Effects of Consonantal Context on the Pronunciation of /æ/ 87of these ¿ ndings except the pre-nasal raising of /æ/ agree with results found in
studies on laboratory phonology and therefore suggest universal rather than
area-speci¿ c effects. Findings on pre-nasal /æ/ raising were not possible in the
Hillenbrand et al. (2001) study, since the nasal environment was not tested.
In general, therefore, no previously unattested coarticulatory effects on
the production of /æ/ are apparent in the speech of Mexican Americans in
Lansing who are native speakers of English. However, there is evidence of
complex social strati¿ cation in one local feature. Very low F1 in tautosyllabic
pre-nasal /æ/—ubiquitous in the local Anglo population—is present among
those interviewed for the present study only in women under 25 who were
born in Lansing and speak English as a native language. This supports the
hypothesis that young women are the leaders of change, but it also raises the
question of whether raising in the production of /æ/ before nasals is a distinctly
female marker in this group. Given the accommodation to NCS /æ/ found in
these young women—as evidenced by their pronunciation of both raised pre-
nasal /æ/ and the centralizing off-glide that is distinct in the pronunciation
of NCS /æ/, it is apparent that lack of assimilation to the pronunciation of
vowels other than /æ/ in these speakers cannot be attributed to lack of contact
with the Anglo community or lack of perceptual acuity. Subtle and accurate
assimilation to local mainstream norms appears to be occurring in only some
aspects of the phonetics and phonology for other reasons. Communicative
competence in Lansing’s Mexican American community clearly involves
more than just the ability to assimilate to Anglo speech characteristics.
Notes
1 This study also included a perceptual experiment, which found unexpected results
when production and perception were compared. The tokens that were most often
confused in the perceptual experiment were not those that were most acoustically
displaced due to phonetic conditioning. The authors offer an explanation based on
relative distance of the token from a prototype.
2 Stevens and House (1963) does not indicate whether or not the three male speak-
ers recorded were linguistically trained or not. The speakers were asked to read
words in the form of “bisyllabic nonsense utterances in iambic form” (112), but
it is not clear if they were presented as phonetic transcription or using regular
orthography.
3 “The relative degree of advancement is inÀ uenced by the manner of articulation
of the following segment, in the order nasals > voiceless fricatives > voiced stops
voiced fricatives > voiceless stops” (Labov 1994: 100).