Interrogative inversion as a learner phenomenon in English contact varieties

The phenomenon of the non-standard use of inversion in both main clause and embedded interrogatives has been mentioned for several varieties of English, as well as for individual second language acquisition (SLA). There have been varying explanations, however: embedded inversion has been attributed to either L1 influence from either Gaelic or conservative British English for Irish English (Filppula 2004); to a general learner phenomenon for Singapore English (Gupta 1994); to overgeneralization for individual SLA (McDavid & Card 1973); or its patterning has left authors short of any explanation at all (Bhatt 2004 for Indian English). In addition, it is one of the examples often quoted in the discussion of so-called “Angloversals” (term attributed to Mair 2003), i.e. phenomena occurring across a range of historically and geographically unrelated vernacular varieties of English (Kortmann & Szmrecsanyi 2004).This paper aims at questioning these divergent explanations of what looks like one and the same phenomenon, by analyzing the syntactic factors that govern the occurrence or non-occurrence of inversion in interrogative clauses in three varieties of English: Indian English (IndE), Singapore English (SingE) and Irish English (IrE). The paper will show that what looks similar is not necessarily the same phenomenon at all and will single out a factor that unites contact varieties of English and individual learner Englishes, a factor which has often been neglected in quantitative studies: the frequency of specific lexical chunks, or “formulaic language”.The data that provides the basis for the quantitative analysis is taken from the International Corpus of English (ICE), more precisely the spoken private conversations sections of its Indian and Singaporean subcorpora. The data for Irish English is taken from the Hamburg Corpus of Irish English (HCIE).