Review of Linking Elements in Compounds
The focus of this book is on linking elements in Dutch compounds and the regional variations of these linking elements in speech perception and production. The purpose of the book is to investigate the relation of the linking element -en- in spoken Dutch compounds and the homophonous plural affix -en in Dutch. Earlier works have shown that these two elements are similar in written Dutch (Neijt et al, 2004). There are two questions central to this work: the first is whether speakers of standard Dutch and speakers from other regions of the Netherlands interpret the variants of the linking -en- in noun-noun compounds as a plural marker; and the second is how rhythm and the interpretation of written and spoken Dutch compounds influence the linking -en-. The book is presented in seven chapters. Each chapter is self-contained and a complete study, and each, except Chapter 7 has its own abstract at the beginning of the chapter and references at the end.
Chapter 1 explains the aim and objective of the study, and gives a brief history of the study of the homophonous plural -en and linking element-en-. The author, Hanssen, sets out to investigate two things: the question of whether the linking -en- in spoken Dutch causes a plural interpretation, regardless of spelling; and the rhythmic function of the linking -en-.
Chapter 2: ‘The similarities of plural endings and linking elements in regional speech variation in Dutch’
Here, Hanssen shows that speakers from different regions of the Netherlands pronounce the Dutch plural suffix differently. However, in the study conducted by the author, participants from 5 regions of the Netherlands showed systematic relations between the pronunciation of the plural suffix -en and the linking element -en-. The results of the study showed that Dutch speakers in general often do not distinguish the plural form -en from the linking -en- and that they regard both, possibly, as the same morpheme.
Chapter 3: ‘Morphological differences in Frisian-Dutch bilinguals: (dis)similarity of linking elements and plural endings’
In this chapter, Hanssen shows the homography of the linking -en- and the plural suffix -en in standard Dutch but states that this homography does not exist in Frisian, which is closely related Dutch. The question of whether Frisian-Dutch bilinguals keep these systems separate during speech production was tested in two sets of bilinguals: Dutch-Frisian bilinguals and Frisian-Dutch bilinguals. The Dutch-Frisian bilinguals were found to maintain the homophony of the plural -en and the linking -en- when speaking Frisian. However, the Frisian-Dutch bilinguals maintained the system lacking homophony and kept the two separate. The conclusion is that most Frisian-Dutch bilinguals distinguish the plural ending -en from the linking -en- when they speak Frisian, but not when they speak Dutch. The author is able to show, through these studies, that the linking -en- and plural -en are homophonous for Dutch speakers but not always for Frisian-Dutch bilinguals.
Chapter 4: ‘Regular noun plurals as modifiers in spoken Dutch compounds’
This chapter investigates the interpretation of the linking -en- in Dutch nominal compounds. So far, the relation between the plural -en and the linking -en- has been investigated in writing. This chapter investigates the same relation in speech. The study in this chapter was carried out using speakers from four regions of the Netherlands: North, Northeast, Middle and South. The study shows that, given the fact that plural formation is regular in Dutch and the homophonous linking -en- shows the plural interpretation in written compounds, speakers of Dutch from these four regions pronounce the -en in phrases and in compounds identically. This gives rise to the plural interpretation of the linking -en- by speakers of Dutch, leading to the conclusion that regular plurals can occur as modifiers in Dutch compounds, contrary to Kiparsky's (1982) and Pinker's (1999) theories that compound formation is constrained by regular inflection.
Chapter 5: ‘Regional origin affects the interpretation of linking elements in spoken Dutch compounds’
The author shows in this chapter that a speaker's regional origin affects the interpretation of regional speech variants. The linking element -en- has the regional variations [ә], [әn], and the syllabic velar nasal in spoken Dutch compounds, which produce interference for speakers from the North, Northeast and South of the Netherlands. This means that they invariably interpret all regional speech variants of the Dutch linking element as a plural marker. Speakers from the Middle region showed interference effects for the variants [әn] and the syllabic velar nasal but not for the linking [әn], even though they realize both plural -en and linking -en- as [ә]. According to the author, this may be because, for this category of speakers, the linking [әn] in Dutch compounds performs a rhythmic function within the compound, thereby creating an alternation of stressed and unstressed syllables. The greatest interference effect was shown by speakers from the North, where Frisian is spoken and there is no homophony of the plural -en and the linking -en-. Speakers from the South showed the largest interference effects for the linking syllabic velar nasal, as this is an unfamiliar speech variant for them. This shows that in everyday speech, speakers of different but related linguistic backgrounds often arrive at different interpretations, occasioned by regional differences.
