Elizabeth Harman: Why racist research must be named, but often allowed