cover final Against Sen Against Rawls on Justice Evan Riley* Amartya Sen has recently levelled a series of what he alleges to be quite serious but very general objections against Rawls, Rawlsian fellow travellers, and other social contract accounts of justice. In The Idea of Justice, published in 2009, Sen specifically charges his target philosophical views with what he calls transcendentalism and procedural parochialism, and with being mistakenly narrowly focused on institutions. He also thinks that there is a basic incoherence—arising from a version of Derek Parfit’s Identity Problem—internal to the Rawlsian theoretical apparatus. Sen would have political philosophy pursue inter-societal comparisons of relative justice more directly and in the manner of social choice theory. Yet the positive argument that he develops in support of this method is quite thin. That aside, Sen’s polemical strategy of inflicting death by a thousand cuts is ineffective against the Rawlsian paradigm. For, as I show herein, none of these criticisms has the force we might be led to expect. Keywords: The Idea of Justice, Amartya Sen, Transendental versus Comparative justice Among his many contributions to morally engaged intellectual life, Amartya Sen has long been a trenchant and sympathetic critic of John Rawls's work.1 One of his seminal efforts in that regard is his charge that Rawls's original emphasis in A Theory of Justice on shares of primary social goods, and in particular, on the socially sanctioned all purpose means of fulfilling one’s legitimate ends— for example, monetary income and wealth—is perhaps distorted (see Sen, 1980). Recall that for Rawls therein, the primary social goods are broadly categorized as the “rights and liberties, opportunities and powers, income and wealth” that a rational representative behind the veil of ignorance is presumed to want, and to want more of, whatever else they may want.2 On Rawls' view, it is the fair distribution of these very goods that the just basic structure of society is properly concerned with. As Sen pointed out, this classification may obscure the fact that what we ought to care about in this connection is not the largest fair share of stuff as such, even such potentially useful stuff as money. Instead, the truly basic emphasis should be on fairly fostering the development of capabilities of individual human beings to realize valuable forms of functioning. It is the production and distribution of this sort of thing that properly constituted societies ought to be concerned with. This is a valuable insight, and more of a friendly emendation than a devastating criticism.3 First, by abstracting away from monetary wealth, Sen’s line of thought broadens the potential scope of the Rawlsian-style argument to naturally include more social worlds and more kinds of individual lives. Second, by placing the emphasis squarely on the realized capabilities of human beings, Sen encourages us not to distort our reflections about social justice in the direction of thinking that it is merely a matter of securing the negative liberties. I say ‘Bravo!’ to this. Indian Journal of Human Development, Vol. 5, No. 1, 2011 * Visiting Assistant Professor of Philosophy, The College of Wooster, Ohio, USA. Email: eriley@wooster.edu 212 Indian Journal of Human Development Sen has more recently been arguing for a full-scale recasting of academic philosophical theorizing about social justice.4 I think that this new line of argument is neither remotely convincing in its attack on Rawls nor insightful in its intended basic aim. In an article written in 2006, “What Do We Want from a Theory of Justice?” (see Sen, 2006), he defended this line and strikingly concluded that political philosophers should simply stop pursuing, in the style of Rawls, the grand question ‘What is a just society?’. In presumptive contrast with this dominant approach, Sen argues that political philosophers should be centrally concerned with the making of comparative judgments of justice in a way inspired by the social choice tradition and by Adam Smith, among others. This line of thinking has been redeployed, along with several other arguments that are critical of Rawls, and of the Rawlsian paradigm, in Sen’s recent book, The Idea of Justice (see Sen, 2009). Therein Sen alleges that Rawls' exhaustively discussed and enormously influential account of social justice in A Theory of Justice is vitiated by a host of very serious faults that have not been properly appreciated to date. Chief among these are transcendentalism, procedural parochialism, institutionalism, and populational plasticity. The faults are serious enough, according to Sen, to render the theory and its many relatives and descendants distinctly unhelpful as tools for “enhancing justice and removing injustice” or for serving as “the basis of practical reasoning” in that regard (see Sen, The Idea of Justice, 2009, p. ix). A new approach is recommended: call it comparativism. Here, I argue that there are basic difficulties with the critical side of this line of thought. First, it appears that either these alleged faults are not both correctly attributable to Rawls and genuine problems, or if they are genuine problems, are not specific to the Rawlsian problematic. Second, it is not clear that every such theory of justice is properly supposed to provide us with concrete plans of action for its own realization, and Sen does not make that case. Herein, I am mostly occupied with a critical discussion of the alleged problems mentioned above. I close with some very brief reflections on theories of justice and the actualization of justice. First, what is the character of the alleged problem of transcendentalism, said to be characteristic of what Sen calls “transcendental institutionalism” or in some places, the “contractarian approach” (see Sen, The Idea of Justice, 2009, pp. 5 and xvi, respectively)? Sen suggests that the Rawlsian-style pursuit of the grand question— ‘What is a just society?’