Eric Hoekstra
Review of: Morphological Atlas of Dutch Dialects
(G. de Schutter, B. van den Berg, T. Goeman and T. de Jong, 2005, MAND. Morfologische Atlas van de Nederlandse Dialecten. Deel I. Meertens Instituut. Koninklijke Academie voor Nederlandse Taal- en Letterkunde. Amsterdam University Press, Amsterdam.")
The Meertens Institute is currently working on three atlas projects:
The SAND (Barbiers et al 2005)
is a syntactic atlas of Dutch dialects ("Syntactische Atlas van
de Nederlandse Dialecten", ). It will consist of two volumes.
Volume 1 was reviewed in Hoekstra (2006).
The FAND (Goossens, J., J.
Taeldeman en G. Verleyen 1998, 2000; Goossens 2002) is a
phonological atlas of Dutch dialects ("Fonologische Atlas van
de Nederlandse Dialecten").
The MAND (De Schutter, Van den
Berg, Goeman and De Jong 2005) constitutes the morphological
counterpart to the SAND and the FAND.
According to the text on the paper cover of the MAND, the three atlases are based on recent field work. However, the MAND does not specify which time span is covered by the recordings on which the MAND is based. It is a long-term project of the Meertens Institute, having its origins in the eighties. Sociolinguists would have liked to know in which years the recordings underlying the MAND were made, so as to be able to ascertain whether the maps reflect the state of affairs in Dutch dialects at a given time, or whether it represents a sketch based on developments in the past twenty years. Similarly, I also could not find information regarding the background of the speakers. Were they old or young, male or female? How representative were their speech data for the area which they represent? How were these speakers selected? Thus the atlas is less valuable both from a quantitative point of view and from a sociolinguistic point of view. Later I found out that this information, which is missing in the printed atlas, is available on the website of the MAND: http://www.meertens.knaw.nl/projecten/mand/. The website informs us that the recordings were collected between 1980 and 1995. Speaker data can also be gleaned from the website. The website is really an excellent piece of work, providing lots of information in a way that can be easily retrieved, so it is quite user-friendly. More of the information presented on the website could have been included in the atlas and/or its commentary, especially since the pair (atlas + commentary) is not cheap. Volume I is 95 euro.
Volume I of the MAND is devoted to nominal morphological properties.
Chapter 1 deals with the morphological expression of plurality on nouns ("Meervoudsvorming in het Nederlands).
Chapter 2 deals with the formation of diminutives ("Diminutiefvorming in het Nederlands").
Chapter 3 deals with the expression of gender on articles, adjectives and nouns ("Genus (woordgeslacht) van Zelfstandige Naamwoorden, Lidwoorden en Bijvoeglijke Naamwoorden").
All three chapters have a strong historical bias. Data, maps and interpretation are primarily determined by their relevance for historical linguistics. Thus, frequent reference is made to reconstructed Old Germanic stem forms and their subsequent development, as can be gleaned from documents composed in the written language Middle Dutch. However, the bulk of Middle Dutch consists of documents written in one of three southern dialects: Flemish, Brabantian and Limburgian. The Dutch standard language still reflects, to some extent, its southern origin. Thus the historical data from Standard Dutch, if anything, are far more relevant to southern dialects than to the coastal dialects of The Netherlands (Zeeland, Randstad) or to the northern dialects (Noord Holland, Friesland, Groningen) or the eastern dialects (Drente, Overijssel, Gelderland). Historical developments for the northern dialects have to reckon with a strong Frisian substrate, and the eastern dialects derive historically from Low Saxon. Thus, historical developments for the northern dialects should reconstruct the Frisian substrate on the basis of Old Frisian and on the basis of a theory about the nature of the substrate. In the case of Frisian (province Friesland), the dialects derive directly from Old Frisian, in the case of Noord Holland and Groningen, the Frisian substrate is certainly relevant (Hoekstra 2001 and others). The eastern dialects should have been related to Old Saxon, given this historical bias.
I also found it a pity that rather a big part of the historical knowledge was presupposed. As a result, the atlas addresses itself in the first place to an expert audience of historical phonologists. If the writers would have condescended to the interested outsider, they might have widened their targeted group of readers / buyers considerably. Nevertheless, the maps are also useful for the linguist doing synchronic research, though they are not always easily accessible.
