Desktop version

Home arrow Language & Literature

  • Increase font
  • Decrease font

<<   CONTENTS   >>

Listener Participants

There were fifty-four listener participants who completed this task. Because of technical issues, we had to discard eleven of these participants’ results, leaving forty-three participants whose data we used in the final analysis.4 The mean age of listeners was 22.8 years. As in the explicit measures task, all listeners completed a demographic questionnaire that included questions about their age, level of formal education, region of origin, the amount of time they have spent in the San Juan metropolitan area, and their linguistic background.


We used two sets of stimuli in this task. The first set, used in the priming phase, was made up of the same six article-noun combinations used in the explicit methods perception task (described earlier).

The second set, used in the test phase, was produced using an adaptation of a matched-guise design (Lambert et al. i960). Speakers produced three variations of the stimuli containing the variable of interest, the /s/ pronunciation. The test phase stimuli included two phrases:

(1) Estos chicos no vienen.

‘These boys aren’t coming’

(2) Estos coches no van.

‘These cars don’t run/don’t go.’

The speakers were instructed to read the phrase several times pronouncing each /s/ several times as [s], as [h], and by eliding all /s/. This method resulted in more natural sounding tokens as compared to digitally manipulated stimuli.

Using the mean scores from the explicit measures perception task, we identified the seven speakers evaluated as the most stereotypically straight-sounding and the seven speakers evaluated as the most stereotypically gay-sounding. We used these speakers’ stimuli as the test items in the voice recognition experiment; we used the stimuli that had been recorded by speakers who were not in either one of these groups as practice and distracter items.

<<   CONTENTS   >>

Related topics