Open Access

Julia Degenhardt, Tobias Bernaisch

• Despite notable recent exceptions (for instance Funke, 2020), corpus-based research into South Asian Englishes has so far concentrated on structural features. Thus, empirical pragmatic research in said varieties is still sparse, although there has been increasing interest in variational pragmatics in world Englishes (Schneider & Barron, 2008). For instance, requests have been examined in both native and non-native varieties of English. However, studies on apologies have largely focussed on first-language varieties of English (Deutschmann, 2003). Against this background, the present study investigates apology patterns in two South Asian second-language varieties of English, Indian and Sri Lankan English, and their historical input variety British English. With the help of the spoken parts of the respective components of the International Corpus of English, multifactorial analyses—including an improved form of random forests that explicitly takes interactions between several predictorsDespite notable recent exceptions (for instance Funke, 2020), corpus-based research into South Asian Englishes has so far concentrated on structural features. Thus, empirical pragmatic research in said varieties is still sparse, although there has been increasing interest in variational pragmatics in world Englishes (Schneider & Barron, 2008). For instance, requests have been examined in both native and non-native varieties of English. However, studies on apologies have largely focussed on first-language varieties of English (Deutschmann, 2003). Against this background, the present study investigates apology patterns in two South Asian second-language varieties of English, Indian and Sri Lankan English, and their historical input variety British English. With the help of the spoken parts of the respective components of the International Corpus of English, multifactorial analyses—including an improved form of random forests that explicitly takes interactions between several predictors into account—model the choice of sorry as opposed to other apology forms. Findings suggest quantitative differences in the use of sorry which are influenced by factors such as type of apology, topic and age or combinations of said factors. In sum, this study suggests that apology forms and frequencies are sensitive to the speakers’ regional background and sociobiographic factors as well as to structural and contextual parameters.…