Lecture | LUCL Colloquium - Fall 2017
How many probabilistic grammars does it take to model syntac¬tic variation in varieties of English around the world?
- Benedikt Szmrecsanyi (KU Leuven)
- Date
- Friday 29 September 2017
- Time
- Series
- LUCL Colloquium
- Location
-
Lipsius
Cleveringaplaats 1
2311 BD Leiden - Room
- 227
Abstract
Work is underway in Leuven to explore the probabilistic grammar(s) of World Englishes. The project (see here) is situated at the crossroads of research on English as a World Language, usage-based theoretical linguistics, variationist lingu istics, and cognitive sociolinguistics. We investigate, among other things, how the effect direction and magnitude of language-internal constraints (e.g. constituent length – language users tend to place heavier constituents after lighter ones) on a number of syntactic alternations (e.g. we looked the word up versus we looked up the word) differs across some nine international varieties of English. One of the key findings emanating from the project is that the probabilistic grammars regulating variation are overall remarkably similar, and we seem to be dealing with a rather solid common core of the grammar of English. Against this backdrop, I discuss a number of ways to measure the homogeneity of probabilistic variation grammars. On the qualitative plane, I will review constraints on variation that seem to be particularly malleable across regional varieties, as opposed to those that aren't.