Abstract
The present study seeks to contribute to two sparsely examined areas of World Englishes research by (i) quantitatively evaluating two potential linguistic epicenters in Asia (Indian and Singapore English) while (ii) investigating the English genitive alternation in a cross-varietal perspective. In a corpus-based bottom-up approach, we evaluate 4,200 interchangeable genitive cases of written English from Great Britain, Hong Kong, India, the Philippines, Singapore and Sri Lanka, as represented in the International Corpus of English. We use a new method called MuPDARF, a multifactorial deviation analysis based on random forest classifications, to evaluate to what extent and with which factors the Asian varieties differ from British English in their genitive choices. Results show conspicuous differences between British English and the Asian varieties and validate the potential epicenter status of Indian English for South Asia, but not unanimously that of Singapore English for Southeast Asia.