Chapter 1 on plurality opens with a bird-eye view of Old Germanic distinctions between nominal plurals which may be relevant in the modern dialects, such as:
strong versus weak plurals (vocalic versus consonantal);
interaction between -E as marker of singular and -E as marker of the plural, for example “ziekte-SG, ziekte-PL” (illness);
vowel lenghtening in the plural in nouns ending in the combination of a short vowel and a single final consonant, such as “vat-SG”, “vaten-PL” (cask, vessel), the latter involving vowel-lengthening;
nouns having a plural involving –ER- such as ‘ei-SG”, “eieren-PL (egg).
These phenomena may or may not be relevant for a synchronic description of Dutch nouns, but they have been chosen because they reflect nominal distinctions going back to Old Germanic. Thus, the atlas is (of course) not a theory-neutral atlas. Phenomena have specifically been selected on the basis for their relevance for the reconstruction of Old Germanic. As a result, phenomena which are interesting from a synchronic point of view, but not from a historic point of view, have been neglected. I found this a pity. For example, a phenomenon like linking sounds in nominal compounds was not investigated, presumably because it is historically uninteresting. Examples of such linking sounds are the –E- in “boekenkast” (book shelf). However, the phenomenon is synchronically relevant for deciding whether this linking sound has anything in common with the plural (see Hoekstra 1996 and the references cited there). It has wide social repercussions as this linguistic question bears on the controversial spelling reform which has recently been introduced. This spelling reform is in fact so controversial that some newspapers and editors have set up an alternative spelling. The present atlas says nothing about these matters, but of course it is perfectly legitimate to limit one’s range of interest.
Each chapter concludes with maps focusing on tonal phenomena which play a role in the southern part of the province of Limburg. In those dialects, tone is phonematic, like it is in Chinese. On the whole, the discussion of the data focuses more on the southern dialects of The Netherlands than on the northern ones. This is presumably the case as the Dutch standard language derives from the southern dialects. Hence an atlas like this, with a strong historical bias, naturally focuses on those dialects. Data from Frisian dialects, and from dialects with Frisian substrate (provinces Noord Holland and Groningen) were included in the map, but not discussed in their historical context, that is, related to Old Frisian. I think that this is an adequate compromise. After all, the Meertens Institute lacks knowledge of the historical development of Frisian, hence the northern data could not easily have been put in their historic context [note 1], as was done with the southern data. Furthermore, it might be a nice task for the Frisian Academy to take the northern data from the MAND and provide the historic discussion and interpretation that is absent from the MAND.
I will now go on to discuss some minor points which attracted my attention. The discussion in the first chapter raises the mystery of the S-plurals. How come these forms spread in English and in the Dutch dialects, especially the coastal dialects? The text suggests these forms could be so-called “ingwaeonisms”, some substratum present in England and in the coastal dialects. However, the text fails to raise an important problem with this hypothesis: if it is a substraum common to English and Dutch coastal dialects, how come it is absent in Old Frisian, a fact which was pointed out in Philippa (1987). The nature of the S-plurals (words from the informal sphere) supports the idea that substratum is involved.
The chapter on diminutives notes that the reconstructed suffix <kin> is the only Old Germanic suffix beginning with a sound that is not coronal. I made the same observation in Hoekstra (2000), and doubtlessly it has been made before. Inflectional morphology is especially prone to be coronal. It is a very interesting observation, since linguistic theories generally assume the relation between form and function/meaning to be arbitrary. If the observation holds water, then perhaps the relation between form and function is not arbitrary.
Chapter 3 on gender claims that adjectives lack an inflectional ending –E in standard Dutch in case they are preceded by the indefinite determiner “een” (a), just in case the following noun is neuter:
“het mooie boek” (the beautiful book)
“de mooie film” (the beautiful movie)
“een mooi boek” (a beautiful book)
“een mooie film” (a beautiful movie).
However, it is not only the indefinite determiner “een” (a) which triggers this behaviour. In Standard Dutch the following quantifiers, among others, trigger the same absence of flection in case the following noun is neuter (Haeseryn, Romijn, Geerts, De Rooij en Van den Toorn 1997:405):
“welk” (which), “elk” (each), “geen” (no), “veel” (much), “weinig” (little), “wat” (what a), “ieder” (every), “menig” (many a), “zo'n” (such a), “zulk een” (such a)
These are all quantifiers compatible with a singular noun. This matter could have been explained in more detail. Furthermore, as some of the quantifiers themselves also show inflection depending on gender, they might have been included in the investigation. Or will it be part of the syntactic atlas (SAND)?
