The role of speech accommodation and crime type in attribution of guilt

In this experiment we examined the attribution of guilt as a function of speech accommodation and crime type. The subjects (129 White, English-speaking, South African students) listened to tape-recorded exchanges between a Coloured, Cape Afrikaans-speaking criminal suspect and a White, English-speaking interrogator. The subjects were asked to rate the suspect on a 7-point scale ranging from innocent (7) to guilty (1). The independent variables were (a) type of language shift (convergence, partial divergence, complete divergence) and (b) type of crime (blue-collar, white-collar). As we predicted (based on speech accommodation theory), the suspects who converged into English were rated as significantly less guilty than those who diverged into Cape Afrikaans. Furthermore, the suspects accused of blue-collar crimes were more often deemed guilty than were those accused of white-collar crimes (cf. Seggie, 1983).