Section 18.2.1.1, sub I, has shown that the referential possessive pronouns have a strong and a weak form. Normally these pronouns are used to refer to some discourse entity, but this subsection will show that the weak forms can also be used as a functional element (syntactic connective) when the possessor is expressed by a proper noun or a complex noun phrase.

The primeless examples in (481) show that when a prenominal possessor is a proper noun or a complex noun phrase, it can be inflected with an -s ending, which may be diachronically related to, but not identical with, the medieval genitive marker; cf. Weerman & De Wit (1998/1999: §3) and Booij (2010: §9.2) for discussion. Although the use of the inflected form is the norm in written language, it is not always the preferred option in speech. If the possessor is a proper noun or a singular noun phrase, the possessive relation is rather expressed by a weak possessive pronoun that agrees in gender and number with the possessive noun phrase, as in (481a'&b'). The judgments on the (c)-examples show that the inflected form is the preferred one when the possessor is plural, which may be related to the fact that the possessive pronoun hun does not have a weak form; the use of % signals that speakers have different judgments on examples such as (481c'), ranging from marked to perfectly acceptable.

Since the proper noun or complex noun phrase in the primed examples in (481) is a referential expression, the weak possessive pronouns have a function similar to that of the -s ending in the primeless examples. This suggests that these pronouns are not referential but purely functional, which is also supported by the fact that the kind of gender agreement found in the primed examples does not occur in earlier stages of the language; cf. Weerman & De Wit (1998/1999: §4). Because of their similarity to the genitival constructions in the primeless examples, the primed examples will be referred to as the semi-genitival construction.

A complication concerning examples with plural possessors is that the acceptability of (481c) with the inflected plural noun phrase mijn ouders may be the exception rather than the rule. This is clear from the fact that in speech the homophonous examples mijn zusters/broers boeken in (482) can only have the singular interpretation of the primeless examples, not the plural interpretation of the primed example; in writing the two readings can be distinguished by using an apostrophe after the possessor.

This means that the plural reading can only be expressed by the semi-genitival construction or, for speakers who disfavor this construction, by definite noun phrases with a postnominal van-PP, as in (483). Example (483a) shows that this construction can also be used as an alternant to (481c).

The above examples involve plural possessors that form their plural in -s. If the possessor has a plural in -en or -eren, interpretive problems like those in (482) do not arise, and we might therefore expect the genitival construction to be fully acceptable, regardless of the number of the possessor. However, this expectation is not borne out: the unacceptability of the primed examples in (484) shows that genitive constructions with plural possessors are also unacceptable in this case (although we should immediately add that we have found a number of examples of the type mijn kinderens N on the internet). What is even more remarkable (at least in view of the acceptability in written language of the primed examples in (482)) is that the primed examples in (484) are also unacceptable in written language, where we find only the form het boek van mijn vrienden and de kamer van mijn kinderen. These idiosyncratic restrictions on the use of the genitival construction (even in writing) suggest that the genitival construction is only a remnant of an older stage of the language; in the present-day language, the productive forms are the construction in (483) with a definite noun phrase and a postnominal van-PP and (at least for those speakers who allow it) the semi-genitival construction.

There are more restrictions on the genitival and, to a lesser extent, the semi-genitival constructions in (481). First, the possessor in these examples is typically a proper noun, as in (481a&b). If the noun phrase is headed by a [+human] relational noun preceded by a possessive pronoun, as in (484a) and (485a), both the genitival and semi-genitival constructions are acceptable. If the noun phrase is headed by a [+human]/[+animate] non-relational noun, as in (485b&c), the genitival construction decreases in acceptability and the semi-genitival construction becomes the highly preferred option. The (d)-examples show that a noun phrase referring to a [-animate] entity yields a strange result in both constructions; in this respect the (semi-)genitival construction is no different from those in (451c'&d') with possessive pronouns. Note that in all cases it is possible to express the possessive relation with the postnominal van-PP; cf. het wiel van de brommerthe mopedʼs wheel. Note in passing that English constructions such as yesterday’s newspaper cannot be rendered in Dutch with a genitive possessor, but only with a (non-possessive) postnominal van-PP or an adverb: de krant (van) gisteren; this construction will be discussed in more detail in Section 16.3.6, sub I.

The acceptability of examples like (484a) and (485a) is problematic for the claim in Weerman & De Wit (1998: §4) that the genitive possessor cannot contain a determiner. Furthermore, the contrast in acceptability between these cases and the cases in (485b-c) cannot be explained by appealing to the kind of determiner used in the genitive possessor (as suggested in the 1999 version of the paper), since the examples in (485b-c) do not improve significantly when we replace the definite articles with possessive pronouns; cf. ??mijn bakkers auto. This means that this contrast cannot be explained in purely syntactic terms without making ad hoc provisos for examples like (484a) and (485a). However, we can adopt Weerman & De Wit’s description that the genitive possessor must be a proper noun or headed by a common noun that can be used as a form of address. For completeness’ sake, note that the fact that the possessive pronouns in (484a) and (485a) cannot be replaced by the definite article dethe is not relevant for the present discussion, because it is due to the fact that the possessive pronoun is an obligatory argument of the relational noun; cf. Section 14.2.3.

