Review of Morphological Atlas of the Dutch Dialects
|EDITORS: Goeman, Ton; van Oostendorp, Marc; van Reenen, Piet; Koornwinder, Oele;
van den Berg, Boudewijn; van Reenen, Anke
TITLE: Morphological Atlas of the Dutch Dialects Volume II
PUBLISHER: Amsterdam University Press
Sebastian Kürschner, Lehrstuhl für Germanistische Sprachwissenschaft,
The Morphological Atlas of Dutch Dialects (MAND) forms part of a large project
in which geographically determined grammatical variation is documented for the
whole Dutch language area. The morphological part of the project has been
published in two volumes, supplemented by four parts of an atlas focusing on
phonology, and two volumes of an atlas focusing on syntax. In line with the
acronym MAND (from Dutch Morfologische Atlas van de Nederlandse Dialecten), the
phonological atlases are called FAND, and the syntactic ones SAND. The data used
in the morphological atlas, as in the phonological one, were gathered from
1979-1995 in the so-called Goeman-Taeldeman-Van Reenen-project (GTRP, cf. Goeman
& Taeldeman 1996). In this project, dialect speakers from 613 locations were
asked to take part in a questionnaire-based interview. The locations are “more
or less evenly spread geographically on the basis of a hexagonal grid with one
location per hexagon” (commentary, p. 7) over The Netherlands, the Flemish part
of Belgium, and the region of France which belonged to the county of Flanders
(Frans-Vlaanderen). The interviews were recorded, and the data was transcribed
in a computer-readable form of the IPA. The transcriptions were carefully
validated (but see Wieling et al. 2007 on the comparability of the Netherlandic
and the Flemish transcriptions).
MAND comes in two volumes. The first volume deals with nominal morphology,
covering plural formation, the formation of diminutives, and the category of
gender. Volume II -- under review here -- deals with the inflectional morphology
of adjectives (i.e., comparison, Chapter 1), pronouns (Chapters 2-3), and verbs
(Chapters 4-5). The volume consists of a book containing the maps and a booklet
containing the commentary. The booklet is detachable from the actual atlas, so
reading the comment while looking at a map is convenient. Both the atlas and the
booklet start with a table of contents. The booklet additionally ends with an
index, enhancing searches for information on specific linguistic phenomena,
toponyms, questionnaire items, and authors.
The atlas and the commentary both have introductory section. The commentary
begins with general background on the subject matter, the types of maps used,
and the data collection and preparation process. The atlas starts with a basic
map of the Dutch-speaking area, which is presented marking both each single
location (p. 10) and the division of the countries into provinces and regions
(p. 14). All locations and regions are also presented in alphabetical order with
a numeric code representing them in the maps. The map of the locations is a
fold-out page, so it is possible to identify the locations when studying the
linguistic maps. Since MAND maps the West Frisian dialects in addition to Dutch
ones, specific attention is paid to the province of Friesland in the northern
Netherlands. A map zooming into this area identifies non-Frisian (i.e., Dutch)
dialects and where some Frisian dialects outside the actual province are located.
The following chapters contain maps and comments on the linguistic phenomena
studied in the atlas. All chapters of the commentary start by describing the
general linguistic phenomena treated and the construction and structure of the
respective maps. After that, each map found in the atlas is commented on.
Endnotes are included in each chapter. The introduction to Chapter 1 covers
methods and criteria of classification and some technical details on map
production. All maps consist of a title starting with the number of the relevant
chapter in the commentary, a legend, and tags identifying from which part of the
questionnaire the mapped data stems.
Chapter 1 describes the formation of comparative and superlative forms of
adjectives. Both variation of the stem and suffixation are considered from
several perspectives. Concerning the stem, the focus is on variation of the
vowel, covering changes of vowel quality, moraic length, and palatalization.
Since tone accents only exist in the dialects of Limburg, tone alternations in
the formation of the comparative paradigm are mapped especially for this area.
