Linguistic Security, Ideology, and Vowel Perception 257is implicit, rather than explicit. An even more dramatic demonstration of the
fact that knowledge about speaker variation can be completely implicit comes
from work on what Vitevitch (2003) calls “change deafness.” In this set of
experiments, he asked subjects to shadow a speaker producing a relatively
long list of lexical items, some of which he labeled “hard” and some of which
he labeled “easy” (based on, among other things, the amount of phonological
distracters an item had). In the middle of the shadowing task, he gave subjects
a rest-break, and upon their return from the break, they continued the shad-
owing task. Some of the subjects, however, were now shadowing a different
speaker. In Figure 11.4 are the results of the reaction times after the rest break.
What Vitevitch found was that, regardless of whether the subjects detected
the change in speaker, the reaction times were longer. In other words, even
without explicit awareness that the speaker had changed, the subjects reacted
as if they had implicit awareness of this fact (particularly with “hard” lexical
items), by showing a statistically signi¿ cant increase in reaction times, com-
parable to those subjects who did detect the speaker change, rather than those
subjects for whom the speaker was not changed.
Finally, brain-imaging studies suggest that exposure to different phone-
mic contrasts actually causes different neural responses. In Conrey, Potts, and
Niedzielski (2005), neural activity of speakers with and without the front lax
Figure 11.4 Shadowing task reaction times, by awareness of speaker change (Vite-
vitch 2003).