Morphological agreement in xdery
In this section it is hypothesised that the nouns which occur in y must agree morphologically with der. It is suggested that this is the result of discontinuous chunking (see section 3.2), through which the genitive marker der became associated with particular derivational suffixes. The accessibility of the internal morphological structure of the nouns involved to the x der y construction, and the absence of this relationship elsewhere in modern Dutch, suggests that the agreement between der and the noun in y is an example of what Booij (2010: 211) terms construction-dependent morphology, in which a morphological marker, otherwise lost from a language, plays a new role within a particular construction (see also Scott 2012a: 99-101).
Originally the relation between der and y was number-based for plural nouns, and gender-based for singular nouns. Thus, a noun could appear in y if it was plural or if it was singular and lexically feminine. The occurrence of plural nouns in y appears to be a straightforward continuation of the original adnominal genitive construction; the occurrence of singular nouns in y in present- day Dutch is trickier to explain. The conclusion of section 3.3.1 was that a noun is placed in y on the basis of that noun’s morphological structure, not its semantics. Furthermore, the data show the occurrence of morphologically simplex words to be relatively rare in x der y. It is therefore assumed here is that a noun’s morphological structure is the key to its occurrence in x der y.
As a continuation of the discontinuous chunking that led to its survival (as proposed in 3.2), the use of the adnominal genitive construction is still based on discontinuous chunks (sub-schemata), each one specifying nouns ending in a particular suffix. Accordingly, a present-day Dutch speaker, having encountered sufficient tokens of x der y involving nouns ending in the relevant suffixes (the input) will, on the basis of the usage-based principles summarised in 1.2, generalise those tokens to identify the various x der y sub-schemata; the possibility of including an adjective (termed z in the formalisations that follow) to modify the noun in y will also be noted. This is illustrated in (16) using the example of the -heid-sub-schema. These sub-schemata can then be used to produce novel formations; at the same time, the speaker notices that there are certain register- based restrictions which characterise the adnominal genitive construction: these are the topic of section 4.
de spiraal der eenzaamheid ‘the spiral the.GEN loneliness’
de engel der gerechtigheid ‘the angel the.GEN justice’
het imago der onnozelheid ‘the image the.GEN silliness’
de grenzen der redelijkheid ‘the borders the.GEN rationality’
de wet der waarschijnlijkheid ‘the law the.GEN probability’
de zaak der elektrische veiligheid ‘the matter the.GEN electric security’
het sobere deel der kenbare waarheid ‘the sober part the.GEN known truth’
(Sources, respectively: INL 27 Mil., April 1995; INL 27 Mil., January 1994;
INL 27 Mil., May 1994; Eindhoven, spoken, 40722; INL 27 Mil., March 1994;
INL 27 Mil., February 1994; INL 27 Mil., March 1994)
[[xNP] [der [z]ADj_heidNP]] (i.e. any noun ending in -heid may follow der)
Notes: x is a string, i.e. a whole NP, of any complexity
an adjective z may be inserted between der and the noun
The identification of structures through experience and repetition is also the basis of the sub-schema for plural nouns: once sufficient plural nouns have been encountered following der, the pattern is generalised to apply to any plural noun and used as the basis of further novel formations.
The identification and entrenchment of the various sub-schemata of x der y through discontinuous chunking does not necessarily preclude non-discontinuous (i.e. prototypical) chunking of der along with certain derived nouns which occur frequently in position y. Although these nouns could be identified as suitable for x der y on account of their ending, the frequency of nouns such as bevolking ‘population’, democratie ‘democracy’, natie ‘nation’, techniek ‘technology’ and waarheid ‘truth’ in the adnominal genitive construction suggests that der bevolking, der democratie, etc., become chunked in their own right to form sub-schemata such as those in (17), in which only position x is free for the insertion of a noun phrase.
(17) [[xNP] [der bevolking]] ‘x the.GEN population’
[[xNP] [der waarheid]] ‘x the.GEN truth'
This notion also explains the occurrence of simplex nouns in position y since, clearly, discontinuous chunking is impossible with simplex nouns. A number of (formerly feminine) simplex nouns recur in x der y, such as aarde ‘earth’, eeuw ‘century’, kerk ‘church’, wereld ‘world’ and zee ‘sea’. Accordingly, chunks such as der aarde, der kerk, and so on, can become entrenched to form sub-schemata such as those listed in (18), in which only position x is open. Some sub-schemata have specific properties: the sub-schema for eeuw allows the placement of an ordinal number before eeuw to specify the century, while those for kerk and zee allow for the placement of a lexeme before the noun to specify the church or sea, respectively.
(18) [[xNP] [der aarde]] ‘x the.GEN earth’
[[xNP] [der ord.num eeuw]] ‘x the.GEN_st/nd/rd/th century’
[[xNP] [der ADJ kerk]] ‘x the.GEN ___ church’
General usage-based principles can account for the productive use of x der y; the structure of morphologically complex nouns serves as the basis for their occurrence in the construction. Given the rarity of simplex nouns in position y, the continued use of the adnominal genitive must be primarily morphologically governed. This is the basis of the formal account of the x der y construction which is put forward now in section 3.3.3.
-  For now, the adjective z has simply been placed between der and the noun to indicate thepossibility of its occurrence; its hierarchical position in the binary structure of the noun phraseis accounted for more precisely in (19).