Variation in the form of suffixes (Standard Dutch comparative -er and
superlative -st) is presented in eight maps. Two maps additionally describe the
occurrence of final schwa throughout the paradigm and the addition of particles
in the superlative (cf. het grootst 'the greatest').
Possessive pronouns are subject of Chapter 2, with a focus on two aspects.
Firstly, internal morphology is considered, i.e. the question of whether forms
are coherently marked as belonging to the same paradigm for person and number.
In Standard Dutch mijn 'my', for example, m- might be viewed as a marker of
first person singular when compared with third person singular (masculine or
neuter) z- in zijn 'his/its'. Secondly, some maps deal with the number and
gender inflection of possessive pronouns (cf. Standard Dutch ons huis 'our
house' - onze taal 'our language' - onze huizen/talen 'our houses/languages').
19 maps manage the difficult task to illustrate the dialectal variation in the
suppletive paradigms considering both these aspects, sometimes in combination,
Chapter 3 deals with personal pronouns. The first part considers the dialectal
variation of subject pronouns, the second part that of object pronouns. Close
attention is paid to the variation of strong vs. weak forms of pronouns. Strong
pronouns are characterized by full vowels (cf. the diphthong in Standard Dutch
jij 'you', hij 'he') whereas weak pronouns contain reduced (cf. schwa in je, he)
or even fully elided vowels ('k compared with strong ik 'I'). In stressed
positions, strong pronouns are more often used than weak ones, whereas weak
pronouns tend to occur more frequently in unstressed positions. Some
morphosyntactically determined variation between weak and strong forms is mapped
as well. For example, the form of a subject pronoun varies according to its
position before or after the verb (e.g. jij/je breekt 'you break' vs. inversion
in breek jij/je). In addition to the distinction between strong and weak forms,
lexical variation is accounted and sometimes traced back to pragmatic aspects.
For example, dialectal forms of doe (2nd person singular subject pronoun, by
contrast with Standard Dutch jij) are used ''as a form of address, sometimes to
express solidarity, sometimes to express deference for children, the sick, and
for the housewife or old parents'' (commentary, p. 33).
In chapter 4, which is considerably longer than the other chapters,
person-number marking of verbs is mapped. The first section deals with present
tense paradigms and the second with past tense paradigms. Concerning person and
number marking, Standard Dutch verbs are more strongly inflected than English
verbs, but the paradigms show numerous syncretisms. For example, Standard Dutch
has a uniform plural marker in present tense (wij/jullie/zij leven 'we/you/they
live') and uniform singular and plural markers in past tense (ik/jij/zij leefde
'I/you/she lived', wij/jullie/zij leefden 'we/you/they lived'). In many of the
Dutch dialects verbal paradigms are characterized by fewer syncretisms than
those in the standard variety. However, some other dialects also have more
syncretisms than the standard variety. For example, dialects in Zuid-Holland
have a uniform present singular marker in addition to the syncretisms known from
Standard Dutch, resulting from t-apocope in 2nd and 3rd person forms. In
addition to inflectional suffixes, stem vowel alternations such as umlaut,
breaking, and quantitative alternations are characteristic of the verbal
morphology of Dutch dialects. Chapter 4 ends with maps of tone alternations
specific to Limburgish dialects.
Two phenomena are subject matter of Chapter 5. The first part deals with the
formation of past participles. A circumfix consisting of ge- and either -d/-t
or -en is mostly used in the standard language, cf. ge-dans-t 'danced',
ge-gev-en 'given'. The atlas deals with variation of the prefixed part, i.e.
variants of ge-. Three major types are distinguished. The first two are prefixed
participles formed by CV- vs. V-structures. Participles of the third type
(characteristic of the northern Netherlands) have no prefix at all. The second
part of Chapter 5 deals with the tense allomorphy of strong verbs, i.e. stem
alternations marking tenses. Alternation patterns are inherited from West
Germanic ablaut classes, but have changed considerably in the dialects of Dutch.
