• Dutch
  • Frisian
  • Afrikaans
Show full table of contents
Synthetic compounds

Synthetic compounds are complex words that appear to be formed by the simultaneous application of compounding and derivation. A clear example is the adjective blauwogig blue-eyed. There is no compound *blauwoog from which this word could have been derived, nor a derived word *ogig that could function as the head of the compound.  Other examples are:

Example 1

a. kortademig
short of breath
b. langbenig
c. tweelettergrepig

A second class of synthetic compounds consists of compounds headed by a deverbal noun or adjective:

Example 2

a. initieafnemer
b. tijdrovend

Deverbal nouns such as *nemer do not occur by themselves, but only in compounds. Hence, it appears that in the formation of these complex words compounding and derivation are combined.


Adjectival compounds of the type breedgeschouderd broad-shouldered can only be analyzed as adjectival compounds with a possible word as head: *geschouderd does not exist by itself, nor its base, the verb *schouderen. So it appears that the conversion of the noun schouder shoulder to a verb which is the formal basis of the participial adjective is dependent on this specific morphological context. This can be expressed by the following template that indicates the implied conversion of a noun to a verb:

[A [[ge [[N](V)]d](V)](A)](A)

[hide extra information]

The analysis of synthetic compounds has been the subject of a debate among students of Dutch morphology (Botha 1984; Booij 1988; Hoeksema 1984; Van Santen 1992; Hoekstra and Van der Putten 1988). One analysis that is clearly incorrect is the derivation of blauwogig from a NP blauw oogblue eye. This option has been discussed with respect to similar compounds in English, and is meant to express the fact that the adjective blauw does not modify *ogig but oogeye. However, this solution is impossible because the adjective in such complex words is never inflected: a common gender noun such as voet foot always requires an adjective ending in -e; nevertheless, we do not find inflection in snelvoetig fast-footed.

What remains open as an analytical option is the interpretation of such adjectives as being derived from a possible, but non-existing compound. In other words, we then claim that one word formation process, the addition of the suffix -ig to a noun, triggers the application of another word formation pattern, that of nominal compounding. That is, there would be a word formation template of the following form:   [[A N](N)ig](A).

An alternative analysis also proposed in the literature mentioned above is that blauwogig is an adjectival compound of which the head is a possible, but non-existing derived adjective. The reason why *ogig does not exist by itself is semantic: there is no use for such a compound because all human beings have eyes, and hence such an adjective cannot be used to refer to a distinctive property. It is only in combination with blauw that a distinguishing property is referred to.

As to the special semantic property that the adjective has to be interpreted as a modifier of the base of the complex adjectival head, there is independent evidence that modifiers may have this kind of restricted scope. Dutch has a productive class of adjectives of which the head has the form of a past participle which is not necessarily an established word for the user of such compounds: an adjective such as breedgeschouderd can be used without having ?geschouderd in one's mental lexicon.

Example 3

a. breedgeschouderd
b. kortgerokt
c. witgejast

In these words, the adjective does not modify the whole head, bot only the nominal base. That is, the adjectival non-head may have restricted semantic scope with respect to the head. Such a restricted semantic scope can also be assumed for blauw in blauwogig.

The same reasoning can be applied to compounds such as tweejaarlijks biannual and driewekelijks three-weekly. The suffix -lijks is unproductive, and only occurs with nouns like week week, maand month, and jaar year. Hence, tweejaarlijks must be analysed as twee + jaarlijks instead of tweejaar + lijks. However, it is clear that the scope of twee is over the nominal base of jaarlijks, jaar. Similarly, in the compound adjective driedimensioneel three-dimensional, the quantifier drie pertains to the base noun of dimensioneel, dimensiedimension (see also relational adjectives).

  • Booij, Geert1988The relation between inheritance and argument structure: deverbal -er-nouns in DutchEveraert, Martin, Evers, Arnold, Huybregts, Riny & Trommelen, Mieke (eds.)Morphology and modularity. In honour of Henk SchultinkDordrechtForis Publications57-74
  • Botha, R.P1984Morphological mechanisms. Lexicalist analyses of synthetic compoundingPergamon Press
  • Hoeksema, Jack1984Categorial morphologyGroningenPh. D. dissertation, University of Groningen
  • Hoekstra, Teun & Van der Putten, Frans1988Inheritance phenomenaEveraert, Martin, Evers, Arnold, Huybregts, Riny & Trommelen, Mieke (eds.)Morphology and modularity. In Honor of Henk SchultinkDordrechtForis163-186
  • Santen, Ariane van1992Produktiviteit in taal en taalgebruikLeidenUniversity of Leiden