Chapter 6: ‘Semantic and prosodic effects of Dutch linking elements’
Three studies were carried out in this chapter in order to establish the semantic and prosodic effects of the Dutch linking element. The first study investigated two types of compounds with linking -en-: pseudo-compounds and novel compounds, that is, new combinations of existing words. The purpose was to establish the role of rhythm on pseudo-compounds. The second test looked at written existing compounds with and without linking -en-, with the aim of finding out the role of plural semantics and rhythm in written existing compounds. The third study investigated the role of plural semantics and rhythm in spoken existing compounds with and without the linking -en-. There are two views reported by Hanssen regarding the function the Dutch linking-en-: as a marker of plurality and as having a phonological effect. According to Krott (2001), the stress pattern of the left constituent does not reliably predict the occurrence of the linking-en- in existing compounds. On the other hand, Neijt and Schreuder (2007) state that rhythm does influence the choice of linking -en-, i.e., linking -en- is used when this improves the overall rhythmic pattern of the compound. In terms of conceptual plurality, the results of the studies show that a linking-en- in existing compounds increases estimated conceptual plurality, i.e., a compound with a linking -en- is considered 'more plural' than one without a linking -en-. One major problem of this conclusion is that the results violate the words-and-rules theory of Pinker (1999), which predicts that irregular plurals like ‘mice’ and ‘men’ can serve as input for compounds but regular plurals like ‘boys,’ ‘cats’ and ‘dogs’ cannot ( mice-eater, but *rats-eater). Studies from Dutch and German (Koester et al. 2007) show that plural modifiers can appear in a compound. Finally, rhythmic effect was shown to be present in pseudo-compounds but not in written novel and existing compounds. This is explained in terms of processing. Phonology was not necessary to grasp the correct meaning of written novel and existing compounds since there is a direct relationship between the constituents of the compound. Rhythm is therefore not likely to play a role as it does in spoken compounds and in pseudo-compounds. There is no relationship between the constituents in pseudo-compounds, and thus, they would have to rely on phonology for their interpretation.
Chapter 7: ‘Summary and conclusion’
In this chapter, the author summarises the purpose and the major findings of the work. To test the (dis)similarity of the Dutch plural -en and linking element -en- in Dutch compounds, various tests were carried out with participants from five regions of the Netherlands regarding: (a) the pronunciation of the homophonous -en-; (b) its perception (i.e. does -en- induce a plural interpretation of the leftmost constituent of a compound); and (c) the influence of rhythm on interpretation. The author then discuses the results in the light of current research on morphology and proposes topics for further research.
In the author's own words, the book is a thesis. Each chapter between 2 and 6 represents an independent study and is self-contained such that it is possible to read one chapter without reference to the others. Each has an abstract, an introduction to and a description of the specific task in the chapter, along with an impressive list of references at the end. What is commendable about the work is how the author is able to tie the results of all the studies together into a coherent whole, as demonstrated in Chapter 7.
This work has made a great contribution to the on-going debate about the nature, production and interpretation of the plural suffix -en and the linking element of noun-noun compounds. One of the major merits of the book is that Hanssen extended the study of the linking elements in written compounds to spoken compound words and then compared the production of these elements in the speech of Dutch-Frisian and Frisian-Dutch bilinguals. Hanssen thus elaborates on the study of linking elements in Dutch compounds from psycholinguistic and sociolinguistic perspectives.
As a non-speaker of Dutch, I find many of the arguments in this work fascinating. In Chapter 4, Hanssen showed that the plural-en and linking -en- are homophonous for speakers of Dutch, which usually leads to the plural interpretation of the linking -en- in compounds. On the other hand, Koester et al. (2004) establish in their study that Dutch speakers do not process the linking elements that are homophonous with plural morphemes as plural morphemes. Prosodic cues differentiate single nouns and compounds, and are used to disambiguate linking elements and plural morphemes. Neijt (2003) gives a few examples of words which may or may not have a linker, and states that the linker does not in fact solve stress clashes: gordijnwinkel / gordijnenwinkel (‘curtain shop’), kamelenhaar – kameelhaar (‘camel hair’), and oogpotlood – ogenpotlood (‘eye pencil’). Although I am hard-pressed to find an explanation for why this should be, it seems to me that since both are acceptable, the only reason to choose one rather than the other is whether or not the left constituent is interpreted as plural or singular. Hanssen et al.’s (2013) Dutch native speakers and second language learners of Dutch show a preference for the linking -en- in compounds. However, whether or not they interpret it as a plural marker or a linker is another matter. The implication for level ordering is clear; a regular plural noun cannot serve as input to a compound. In Dutch, however, it can. In this case, Dutch, (and perhaps German) is an exception to this rule. On the other hand, the minimalist programme allows each language to set its own parameters. Further research should go in this direction.
