The use of speech in racial identification has important legal consequences. Many of my students are shocked when statements about the racial identity of unseen speakers is allowed in court testimony (e.g. the OJ Simpson trial). In my experience, though, they are more accepting when hear that the Fair Housing Office manipulates racial and ethnic vocal qualities in order to determine whether or not a real estate group is screening their phone calls.
Although there have been very few legal cases that have employed linguistic experts when subjects like these come up, the standards seem to be pretty obvious. If you are going to allow legal testimony about the racial/ethnic vocal qualities of a speaker, you need to determine (1) what the person testifying explicitly and implicitly knows about those speech patterns of that racial/ethnic group AND (2) whether the speaker in question follows those patterns.
In the past, most of the linguistic research that could inform this issue has focused on the patterns of three very broad ethnic groups: Whites, Blacks, and Latinos. A new article by Newman and Yu promises to contribute to another very broad category, that of “Asians.” A link to this article, which appears in one of the latest editions of American Speech, can be found below: