University of Pennsylvania Working Papers in Linguistics


Edwin Ko


Recent investigations have shown an effect of geography on syntactic variation among dialects in a handful of languages. Developments in corpus-based dialectology have introduced sophisticated analysis techniques for studying dialect differences. However, not much is known about syntactic variation among socially and geographically non-contiguous varieties such as World Englishes, and the effectiveness of modern dialectometric techniques in these types of studies. Following the findings of Grieve 2012, which identified the dialects of the Northeast against those in the Southeast and South Central states by examining the positions of adverbs from a corpus of written Standard American English, this study employs a similar approach to spoken and written corpora of World Englishes, drawing from the International Corpus of English (ICE). Applying spatial autocorrelation techniques to identify similarities and differences due to geographical space shows significant overall spatial clustering in the spoken data, but not in the written data. English varieties in India, Hong Kong, Singapore and the Philippines were identified as ‘hot spots’ for placing certain adverbs before the verb, whereas varieties in Canada and Ireland acted as hot spots for positioning adverbs after the verb.