A range of phenomena - from wrongful convictions to family vendettas – suggest that blame and punishment sometimes target individuals who are not involved in wrongful acts. However, dominant causal theories of blame often fail to explain this. In the present research we go back to Heider’s (1958) proposal that association is the first, most-basic mechanism for attribution of responsibility, examining whether the existence of social association on its own could allow for blame to spread. We presented 101 university students with stories featuring a character who was in the same group as the perpetrator of a harmful deed and a character who was not. Both characters were depicted as uninvolved and unaware of the harm. A third group member was an accomplice of the perpetrator. Groups were either presented as being close knit (high in entitativity) or organized ad-hoc (low in entitativity), allowing us to examine how strength of affiliations influence on the spread of blame. Participants rated the victim's anger towards each character. Our key hypotheses posited that (a) uninvolved group members would incur more blame for a harmful act than uninvolved non-group members, (b) this difference in blame attribution would be greater in groups with high entitativity. As expected, the accomplice was blamed less than the perpetrator, and the group member and the non-group member were blamed less than the accomplice. However, consistent with the associative proposal advanced by Heider, participants attributed more blame to the uninvolved group member than the equally uninvolved non-group member. The difference was significant in the high but not in the low entitativity condition. These findings provide strong evidence for the role of non-causal social association in the spread of blame, calling for the expansion of current models of blame attribution to incorporate purely associative pathways.
Culbert et al. (Thu,) studied this question.