On-line processing of social stereotypes during spoken language comprehension.

Four experiments investigated stereotype processing during the on-line comprehension of spoken sentences. Participants listened to a critical prime (terrorist) embedded in a sentence and then made lexical decisions to visually presented standard-related (bomb) or stereotype-related targets (Arab) or unrelated controls for both conditions. For Experiments 1A and 1B, the preceding context provided no contextual information about the critical prime. Targets were presented at prime offset (Experiment 1A) or 300 ms after prime offset (Experiment 1B). In Experiments 2A and 2B, the preceding context was biased towards the stereotype meaning of the critical prime. Visual targets were presented at prime offset (Experiment 2A) or 300 ms after prime offset (Experiment 2B). Experiment 1A revealed a priming effect, but only for the standard meaning (i.e., bomb) of the critical prime. This effect was further replicated in Experiments 1B, 2A and 2B; however, for these experiments, unlike Experiment 1A, the results revealed lexical inhibition for the stereotype meaning of the critical prime. That is, lexical decisions to the stereotype-related targets were actually slower than the stereotype-unrelated controls. The results are discussed in terms of current social cognitive models of stereotype processing.