As with the referential personal pronouns discussed in Section 19.2.1, the form of referential possessive pronouns depends on person, number, and gender. With the exception of the second-person honorific uw, the singular forms can be either strong or weak. In the first case, the nucleus of the pronoun is a long vowel or a diphthong and can be stressed, whereas in the second case, the nucleus is a schwa and therefore necessarily unstressed. The plural forms, on the other hand, do not have a weak form, with the exception of the second-person plural je: we will turn to this form in Section 19.2.2.3, where we will discuss the distinction between weak and strong forms in more detail. Table 13 lists the full set of strong and weak referential possessive pronouns.

Haeseryn et al. (1997:290) suggests that the third-person possessive pronoun d’r/’r is also used as a weak plural, but no examples are given. We have reason to believe that this use of d’r/’r is not possible in standard Dutch; the strong pronoun hun in Die meisjes hebben hun\`1#d’r boeken verkocht those girls sold their books can take the subject of the clause as its antecedent, while the weak form d’r cannot and can only refer to another female person in the discourse domain. The singular weak forms me and ze, ending in schwa, are considered substandard by some speakers, but are often heard in speech (and sometimes seen in informal writing); cf. Hoeksema (2013).

Section 19.2.1.1, sub I, has shown that the indefinite/generic personal pronoun menone can only be used as a subject, and thus has no possessive counterpart; there is no form like English one’s. This section has also shown that men can serve as an antecedent of the third-person possessive pronoun zijn, as in (445).

However, example (446a) shows that the weak singular second-person possessive pronoun je can be used generically, just like the weak second-person referential personal pronoun je in (446b). This reading is not available for the strong possessive form jouw your’, which is to be expected because the strong personal pronoun jijyou cannot be used generically either.

Referential possessive pronouns usually refer to [+animate] entities. This is true not only for first-person and second-person pronouns referring to (sets including) the speaker and the listener, respectively, but also for third-person pronouns. Thus, while the examples in (447a&b) clearly have counterparts involving a possessive pronoun, using the pronominal counterparts of (447c&d) can lead to interpretive problems.

The reason for the contrast between the primed examples in (447) is that, out of context, the possessive pronoun zijn is interpreted as [+animate], or even [+human]. This seems related to the fact that the postnominal pronominal PP ervanof it in the primed examples of (448) must be interpreted as referring to a [-animate] referent.

This does not mean, however, that the possessive pronouns never have [-animate] antecedents, only that this use is more restricted. Haeseryn et al. (1997:291 ff.) correctly point out that the best result is obtained when the antecedent is an argument of the minimal (i.e. smallest) clause containing the possessive pronoun, as in (449a). We could express this more precisely in terms of binding by saying that the result is fully acceptable if the possessive pronoun is bound by a [-animate] antecedent within its local domain. Note also that possessive pronouns with an [-animate] antecedent are normally weak, just like their referential counterparts; cf. Section 19.2.1.1, sub VC.

When the possessive pronoun and its antecedent are in different clauses, there is a certain preference to use a definite article instead of the possessive pronoun, as in (450): if the reader wants to be very explicit, the noun is modified by the pronominal PP ervanof it.

Although the use of a possessive pronoun is still acceptable in example (450b), in many other cases the result may become highly questionable; cf. (451b), adapted from Haeseryn et al. (1997:292). It is tempting to assume that the difference between (450b) and (451b) is related to the common ground: while it is usually assumed that every car has an engine, there is no corresponding assumption that every problem has a solution. We leave this suggestion for future research.

Unlike the referential possessive pronouns, the remaining possessive pronouns are all derived from other pronominal forms. For example, the interrogative and relative possessive pronouns are old genitive forms of the interrogative personal pronoun wie. First, consider the interrogative examples in (452).

Given that the Dutch case system is archaic, it is not surprising that examples like (452a&b) are quite formal; more colloquial ways of expressing the same question are given in (453); these forms are also found in writing.

Nevertheless, the case-marked forms are still used productively when the noun phrase to which they belong functions as the complement of a preposition, as in (454a), which may be due to the fact that the more colloquial alternatives in (454b) and especially (454b') feel awkward.

The examples in (455) show that case-marked forms can also be found as relative pronouns, especially in the formal register; cf. Section 17.3.2.2, sub II. Note, however, that in examples such as (455b) the feminine form wier is often replaced by the masculine form wiens. A Google search (March 11, 2021) for the strings [de vrouw wier man] and [de vrouw wiens man] yielded 119 cases of the former (including several linguistic sources) and 115 cases of the latter. This shows that the two forms are chosen more or less randomly and strongly suggests that the genitive forms are no longer part of the core grammar.

