allowing for a distinction to be made between political philosophy and political theory, these prescriptions seem to me to be unduly constraining. Political theory is a diverse field, accommodating highly abstract inquiries that aim to identify fundamental normative principles by putting to one side facts about our natures and the circumstances in which we live, as well as more grounded forms of reflection that begin by paying close attention to the political problems we face and then seek guidance from the normative principles that are immanent in our practices. In response to attempts to draw limits to what kinds of political theory should be done, or to elevate some forms above others, I shall defend a qualified pluralism that recognizes value in a variety of approaches and resists arguments that purport to show that one particular approach should occupy a privileged position. Against realists, I 2 argue that abstract analyses of political values that bracket a wide range of facts about people and their circumstances can be both coherent and important, whereas I argue against those who think "ideal theory" or the identification of ultimate principles should come first, that the case for always giving it priority is weak. 3 I focus on the notion of justice in particular. 4 I examine three claims that pull in different directions, but any one of which, if it were correct, would alone threaten to place a serious limit on legitimate theorizing about justice. The first is that a socalled "ideal theory of justice" is both useless and misconceived because it is founded on the assumption that citizens agree on principles of justice, thereby depriving it of practical relevance and displaying a misunderstanding of the nature of the political.The second claim is that we cannot justify a principle of justice without taking into account what the effects would be of adopting it in the particular context that it is to govern. And the third one is that we cannot justify conclusions about what justice requires in the circumstances we face, with the constraints they impose on us, unless we know what justice would require in the absence of these constraints. This third claim comes in radical and moderate versions, both of which I shall examine. Neither denies the legitimacy or importance of theorizing about justice that takes into account local facts about what is feasible. But one version maintains that if we are to justify our practical conclusions fully, we have to abstract from all facts about what is feasible in order to identify and then defend the ultimate principles of justice that (it is claimed) must underlie the arguments for these conclusions; the other version maintains that in order to justify these conclusions fully, we need to develop an account of justice that involves putting to one side at least some "soft" constraints that may change over time, while keeping in place "hard" constraints that derive from 3 physical laws, unchangeable facts about the natural environment, or aspects of human nature that are fixe...