The DIG Mandarin Conversations (DMC) Corpus

Abstract

This paper introduces the DMC Corpus – a newly collected dataset of 150 mundane cell phone calls from Mainland China in Mandarin Chinese (audio and detailed transcripts) – which is now publicly available for use in research and teaching. In this report, we first describe the constitution and current contents of the DMC Corpus, as well as instructions for access. Additional calls will be added periodically to the Corpus, and so the quantitative overview presented here should be considered conservative. We then provide concrete examples of the sorts of phenomena that might be explored with these new data, underscoring how the Corpus offers researchers the ability to build systematic collections for analysis – no matter whether researchers prefer to begin with ‘forms’ (e.g., utterance-final particles), with ‘functions’ (e.g., complaining), and/or with the temporal organization of interaction itself (e.g., preference organization, repair). The paper concludes with an explicit call for increased research on Mandarin conversation, to which we hope the materials in the DMC Corpus will contribute.