Motivational changes due to individual differences and situational variations in ingroup identification can influence accessibility of memories of ingroup violence, victimization and glories. In Study 1, high identifiers recalled fewer incidents of ingroup violence and hatred than of ingroup suffering. As well, they recalled fewer incidents of ingroup violence and hatred than did low identifiers. In Study 2, a manipulation of ingroup identity produced shifts in memory. Relative to those in the low identity condition, participants in the high identity condition recalled fewer incidents of violence and hatred and more good deeds by members of their group. Participants in a control condition recalled more positive than negative group actions; this bias was exaggerated in the high identity condition and eliminated in the low identity condition. With respect to memories of ingroup tragedies, Studies 3 and 4 demonstrated that experimental reminders of ingroup suffering enhanced participants' sense of connectedness to the ingroup. The findings suggest that memories of ingroup aggressions threaten ingroup identity whereas memories of ingroup suffering enhance ingroup identity. Societal implications of the findings are discussed. The present research informs the literature on reconstructive memory by extending previous findings on the flexibility of personal memories to historical memory.

Motivational changes due to individual differences and situational variations in ingroup identification can influence accessibility of memories of ingroup violence, victimization and glories. In Study 1, high identifiers recalled fewer incidents of ingroup violence and hatred than of ingroup suffering. As well, they recalled fewer incidents of ingroup violence and hatred than did low identifiers. In Study 2, a manipulation of ingroup identity produced shifts in memory. Relative to those in the low identity condition, participants in the high identity condition recalled fewer incidents of violence and hatred and more good deeds by members of their group. Participants in a control condition recalled more positive than negative group actions; this bias was exaggerated in the high identity condition and eliminated in the low identity condition. With respect to memories of ingroup tragedies, Studies 3 and 4 demonstrated that experimental reminders of ingroup suffering enhanced participants' sense of connectedness to the ingroup. The findings suggest that memories of ingroup aggressions threaten ingroup identity whereas memories of ingroup suffering enhance ingroup identity. Societal implications of the findings are discussed. The present research informs the literature on reconstructive memory by extending previous findings on the flexibility of personal memories to historical memory.