by Catherine Travis

The links between language and society have been well recognised in over 50 years of research in sociolinguistics. This endeavour was spearheaded by William Labov, who established a role for social meaning in sound change in his classic study of vowel realisations on the island of Martha’s Vineyard in Massachusetts (Labov 1963). This work also shows the clear affinities between variationist sociolinguistics and corpus linguistics — it was one of the first corpus-based studies, and the actual content of the corpus collected was drawn on to interpret the social meaning of the linguistic patterning observed.

Such is the motivation behind a large-scale sociolinguistic project currently being conducted in Sydney, the most ethnically and linguistically diverse city in Australia. The Sydney Speaks project is exploring variation in Australian English, based on a spontaneous-speech corpus of over 1.5 million words, from around 150 hours of speech from some 250 Sydneysiders of diverse ages and social backgrounds. Participants include Anglo Australians, as well as second-generation Australians of Chinese, Greek and Italian background, some of the largest migrant communities in the country. Occupations range from plumbers and hairdressers, to university students and teachers, to lawyers and doctors. And the corpus includes recordings made at two time points (1970/80s and 2010s), with speakers of different ages, born from as early as 1890 to as recently as 1990, thus capturing the speech of generations of Australians since federation, as depicted in Figure 1.

A key defining feature of the Sydney Speaks corpus is the stories the participants tell, allowing their speech patterns to be interpreted in relation to their lived experiences. In the Bicentennial Oral History corpus, we hear stories that are very familiar to us today, about restrictions on social gatherings and mask wearing; this is not about COVID-19, however, but about the Spanish flu one hundred years ago. The ethnic minorities recount experiences of discrimination, in particular at school, being teased because of their surname, or their sandwich fillings. Salami or mortadella may be common now, but in the 1970s and 1980s “it was not a trendy thing then, when everybody had their sliced bread, and your devon and tomato sauce and whatever else.” Discussions about ethnicity reveal subtle complexities, with participants showing different degrees of orientation and attachment towards their ethnic heritage and towards Australia. Many of the Chinese-Australians opt for the term “ABC” (Australian-Born Chinese) to describe themselves, which seems to capture the essence behind the words of one participant: “I don’t really feel Chinese but I don’t really feel Australian”.

Through these accounts, we also gain a wealth of linguistic data that allow us to track change in Australian English over time, and variation across the community. One particularly interesting picture emerging is a reduction in social class differences over time. For several features, we find that marked distinctions across social classes in the 1970s/80s recordings no longer exist today. We have described this change in recent work, including for vowel pronunciation, in particular, the diphthongs in words such as people, say, know, now and like; lengthening of -er on the end of words like teacher and remember; and the pronunciation of -ing vs. -in’ in words such as something, doing, and interesting. You can hear differences in the way Australian English is spoken over time in voices from the Sydney Speaks corpus available on the “Sydney Speakers” link on the project web page.

We also find that class and ethnicity are closely linked. For example, in these same studies we report that the speech of the Sydney Speaks Italian participants (and in particular the 1970s Teenagers and the 2010s Adults) tends to be associated more with Working Class Australian pronunciation, while the speech of the second-generation Chinese Australians tends to be more similar to the pronunciation of the Middle Class. This is no doubt tied to the social status of these communities; second-generation Chinese Australians tend to be highly educated, work in white collar occupations, and live in wealthier areas in Sydney (primarily in the north), while Italian Australians tend to be situated more in the less privileged southwest, and this is reflected in the distribution of the Sydney Speaks participants across Sydney, as shown in Figure 2 below.

Nevertheless, the information we glean about the participants from the stories they tell allows us to contextualise their speech patterns in relation to the associated social meaning of these pronunciations. This leads us to hypothesise that the Italian Australians are orienting to norms of covert prestige associated with non-standard language use, while the Chinese Australians are orienting more towards norms of overt prestige, associated with the high prestige standard. We look forward to testing this hypothesis further with other linguistic features, and expanding it to encompass other ethnic communities as the Sydney Speaks corpus continues to develop.

Sydney Speaks is led by Catherine Travis at the Australian National University, and the research reported on here was conducted in conjunction with James Grama and Simon Gonzalez. The project is funded by the ARC Centre of Excellence for the Dynamics of Language.