The investigation into linking elements in Dutch has implications for cross-linguistic studies of similar elements. In Yoruba, for example, a language within the Niger/Congo family, there are linking elements whose presence in a compound affects the semantics of the compound whether or not the linker has a dictionary meaning.
ki= ? ilé 'house' ilékílé ' any house' ; omo 'child' omokomo ' any/useless child'
In some cases, when a linking element is introduced into a compound, the result is a phrase:
erú u obinrin
'slave POSS woman ' the woman's slave'
erúbinrin ' a female slave'
(Ogunkeye 2002, 2004)
Such linkers normally assume the shape of the preceding vowel and often retain the original function as a genitive morpheme.
Compounding is a very productive word-formation process and is widely studied cross-linguistically. The book presupposes familiarity with issues in compounding, especially concerning noun-noun compounds in Dutch. The author presents an impressive amount of data covering existing noun-noun compounds, novel compounds and pseudo-compounds. I would have liked to see more examples of noun-noun compounds where the first constituent is a mass or uncountable noun (e.g. ‘waterpot’ or ‘skyscraper’) to see what implication that would have for the choice of linking element. Nevertheless, the effort of the author in this work is highly commendable. The book will appeal to linguists who are interested in word-formation processes and are keen on adequate description of derived words, psychologists and cognitive scientists who are interested in language and thought as well as sociolinguists who are interested in regional language variations.
Hanssen. E., Neijte A., Schreuder, R. 2013. Preference for linking element -en- in Dutch noun-noun compounds: native speakers and second language learners of Dutch Morphology Volume 23, Issue 1, pp 33-56
Kiparsky, P. 1982. Lexical morphology and phonology. In I. S. Yang (Ed.), Linguistics in the morning calm (pp. 3-91). Seoul: Hansin
Koester, D., Gunter, T. C., Wagner, S., & Friederici, A. D. 2004. Morphosyntax, prosody, and linking elements: The auditory processing of German nominal compounds. Journal of Cognitive Neuroscience, 16(9), 1647-1668.
Krott, A. 2001. Analogy in morphology. The selection of linking elements in Dutch compounds. Nijmegen: PhD Dissertation, MPI Series in Psycholinguistics
Krott A. 2002 Effects of rules and analogy on processing Dutch compound words “www.sciencedirect.com/science/article/pii/S0093934X01925581”
Neijt Annette 2003. Linking schwa in Dutch compounds: a phonomorpheme?
“www.let.rug.nl/koster/DenBesten/Neijt.pdf” Retrieved 16 October 2013
Neijt, A., Schreuder, R., & Baayen, R. H. 2004. Seven years later. The effect of spelling on interpretation. In L. Cornips & J. Doetjes (Eds.), Linguistics in the Netherlands, 2004 (pp. 134-145). Amsterdam/Philadelphia: John Benjamins Publishing Company.
Neijt, A., & Schreuder, R. 2007. Rhythm versus Analogy: prosodic form variation in Dutch compounds. Language and Speech, 50, 533–566.
Ogunkeye, O.M. 2002. A Lexicalist Approach to the Study of Aspects of Yorùbá Morphology. PhD Thesis. University of Ibàdàn.
Ogunkeye, O. M. 2004 The Lexical Organisation of Yorùbá Compounds. Journal of Nigerian Languages and Culture. Volume 6 No. 1 pp 88-99
Pinker, S.1999. Words and rules: The ingredients of language. New York: Basic Books
| ABOUT THE REVIEWER:
I received my M.A and PhD in Linguistics from the University of Newcastle-Upon-Tyne, England and the University of Ibadan, Nigeria respectively. I teach Linguistics at the University of Jos, Nigeria. My research interests include generative morphology, lexicography and the interfaces between morphology/syntax, and syntax/semantics. I have worked on aspects morphosyntax of Yoruba as well as intercultural communication.