The most common suffix used to form abstract deverbal nouns is -ing. This suffix is often used to derive a noun denoting the same state of affairs as the input verb. Like inf-nouns, ing-nouns can be taken to inherit the arguments of the input verbs. Two examples, based on the transitive verb vernietigento destroy and the unaccusative verb stijgento rise, are given in (137).

Although many verbs have a corresponding ing-nominalization, the process cannot be applied indiscriminately to all verbs, i.e. unlike inf-nominalization, ing-nominalization is not fully productive. Furthermore, the form of the resulting nominalization is not fully predictable; the set of ing-nouns includes the forms in Table 11. With the exception of the class of nouns ending in -ing, all classes are the result of unproductive processes, with the endings -age and -atie typically attached to verbs of non-Germanic origin. Although they do not end in -ing, we count these nouns as ing-nominalizations because of their denotation (state of affairs) and their syntactic behavior (distribution, complementation, etc.).

Semantically, the forms in (138) also seem to belong to the class of ing-nouns. However, since they formally correspond to either the stem or the infinitival form of the verb, it seems difficult to determine whether they are derived from the verbs or whether the verbs are derived from them.

Finally, there are completely idiosyncratic nominal forms like those in (139). Since these forms cannot plausibly be derived from the verbs in the first column of the table in (139), it seems reasonable to assume that these verbs cannot be the input of ing-nominalization to derive nouns with the same meaning due to lexical blocking (although bieding is used in the domain of e.g. house agency). Nevertheless, we will treat the idiosyncratic nominal forms in (139) on a par with the ing-nominalizations.

Not all nouns ending in -ing belong to the category of ing-nouns. Many nouns ending in -ing have acquired specialized meanings that, while still related to the meaning of the input verb, are associated with the nominal rather than the verbal aspects of the nouns. This particular group of derived nouns is often referred to as “result” nouns, in contrast to the “verbal” nouns in Table 11: rather than referring to the event in question, result nouns denote the (concrete or abstract) result of that event. Examples of result nouns denoting concrete objects are given in (140): the nouns do not refer to the activity of collecting, discovering, damaging, and translating as such, but to the result of these activities.

Although perhaps less clearly so, the same phenomenon is also illustrated by verenigingsociety/club, in that a club can be seen as the result of the union of a group of people. Example (141), taken from Dik (1985a), further shows that the noun ondernemingenterprise, although related to the state of affairs denoted by the input verb ondernemento undertake, has acquired a specialized meaning which makes it impossible to refer to the actual event itself with this noun; cf. (141). The noun onderneming is therefore not an ing-noun in the sense intended here.

Abstract result nouns, although intuitively closely related to the input verb, are not true ing-nouns either. Again, they fail to denote the state of affairs denoted by the verb. Examples are the lexicalized nouns in (142).

The (a)-examples in (143) show that these nouns can be modified by postnominal van-PPs and prenominal genitive noun phrases and possessive pronouns. However, in contrast to the inf-nominalization in (143b), these modifiers are not primarily interpreted as arguments of the head noun. Thus, in (143b) the jury has a hard time reaching a verdict: it is the act of convicting, a state of affairs, that poses problems. In the (a)-examples, on the other hand, it is the sentence, i.e. the result of an act of convicting performed by someone else, that the jury finds difficult to accept. This implies that the relationship between the head nouns and their modifiers in the (a)-examples is one of possession (in addition to that of noun-theme or noun-agent).

For the sake of completeness, let us add that some of the nominals ending in -ing mentioned above are ambiguous between an abstract and a concrete reading. In (144) this is shown for the nouns uitvindingdiscovery and verenigingsociety/club: the primeless examples exemplify their (concrete) result reading, the primed examples their use as ing-nouns; cf. Grimshaw (1990).

There are also person and object denoting nouns ending in -ing that are semantically related to the verb from which they derive, but do not denote the result of the state of affairs denoted by the verb. Examples of such nouns are person nouns like beschermelingprotégé, zuigelingbaby, leidingmanagement and object denoting nouns like leuningrailing, sluitingfastener, leidingpipe/wire. These nouns behave like normal simplex nouns: not only do they have all the nominal properties like (in)definiteness, pluralization, etc., but they also lack an argument structure: despite their obvious relation to some verb, there is no inheritance of arguments. This is illustrated in (145) and (146) by comparing leidingmanagement and leuningrailing with the corresponding det-inf nouns.

Nouns ending in -ing can have a number of denotations, which are listed in Table 12. In the remainder of this section, we will only be concerned with what we have called ing-nominalizations, i.e. with deverbal nouns denoting a state of affairs.

