SpiCE: A New Open-Access Corpus of Conversational Bilingual Speech in Cantonese and English

Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), pages 4089–4095 Marseille, 11–16 May 2020 c⃝ European Language Resources Association (ELRA), licensed under CC-BY-NC 4089 SpiCE: A New Open-Access Corpus of Conversational Bilingual Speech in Cantonese and English Khia A. Johnson, Molly Babel, Ivan Fong, Nancy Yiu Department of Linguistics, University of British Columbia 2613 West Mall, Vancouver, BC Canada V6T 1Z4 khia.johnson@ubc.ca, molly.babel@ubc.ca, ivan.tw.fong@gmail.com, nancyyiu64@hotmail.com Abstract This paper describes the design, collection, orthographic transcription, and phonetic annotation of SpiCE, a new corpus of conversational Cantonese-English bilingual speech recorded in Vancouver, Canada. The corpus includes high-quality recordings of 34 early bilinguals in both English and Cantonese—to date, 27 have been recorded for a total of 19 hours of participant speech. Participants completed a sentence reading task, storyboard narration, and conversational interview in each language. Transcription and annotation for the corpus are currently underway. Transcripts produced with Google Cloud Speech-to-Text are available for all participants, and will be included in the initial SpiCE corpus release. Hand-corrected orthographic transcripts and force-aligned phonetic transcripts will be released periodically, and upon completion for all recordings, comprise the second release of the corpus. As an open-access language resource, SpiCE will promote bilingualism research for a typologically distinct pair of languages, of which Cantonese remains understudied despite there being millions of speakers around the world. The SpiCE corpus is especially well-suited for phonetic research on conversational speech, and enables researchers to study cross-language within-speaker phenomena for a diverse group of early Cantonese-English bilinguals. These are areas with few existing high-quality resources. Keywords: Speech Corpus, Conversational Speech, Bilingualism, Less-Resourced Languages, Cantonese 1. Introduction Most of our knowledge about spoken language and speech processing comes from monolingual individuals produc- ing scripted speech in laboratory settings. Monolingual lab speech allows for researchers to exercise tight control over the linguistic backgrounds of the speakers and the linguistic material (e.g. reading or repeating sounds, words, or sen- tences). While highly informative, these controlled mono- lingual speech samples are not representative of spoken lan- guage in the real world. Multilingualism is the norm, not the exception, and individuals regularly make creative linguis- tic choices. Crucially, corpus-based research with conversational or spontaneous speech is important in the fields of phonetics and psycholinguistics, as the research conclusions drawn from corpus and lab-based experiments do not always coin- cide (Gahl et al., 2012). Conversational speech allows for a more accurate empirical description of spoken language, as it represents more realistic and natural productions than scripted laboratory speech, even compared to scripted con- nected speech. Conversational speech also crucially per- mits for field testing of speech production theories (Bell et al., 2009; Gahl et al., 2012) in their natural habitats. The discrepancies between results for conversational and lab speech have been found for monolingual (English) speech, but are likely be found with bilingual speech as well. Resources to query bilingual conversational speech are limited, however, as the necessary resources permitting this type of inquiry are rare. As a step towards filling this gap, …

Content truncated to 500 words