80 Rebecca Roeder
Table 3.2 indicates, for example, that /æ/ in a word such as gamble—where
it is preceded by a velar consonant and followed by a nasal consonant—should
be more raised and fronted than /æ/ in any other environment.
Eight ANOVA’s were run for each individual. The dependent variables
were F1 and F2, analyzed for the following environmental factors: preced-
ing manner and voice, preceding place of articulation, following manner
and voice, and following place of articulation. Unlike previous studies,
manner and voice were separated from place of articulation for preced-
ing environment, so that every possible combination of features could be
tested. In accordance with the ¿ ndings of previous studies, the inÀ uence
of following place of articulation was minimal. No signi¿ cant difference
between the effects of any two following place of articulation features was
found in more than two speakers, so those results are not presented here. In
addition, only results for other environments that are signi¿ cant in at least
four speakers will be discussed. As mentioned before, the discussion notes,
in particular, the number of Group A speakers for whom these categories
are signi¿ cant, since that may shed light on the relevance of each feature
to the NCS.
3.3 Following Manner and Voice
Five following manner and voice environments were tested: voiced and voice-
less stops, voiced and voiceless fricatives, and nasals. Table 3.3 lists words
used by environment.^5
Table 3.3 Following Manner and Voice Features and Words
Manner/Voice Words
Voiced Stop tab, cabin, dad, Saginaw, brag, rag
Voiceless Stop apple, nap, zap, pat, mattress, rack, black
Voiced Fricative have, has, jazz
Voiceless Fricative Bath, laugh, ask, past, cash, mash
Nasal Sam, Lansing, plant
Excluded (singleton): pal (liquid); badge (affricate); gamble, gang, thank, banker (velar)The only feature that is signi¿ cant in at least two speakers is following
nasal, and the effects on F1 in this environment are dramatic. Table 3.4 indi-
cates the effects of following nasal environment on the production of /æ/ for
each individual.