Abstract
The interaction of linguistic variability and social factors has often been demonstrated in sociolingusitic research. Similar patterns have been found in studies done with different populations and even with more than one data set drawn from the same sample, but never have similar patterns been reported from repeated sampling from the same populaton. In the study reported here, the interaction between class and race and the pronuciation of the vowel nucleus in the word nine was found to be replicable at a high degree of statistical significance in twelve samples drawn from speakers in Washington, D.C. over a six year period. At the same time, no significant patterns were observed in the samples when sex and age were correlated with the same pronunciation.
Notes
An earlier version of this paper was presented at the Linguistic Society of America annual meeting. Chicago, Illinois, December 1977.