For completeness, note that there are also nouns ending in -ing that are not derived from verbs, such as dorpelingvillager and ellendelingwretch (which have a nominal base), stommelingfool and zwakkelingweakling (which have an adjectival base), and tweelingtwins (which has a numeral base). Obviously, they will not be included in the following discussion.

Ing-nominalizations can be both indefinite and definite, and can co-occur with various definite [-neuter] determiners like the definite article de, the demonstratives deze/diethis/that and possessive pronouns. They can also be modified by quantifiers like elke/iedereeach/every, alleall, veel/weinigmany/few and cardinal numerals. Some examples are given in (147).

Ing-nouns can also be preceded by interrogative determiners like welke, and (148a) shows that they can be wh-moved as a result. Example (148b) shows that they can also be topicalized.

Pluralization of ing-nouns is possible, but often leads to a marked result. For instance, in the examples in (149), where the theme argument is explicitly mentioned, Dutch seems to prefer the use of a compound noun.

In contexts with implied (e.g. contextually recoverable) arguments, or with adjectivally modified ing-nouns, pluralization is perfectly acceptable; this is shown by the examples in (150).

Generic contexts also allow pluralization of ing-nouns, as shown by example (151). Recall that the noun overname also counts as an ing-noun due to its abstract denotation; cf. Section 15.3.1.3, sub IA.

Finally, note that pluralization of result and object denoting nouns ending in -ing like ondernemingcompany/enterprise and leuningrailing (cf. Section 15.3.1.3, sub IB/C) is often possible. This may be an additional justification for not including these nouns in the set of ing-nouns.

Ing-nouns also behave like nominals with respect to adjectival modification: the obligatory presence of the -e suffix on the prenominal adjectives in (152a&b) shows that we are dealing with adjectival modification of a nominal, not adverbial modification. Note, however, that the primed examples show that modification by adjectives expressing frequency or duration is also possible, which still may be indicative of a verbal meaning aspect in these nominals.

An important difference between inf and ing-nominalization is that the former is not fully productive. This is immediately clear from the fact that ing-nouns cannot be derived from many intransitive verbs: the intransitive verbs given in the primeless examples of (153) have no corresponding ing-noun. Note that the deverbal nouns in the primed examples preceded by the number sign “#” do exist, but not with the intended meaning, i.e. they do not denote the same state of affairs as their verbal stem.

A possible counterexample might be the ing-noun aarzelinghesitation; cf. Jan aarzeltJan hesitates and Jans aarzelingJans hesitation’. Note that the restriction does not extend to intransitive verbs that take a PP-complement; cf. Subsection E. Thus, on the basis of the primeless examples in (154), it could be argued that the verb aarzelento hesitate takes an optional CP or PP-complement, and as such does not belong to the class of true intransitives.

For now, we can conclude with some certainty that verbs that do not have an internal argument do not allow ing-nominalization.

An example of ing-nominalization of a transitive verb is given in (156), where the deverbal noun behandelingtreatment inherits the argument structure of the monotransitive verb behandelento treat, in the sense that the arguments of the input verb can be optionally expressed in a postnominal PP. The derived form is given the category ing-n, rather than N, to express its special nature, with its combination of nominal and verbal features.

The examples in (156) show that the agent and the theme argument can be realized either by a prenominal genitive (or possessive pronoun) or by a postnominal PP if they have the form of a [+human] proper noun.

In contrast to what is the case with inf-nominalizations, however, the theme argument of an ing-noun never appears in the form of a prenominal accusative noun phrase: the ing-nominalization *De Jan/patiënt behandeling is not possible although this may be obscured by the fact, discussed in Subsection F, that the compound patiënt(en)behandeling is attested. The above examples have shown that ing-nouns form a hybrid category: they retain the argument structure of the input verb, but seem to lose the ability to assign structural case. We refer the reader to Section 16.2.3.3, sub I, for more details.

Ing-nominalization of ditransitive verbs such as uitreikento present also preserves the argument structure of the input verb in the sense that the arguments of the input verb can be expressed in a postnominal PP. The arguments are normally optional, although the expression of the agent or recipient seems best when the theme is also expressed. Expressing all three arguments, as in (157), is acceptable but not common.

Again, in contrast to inf-nominalizations, the theme argument of an ing-noun never appears in the form of a prenominal accusative noun phrase: the ing-nominalization *De (winnaars/hun) prijzen uitreiking is not possible, although this is again somewhat obscured by the fact that the compound prijsuitreikingprize-giving ceremony is attested. The above examples have again demonstrated that ing-nouns form a hybrid category.

