The spoken L2 corpus represents present-day spoken Chinese (Putonghua) used in mainland China. It comprises L1-L2 conversational interactions between L2 speakers of Chinese and a native Chinese speaker (the corpus builder) in non-academic settings. This corpus contains 220,792 words of transcribed interaction gathered in 2018, featuring 14 L2 speakers of Chinese in 34 recordings. All the L2 speakers of Chinese are native English speakers of non-Chinese ethnicity (12 New Zealanders and two Australians) who were at intermediate to advanced proficiency levels.
This is the downloadable version of the L2 component of the spoken Chinese corpus of informal interaction - the spoken L2 corpus. Its contents are listed below.
- spoken/untagged
- orginial text files of L1-L2 conversations. No headers.
- original text files of L2 speech only. No headers.
- spoken/metadata
- text files containing speaker and recording metadata.
- transcription guidelines
- the spoken Chinese corpus transcription scheme. pdf
- License
- Version 1.0 - the spoken L2 corpus release as downloadable text files in March 2021
The corpus user is required to acknowledge the spoken Chinese corpus in any publication or presentation arising from the use of the corpus data by citing the following reference:
- Li, L. (2021). A spoken Chinese corpus: Development, description, and application in L2 studies [Unpublished Doctoral dissertation]. Massey University. https://github.com/blculyn