ICC Programming

Analysis Topics

5602.1 - Visualising and Analysing the Impact of Geographic Factors on Linguistic Variation in Dialects

Wednesday, July 5
2:50 PM - 3:10 PM
Location: Delaware B

Delineating and representing dialect areas has been a fundamental question of dialectology since the first dialectal surveys at the end of the 19th century. Today dialectologists agree (Wieling & Nerbonne 2015) that dialect areas mostly are not separated by crisp boundaries, but rather by smooth transitions between core areas of dominant dialectal variants. Areas of dialect types overlap, coexist and typical features of multiple dialect areas can be found at a single survey site (Pickl 2016).

We propose methods to quantitatively assess the potential impact of geographic factors on the variation of Swiss German dialects, as represented in a language survey database. The survey for the “Syntactic Atlas of German-speaking Switzerland” (SADS), intended to capture diversity of dialect syntax (i.e. sentence structure), was conducted between 2000 and 2002 among over 3000 respondents answering 118 questions about syntactic phenomena in one quarter of the German-speaking municipalities of Switzerland (Bucheli & Glaser, 2002). For each question, the survey recorded the answers of multiple respondents per survey site.

In the first part of our study, two opposing conceptualizations of dialectal boundaries were assessed. On the one hand, the isogloss model assumes the occurrence areas of different variants corresponding to a linguistic phenomenon to be separated by a crisp boundary (i.e. the isogloss). The conceptual model of the dialect continuum, on the other hand, assumes smooth transitions between the areas of prevalent dialect types as well as individual dialect phenomena. Methods have been developed to quantitatively model the relationship between variants of individual linguistic phenomena. With multiple answers per survey site, co-occurrence of different variants per site is present. About 40% of the SADS phenomena feature two main answer variants that are „competing” in the dialect landscape, some of them gradually blending into each other, others exhibiting a rather steep gradient between their dominance areas. For an initial exploration of the transition patterns, we used various visualisations, including area-class maps using Voronoi polygons, as well as 3-D plots and cross-sections through the investigation area. We then calculated different characteristics of spatial patterns, based on which we were able to classify the different dialect phenomena regarding whether the transition can best be described by an isogloss or a dialect continuum model, respectively (Seiler, 2005). The latter model was tested by calculating residuals to benchmark trend surfaces, while isoglosses were tested using logistic regression due to the assumed binomial patterns. Moreover the transition patterns of the phenomena were quantitatively compared (Jeszenszky & Weibel 2016; Jeszenszky, Stoeckle & Weibel, in prep.). We thus were able to demonstrate that by dividing the study area into two dominance zones and a transition zone, we can best account for the nature of the transition, advocating an alternative conceptual model of spatial dialectal variation, positioned midway between the isogloss and the dialect continuum model.

Additionally, the effect of geographic distances was quantified. Euclidean distances and travel times were used to capture the possibility of language contact. Based on the SADS, a syntactic distance measure between survey sites was devised. It was shown that geographic distance is responsible for, and thus explains, the majority of the variance found in Swiss German syntax, as represented in the SADS data. Travel times (for the years 2000, 1950 and 1850) correlate with the syntactic spatial variation significantly better than Euclidean distance. Travel times of older years yield higher correlations than newer ones, but the difference is not significant. Correlation analysis of different (spatial) subsets of the study area was conducted to demonstrate the effect of different topographic constraints and contact possibilities on the linguistic variation (Jeszenszky & Weibel 2014; Jeszenszky, Stoeckle & Weibel, submitted).

Péter Jeszenszky

PhD Candidate
Department of Geography, University of Zurich

PhD Student, University of Zurich, Department of Geography, GIScience Center; grant by the Swiss National Science Foundation (SNF).
ICC title: Visualising and Analysing the Impact of Geographic Factors on Linguistic Variation in Dialects.
Dissertation title: Quantitatively modeling the effects of geographic factors on Swiss German dialectal variation (Supervisor: Robert Weibel, Elvira Glaser),
Project "SynMod" (http://www.spur.uzh.ch/en/research/associated/synmod2.html), a joint project between the Department of Geography and the German Department.
Main research: Developing quantitative processes to account for the role of different geographic factors in the variation patterns in linguistic data. Testing hypotheses posed by our linguist colleagues in the project. Linguistic survey data, geostatistics, spatial analysis and variationist theories contribute to my research.
As a cartographer by training I was always interested in spatial distributions of differences in the world. I was editor-in-chief of a map series presenting touristic and ethnic data. After visualising variation, during my PhD I turned onto investigating the spatial factors underlying variation of dialectal data. One topic of my research is the investigation of correspondences of geographic distances and extralinguistic differences with linguistic differences (dialectometry). Another topic is the development of methods using GIScience and geostatistics to quantitatively describe language-internal gradual boundaries (quantitative dialectology). This is done on the one hand by quantitatively modeling linguistic concepts and on the other hand in a data-driven, automatic way to visually and quantitatively represent boundaries in the data, using DBSCAN clustering.
I am also involved in the Language Contact Group at the University of Zurich, aiming to write a comprehensive volume on extralinguistic (geographic among others) factors that impact language contact.
In the near future of my academic carreer I am planning to conduct comparative GIScience research on Swiss German and Japanese dialects with regards to the spatiotemporal evolution of variants.


Send Email for Péter Jeszenszky

Philipp Stoeckle

Austrian Academy of Sciences Vienna

PhD at the Universität Freiburg im Breisgau. 2013 Title of Dissertation: "Subjektive Dialekträume im alemannischen Dreiländereck" (summa cum laude). Research interests:
Sociolinguistics, (Ethno-)Dialectology, Syntax, Linguistic Geography und Geographic Information Systems (GIS), Dialektometry, Language Change, Spoken Language, Language Contact, Language and Media. Involved in the "SynMod" project: Modelling morphosyntactic area formation in Swiss German.


Send Email for Philipp Stoeckle

Elvira Glaser

German Department, University of Zurich

Professor at the German Department of the University of Zurich. Initiator and leader of the linguistic survey project "Syntactic Atlas of the German-Speaking Switzerland". Co-leader of project "Synmod": Modelling morphosyntactic area formation in Swiss German


Send Email for Elvira Glaser

Robert Weibel

Department of Geography, University of Zurich

Professor of Geographic Information Science at the Department of Geography at the University of Zurich, GIS Center. Chair of ICA Commission on Map Generalization and Multiple Representation, 1991-2003. Regional Editor for Europe and Africa, International Journal for Geographical Information Science. Principal investigator or co-investigator of about 50 projects since 1992. Current research interests: Digital cartography and map generalization; computational movement analysis; spatial analysis in linguistics.


Send Email for Robert Weibel

Samantha T. Arundel

Research Geographer
U.S. Geological Survey

Samantha received her Ph.D. in geography from Arizona State University in 2000, and was an assistant and associate professor at Northern Arizona University until 2009, when she joined the USGS. There she has served as elevation and hydrography specialist for the Applied Research and Technology Branch. During this time, she led the development team in automating contour production for the USTopo product; and most recently, served as the program manager for the automation of the National Elevation Dataset production, in its transition from EROS to the NGTOC. Recently Sam moved to the Center of Excellence in Geographic Information Science, where she is a research geographer conducting terrain semantics studies.


Send Email for Samantha Arundel


5602.1 - Visualising and Analysing the Impact of Geographic Factors on Linguistic Variation in Dialects

Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Visualising and Analysing the Impact of Geographic Factors on Linguistic Variation in Dialects