Your ns are numbered: On linking morphemes in Dutch

Proceedings of NELS 50, 2020

This paper provides a unified account of inflectional markers and linking morphemes in Dutch and German. We show that supposed differences between the two are a matter of degree and not categorical. We argue that both linking morphemes and the corresponding inflectional markers are instances class markers that interact with the inflectional system.

2008

Studia Linguistica, 2015

In this article we provide empirical evidence against the claim that morphology contrasts with syntax in dealing with items that are listed in the lexicon. (Jackendoff 1975, Aronoff 1976, Jensen and Stong‐Jensen 1984). More specifically, we distinguish between three types of ANN compounds in Dutch. We show that the structural properties of these types do not show a one‐to‐one mapping with lexical properties, such as having a listed or even idiomatic meaning (see DiSciullo & Williams 1987). On the basis of this, we argue that conclusions on the structure of certain morphologically complex word‐forms should be based on structural properties and not on lexical properties such as idiomaticity or being lexicalized. We propose a syntactic derivation for all types of ANN compounds in Dutch (pace Ackema and Neeleman 2004). Structural differences follow from the level of merge: what we traditionally call morphology is syntax below the functional domain.

1999

2003

In Dutch, compounds are formed with or without linking elements, cf. zin+s+bouw ‘sentence structure’, woord+en+boek ‘dictionary’ (lit. ‘word book’) and woord+bouw ‘word structure’. The use of linking elements has been the subject of investigations since the dissertation on Dutch compounds by van Lessen (1927), who concludes that linking elements are historic relics of stem allomorphy and case. Rule-based approaches taking a synchronic point of view (Mattens 1970, 1984 and 1987 and van den Toorn 1981 and 1982) conclude that no strict rules, but only tendencies can be formulated, and Krott (2001) shows that the combined effort of these tendencies explains only 32% of the distribution of linking elements in the compounds found in the CELEX database (Baayen, Piepenbrock and Gulikers 1995). Still, speakers of Dutch know how to use linking elements in existing compounds and although variation in the use of linking elements exists, it is not dominant; for most compounds only one form is in...

1991

This paper presents a survey of the patterns of compounding in Dutch. Nominal and adjectival compounding are productive, verbal compounding is not. Inflection and derivation can both precede and follow compounding, and hence they cannot be ordered in terms of levels in the lexicon. Phrases are allowed in non-head position, and thus there is no complete separation of morphology and syntax. The head constituent is always in the right position. Word-internal coördination is also possible. de voetbal 'the football' het avondbal 'the nightball'

Journal of Child Language, 1980

ABSTRACTThe acquisition of the morphological rules for plural, agentive, and diminutive suffixes in Dutch was studied. Subjects included 7- and 12-year-old native speakers, and second-language learners in three age groups (5–10 years, 12–18 years, and adult). The first- and second-language learners showed very similar orders of acquisition for the rule systems governing plural and diminutive, but the second-language learners showed a subtle form of interference from their first language in acquiring the agentive. The findings suggest that morphological acquisition proceeds piecemeal, with the learning of specific word ending + allomorph sequences, and that generalizations at the level of morphological rules may not be made even after several years of correct performance with the allomorph in question.

Folia Linguistica, 2000

As in many other languages, the constituents of nominal compounds in Dutch are often separated by a linking element. This study investigates to what extent form and semantic properties of the right constituents in Dutch compounds affect the choice of the linker Using both lexical statistics and experimentation, we show that the left and right constituent families affect the choice of the linker independently of the semantic categories of the left and right constituents themselves. We also show that the choice of the linker is co-determined by the animacy and concreteness of the left constituent. No role for the semantic class of the head constituent was observed in the experiment. Apparently, linkers are non-canonical Suffixes in the sense that their occurrence is codetermined by the form properties of the constituent to their right.

Brain and Language, 2002

In this study, we use the association between various measures of the morphological family and decision latencies to reveal the way in which the components of Dutch and English compounds are processed. The results show that for constituents of concatenated compounds in both languages, a position-related token count of the morphological family plays a role, whereas English open compounds show an effect of a type count, similar to the effect of family size for simplex words. When Dutch compounds are written with an artificial space, they reveal no effect of type count, which shows that the differential effect for the English open compounds is not superficial. The final experiment provides converging evidence for the lexical consequences of the space in English compounds. Decision latencies for English simplex words are better predicted from counts of the morphological family that include concatenated and hyphenated but not open family members.

2001

Implementation (NSDI '16) is sponsored by USENIX.