256 Nancy Niedzielski
The eye-tracker measures the amount of time that the subjects look at
the words printed around the subjects. Figure 11.3 displays the results: the
perceived age of the speaker affects the amount of time spent looking at the
competitor (“rent”), so that a photograph of an older woman causes the sub-
jects to look at what would more likely be a homophonous lexical item in an
older person’s dialect. Thus, while Houstonians do not claim explicit knowl-
edge of the correlation between age and the unmerging of these vowels, the
eye-tracker reveals that they have implicit knowledge of the correlation.
I suggest that the fact that Houstonians fail to offer age a factor in the
unmerging of these vowels as evidence that the knowledge of this correlation
Figure 11.2 Screen subjects view as they hear “rinse” presented aurally. In the cen-
ter is one of three photographs of women varying in age.
Figure 11.3 Number of ¿ xations on the competing lexical item, by actual and per-
ceived age of speaker (Koops, Pantos and Gentry 2006).