There are three quantificational possessive pronouns, iederseveryoneʼs, iemandssomeoneʼs, and niemandsno oneʼs. These are shown in the primeless examples in (456), which alternate with the primed examples.

We have the impression that the use of ieders is somewhat formal compared to the use of the postnominal PP van iedereen, while the use of (n)iemands is more common than van (n)iemand. The first assertion cannot be supported by the frequency of the strings of the (a)-examples on the internet: the string in (456a) occurs more often than (456a') according to our Google search (December 1, 2020), but this is not decisive because our search does not provide insight into the registers involved. However, it may be supported by the fact that ieders is regularly replaced by the less common but regularly derived form iedereens. The second is supported by a Google search (December 1, 2020): the string [iemands recht] yielded 160 hits, whereas the string [het recht van iemand] yielded only 70 hits, including many cases in which the pronoun is modified, like het recht van iemand andersthe right of someone else and het recht van iemand die ...the right of someone who .... The string [niemands recht] returned 140 hits, whereas the string [het recht van niemand] returned no relevant results.

The examples in (457) show that a striking difference between personal pronouns and possessive pronouns is that the latter do not have reflexive forms, although there is a reciprocal form.

The lack of reflexive forms is probably due to the fact that referential personal pronouns such as hemhim cannot be bound by a co-argument, whereas the referential possessive pronoun zijnhis can be bound by a co-argument of the noun phrase of which it is part; a special reflexive form is therefore not needed. The reciprocal form of the possessive pronoun in (457c), on the other hand, is useful because this pronoun adds a reciprocal meaning aspect. Note that such a functional “explanation” for the lack of a reflexive possessive pronoun is not very strong, since such forms can be found in other languages; cf. Section 23.3, sub IB, for an illustration from Norwegian.

Possessive pronouns also have a demonstrative form: the genitive masculine form diens. The feminine counterpart of this form is dier, but it seems that this form is completely obsolete: it is certainly less common and feels extremely formal and artificial. All occurrences of diens can in principle be replaced by a referential possessive pronoun, but the reverse is not the case: diens can never refer to a subject, regardless of whether that subject is part of the same clause, a higher clause, or even another sentence (Postma 1984). While the pronoun zijn can be bound by Jan in the examples in (458), diens can only be used to refer to some other salient discourse entity. Coreference is indicated by means of italics.

This difference between the referential and demonstrative possessive pronouns explains why, despite its highly formal nature, the use of diens is still popular in written language; it resolves certain ambiguities that may arise when we use the referential possessive pronoun. This is clear from the examples in (459): whereas the referential pronoun zijn can be interpreted either as coreferential with the subject de vader van Jan or with the proper noun Jan embedded in the subject, the possessive pronoun diens has only the latter option. Note that the question mark in (459a') indicates that this is simply a less prominent reading.

The examples so far might wrongly suggest that diens behaves like referential noun phrases in that it cannot have a c-commanding antecedent. That this is actually possible is shown by the examples in (460): in (460a) a nominal indirect object functions as the antecedent of diens embedded in a direct object, in (460b) the direct object functions as the antecedent of diens embedded in a periphrastic indirect object, and in (460c) the direct object is the antecedent of diens embedded in an adverbial phrase. This means that the correct generalization is indeed the one given earlier, viz. that diens cannot be bound by a subject. Note that the use of diens can again be given a functional explanation, since all examples become ambiguous if we replace diens by the possessive pronoun zijnhis.

In (461) we give some examples involving coordination, with coreference expressed by co-indexing. The (a)-examples show that while the possessive pronoun zijn can be interpreted as coreferential with either the subject of the clause or the first conjunct of the coordinated phrase Piet en zijn dochter, this ambiguity is resolved when we use diens, which cannot be bound by the subject noun phrase Jan. The (b)-examples provide similar cases with coordinated sentences. For further discussion, see Postma (1984) and onzetaal.nl/advies/diens.php.

Finally, note that complex noun phrases and proper nouns with the genitive ending -s can alternate with possessive pronouns; these complex noun phrases usually refer to [+human] entities. Of course, these noun phrases do not function as determiners in the same sense as the possessive pronoun: they are phrases, not just words, and therefore cannot be placed in the D-position of the DP. We refer the reader to Section 19.2.2.4, sub I, for a more detailed discussion of the restrictions on the use of these prenominal genitive phrases. -