A Corpus-Based Analysis of Mixed Code in Hong Kong Speech

We present a corpus-based analysis of the use of mixed code in Hong Kong speech. From transcriptions of Cantonese television programs, we identify English words embedded within Cantonese utterances, and investigate the motivations for such code-switching. Among the many motivations observed in previous research, we found that four alone account for more than 95% of the use of English words in our speech data across genres, genders, and age groups. We performed analyses over more than 60 hours of transcribed speech, resulting in one of the largest empirical studies to-date on this linguistic phenomenon.