The social life of the sounds of English : comparing the intelligibility and acceptability profiles of Hong Kong English accent features

The constructs of intelligibility and acceptability are well-known in studies of how spoken language is processed and perceived. Both are complex and context-dependent, but intelligibility - defined here as ‘actual understanding’ or the extent to which listeners’ perceptions match speakers’ intentions - can be seen as somewhat more objective. Acceptability relates to various forms of subjective evaluation by listeners. It can be performed by different listeners, be global or focused in scope, and range across linguistic, sociolinguistic and pedagogical domains. The two constructs are of central importance in assessing accents, varieties and speakers. For example, in selecting accent samples or models for pedagogical purposes, intelligibility without acceptability to local stakeholders may compromise the uptake of educational reforms. However, acceptability without both local and international intelligibility may compromise the well-rehearsed arguments for local models in language teaching. This talk focuses on the relationship between the two constructs, as this has seldom been elucidated in research studies. I explore the relationship by comparing the findings from two of my earlier studies. The intelligibility data were collected by asking 91 Hong Kong student listeners to transcribe a range of Hong Kong English accent samples. The targeted aspect of intelligibility was therefore intra-community or ‘local’ (as opposed to international) intelligibility. The acceptability data come from questionnaires completed by 52 student listeners who evaluated some of the same accent samples. The targeted aspect of acceptability was pedagogical acceptability, and more specifically the acceptability of local English accents for use in local English classrooms. Comparisons between intelligibility and acceptability data yield several interesting results. Segmental features, and especially consonantal features, appear to be important influences on both constructs. Although there is some overall correspondence between the two, so that features which reduced intelligibility also reduced acceptability, exceptions to the pattern are also visible. As part of an overall explanatory approach, I adopt the sociolinguistic concept of ‘sound symbolism’ (Eckert 2012). From this perspective, while hypoarticulated forms may not actually reduce intelligibility, if noticed they can trigger symbolic associations that link them with qualities such as a lack of carefulness and precision. The approach demonstrates the value of combining linguistic and sociolinguistic perspectives on accent features, and I conclude by outlining some of the general indications of the study for pronunciation teaching in the Hong Kong context.