Many people have contributed to the development of this article-not the least by restraining its excesses. Conversations with Richard Nisbett and Jeffrey Paige were especially helpful. The influence of other members of the social psychology community at Northwestern University working on related problems-especially Ronnie Janoff-Bulman, Christine Dunkel-Schetter, Roxane Silver, and Camille Wortman-will be apparent at many points. Thoughtful responses to earlier drafts were also received from