Evaluating English Accents WorldWide: Current Research
The basic data presented for each sample of listeners in the Results section provides answers to some of the questions outlined in Aims (recognition and distinctiveness of accents, association of traits with them). However, research is also under way using more sophisticated analyses of the data.
A paper by Bayard, Sullivan, and Kügler was given at the 14th conference of the Linguistic Society of New Zealand in September 2001. It presented the results of surveys in three non-Anglophone European countries to investigate the influence of English-language television broadcasts on accent ratings (see Influence). An overall factor analysis of all three nations revealed the same clear-cut division of the 22 personality traits used into the basic dimensions of status, power, solidarity, and competence that was so evident in the results from the three Anglophone nations (Bayard, Weatherall, Gallois, and Pittam 2001). However, data was also gathered on English-language television viewing habits: programmes watched and number of hours per week. Two of the nations surveyedSweden and Finlandbroadcast all English-language programmes in their original accents, as the relatively low number of speakers of each of the native languages made dubbing uneconomic. However, the German sample reflects a country in which all English programmes are dubbed into German.
Although there is considerable evidence suggesting that the NAm accent is replacing RP as preferred standard model in Western Europe (van der Haagen 1998, Mobärg 1999but cf. Ladegaard 1998), no significant differences were found between English television viewing hours and NAm vs. RP accent ratings; however, much more study is needed, with a larger sample of "dubbing" and "non-dubbing" nations. It does seem safe to conclude that influences from other media (films, pop music, advertisements, etc) also play an important part in the rise of NAm as a global standard.
Bayard and his research assistant James Green have begun the investigation of national and ethnic variation in the emphasis on the four basic dimensions of status, power, solidarity, and competence (see Cross-cultural research). Sullivan and Bayard are examining the effects of speaker intonation on listener evaluation by the digital manipulation of the eight stimulus voices to produce two new tapes: one of these has a "flat" F0 intonation pattern, and the other uses the lively, tonally variable speech styles of the two NAm voices as a model for the other six voices. This provides us with one stimulus having minimal tonal variation and a second with the maximum tonal variability found in our eight stimulus voices. They have administered these to groups where we already have a large sample of responses to the "normal" stimulus tape, namely Sweden and New Zealand. For some results see The Impact of Intonation).
France Mugler (University of the South Pacific) is looking at the differences in accuracy of accent identification between Fijians and Indo-Fijians and at the effect of speaker gender on the attribution of social status and personality traits.