For Modern Dutch, the patterns are illustrated by means of three forms, the
infinitive, the singular past, and the past participle. Alternations are
described both by means of the actual vowels and with the help of an abstract
number notation. This number notation lists if the vowel alternates in each of
all three forms (123, cf. breken -- brak -- gebroken 'break'), or if the vowels
are identical in two of the three forms (e.g., 121, cf. lopen -- liep -- gelopen
'walk'; alternatively 112, 122). 17 verbs are examined to illustrate verb stem
alternations. In certain cases identical verbs also vary between strong and weak
Throughout the atlas, variation is most often illustrated using symbol maps,
representing each variant by means of a specific symbol, which is then located
at each of the places where this variant occurs. Two other types of maps are
used in addition to symbol maps. Firstly, the geographic distribution of certain
verbal paradigm types is illustrated in two isogloss maps in Chapter 4.
Secondly, contour maps based on a computational technique modeling ''cohesion
within a locality's neighbourhood'' (commentary, p. 6) are found. On these maps,
the dark-light continuum is used to map the clustering of variants in certain
areas. In this manner, the continual character of geographical variation is
MAND II is excellent work of great value to anyone concerned with Dutch
linguistics and geographically determined variation in continental West
Germanic. The data collected in the GTRP-corpus are exceptional since they cover
the entire Dutch-speaking area based on a well thought out questionnaire. Both
with respect to phonological and morphological research, the analysis has
resulted in atlases providing information on a wide range of issues. The second
volume of MAND, dealing with adjectives, pronouns, and verbs, offers a broad
survey of the variation accounted, and the commentary provides the reader with
necessary background information and a deeper understanding of what is illustrated.
Most of the maps cover the whole Dutch-speaking area, i.e. The Netherlands,
northern Belgium, and a small part of France. This is fascinating because they
offer clear pictures of the variation for a very large area. To facilitate user
orientation, both the biggest rivers and the political borders of provinces and
countries are found in the background of each map. Nevertheless, for a user not
familiar with the geography of this area, it might take a while to get
acquainted with the basic structure of the maps, since rivers are represented by
thin blue lines of the same kind used to represent political borders. While one
might wish to have at least the borders between the three countries represented
by a different kind of line for better orientation, the mode of presentation
found in the atlas also has advantages. For example, political borders often
play no role in linguistic variation, i.e. clusters of symbols are only rarely
split along the border between, e.g., The Netherlands and Belgium. Quite the
contrary, some areas sharing many linguistic features (termed 'super regions')
actually cross political borders. For example, the Limburg super region consists
of the Dutch and the Belgian provinces of Limburg. A more border-oriented
marking on the basic map would thus have had the dangerous effect of visually
supporting political borders which are by and large not reflected in the
Including Frisian dialects in an atlas of Dutch dialects is not unproblematic.
Frisian is a closely related language, but no dialect of Dutch, which means that
the Frisian dialects expectably provide variation patterns different from the
Dutch dialects. Although it would make sense to provide special symbols, a
different color or other means of illustrating special regions, the special
character of Friesland remains unmarked in the maps. There are, nevertheless,
also reasons to include the Frisian-speaking area without any special marking.
The authors are aware that what they collected were dialectal varieties in the
process of regionalization and standardization (cf. Goeman/Taeldeman 1996, p.
52). Since speakers of Frisian in the Netherlands today usually also speak
Standard Dutch, the Dutch standard variety has influence on the Frisian dialects
as well, and this may be visualized in the maps. Another reason for no special
marking of the Frisian dialects is the fact that the dialects of North-Holland
have a Frisian substrate, so continual linguistic relationships can be traced
back to this language contact situation.
The three types of maps provide excellent exemplary illustrations of
morphological variation. Geographic patterns are normally clearly visible from
symbol maps because of an intelligent grouping of variants and the choice of
clearly differing symbols. The contour maps use dialectometric techniques to
visualize the cohesion of variants in geographical neighborhoods. These maps are
often used to bundle general information on morphological systems, instead of
dealing with variation in single words, as most of the symbol maps do. For
example, the general tendency of changes in the vowel quantity and in stem
palatalization is summed up in the first two maps (atlas, p. 16).