—involves a pretension to extremely demanding theoretical comprehensiveness concerning its object. Indeed, Sen takes the contractarian tradition generally to be so focused on a kind of theoretical perfection that it is blocked from developing the materials for making comparative judgments of relative justice and injustice. It does seem clear that making such comparative judgments is a necessary condition of practically reasoned social reform. For in the absence of some actual comparisons, in point of justice, between say present conditions on the ground and an achievable reform, a desire to seek and support justice would be without reasoned practical import. Thus, we are invited to think that pursuing the question in a Rawlsian—or more generally contractualist—fashion will militate against realizing needed social reform. Against Sen Against Rawls on Justice 213 Two questions naturally arise. First, what exactly is the nature of the problem for the comparison posed by the relevant theoretical perfectionism? Second, is the Rawlsian framework perfectionist or transcendentalist in that sense, and is it thereby worse than useless for thinking comparatively about social justice here and now? Sen’s answer to the second question is ‘Yes’. I will take this up below and show that this is a mistaken answer. Sen’s answer to the first of these questions appears to be as follows: first, any theory that aims at completely thoroughly describing the just society and at making out an ideally complete and useful set of practical principles for realizing and maintaining justice is perfectionist (or ‘transcendental’) in the relevant sense. And, second, aiming at this terribly ambitious set of goals is substantively mistaken. There seem to be two considerations that Sen is relying on in support of this second claim just above. On the one hand, he thinks that social justice is intrinsically pluralistic in character, so it will often make inconsistent demands. He claims that there are “genuinely plural, and sometimes conflicting, general concerns that bear on our understanding of justice.” (See Sen, The Idea of Justice, 2009, p. 57). And the fact that there is a sort of deep pluralism in the demands of justice is the lesson naturally drawn from Sen’s parable of the flute and the three children (see Sen, The Idea of Justice, 2009, pp. 12-13 and p. 106). If that is correct, there simply will be no coherently unified story to tell about those demands, even where, as in the social contract model, everyone’s impartial reasons are being fully expressed and fairly consulted. One might presume against this that there is some unified answer to the question of what social justice requires in and perhaps between contemporary societies, albeit perhaps expressible only as a very abstract set of norms, in some order of priority. This is the assumption of most philosophers who inquire into the nature of social justice. There are two issues worth considering just here. First, Sen does not give a decisive argument for the controversial view that justice is intrinsically pluralistic even at a quite general level, and the jury appears to be out, maybe permanently so, on that question. Second, if he were right about this, that should not be particularly comforting for the reform-minded theorist of justice of the proposed new comparative style. For in that eventuality, things would be even murkier than on the competing monistic picture. We would be in the unenviable position of weighing multiple competing demands of the various justices along with all the familiar other possibly competing demands imposed on us by such alleged social virtues as: security against invasion, the very efficient provision of basic material goods, high rates of growth in GDP, high average living standard, or, in a more philosophical gear, maximal well-being. To put it simply, if social justice itself genuinely pulls us in multiple and even opposed directions, there may be no grounds for making even rough comparisons of relative justice and injustice as such. Pointing out that we can and do make choices in circumstances wherein we weigh goods that cannot themselves be made fully commensurate is not a good defence here (see Sen, The Idea of Justice, 2009, p. 241). In short, if Sen leans too heavily on the alleged fact of robust pluralism, then he undermines the grounds for thinking that there could be a comparative theory of justice at all. 214 Indian Journal of Human Development Sen’s other reason for thinking that the complete description of societal justice (and its realization) is an inapt aim has to do with the thought that it would be such a difficult task to complete. This is quite clear in the following: How complete should the assessment be? In the “totalist” approach that characterizes the standard theories of justice, including Rawls' incompleteness tends to appear as a failure…Indeed, the