Influential alternative accounts conceive of justice either as broadly applicable but not as stringent as I do or as similarly stringent but not as broadly applicable as I do. Ernst Tugendhat exemplifies the former approach, Immanuel Kant the latter. Their approaches generate an objection to mine: either distributive justice can be defined, as Tugendhat does, in terms of the kind of situation to which it applies and then plausibly applied to a broader range of contexts than my theory proposes, or distributive justice can be defined in terms of its stringency, as Kant does. I argue that it is sensible to limit justice to a narrower range of situations than Tugendhat allows, while thinking of the stringency of justice in such a way that a broader class of cases is covered than Kant allows. This takes us to something like my view.