The quantificational pronouns iets'something', niets'nothing' and wat'something' in (34) are the most common nominal elements in the partitive genitive construction.

Section 7.2.1 has already shown that the partitive genitive construction can be used in the same syntactic environments as the quantificational pronoun in isolation. Partitive genitive constructions also behave like pronouns in other respects. First, the examples in (35) show that the pronouns iets, niets and wat are never preceded by a determiner (although het niets is acceptable if niets is interpreted as a noun meaning “nothingness").

Second, the examples in (36) show that modification of the quantificational pronoun by an adjective or an intensifier is excluded regardless of whether the pronoun is followed by the partitive genitive or not. Note, however, that more or less fixed phrases like een aangenaam iets'a pleasant thing' can be encountered; cf. the discussion of (15).

Example (37) shows that, although the +animate pronouns iemand'someone' and niemand'no one' belong to the same class as the pronouns in (34), they are nevertheless excluded in the partitive genitive construction. Note that Section 7.4 will discuss a construction with (n)iets and (n)iemand that can easily be confused with the partitive genitive construction.

Quantifier nouns like (hele)boel'lot', massa'mass' and hoop'lot' in (38) constitute the second class of nominal expressions that can occur in the partitive genitive construction, although speakers tend to have somewhat varying judgments about the acceptability of these cases. Just like the -animate existential quantificational pronouns, these nominal expressions can be used without the partitive genitive. Observe that the quantifier noun is obligatorily preceded by the indefinite determiner een'a'.

      The quantifier nouns in (38) may also be combined with plural or substance nouns to form quantificational binominal constructions like those in (39); cf. Section N4.1.1.

It seems that quantifier nouns can only occur in the partitive genitive construction if they can be combined with a substance noun in the binominal construction: quantifier nouns that co-occur with plural nouns only give rise to an unacceptable result in the partitive genitive construction. This is shown in (40).

Binominal constructions also occur with measure nouns like kilo'kilo', container nouns like pak'pack', part nouns like stuk'piece' and collective nouns like berg'pile', but these never occur in the partitive genitive construction, not even if they can be combined with a substance noun. Of course, examples such as een kilo/pak/berg lekkers'a kilo/pack/pile of sweets' are acceptable but this is due to the fact that lekkers also occurs as a nominalized form with the specialized meaning “sweets".

      The quantifier nouns heleboel, massa and hoop in (38) are obligatorily preceded by the indefinite determiner een. Using the plural form of the noun (provided that it has one) decreases the acceptability of the construction considerably. The same thing holds if the quantifier noun is modified by an adjective. This will become clear by comparing the examples in (42) to example (38c). The decreased acceptability is probably due to the fact that the nouns are more referential in these cases; een grote hoop boeken no longer indicates an unbounded quantity of books but is interpreted literally as “a big pile of books".

Quantifiers like veel'many/much', meer'more', weinig'few/little', minder'less', genoeg/voldoende'enough', hoeveel'how many/much', and evenveel'as many/much as' are generally used as modifiers of nouns, and the examples in (43) show that many of them can also be used without the noun, in which case they have the same distribution as the corresponding full noun phrases.

These quantifiers may also act as the nominal part of the partitive genitive construction. Some examples are given in (44).

The primed examples in (44) show that these quantifiers can be combined both with plural and substance nouns. The examples in (45) show that quantifiers like alle'all', elke/ieder'every' and cardinal numerals like vier'four', which cannot co-occur with substance nouns, cannot be used in the partitive genitive construction either.

Given that we have seen a similar distinction in the previous subsection, the examples in (44) and (45) suggest that the notion of non-countability seems to be a crucial factor in the delimitation of the set of nominal elements that may occur in the partitive genitive construction. It should be noted that the universal quantifier alle poses a potential problem for this claim: although it cannot be combined with neuter substance nouns like water in (45a'), most speakers do accept the combination of alle and a non-neuter substance noun like wijn 'wine' ; cf. N6.2.2, sub IG . Note also that elk 'each' can be used with noun like brood 'bread' or wijn 'wine' , which are normally used as substance nouns, but in this case the quantifier triggers an interpretation as a common noun; such cases can therefore be put aside as irrelevant.
      An alternative way of delimiting the relevant set is to appeal to the fact that whereas the quantifiers in (44) can be used without a following noun or partitive genitive (cf. (43)), the quantifiers and numerals in (45) cannot. The examples in (46) are a problem for such an approach, given that the more or less archaic forms allerlei/allerhande 'all kinds of' and velerlei 'many' may enter the partitive genitive construction but cannot be used without a following partitive genitive or noun; cf. Kester (1996:306). For completeness’ sake, note that some speakers tend to judge both the partitive constructions in the primeless examples and the primed examples with a substance noun as marked.

      Example (47a) shows that the quantifier cannot be preceded by a determiner, and the two (b)-examples show that modification of the quantifier by an intensifier is possible in the partitive genitive construction, provided that this is also possible if the quantifier modifies a noun.

The sequence wat voor'what sort of', which is discussed extensively in Section N.4.2.2, is often combined with a noun, which is either bare or preceded by the (spurious) indefinite article een. Being interrogative, the wat voor-phrase is generally moved into the clause-initial position, but it can also be split. In the latter case the interrogative element wat occupies the clause-initial position, the remnant phrase voor (een) N occupying a clause-internal position.

The string wat voor also occurs as the nominal part of the partitive genitive construction, and, again, both the unsplit and the split pattern occur. Modification of the wat voor-phrase by means of an adjective or an adverb is not possible.

The use of the spurious article een seems to give rise to a somewhat marked result, although it is easily possible to find examples of that sort on the internet. The table in (50) gives the number of hits that resulted from a Google search (May 2009). For comparison, it can be noted that a similar search showed that the string [wat voor] is only twice as frequent as the string [wat voor een]; the former resulted in 3.49 million and the latter in 1.77 million hits.

The constructions in (49) also seem to alternate with partitive genitive wat voor constructions that contain the quantificational pronoun iets'something'. It is not clear to us whether (49) can be analyzed as the counterpart of (51) with deleted iets. Table (52) shows that these constructions never feature the spurious article een.

      The wat voor construction can also be used as an exclamative. As is shown in (53a&b), both the unsplit and the split pattern are possible then. For completeness’ sake, observe that the finite verb of the exclamative clause must appear in clause-final position; the primed examples in (53) are excluded.

      To conclude this subsection we want to note that the fact that the wat voor-phrases in (48) may contain a substance noun, but become ungrammatical if the noun is omitted, supports the suggestion from the previous subsection that it is the notion of non-countability that functions as the crucial factor in the delimitation of the set of nominal-like elements that can enter the partitive genitive construction, and not whether that element can be used independently of the following adjective.

This section has shown that the four classes of elements in Table 1 can be used as the nominal part of partitive genitive constructions; some examples are given in the second column. Recall that it is only a subset of the quantifier nouns and quantifiers that can appear in the partitive genitive construction: those that cannot be followed by a substance noun give rise to an unacceptable result. The third column indicates whether the nominal part can be used independently as a subject or a direct object argument or whether it obligatorily enters a binominal or partitive genitive construction. The fourth and fifth columns indicate whether the nominal part can be followed by a noun, and, if so, whether it can be a substance noun in that case. The final column indicates whether the nominal part can be modified by means of an intensifier. The table does not show that modification of the nominal part by means of an attributive adjective is never possible.