Listeners are remarkably sensitive to nonnative patterns of speech, whether they are presented with full sentence productions, or with very short or severely degraded speech, including noisy, filtered and temporally disrupted utterances. Furthermore, even phonetically unsophisticated listeners can reliably scale accents. From the standpoint of second language (L2) users, speaking with a detectible accent has important social consequences. One is that L2 speech is sometimes less intelligible or may require listeners to allocate more processing resources than does native speech. However, evidence also indicates that some salient phonological markers of L2 status have little or no impact on listener comprehension. Distinguishing these from accent features that reduce intelligibility is a critical concern in language pedagogy. Another consequence noted in empirical research is that accent stereotyping and negative social evaluation of accented speakers are linked to discrimination in remuneration, employment, and services. They may also be implicated in the exploitation of L2 speakers through questionable claims about the value of “accent reduction.” This synthesis of previous and new research findings highlights a number of issues concerning methodology and interpretation in L2 speech research that are relevant to language teaching and assessment, and to human rights litigation.

Return to Invited Speakers page