survival of incompleteness is sometimes seen as a defect of a theory of justice, which calls into question the positive assertion such a theory makes. In fact, a theory of justice that makes systematic room for incompleteness can allow one to arrive at quite strong… judgments, without having to find highly differentiated assessments of every political and social arrangement in comparison with every other such arrangement (for example addressing such questions as: exactly how much tax should be put on the sale of petrol in any particular country, for environmental reasons?) (Sen, The Idea of Justice, 2009, p. 103, emphasis added). Sen seems to be thinking that a complete theory of justice is one that will address every practical scenario generating a need for collective or cooperative decision that might come up in the human circumstances of justice. Such a theory would not only descry a coherent system of social institutions but would provide a detailed conceptual structure suitable for determining what justice calls for, in actions, institutions, and policies given whatever circumstances we find ourselves in, and with whatever preferences we might have. It seems that the theory would contain both a highly detailed model of the just society, and also reliable procedures for getting there and for maintaining justice once it is achieved. Sen suggests that this sort of (hopelessly) extreme comprehensiveness is precisely what the transcendental approach, as he calls it, aims at, that is, it aims at “a complete resolution of all the existing decisional issues about societal organization.” (See Sen, “What Do We Want from a Theory of Justice?”, 2006, p. 236.) This is a grand aim indeed. I am inclined to accept, with Sen, that this aim is indeed far too grand to be seriously pursued, and that it reflects confusion. But is one stuck with this aim merely in pursuing the question in view in the relevant fashion? That does not seem to be the case. Let us now consider Rawls' work in this connection, for it is Sen’s paradigm case. Rawls' dominant focus was indeed the topic of social justice in modern mass democracies and Sen is on solid ground to count him as pursuing a version of the grand question by appealing to a development of the idea of the social contract. Yet, in several places, Rawls also explicitly denies that he has the aim of settling in advance, by means of his theory, every practical question that might possibly arise in the circumstances of justice, even for those imperfect contemporary democracies that his theory speaks to most directly. He simply is not aiming at that sort of absolute theoretical comprehensiveness. Consider the following late discussion of the political conception of justice and the conception of the basic structure. The role of a political conception of justice however is not to say exactly how these questions are to be settled, but to set out a framework of thought within which they can be approached. Were we to lay down a definition of the basic structure Against Sen Against Rawls on Justice 215 that draws sharp boundaries, not only would we go beyond what that rough idea could reasonably contain, but we would risk wrongly prejudging what more specific or future conditions may call for, thus making justice as fairness unable to adjust to different circumstances. For our judgments to be reasonable, they must usually be informed by an awareness of those more specific circumstances (See Rawls, Justice as Fairness: A Restatement, 2001b, p. 12). Rawls, as we can see, is indeed pursuing the grand question, while self-consciously explicitly abjuring the ambitious theoretical aim. Sen, however, takes this aim to follow from that pursuit. It follows that if Rawls is correct about the nature of his own theory, and if his project is roughly coherent (as it seems), Sen is wrong. It is possible to pursue the grand question without aiming at absolute theoretical comprehensiveness. Perhaps it will be thought that since the above passage appears relatively late in Rawls's career, it may not reflect the truth about the view in its dominant form. Perhaps, as Sen puts it, Rawls has telling “second thoughts” on this score, the implications of which are not fully worked out (see Sen, The Idea of Justice, 2009, p. 58). I am not convinced. It seems clear from the canonical expression of the view in A Theory of Justice that Rawls never embraced the sort of theoretical comprehensiveness that Sen attributes to him.5 Moreover, I think that there is general reason to doubt that pursuing the grand question requires one to embrace the aim of total (practical-) theoretical comprehensiveness. Suppose that some philosopher embraces a substantial theoretical modesty when it comes to ethics and political philosophy, or, in general, the demands of practical reason. We might imagine that she has become convinced that practical wisdom is uncodifiable, and accordingly also convinced that the relevant idea of a complete theory of perfect social justice is a misleading chimera.6 Do we insist that this philosopher somehow contradicts herself if she nonetheless vigorously pursues the grand question in a contractual vein? Is there something self-defeating about combining moral-theoretical modesty with the pursuit of the grand question of justice in the style of Rawls? I see no reason to think so. Such a philosopher perhaps will insist on the importance of attending to details if