376 Renée van Bezooijen and Vincent J. van Heuven
Methodologically, although cross-speaker and especially cross-gender
comparison of acoustic measures of vowel quality are hazardous in principle,
the pro cedure that we applied in our study, i.e., recording reference vowels
and performing partial extrinsic speaker normalization on Bark-transformed
formant measurements, affords useful comparison across vowels produced by
male and female speakers. As far as we have been able to ascer tain, we are the
¿ rst to have adopted this speci¿ c normalization procedure, which is a mixture
of extrinsic and intrinsic normalization. It bears a resemblance to Gerstman’s
(1968) end-point normalization, but differs from it in two details: (i) our pro-
cedure speci¿ cally looks for the front-most /i/ and the open-most /a/ in the
front vowel continuum only, while the Gerstman procedure indiscriminately
adopts the lowest and highest F1 and F2 values in an entire vowel set as the
end-points, and (ii) our procedure is applied after Bark-transformation, which
is a form of intrinsic normalization.
It should be reiterated that our acoustic procedure should only be used
with caution when making comparisons between steady-state vowels and
dynamic vowels such as the diphthongs in our study. We suspect that onset
of a closing diphthong is heard with a more open vowel quality as the diph-
thongal gesture is larger. This effect, if indeed it can be shown to exist in a
full-scale psychophysical experiment with static and dynamic vowel sounds,
should be modeled in future vowel normalization procedures. Only then can
acoustic measurements be used as a fully adequate substitute for (expert)
human perception of vowel quality and quality change in diphthongs.
The evaluation study, in which samples of Avant-garde Dutch mainly
characterized by the typical realization of /͑i/ were judged by groups of lis-
teners, con¿ rmed the hypothesis that young women are more positive towards
Avant-garde Dutch than other listeners. Generally speaking, young women
place Avant-garde Dutch on an equal footing with Standard Dutch, giv-
ing these two varieties the highest ratings, whereas other listeners place it
together with Randstad Dutch in second position. Only with respect to pol-
ished do the young females place Standard Dutch above Avant-garde Dutch.
So our experiment lends strong support to the real-life observations by Stroop
(1998). It indeed appears to be the young Dutch women who favour Avant-
garde Dutch. In fact, there is no evidence in our results that men or older
women even distinguish Avant-garde Dutch from regional accents from the
Randstad, since it is judged signi¿ cantly different from these on none of the
scales. There is little evidence that Avant-garde Dutch would be spreading
among young males.
Also in accordance with Stroop’s (1998) views is the ¿ nding that people’s
regional origin is irrelevant. The same attitudes towards Avant-garde Dutch