260 Nancy Niedzielski
However, as many sociophonetic studies have shown, the vowel systems of
speakers from this area are rapidly diverging from what most Americans—in-
cluding most Michiganders—consider “standard.” As participants in the North-
ern Cities Shift (NCS), which encompasses speakers from not only Michigan
but Ohio, Illinois, Wisconsin, and New York, their vowels have shifted dramati-
cally away from values offered as those found in “standard American English”
(SAE), for example those offered in Peterson and Barney (1952).
Speakers in Michigan appear to have no explicit awareness of this, and
they reveal this even in perceptual tasks. Tables 11.1 and 11.2 present the
results of a perception task given to 42 Michigan residents, in which they were
asked to choose from a list of synthesized vowels the vowel that best matched
a vowel that they heard in a Michigan speaker. Speci¿ cally, after hearing a
Michigan speaker produce a sentence containing the words “last” and “pop”
(both of which contain a shifted vowel: /æ/ and /a/ respectively), the subjects
were asked to listen to a set of six synthesized tokens, and to choose the token
that they thought the speaker produced. They were given the option of listen-
ing to the sentence as many times as they wished (although few listen more
than once), and were then to record their choice.
Ta ble 11.1 Formants of Tokens of “last” Played for Respondents (N=42) and
ResponsesToken # F1 F2 Label of Token # and % of Retspondents Who Chose Each oken1 900 1530 hyper-standard 4 (10%)
2 775 1700 canonical /æ/ 38 (90%)
3 700 1900 actual vowel produced 0Ta ble 11.2 Formants of Tokens of “pop” Played for Respondents (N-42) and
Responses
Token # F1 F2 Label of Token # and % of Respondents
Who Chose Each Token
1 770 1050 hyper-standard 4 (10%)
2 900 1400 canonical /a/ 36 (86%)
3 700 1600 actual vowel produced 2 (5%)These tables reveal that few of the respondents chose the token that was
the actual vowel the speaker produced. In fact, none of the respondents chose