This is an attempt to show structured heterogeneity of English stress variants by applying a Labovian quantitative study to suprasegmental variation. Compared with the relatively frequent occurrence of segmental variation such as [IN].[In] in -ing, the low frequency of suprasegmental variation makes it difficult to observe differences of pronunciation in different styles. A detailed analysis of the recorded speech of twenty-one subjects suggests that the apparent anomalous distribution of stress variants is by no means random, and acoustic measurements reveal two kinds of strategies available to the speaker for avoiding stress clash.
Bibliographic reference. Hattori, Noriko (2008): "Structured heterogeneity of English stress variants", In INTERSPEECH-2008, 544.