The semantic characterization of the (syllogistic) relations of contrariety and subcontrariety is problematic, as the present discussion illustrates here by attending to some suggestions of D. H. Sanford.