Ge-nominalizations can be indefinite, definite, and quantified. Indefinite ge-nominalizations are often article-less (e.g. Ik haat gezeurI hate nagging) but in certain contexts they can also be introduced by the indefinite article eena (e.g. een oeverloos gepraat in (175a)). Example (175b) shows that definite ge-nominalizations allow all kinds of definite [+neuter] determiners: the definite article het, the demonstratives dit/datthis/that and the possessive pronouns. Example (175c) shows that definite ge-nominalizations can contain a wide variety of quantifiers, including elke/iedereeach/every, alleall, and veel/weinigmany/few, etc.

ge-nouns can also be preceded by interrogative determiners like welke, and (176a) shows that they can be wh-moved as a result. Example (176b) shows that they can also be topicalized.

Pluralization of ge-nouns is not possible. We would not expect this, given that ge-nouns are substance ([-shape][-set]) nouns; cf. Section 15.2.2.1. Recall that this can be used as a reliable test for distinguishing ge-nouns from concrete and lexicalized ge-nouns: the mere fact that gebouw(en)building(s), gedicht(en)poem(s), geschil(len)dispute(s) and gewaad/gewadengown(s) allow pluralization confirms the view that they are not ge-nouns.

Ge-nouns also behave like nouns with respect to adjectival modification: the fact that the adjectives in the definite constructions in (177b) prefer the suffix -e shows that we are dealing with attributive modifiers, not adverbial phrases. Nevertheless, the primed examples show that modification by adjectives expressing duration or frequency is possible, which underlines the verbal quality of these nominals.

Example (178) provides a ge-nominalization with an intransitive input verb: both the verb wandelento stroll and the ge-noun gewandelstrolling have an argument structure with a position for an agent argument. The (b)-examples show that the agent can appear either postnominally as a van-PP or prenominally as a possessive pronoun or genitive noun phrase. Note that the derived form is labeled ge-n, rather than N, in order to express the special nature of the derived noun, with its combination of nominal and verbal properties.

Ge-nominalization can also take transitive verbs such as treiterento badger as input. Despite the change in syntactic category, the argument structure of the verb is inherited by the derived form getreiter in an essentially unchanged form: the number of arguments and their thematic functions remain the same. However, the verb and the derived noun do differ in that the arguments are obligatory in the former, but can be left implicit in the latter. The (c)-examples in (179) further show that the theme argument of a ge-noun can only be realized in the form of a postnominal van-PP; it cannot appear in the form of an accusative prenominal noun phrase (as opposed to inf-nominalizations), nor in the form of a possessive pronoun or genitive noun phrase (as opposed to er and ing-nominalizations).

The examples in (180a&b) show that ge-nominalizations of ditransitive verbs also leave the argument structure essentially unchanged, although instances of such nominalizations in which all three arguments are expressed are quite rare. The (c)-examples show that the recipient, like the theme argument, must be expressed as a postnominal PP; it cannot be realized as a dative prenominal noun phrase or as a possessive pronoun. According to Dutch tradition, Saint Nicholas brings gifts to all good children on the eve of his memorial day (December 6).

Ge-nominalization is usually not possible with unaccusative verbs; cf. Hoekstra (1984a) and Knopper (1984). This holds not only for dyadic unaccusative (object-experiencer) verbs like ontgaanto escape, bevallento please and lukkento succeed, which resist any kind of nominalization (cf. Section 15.3.1.1, sub IIID), but also for monadic unaccusatives, which do allow inf and ing-nominalization. This means that the nouns in (181) are all unacceptable.

This conclusion is supported by the behavior of motion verbs like wandelento walk which have both an intransitive and an unaccusative use, which auxiliary selection can bring out: intransitive verbs select hebbento have in the perfect tense while unaccusative verbs select zijnto be. Now look at the examples in (182). Example (182a) shows that the verb wandelento walk normally functions as an intransitive verb when combined with a locational prepositional phrase, while it functions as an unaccusative verb when combined with a postpositional phrase. That the ge-noun gewandel in the primed examples can be combined with a prepositional phrase but not with a postpositional phrase follows directly from our claim that intransitive but not unaccusative verbs can be the input for ge-nominalization.

The primeless examples in (183) with the unaccusative verbs stervento die and trouwento marry again support our conclusion above. However, the primed examples show that an exception should be made for generic contexts. In such cases, the ge-nouns acquire an iterative aspect that is not present in the primeless examples.

Ge-nouns can be formed on the basis of verbs that select a PP-argument. Example (184b) shows that the preposition selected by the verb is inherited by the ge-noun.