A nice feature of the symbol maps is to mark dialectal variants in blue if they
are identical (or close to identical) to forms in the standard variety, whereas
dialectal variants different from the standard variety are marked in black.
However, this feature should not be misunderstood as providing information about
the degree of standardization in specific dialect areas. Some dialects (mostly
from Holland) form the historical basis of the standard variety, so it is no
coincidence that they still show some similarities. Some forms may also be
widespread in the Dutch-speaking area. Even if we cannot draw conclusions about
the standardization process, a distinction between variants close and distant
from the standard variety just makes it easy to access the maps (at least for
users who know Standard Dutch). In the future, the atlas will provide a good
basis for studying dialect changes (including standardization). For example, the
'Reeks Nederlandse Dialectatlassen' (Blancquart/Pée 1925-1982) provides a large
collection of older dialect data which is in part comparable with the data
mapped in the new dialect atlas. In such comparisons, the blue marking of
standard forms in MAND will probably be a great help in identifying where
standard forms are increasingly used. Additionally, historical observations are
found in some maps in which the new material is compared to data from a corpus
on 14th-century charters.
It is not always clear to me why the authors chose the given particular words
they did to illustrate subjects s. Of course, it is necessary to restrict
oneself to a certain number of relevant items when making an atlas, both because
mapping morphological variation on all words would result in redundant maps
showing the same geographical patterns with different words, and because many
phenomena are word-specific and would be lost in maps based on aggregate
analyses. Still, a user expects that most of the maps illustrate general
patterns of variation by one example. The authors often leave us unaware of why
they chose the specific words mapped. For example, maps of strong verb
alternations are found in Chapter 5. The commentary tells us that ''17 verbs out
of more than 100 available for this purpose'' (p. 56) were selected to illustrate
the patterns, but it does not tell us why those specific verbs were chosen. From
a synchronic point of view, one could expect that they were chosen to cover the
abstract vowel alternation patterns introduced above (i.e., 123 vs. 112, 122,
121). From a diachronic point of view, one might guess that the atlas provides
at least one word representing West Germanic ablaut classes. From the words
mapped and the commentaries on each map, it seems that both played a role in the
actual selection, but it would be preferable to have this made explicit.
Dutch is one of the few languages in which work on a linguistic atlas can
benefit from a long dialectological tradition. The authors of MAND have used
this knowledge base to create a highly interesting collection of maps using
modern technology capable of dealing with large amounts of data. MAND offers an
easily accessible and well-informed presentation of maps and commentaries.
Although I raised some minor points of critique, the publication of MAND II (and
its accompanying atlases in the MAND-, FAND-, and SAND-family) is of great value
for the systematic description of Dutch dialects. The whole family of atlases
will not only be relevant to linguists of Dutch, but also to researchers
concerned with the dialect geography of neighboring Germanic language areas,
most of all Frisian, German, and Luxembourgish. MAND will be a standard
reference to dialectal variation in Dutch, and justifiably so.
Blancquart, Edgard & Willem Pée (1925-1982) Reeks Nederlandse Dialectatlassen.
Antwerpen: De Sikkel.
Goeman, Ton & Johan Taeldeman (1996) Fonologie en morfologie van de Nederlandse
dialecten. Een nieuwe materiaalverzameling en twee nieuwe atlasprojecten, Taal
en Tongval 48, 38-59.
Wieling, Martijn, Wilbert Heeringa & John Nerbonne (2007) An aggregate analysis
of pronunciation in the Goeman-Taeldeman-Van Reene-Project data, Taal en Tongval
ABOUT THE REVIEWER
| ABOUT THE REVIEWER:
Sebastian Kürschner holds a tenure track position ('Juniorprofessur') in
variationist linguistics and language contact at the University of
Erlangen-Nürnberg, Germany. He studied German, Scandinavian, and
computational linguistics at the Universities of Freiburg and Copenhagen
and received the doctoral degree from the University of Freiburg in 2007.
From 2007 to 2009, he had a postdoc position at the University of
Groningen. His research interests include language variation, language and
dialect contact, historical linguistics, and morphology.