 |
 |
Dear all, Below are details of the May meeting of the Merseyside local group of the RSS. Anyone requiring directions or a parking permit should contact Helen Wilson ([log in to unmask]). All welcome! Helen Wilson Merseyside RSS Local Group Secretary ------------------------------------------------------------------------------ Date and time: Tuesday 4th May, 5pm (tea in Coffee Room, 3rd floor, Maths and Oceanography Building, at 4:30pm) Venue: Room 2.11, Maths and Oceanography Building, University of Liverpool Speaker: Mario Cortina Borja (Department of Statistics, University of Oxford) Title: Representations of linguistic dissimilarities Abstract: Many instances of linguistic data are available in the form of dissimilarity matrices. The best known example is the percentage of cognates, out of a list of words, shared by pairs of languages. Common choices of words expressing basic concepts, present in almost every culture, are the 100 and 200-word lists compiled by Swadesh. Using these lists linguists can assess the genetic relationship of languages using the comparative method of reconstruction. Another example concerns the analysis of phonological dissimilarities. There are a wide variety of methods useful in analysing and exploring dissimilarity matrices. In this talk we discuss several models for lexical and phonological data based on minimum spanning trees, exodic trees, families of relative neighbourhood graphs and other proximity graphs used to produce representations relevant for classifying the relationships of languages, and their grouping into language families. We also analyse the relation between spatial distance and lexical replacement. We illustrate the methods proposed using data from the Uto-Aztecan and Indo-European families of languages. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%