The social dimensions of the phonological continuum – vowels

Whereas most other researchers have classified speakers as belonging to one of the three varieties, Horvath approached the description of the vowel system from a different perspective. Using a statistical procedure called principal components analysis, she was able to group the speakers from her sociolinguistic study who were similar in their overall linguistic behavior on five vowels (FLEECE, FACE, GOAT, PRICE and MOUTH). In place of a three-way division of the AusE spectrum, she argued for a four way division and simply named them Sociolects 1–4, with Broad corresponding most closely to Sociolect 1, Cultivated to Sociolect 4 and General to Sociolects 2 and 3. As Figure 2 shows, no speaker used only Broad, General or Cultivated vowels but each variety consisted of a mix of all of the vowel pronunciations; the Broad variety used more ‘broad’ vowels and the Cultivated used more ‘cultivated’ vowels, but all speakers often used ‘general’ vowels. No variety existed in a ‘pure’ form. Furthermore, the varieties correlated with social class and gender. At the Broad end of the continuum men and the working class predominated while women and the middle class were associated with the Cultivated end. In fact, at the most Cultivated end of the continuum, there were only women.

The falling diphthongs also show an interesting social and linguistic distribution. Horvath (1985) found that the NEAR vowel was more than twice as often pronounced with the two targeted variant [iə] than it was for the SQUARE vowel. The reverse was true for the monophthongal variant: speakers are more than twice as often heard pronouncing hair as [he:] than they are heard saying [bi:] for beer. The social distribution indicates that the middle class favors some kind of diphthongal realization, either an offglide or a two-targeted variant. The monophthongal variant was associated with working class speakers, older speakers, and men.