Gender-role stereotypes exert strong effects on
behavior. For example, gender stereotypes influence mothers' perceptions of their
children's abilities and children's self-perceptions. Gender-role occupational
stereotypes play a role in job choice, hiring, promotion, and compensation. However,
research on stereotypes and their effects is beset by a problem of measurement:
Although most researchers rely on introspective self-reports, these reports do not
invariably reflect attitudes or beliefs, particularly when expression of the belief is
deemed to be socially inappropriate. It is also conceivable that a person could
maintain a belief but not be consciously aware of it.
We combined ERPs and language-sensitive ERP effects to
chart a new course for studying social stereotypes. This approach was motivated by
work in our lab showing that violations of gender agreement between a pronoun and its
antecedent (e.g., sentence (2) below) elicit a P600 effect. To examine the brain
response to violations of occupational gender stereotypes, we presented sentences like
(1) The man prepared
(2) The man prepared
(3) The doctor prepared
for the operation.
(4) The doctor prepared
for the operation.
Participants made aceptability judgments after reading
each sentence. Sentences like (2) were judged to be unacceptable, but all other
sentences were usually judged to be acceptable. Thus, subjects' self-reports gave
llittle indication that sentences like (4), in which a presumed gender stereotype has been
violated, were perceived to be unacceptable or anomalous.
ERPs to the underlined pronoun in each sentence were of
interest and are plotted in the figure below. The vertical bar indicates onset of
the pronoun in each of the four conditions. Each hashmark represents 100 ms, and
negative voltage is plotted up. As expected, the pronoun in sentence (2), which
disagreed with the gender a definitionally male or female antecedent noun, elicited a
large P600 effect, relative to the condition in which the pronoun and antecedent noun
agreed in gender (sentence 1). The question was what would happen in sentences like
(3) and (4). Interestingly, pronouns that disagreed with the stereotypical gender of
it's antecedent noun also elicited a P600 effect, albeit one with lesser amplitude that
that elicited by the outright ungrammatical disagreement. Clearly, our subjects'
brains were classifying the stereotype violations as anomalous. This continued to
be true even when response-contingent ERPs were plotted: Even on trials on which
subjects said the stereotype-violating sentences were acceptable, the pronouns in these
sentences elicited a P600 effect.
We also observed compelling differences between our female
and male participants: Female participants exhibited a much larger-amplitude
"anomaly response" to both the definitional and stereotypical gender violations,
compared to the male participants.
Osterhout, L., Bersick, M., & McLaughlin, J. (1997).
Brain potentials reflect violations of gender stereotypes. Memory and Cognition,