Literary and Linguistic Computing Advance Access originally published online on September 20, 2006
Literary and Linguistic Computing 2006 21(4):493-506; doi:10.1093/llc/fql043
| ||||||||||||||||||||||||||||||||||||||||||||||||||
Measuring Syntactic Variation in Dutch Dialects
Meertens Instituut, Joan Muyskenweg 25, Postbus 94264, 1090 GG Amsterdam, The Netherlands
Correspondence: Marco René Spruit, Meertens Instituut, Joan Muyskenweg 25, Postbus 94264, 1090 GG Amsterdam, The Netherlands. E-mail: marco.rene.spruit{at}meertens.knaw.nl
This research applies dialectometric methods to purely syntactic dialect data. It will be shown that there is geographic cohesion in syntactic variation when viewed in the aggregate. The amount of syntactic variation which can be accounted for by geography will be determined. Dialectometric techniques will be used to develop an additive measure of syntactic differences. Multidimensional scaling will be applied to visualise the geographic distribution of the Dutch dialects with respect to syntactic variation in the aggregate. The Dutch dialect map based on a syntactic measure will be compared with a dialect map based on subjective judgements and a dialect map based on pronunciation differences to put the syntactic measurement results into perspective. An alternative way to measure syntactic distance will be presented and will provide indications for future research to more accurately quantify syntactic variation.
1 Spruit mentions 135 maps. However, this included SAND1 map 73b which does not contain unique data. It has been left out of the measurement procedures reported on in this work.
2 Syntactic variables are referred to as syntactic features in Spruit (2005).
3 A distance matrix is always symmetric because the distance from dialect A to dialect B is always identical to the distance from dialect B to dialect A. Therefore, only the distances in either the lower left part or the upper right part need to be included in the measurement. Also, all distances from a dialect to itself are excluded from the procedure.
4 The space between dialect locations on the MDS maps is partitioned by using the Delaunay triangulation to obtain a pattern of polygons known as Voronoi polygons or Dirichlet tessalation. This technique for determining dialect areas is also used in Goebl (1982) and Heeringa (2004). Alternatively, an interpolation procedure could be applied to colour the space between dialect locations.
5 The following twenty-one dialects were used in the regression analyses, listed from the north-east to the south-west of the Dutch language area: Nieuw-Scheemda, Spijkerboor, Rolde, Hooghalen, Diever, Staphorst, Wezep, Epe, Hoog Soeren, Lunteren, Geldermalsen, Waspik, Zundert, Ossendrecht, Doel, Koewacht, Zaffelare, Gent, Deinze, Waregem and Kortrijk.
6 Application of Kruskal's Non-metric MDS method results in a nearly identical dialect map. This can be interpreted as a confirmation of the reliability of the SAND1 MDS map shown in Fig. 2.
7 The three Frisian city dialect islands in Fig. 3 which also occur in the SAND are Midsland, Heerenveen and Kollom.
8 Furthermore, the feature nominative is used in the reflexives subdomain to help describe the variation with respect to the syntactic context reflexive pronouns in adverbial middle constructions as shown in SAND1 map 77a.
9 The MDS map visualising syntactic distances with respect to reflexive and reciprocal pronouns is printed in Spruit (2005: 186).
10 No data is available with respect to reflexives for the dialect of Morbecque.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  F. Manni, W. Heeringa, and J. Nerbonne To What Extent are Surnames Words? Comparing Geographic Patterns of Surname and Dialect Variation in the Netherlands Lit Linguist Computing, November 1, 2006; 21(4): 507 - 527. [Abstract] [Full Text] [PDF]
|
|