Another phenomenon which is discussed involves the retention of –N on the determiner “de(n)” (the), in sourthern dialects, before a following vowel /h, b, d, t, r/. Now, the vowel and the /h/ are not a problem, because that is a context characteristically preventing deletion of consonants. But the remaining set of consonants before which –N is retained /b, d, t, r/ is not a natural class. The text remarks that “some phonologists feel uneasy” with that set of contexts. That is an understatement! The text suggests, paraphrasing a suggestion of De Wulf en Taeldeman (2001), that /n/ shares the feature specification <+anterieur>, <-continuant> with (d, t, r, b), and then is non-distinct in having either the feature <coronal> (d, t, r) or <voice> (b). This unnaturally disjunctive approach just describes the problem. Furthermore, it incorrectly fails to exclude /l/ from the set of contexts, which is like /r/ as far as the features mentioned above are concerned. It seems to me that this puzzle remains unsolved. It is fascinating that the restriction does not apply to adjectives ending in –N. Only determiners are sensitive to conditioning by the initial segment of the following word. It reminds me superficiously of soft mutation in Welsh and Irish.
To sum, we have another excellent atlas from the Meertens Institute. It is an atlas of synchronic variation presented and interpreted from a diachronic (Old Germanic) point of view. As a result, the atlas is less accessible to linguists not schooled in the ins and outs of Old Germanic phonology. This objection could have been obviated by a more pedagogical presentation of the material. But I can imagine the authors did not want to include an introduction to Old Germanic in their atlas. Furthermore, the data on which this atlas is based are all available on the website of the Meertens. People who are dissatisfied with this atlas can simply make their own atlas: excellent service!
Notes
1. Arjen Versloot incidentally supplied the authors with some valuable information on Frisian.
Bibliography
Barbiers, S., H. Bennis, G. de Vogelaer, M. Devos, M. van der Ham (2005) Syntactische Atlas van de Nederlandse Dialecten. Meertens Instituut, KNAW. Amsterdam University Press, Amsterdam.
Goossens, J., J. Taeldeman en G. Verleyen (1998) Fonologische atlas van de Nederlandse dialecten. Deel I. Gent. Koninklijke Academie voor Nederlandse Taal-
en Letterkunde.
Goossens, J., J. Taeldeman en G. Verleyen (2000) Fonologische Atlas van de Nederlandse
Dialecten (F.A.N.D.) deel II + III. Gent. Koninklijke Academie voor Nederlandse Taal-
en Letterkunde.
Goossens, J. (2002) Fonologische Atlas van de Nederlandse Dialecten. Deel IV. Gent. Koninklijke Academie voor Nederlandse Taal- en Letterkunde.
Haeseryn, W., Romijn, K., Geerts G.,, De Rooij, J. en M.C. van den Toorn (1997) Algemene Nederlandse Spraakkunst. Martinus Nijhoff, Groningen.
Hoekstra, E. (1996) Iets over eerste leden van samenstellingen. Leuvense Bijdragen 84, 491-504.
Hoekstra (2000) Grammaticale Functies van -E en -EN in het Westfries en het Fries en taalcontactgestuurde veranderingen. Taal en Tongval 70, 136-149.
Hoekstra, E. (2001) Frisian Relicts in the Dutch dialects. In H.H. Munske in collaboration with N. Århammar, V. Faltings, J. Hoekstra, O. Vries, A. Walker and O. Wilts (eds). Handbook of Frisian Studies. Max Niemeyer Verlag, Tübingen, 138-142.
Philippa, M. (1987) Noord-Zee-Germaanse ontwikkelingen. Dissertatie Universiteit van Amsterdam.
Wulf, Chr. De en J. Taeldeman (2001) Apocope en insertie van -n na sjwa in de zuidelijke Nederlandse dialecten: conditionering en geografie. In L. Draye, L., H. Ryckeboer en J. Stroop (eds, 2003) De variabiliteit van de -(e)n in het Nederlands. Taal & Tongval themanummer 14 (2001).