due to the fact that the motion verb lopen is not easily combined with inanimate subjects). Still, it is not impossible to suppress the meaning of the main verb, as is clear from the fact that the semi-aspectual lopen in (157a) can be used without problems even when the speaker cannot observe the referent of the subject of the clause and thus cannot tell whether this referent is actually walking at the moment of speech. Furthermore, example (155b) shows that the semi-aspectual verbs lopen can co-occur with the movement verb rennento run, which would be very surprising if it had preserved the lexical meaning of the main verb.

Finally, we want to point out that Lemmens suggests that the meaning of the semi-aspectual construction in (158a) depends on the choice of the non-main verb: the case with zitten is purely durative in that is simply refers to an eventuality of Jan crying that includes speech time, while the one with lopen is durative and iterative in that we may be dealing with a repeated eventuality within the present-tense domain, which is clearly the reading that would be given to his example in (158b). It is not so clear, however, that a similar interpretation is impossible for the case with zitten in (158a); both versions allow an iterative reading when we add a frequency adverb such as vaakoften, as in Jan zit/loopt vaak te huilenJan is often crying (cf. also Lemmens 2015:§4.4.1).

We conclude that the primary function of the semi-aspectual verbs is to indicate that we are dealing with an ongoing event, which prototypically takes place at speech time (or the virtual speech-time-in-the-past when the sentence is in the past tense); they create a progressive construction comparable (but not identical) to the English progressive construction, which we have therefore used in the translations of our examples.

The lexical projection of the main verb usually denotes an activity, as in the primeless examples in (159); the primed examples show that telic events (i.e. achievements and accomplishments) usually produce less felicitous results, although it is not impossible to encounter cases such as (159b'). The relevance of telicity is shown by the numbers in brackets, which give the results of a Google search (February 4, 2024) on the strings [ligt te rollen], [ligt * af te rollen], [zit/ligt te slapen], and [zit/ligt in slaap te vallen]; cf. also Section 1.5.3.

In general, semi-aspectual verbs cannot be combined with verb phrases denoting states: examples such as (160) are only possible with a very special “pretense” reading, which can be brought out by adding the adverbial phrase weer eensonce again; presumably this special reading makes the event dynamic.

It also seems that the lexical projection of the main verb usually denotes an activity that can be controlled by the subject of the clause. As a result, the subject is typically animate, as can be seen by comparing example (161) with example (159a).

That the subject must be able to control the event can be brought to the fore with the help of the examples in (162): while events denoted by the perception verbs kijkento look and luisterento listen are typically controlled by the subject, events denoted by ziento see and horento hear are not, and this may account for the contrast in acceptability between the two primed examples.

A clear exception to this general rule is when the event involves an involuntary bodily function or natural process, as evidenced by the fact that examples such as (163) are very common. Since no control by the subject is involved, it is not surprising that we often find inanimate subjects in such contexts.

Another possible exception is (164a) with the reflexive psych-verb zich ergerento be annoyed which at first glance seems to denote an involuntary mental state. However, it is not so clear whether it is really the case that events denoted by such reflexive psych-verbs cannot be controlled by the referent of the subject of the clause; example (164b) strongly suggests that the mental state denoted by zich amuserento amuse oneself is consciously brought about by Jan himself.

The examples given in the previous subsections have already shown that semi-aspectual verbs take te-infinitives as complements: the examples in (165) show that omitting the infinitival marker te leads to ungrammaticality.

However, in constructions such as (166), in which the semi-aspectual verbs are themselves infinitives, the infinitival marker te preceding the main verb can be omitted; in fact, leaving te in often leads to a marked result; cf. Haeseryn et al. (1997:970ff).

The use of the infinitival marker te more generally seems to lead to a marked result when the semi-aspectual verb takes the form of an infinitive. For example, in perfect-tense constructions such as (167), in which the semi-aspectual verb appears as an infinitive due to the infinitivus-pro-participio (IPP) effect, the marker te is optional; the construction without te is the more common one, although the corresponding construction with te is certainly acceptable.

Haeseryn et al. adds to the above observations that the marker te is also optional when the semi-aspectual verb is a plural finite form. However, the contrast between the two examples in (168) shows that the omission of te is only possible in embedded clauses, i.e. when the aspectual verb is part of the clause-final verb cluster. In fact, we consider the omission of the marker te to be degraded in both cases, which is why we have marked the omission of te in (168b) with a percentage sign.

For completeness’ sake, note that similar observations with respect to the distribution of the infinitival marker te can also be made (at least diachronically) for other constructions with infinitival non-main verbs, such as the modal verb hoevenneed; cf. Van de Velde (2015:§3.4). Note also that the infinitival marker te is not easily used in nominalizations; cf. the examples in (169).

The above overview suggests that the marker te can always be omitted when the semi-aspectual verb is non-finite, and that this is often even the preferred option. However, it is not entirely clear whether the above judgments about structures with the marker te are representative of the majority of standard Dutch speakers, since Barbiers et al. (2008: §2.3.4) found that speakers in the Netherlands are more favorable to constructions with te than Flemish speakers. Of course, this still leaves the question why the infinitival marker te is obligatory in (165) but can be omitted in the examples in (166) and (167). One option that might be explored is that the availability of the te-less forms is related to the fact that Middle Dutch double-infinitive constructions such as (152a) above, which predate the progressive complex-verb construction, do not feature the infinitival marker te either; in short, the marked cases with te in (166)/(167) are true progressive constructions, while the cases without te are descendants of the older Middle Dutch double-infinitive constructions. We leave this suggestion to future research.

That constructions with semi-aspectual verbs are monoclausal is evident from the fact that they exhibit the IPP-effect. We illustrate this again in the examples in (170).

The monoclausal behavior of such constructions is also clear from the fact that they involve clause splitting/verb clustering; the infinitival main verb can be separated from its dependents by the semi-aspectual verb. The use of the percentage signs indicates that clause splitting is obligatory for northern speakers, while some Flemish speakers accept permeation of the verb cluster as a marked option.

It seems that semi-aspectual verbs obligatorily precede the main verb in the clause-final sequence; since this will be an important issue in Section 6.3.2, we did a Google search (November 15, 2023) for the examples in (172) and found that the search string [te piekeren V], in which V stands for the three semi-aspectual verbs in their third-person singular present-tense form, returned only 2 relevant hits (we checked all instances individually); that the string [V te piekeren] is common supports the conclusion that the order [te piekeren V] is not part of Dutch core grammar.

In clusters of more than two verbs, the main verb is always last in the clause-final cluster. The examples in (173) illustrate this for both main and embedded clauses, and for constructions with and without the infinitival marker te.

The examples in (174) show the same for imperfect constructions with three verbs. We have not given examples with the infinitival marker te, because such examples are marked anyway.