Referential and reciprocal personal pronouns are never used in the semi-genitival construction: instead, we always find constructions with a referential or reciprocal possessive pronoun.

This does not mean, however, that there is a general prohibition against using a personal pronoun as a possessor in semi-genitival constructions. The primeless examples in (487) show that demonstrative or interrogative pronouns can be used as such provided that their referent is [+human]. The primed examples show that these forms alternate with the genitival demonstrative pronoun diens or the interrogative pronoun wiens, discussed in Section 18.2.2.1, sub II and V.

Finally, the examples in (488) show that it is also possible for the quantificational personal pronouns to enter the semi-genitival construction. In all cases the possessive pronoun z’n is used. The universal semi-genitival construction in (488b) seems impossible with the more formal pronouns ieder and elkeen, which may be due to a clash in register. The most common genitival counterpart of the semi-genitival form iedereen z’n is ieders, although iedereens is also common. The semi-genitival forms in (488a&c) alternate with the genitive forms iemands and niemands.

The observation in Section 18.2.2.1 that possessive pronouns cannot occur with determiners applies only to possessive pronouns that modify an overtly realized noun. In the absence of such a noun, the possessive pronoun is obligatorily preceded by a definite article and followed by an -e suffix; Table 14 shows that not all forms are equally acceptable. Note in passing that some Dutch dialects, but not standard Dutch, have similar nominalized forms of demonstrative and interrogative pronouns; cf. Corver et al. (2013:131).

		singular	plural
1st person		de/het mijne	de/het onze
2nd person	colloquial	de/het jouwe	*de/het jullie-e
	honorific	de/het uwe	de/het uwe
3rd person	masculine	de/het zijne	?de/het hunne
	feminine	de/het hare
	neuter	de/het zijne

possible analysis is that we are dealing with elliptical constructions; cf. Section A27.4. This may be supported by the fact that nominalized possessive pronouns are typically used in contrastive contexts such as (489), where the head noun of the subject of the first conjunct clause (i.e. autocar and paardhorse) can be used to content-identify the supposedly elided noun in the second conjunct. A further argument for such an analysis is that these examples also show that the article preceding the possessive is sensitive to the gender of the noun in the first conjunct: auto is non-neuter, and the possessive in (489a) is preceded by the non-neuter article de; paard is neuter, and the article preceding the possessive in (489b) is the neuter article het. See Schoorlemmer (1998) and Corver & Van Koppen (2010) for arguments in favor of an ellipsis analysis.

There are also reasons to assume that the possessive pronoun mijne in (489) is not part of an ellipsis construction, but the head of the noun phrase. The first is that, unlike the regular form mijn, the form mijne cannot be used when it is followed by an attributive adjective. This is shown in (490).

Since an attributively used adjective must precede the noun it modifies, the unacceptability of (490b) is to be expected if mijne is a noun, but not if it is followed by an empty noun (cf. de blauwethe blue one). The force of the argument is rather weak, however, since the adjective cannot precede the nominalized possessive pronoun either. A second reason is that, at least orthographically, the possessive pronoun can be followed by the plural marker -n (the n in the plural suffix -en is usually not pronounced in standard Dutch). The noun phrase headed by the possessive pronoun must then refer to the persons belonging to the referent of the possessive pronoun (especially his family or followers).

We will leave it to future research to determine the proper analysis of the examples in (489). For now, suffice it to say that they do not count as counterevidence for the claim that possessive pronouns are determiners; they simply function as nouns.

There is yet another potential problem with the claim that possessive pronouns are determiners that we have ignored so far, namely that the plural first-person pronoun ons seems to exhibit attributive inflection. This is illustrated in (492).

A comparison of the primeless and primed examples suggests that the inflection of ons is similar to the inflection of an attributive adjective in an indefinite noun phrase (cf. Section 16.2, sub I), which in turn suggests that ons occupies the attributive position in the noun phrase (and not the determiner position). Note, however, that the only thing we can conclude from the primeless examples in (492) is that the possessive pronoun agrees in gender and number with the head noun. In this respect, it is no different from articles and demonstrative pronouns, which agree with the head noun in the same way; what should surprise us is not that ons exhibits agreement with the head noun, but that the other forms do not. Note also that the fact that agreement with ons involves the suffix -e may be merely coincidental, and need not lead to the conclusion that we are dealing with an attributive phrase. That it is only a coincidence is supported by the fact that the attributive ending -e is absent only in indefinite noun phrases, whereas the noun phrases introduced by the referential possessive pronouns in Table 13 are always definite (cf. Section 18.2.2.2); in short, if ons were analyzed as an attributively used adjective, we would erroneously predict (492b) to be unacceptable.

Possessive pronouns appear in several more or less idiomatic constructions. Some of these are discussed in the following subsections.