The ing-noun aankomstarrival in (158) is given an argument structure similar to that of the unaccusative input verb aankomento arrive. The two (b)-examples show that the inherited theme argument can appear either postnominally in the form of a van-PP. If the argument is [+human], it can also appear in the form of a genitive noun phrase or a possessive pronoun.

ing-nouns can also be derived from verbs that select a PP-theme such as jagen (op)to hunt. As can be seen from example (159), the preposition selected by the input verb is inherited by the ing-nominalization. In these constructions the PP-theme can only occur in postnominal position.

It is quite common for the theme arguments of an ing-noun to be incorporated into the noun. As example (160) shows, this is possible regardless of the type of the input verb. As expected, such incorporation results in a reduction of the number of arguments of the derived noun, since the argument slot of the incorporated argument is no longer available.

Note that examples such as (161) are acceptable, but this does not refute the claim that incorporation results in valency reduction, since the noun phrase de benzinethe petrol is clearly not the theme of the construction, as is clear from the contrast between the two (b)-examples in (161b).

Example (162) illustrates that inherently reflexive verbs cannot normally undergo ing-nominalization. This need not surprise us, since Section 15.3.1.2, sub IV, has shown that the reflexive pronoun cannot occur postnominally in inf-nominalizations, but must be realized in prenominal position. Since ing-nominalizations only take postnominal complements, the impossibility of ing-nominalization of inherently reflexive verbs is exactly what one would expect. Note that the addition of the emphatic element zelf, which corresponds to English himself in he himself, does not improve the result of the (162b), which means that in this case we cannot assume that the restriction is of a phonological nature.

The unacceptability of the ing-noun vergissing in (163b), derived from the inherently reflexive verb zich vergissento be mistaken, is in line with what we saw in (162). Note, however, that vergissingmistake exists as a “non-verbal” noun, in which case it does not seem easy to combine it with a PP-complement: Zijn vergissing (?in de weg) bleek fataal His mistake turned out to be fatal.

The ing-noun verbazingsurprise, derived from the inherently reflexive verb zich verbazen (over)to be surprised (at), patterns slightly differently: example (164b) shows that verbazing does not allow the expression of the reflexive pronoun, but does allow the expression of the PP-complement.

Finally, example (165) shows that ing-nominalization is possible when the input verb is not necessarily inherently reflexive. Verbs like verzorgento take care of and verdedigento defend can take either the simplex reflexive zich, which can be seen as a part of the verb, or the complex form zichzelfhimself, which can be seen as a regular argument of the verb, just like the lexical noun phrase Marie. The acceptability of the primed examples is therefore expected.

The classes of verbs in (166) also resist ing-nominalization. This list is far from complete, but it gives an idea of the nature of the restrictions on this kind of nominalization. Again, the nouns preceded by the number mark “#” do exist, but not with the intended meaning, i.e. they do not denote the same state of affairs as the verbal stem.

The abstract nouns haathatred and verblijfstay do exist, but should probably not be seen as nouns derived from the stative verbs hatento hate and verblijvento stay; cf. Sections 15.2.2.2, sub IE, and 16.1, sub V. A large number of verbs that do not allow ing-nominalization do accept this process after prefixation or incorporation of a particle, although the resulting ing-noun often has a specialized meaning. Some examples are aantastinginfringement, beproevingordeal, herzieningrevision, overhoringexamination, bedenkingobjection, overdenkingcontemplation, verdenkingsuspicion, herdenkingcommemoration, ontluisteringdisillusion, opzeggingcancellation, aanbiedingdiscount, toenameincrease, bevliegingwhim, etc.

Even if we take into account the exceptions discussed above, this does not imply a general applicability of the process to all remaining verbs, as illustrated by the unacceptability of the following derivations from transitive verbs: *neming (nemento take), *eting (etento eat), *vergeting (vergetento forget). In fact, there are many more unexpected cases, such as *wachting (wachtento wait) based on a verb selecting a PP-complement. Since the lexicon does not seem to provide an alternative for these forms, we cannot appeal to “lexical blocking” and must assume that these are cases of accidental “lexical gaps”.

Sometimes there are two forms of derived nouns based on the same verb. In all such cases, however, there is a difference in meaning between the two forms. Moreover, in most cases neither form can be considered a proper ing-noun in the sense of denoting a state of affairs. Some examples are given in (167).

For now, it seems we must conclude that the exact nature of the restrictions on the productivity of ing-nominalization remains something of a mystery. For more information, we refer the reader to Van der Wouden (2020).