Daily Archives: November 3, 2010

Does language affect our implicit biases?

Does the language we speak shape our biases?  New research coming out of Banaji’s lab by a graduate student, Oludamini Ogunnaike, seems to say that the answer to that might be “yes.”  They report some interesting findings about how language affects our implicit biases.  Their article, recently published in the Journal of Experimental Social Psychology, demonstrates that bilingual individuals show preference toward groups associated with the language they are being tested in.  For example, in a Moroccan population, Moroccans tested in Arabic showed an implicit preference (using the Implicit Association Test or IAT) for other Moroccan names compared to French names.  When participants took the test in French, this difference disappeared.  In the U.S., bilinguals tested in Spanish (compared to English) showed a greater preference for Hispanic individuals.  These differences disappeared when participants were tested in English.

This research has interesting implications for the study of biases and prejudice (my own area of specialty).  If the same individuals can show such differences in their own attitudes based on what language they are tested in, how much more-so do these differences exist cross-culturally (i.e., people speaking different languages who are monolingual)?  Additionally, one starts to wonder how language particular to a certain group of people (e.g., ethnic, religious, and political groups) might shape biases that prefer or favor one’s own in-group.  In short, how does the way we talk and the language we speak shape our attitudes?  Years of research has demonstrated that using “us” vs. “them” language increases biases, but can speaking just in language relative to one’s own in-group do it as well?  For instance, using religious words like “Spirit,” “Enemy” (to refer to Satan), etc. might increase bias towards non-Christians.  The literature on priming religion seems to indicate that exposure to these words, at the very least, increases bias.  Could testing individuals using religious (or political or any other group) jargon increase bias in the same way that using different languages does?


*An interesting summary of this research can be found in the Harvard Gazette.