Abstract

The Corpus of Australian and New Zealand Spoken English (CoANZSE) is a 190-million-word corpus of Automatic Speech Recognition (ASR) transcripts from YouTube channels of local councils and other governmental bodies in 472 locations in Australia and New Zealand. CoANZSE can be used to examine grammar and syntax in Australian and New Zealand spoken English, and because tokens are word-timed and transcripts are linked to videos, it can serve as the starting point for phonetic or multi-modal studies. Two exploratory analyses demonstrate differences between Australia and New Zealand in the relative frequencies of double modals, a rare non-standard syntactic feature, and show that transcripts from Australia and New Zealand can be distinguished on the basis of common lexical items.