Unlike other forms of nominalization, the process of ge-nominalization can add specific aspects of meaning to the meaning of the input verb; cf. Tálasi (2009), who also discusses the diachronic development of this type of nominalization. Thus, the results of ge-nominalization are durative substance nouns like gestaarstaring, gedraafrunning or gepraattalking, which express that the eventuality expressed by the input verb occurs frequently. Moreover, such nouns often have a negative connotation—they typically express some degree of irritation, condescension or “unfavorable connotation” (Kruisinga 1949) on the part of the speaker. That this pejorative effect is actually a result of the nominalization process and not part of the semantics of the input verb is illustrated in (185), where ge-nominalization has a negative effect on such neutral input verbs as pratento talk, regelento regulate/arrange and wandelento walk. Note that the use of the expressive demonstrative datthat has the effect of strengthening the negative connotation of the deverbal noun.

Of course, the pejorative effect cannot be observed with ge-nouns derived from verbs that already have a negative meaning aspect, like jengelento whine, klagento complain, leuterento drivel, mekkerento yammer, zeurento nag, zwammento twaddle, etc.; however, such verbs seem to be particularly popular as input for ge-nominalization (Mackenzie 1985a). Ge-nouns derived from verbs of sound emission are exceptional in that they lack this negative connotation (except in cases where the input verb already contains such a meaning aspect); the meaning of nouns like gefluisterwhispering, gefluitwhistling, geronkthrobbing, gezoembuzzing, humming, etc., may or may not be negatively affected by the nominalization process.

The inability of unaccusative verbs to function as input for ge-nominalization may be related to the fact that ge-nominalization is also disallowed for verbs that denote events that cannot be controlled by the participants in the event. As a result, verbs expressing an opinion such as menento think (#gemeen), achtento consider (#geacht), and vindento consider (*gevind) are excluded from ge-nominalization. The same is true for such typically [-controlled] verbs as slapento sleep (*geslaap), liggento lie (*gelig), zittento sit (*gezit), kennen/wetento know (*geken/*geweet). It is important to note, however, that ge-nominalization is only excluded in the regular use of these verbs. If the verbs can be given a [+controlled] interpretation, ge-nominalization becomes possible, which explains why the illustrations in (186) typically involve [+human] participants. Due to a clash between the durative meaning aspect of ge-nominalizations and the meaning of the input verbs, i.e. the denotation of some (unwanted) inactivity, a negative connotation is almost inevitable for such examples.

Another group of verbs systematically excluded from ge-nominalization are those with Germanic prefixes like be-, ver-, ont-, her-, which have a participial form without the prefix ge-; cf. Schultink (1978). Historically, we are dealing with the same prefix.

There seems to be a relation between the possibility of ge-nominalization and the form of the past/passive participle; this is suggested by the fact that particle verbs that form their past participles with the prefix ge- also allow ge-nominalization. However, the result is sometimes marked and often has a negative connotation. Some examples are given in Table (188).

The same thing can be illustrated by verbs with non-Germanic prefixes: they also have a past/passive participle preceded by ge-, and in most cases ge-nominalization does not lead to an outright unacceptable result in the way that ge-nouns derived from verbs with a Germanic prefix do. At worst, they are unusual, as evidenced by the fact that the cases marked as fully acceptable in Table (189) are easy to find on the internet, while the cases marked with a single question mark are rare. The ge-noun gerepatrieer is not attested, which may be due to the fact that the verb repatriëren belongs to a higher register. The search string used in our Google search (March 12, 2021) is [het ge-Vstem].

A potential problem for the suggested relationship between the possibility of ge-nominalization and the form of the past/passive participle is that example (190) shows that the ge- prefix can be found with verbs like herhalento repeat; we found more than 35 occurrences of the form [het/dat geherhaal] on the internet. The relative acceptability of this example may be due to the fact that herhalen (unlike herlezen in (187)) is not interpreted as consisting of an input verb (halento fetch) and a prefix her-, but as a monomorphemic verb.

An awkward side-effect of this suggestion is that we would expect the past participle form geherhaald to be quite common as well, but this does not really seem to be borne out: a Google count (March 3, 2012) showed that although there are about 15 unsuspicious cases of Dutch origin to be found on the internet, they are vastly outnumbered by the form herhaald (220 hits).

Ge-nominalization of inherently reflexive verbs is rare, which should not surprise us. First, many inherently reflexive verbs are prefixed with be-, ver-, ont-, her- and are therefore excluded from ge-nominalization: some examples are zich bedrinkento get drunk, zich vergissento be mistaken, zich onthouden vanto refrain from. Second, we have seen that the simplex reflexive zich cannot occur postnominally in inf-nominalizations but must be realized in prenominal position; cf. Section 15.3.1.2, sub IV. Since ing-nominalizations only take postnominal complements, the impossibility of ing-nominalizations of inherently reflexive verbs in (191) is exactly what one would expect.

However, it seems possible to use the corresponding ing-nominalizations if the postnominal PP containing the reflexive is omitted: examples like those in (192), both of which are adapted versions of attested examples, are common on the internet.

Example (193b) shows that with non-inherently reflexive verbs ge-nominalization is possible, because zichzelf functions as a regular argument comparable to Marie.