230 Bartáomiej Plichta and Brad Rakerd
of Lower Michigan; nine (5 men, 4 women) were recruited from the Ishpeming
area of the Upper Peninsula. To participate in this study, an individual had to have
been born and raised in one of these regions and to have never left there for more
than a year. It was also required that the person be a native speaker of English.
4.3 An /a/-to-/æ/ Vowel Continuum
Fronting of /a/ primarily affects the frequency of the second formant (Labov,
Yeager, and Steiner 1972). We therefore generated a continuum of synthetic
vowels that varied from /a/ to /æ/ along the F2 dimension. The continuum was
based on real vowel formant data obtained from two young, middle-class, adult
male talkers. One of these talkers, referred to here as Talk e r L M, was from the
Detroit area, the other, Talk e r U P, was from Ishpeming. Talkers LM and UP
were selected to be matched, so much as possible in physical size, in voice fun-
damental frequency, and in the general characteristics of their vowel systems.
Figure 9.4 shows a comparison of acoustic properties of their vowels obtained
by LPC analysis of 4 pronunciations of 50 vowel tokens in a broad range of
consonantal contexts (see Appendix 9.1). It can be seen that the overall range of
Figure 9.4 Vowel systems of Talker LM and Talker UP.