Welkom bij marco@work, waar ik mijn promotieonderzoek documenteer in het kader van het Determinants of Dialectal Variation project.
Welcome to marco@work, where I document my Ph.D research in the context of the Determinants of Dialectal Variation project.

 Onderdelen
· Startpagina
· Afbeeldingen
· Archief
· Downloads
· Lidmaatschap
· Links
· Onderwerpen
· Statistieken

 Speerpunten
· Meertens pagina
· Video
· Powerpoint
· Paper
· Status
· Literatuurlijst
· Project portaal
· Recentelijk
· Taalhulpjes
· /

 Babylon

 Prijsuitreiking


 marco@work

© 2003-2007
Marco Rene Spruit

Abstract Methods XII workshop 'Progress in Dialectometry: Toward Explanation'
Geplaatst op Zondag 03 oktober @ 22:48:02 GMT+1

Onderzoek This is my abstract for the workshop Progress in Dialectometry: Toward Explanation, [...] a workshop at the Methods XII Conference on Methods in Dialectology, Aug. 1-5, 2005 at the Université de Moncton, New Brunswick. The workshop aims to feature original computational work in dialectology, and most particularly work aimed at explanations of dialectal facts and patterns in various languages. [...]

Measuring syntactic variation in Dutch dialects

Marco René Spruit
Meertens Instituut


In this dialectometric research a measure of syntactic distance is developed and applied to Dutch dialects. It will be shown that this quantitative perspective on syntactic variation provides new insights in the degree of geographical coherence in syntactic variation.

Methods to assign numerical values to linguistic phenomena in order to aggregate individual dialect differences were first described in Seguy 1971 and further investigated in Goebl 1984 and Heeringa & Nerbonne 2002, among others. However, until recently no extensive collection of syntactic data was available, limiting dialectometric research mainly to lexical and phonological data.

Now the Syntactic Atlas of the Dutch Dialects (Barbiers et al. 2005) has become available. It contains a wealth of data with respect to syntactic variation in the left and right periphery of the clause, pronominal reference and negation in 267 Dutch dialects. A subset of left peripheral and pronominal data has been used to obtain the first results described here.

The data in the anaphora subdomain contain 87 variants of 17 syntactic features. A list of geographical location codes is provided for each feature. Th ese lists of locations per feature are transformed into sets of features per location. After this conversion the number of differences can be determined between pairs of locations.

The Hamming distance between each pair of locations is calculated to obtain a measurement based on binary comparisons between feature variants. For each variant the distance between location A and location B is increased by 1 when observed at location A but not at location B, and vice versa. The distance between two locations based on the anaphora data is therefore an integer between 0 and 87. This rudimentary measure of syntactic distance will be compared to more sophisticated measurements which take into account additional information such as the number of occurrences and the number of alternative variants per feature.

Cluster analysis and multidimensional scaling are applied to interpret the resulting distance matrix. These two classification methods are known to complement each other. Three main groups are identified after clustering the anaphora data using Ward's method which corresponds to expert consensus. After multidimensional scaling a continuum is revealed that also confirms the cluster analysis, providing new insights in the degree of geographical coherence in syntactic variation.

 
 Gerelateerde links
· Meer over Onderzoek
· Nieuws door Marco


Meest gelezen verhaal om Onderzoek:
Promotieplan met uitleg


 Score Artikel
Gemiddelde score: 0
Stemmen: 0

Neem even tijd om dit artikel te beoordelen:

Uitstekend
Zeer Goed
Goed
Gewoon
Slecht


 Opties

 Printervriendelijke pagina Printervriendelijke pagina

 Stuur dit verhaal naar een kennis Stuur dit verhaal naar een kennis