Aspects of the Acoustic Analysis of Imitation 385sociolinguistic research). If, however, the normal performance conforms to
the conservative system and the imitation system conforms (or at least in part
conforms) to the Southern Vowel Shift system, then we may conclude that
imitative or performance ability is rather more precise than expected.
2.1 Results
A comparison of Noah’s normal and imitation speech reveals the system-
atic differences between them. Figure 16.3 displays the mean scores for
Noah’s “normal” and WV imitation vowel system where an “X” marks the
position of the mean score of Noah’s vowels in his normal system, and an
arrow points to its position in the imitation system. In general, if we com-
pare the mean scores of the vowels of Noah’s usual speech to a conser vative
“Midwestern” system, such as that represented in Figure 16.1 (or Kurath
and McDavid, 1961), we ¿ nd a great deal of correspondence. For example,
[iy] “peel” is higher and more front than [ey] “paid”; [æ] “apple” is in a
low front position, and [ow] “hope” is mid and back. The only signi¿ cant
deviation is the fronted position of [uw] “food,” which is higher but more
front than [ow]^3.
The imitation system, on the other hand, bears a great resemblance to
the Southern Shift (Figure 16.2), which consists of the backing and lower-
ing of the tense front vowels [iy] and [ey], the fronting and raising of the lax
front vowels [ܼ] and [e], the raising and backing of low vowels, the fronting
of the back vowels [ow] and [uw], and the reduction of the diphthongs [ay],
[aw], and [oy]. We might expect the respondent’s imitation of West Virginia
speech, to reÀ ect some characteristics of the Southern Vowel Shift as Mor-
gantown, West Virginia borders on the region identi¿ ed as belonging to the
Southern Shift (Labov, Ash, and Boberg 2006). As can be seen in Figure
16.3, Noah’s imitation speech demonstrates just such elements. Figure 16.3
not only demonstrates the similarities of his imitation system to the Southern
Shift but also the very wide range of differences between his normal and
imitation systems.
Noah’s normal system mean scores are indicated by boxes. Arrows point
to the imitation mean score.
In order to gauge the perceptual success of Noah’s imitation, a “matched
guise” (Lambert et al. 1960) tape was created with Noah’s “normal” and
“West Virginian” speech and ¿ ve other male speakers; three from West
Virginia (Charleston—in the southern part of the state, Parkersburg—
west of Morgantown but also near the north south division of the state, and