We concluded our recent book on Successful Failure with an unusual call. To improve the fate of children at risk of any labeled failure, “the first and perhaps only step is to turn away from them” (Varenne & McDermott, 1998, p. 217). Counter-intuitive, yes, and surprisingly constructive and respectful to

Ray McDermott Stanford University

Hervé Varenne Teachers College, Columbia University

all children. If the only tools available for helping children in trouble are the diagnostic and remedial preoccupations of American education, it might be best to forget individual children and focus instead on how we have created contexts that make some children-about half of them-so problematic. If schools are for all children to flourish, then the individual child can be our unit of concern, but not our unit of analysis or reorganization. Why should kids be the focus of change when it is the rest of us-the culture that is acquiring them-that arranges their trouble? This conclusion was the systematic product of a cultural analysis applied to the most pressing issue in American schooling: the attribution of success/failure. In this chapter, we restate our conclusion in order, first, to discuss why culture-and neither the individual nor socializing group-is the crucial analytic unit for educational research and, second, to sketch how a cultural analysis leads to a new articulation of major policy issues, in particular, the failure of students and its complex relation to kinds of person by gender and race.1