we are to uncover anything useful for action. She would remind us that the general principles of justice are always subject to revision and often require supplementation, or filling in, in order to express the relevant provisional practical truth, or that reflective equilibrium is always somewhat provisional. But being committed to this sort of theoretical modesty is not in any tension with the vigorous pursuit of the grand question. Despite these considerations, Sen could yet conceivably nonetheless be correct and Rawls incorrect about the Rawlsian view, and more generally about the social contract approach. But if that were so, we would not expect to find principles useful for making comparisons of relative justice nested within, and derivable from, the theories that this approach yields. Yet we do find such principles. And worse still, for Sen’s line of objection, we find principles that manifestly call upon us to make comparisons. Consider Rawls' difference principle, which roughly holds that social and economic inequalities are to be arranged to the greatest benefit of the least advantaged members 216 Indian Journal of Human Development of society (see Rawls, A Theory of Justice, 1971, p. 302). The basic implication is that any society S will count as just only when the social and economic inequalities are such that the position of those worst off relative to the other members of S, are better off in absolute terms than are the worst off members of every feasible alternative: S*, S**, etc. This is not a sufficient condition of social justice, according to Rawls, but it is in his view a necessary one. Now, not only can we not apply the difference principle in assessment of the justice of some society S except by making comparisons with other possible or actual societies, we cannot really understand the difference principle except as calling for comparative judgments. This is so since Rawls views the principle as a guiding choice between feasible alternatives, one of which may be actual. Manifestly then, one can pursue the grand question, and in so doing, develop the conceptual materials for making comparative judgments of justice. Sen grudgingly admits as much at one point in The Idea of Justice (see Sen, The Idea of Justice, 2009, p. 97). Consider next the thought that some explicitly comparative approach abjuring the grand question is thereby to be preferred to the transcendental approach. Sen supports this by arguing that achieving the aim of the transcendental approach is neither sufficient nor necessary for producing the relevant comparisons (see Sen, The Idea of Justice, 2009, pp. 98-102.) Let us consider this. As a matter of logic, it is true that possessing even some ideally complete theory of the just society does not, by itself, rank any given set of societies in point of justice. Even achieving the aim of the transcendental approach, as Sen describes it, were that possible, would not automatically generate comparisons. But this is equally true of accounts of justice explicitly focused on comparison and which are designed to accommodate incompleteness! It is of course only when it is conjoined with some relevant information about those societies—their historical and cultural circumstances, and so on—that any theory can provide the cognitive materials needed to make the relevant judgments. Once conjoined with relevant information, a theory generated by either approach may be quite useful for making comparisons. For illustration, consider again Rawls's justice as fairness with its two principles of justice, in lexical order.7 Recall that the first principle calls for the equal protection of the basic liberties, and that the second demands both protection for equitable opportunity to compete for attractive positions, and that any remaining inequalities serve the advantage of the least well-off members of society, relative to feasible alternatives. Suppose that the conditions on the ground are as follows. Three societies—S, S*, and S**— are each apparently feasible alternatives of the others. In society S, there are barriers to advancement and political participation, entrenched in positive law, directed against the members of a particular ethnic group on the sole grounds of membership in that group. Further, the economic institutions of S are rife with corruption and inefficiency, and shot through with extremes of income and wealth inequality kept in place by pervasive racism. Here, it seems as if neither of Rawls's principles is met. In society S*, let us suppose that the first principle is met, since there is a familiar set of enshrined constitutional protections of a version of the equal basic liberties that are reasonably effectively enforced. But, let us also suppose that the worst off in S* Against Sen Against Rawls on Justice 217 are plainly not as well off as are the worst off in society S**, though better off than S.8 On still further supposition, in S**, both principles are arguably met, for, as in S*, the constitutional protections are in place and enforced. Let us also assume that these are realistic cases in that none of these societies is perfectly well-ordered under any theory of justice. When made aware of these empirical conditions, the Rawlsian theorist is surely in a cognitive position to rank the three societies in point of justice. If she has had her morning coffee, she would rank S** as just, S* as partly just, and S as unjust. Notice