24-25 October 2011, University of Gothenburg
Since the introduction of lexicostatistics in the 1950s, researchers have been investigating methods for the grouping of languages and dialects. The input for these methods are relevant abstractions of linguistic data (most often based on fixed lists of senses, e.g., the Swadesh lists).
The abstractions are traditionally character-based (as in work by Swadesh, Dyen, Ringe, Warnow, Gray and many others), which can be said to define distances on the level of languages. In these approaches, a particular item either is or is not present in a language.
More recently, drawing on advances in computer software and hardware, methods aiming at measuring and aggregating distances among individual linguistic items have been explored (as in work by Kessler, Wichmann, Nerbonne, Kondrak, McMahon and others). In these approaches, individual items are (dis)similar to some quantifiable degree between two languages.
A number of such methods for measuring linguistic differences have been proposed in the literature, but so far, researchers have mostly explored individual methods. Some authors have made explicit comparisons among some of the methods (e.g., Kessler, Kondrak, Nerbonne), but this aspect is still largely unexplored.
There is also a need to contrast the more recent automatic methods with a more traditional historical- comparative methodology (e.g. as recently surveyed and critiqued by Campbell and Poser), which generally draws on a broader range of linguistic phenomena than have so far been used with the automatic approaches, including semantic and grammatical characteristics.
This two-day workshop, arranged by the Swedish Languagde Bank, University of Uppsala and University of Leipzig, is intended as a forum to discuss these issues in more depth.
The workshop will take place in Gothenburg, in the Faculty of Arts building at the University of Gothenburg.
Register at this link at the latest October, 12th.