10 Dennis R. Preston and Nancy Niedzielski
Chapter 15 studies the distribution of and reactions to the lowering of the
onset of the /͑i/ diphthong in an emerging variety here called “Avant-garde
Dutch,” one associated with younger, well-educated women. Van Bezooijen
and van Heuven ¿ rst establish, from talk-show data, that that social distri-
bution is correct and go on to detail the acoustics of the shift. In this stage,
they employ, in addition to a Herz-to-Bark normalization procedure, an “end-
point” normalization routine, in which the extremes of /i/ and /a/ are measured
to determine the relative height of the lowering of the onset of the diphthong
under question for individual respondents. This procedure shows as well a
female preference for lowering, although there is a great deal of intra-personal
variation in both sexes and glide-weakening appeared to accompany some
onset lowering, perhaps an inÀ uence on the ¿ rst stage of the study. Respon-
dents from a variety of regional backgrounds listened to speakers of Standard
Dutch, Amsterdam Dutch, “Randstad” Dutch (the Amsterdam conurbation),
and Avant-garde Dutch in a typical Likert-scale experiment. Young women
showed not only a consistent ability to distinguish Avant-garde Dutch in their
rankings, regardless of regional background, but also had more positive atti-
tudes towards it.
In Chapter 16 Evans explores the idea that untrained respondents can-
not effectively imitate the linguistic details of another variety. Her respon-
dent, “Noah,” is an Inland Northern speaker who spent some time in the
North Midlands-South Midlands area of Morgantown, West Virginia, USA.
Although most of his associates were also Northern during his time there, he
acquired an ability to imitate the local variety, but one that he characterized
as no more ef¿ cient than that of many of his friends. An acoustic investigation
of his imitation speech (from a set passage and word list), however, reveals his
relative mastery of important details of the Southern Vowel Shift when he was
asked to “read like a West Virginian”: reversal of the tense-lax and peripheral
character of the pairs /Ԍ/-/i/ and /ܭ/-/e/, fronting of back vowels, and monoph-
thongization of diphthongs. This sample was then played for local, West Vir-
ginia respondents along with three authentic West Virginia voices and typical
Northern and South Midlands speakers from Michigan and southern Indiana,
respectively; the respondents were asked to rate how likely it was that each
speaker was West Virginian. The imitation variety tied one authentic West
Virginia speaker and bested the two others; only very few respondents felt
that there was “something wrong” with the imitation; one said it was “over the
top,” but most had no inkling that one of the voices was an imitation. Perhaps
most importantly, as the Atlas of North American English clearly shows, the
southern features Noah used in his imitation are, in fact, not all that common
in West Virginia, perhaps another victory for stereotype over acoustic fact.