that in the absence of her grasp of Rawls's specific theory, she might well be unable to rank S* and S** differentially, in point of justice, that is, her grasp of the theory allows her to judge that S** is more just insofar as the inequalities are organized to the benefit of the least advantaged. The grasp of the theory is perhaps usefully thought of as an INUS condition for making the comparative judgments. That is, perhaps it is an individually insufficient but non-redundant part of a set of conditions that are jointly unnecessary but here and now sufficient to produce the comparative judgments (see J.L. Mackie, The Cement of the Universe, 1980). In any event, it is clear that Sen’s original claim that no “transcendental theory of justice” even if “entirely successful in answering that classic question…would …yield—directly or indirectly—a comparative framework… needed for the actual assessment of justice” is false (see Sen, “What Do We Want from a Theory of Justice?”, 2006, p. 236). But perhaps the reader is unimpressed with my hypothetical and terribly abstract example of the Rawlsian apparatus in action. Suppose one thinks that, in practice, social contract philosophy, both generally, and in its specifically Rawlsian version, is likely to lead one down unproductive paths from the point of view of realizing justice. Against this empirical claim, consider the work of Thomas Pogge, a doctoral student of Rawls and an eminent political philosopher in his own right (see Pogge, World Poverty and Human Rights, 2008). Pogge has been working towards realizing a Rawlsian conception of justice for many years. He is particularly interested in issues of global justice, and is sympathetic in places to the idea of a global analogue of the Rawlsian hypothetical Original Position (OP), one that takes individuals—and not peoples—to be the represented parties (see, for example, Pogge, “An Egalitarian Law of Peoples”, 1994, pp. 195-224). Very simply, Pogge takes it that the reasoning exhibited by the representatives in the global OP would license at the very least the justice of a global tax on the extraction of non-renewable resources, with the proceeds intended to benefit the globally worst-off. More recently, he has been working on institutionally realizing the idea of a Health Impact Fund to reduce the burden of preventive disease on the world’s poor, and in general, on issues of global poverty relief. It seems to be obvious that in Pogge’s case, his Rawlsianism has productively and directly informed his work, justifying controversial claims about justice, though it has, of course, not been solely determinative. Sen is, of course, familiar with Pogge’s work, but it gets extremely short shrift in The Idea of Justice. Sen’s critical comment on Pogge that “the idea of one global exercise of social contract for the entire world population would appear to be deeply unrealistic” does not bear much scrutiny.9 218 Indian Journal of Human Development It is obviously true that one need not have a precisely and elaborately detailed complete theory of justice (in the relevant sense) in order to have some basis for making comparative judgments. We should accept Sen’s view that this sort of theory is not a necessary condition for comparison (and so for achieving reasoned reform) here and now. But what does one need and how should one proceed? If one is interested in thinking systematically about the basic demands of justice and then comparing some set of societies’ relative distances from meeting those demands, it remains clear, Sen’s arguments to the contrary notwithstanding, that a good way to begin is with the pursuit of the grand question. Sen is undoubtedly correct in thinking that it is possible to give reasons for a choice, and even to secure some agreement, without presupposing total unanimity of well-ordered preferences, or complete knowledge within a community of choice. But part of the strength of the Rawlsian perspective, and of that afforded to us by the social contract mode generally, is to get some significant traction on the question of social justice that rises above empirically given preferences. These are devices for modelling reflectively sound reasoning. It seems to me that there is no good reason for thinking that the partial social ordering generated by a congruence of, say, a majority of a given group’s set of preference orderings is per se to be given special weight when thinking about what justice requires. In the discussion in the concluding pages of his book, Sen gestures towards considerations of public reasons and impartiality as necessary filters for arriving at a sensible inference (see Sen, The Idea of Justice, 2009, pp. 398-99). Yet this seems to push us towards a Rawlsian model. Allow me to briefly consider some of the remaining charges levelled in the book. First, consider what I call procedural parochialism. In a move partly inspired by his love for Smith’s social thought, Sen argues that an impartial outside observer may be in a better position to sort out what justice requires in some concrete social setting than is a person whose life and livelihood are bound up in that locale. Thus, if an actual group were to embark on an attempt to engage in public reason about the rules of the game for themselves, the views of an outsider might be useful for blocking local prejudice or personal and potentially divisive interest from playing the wrong sort of role (see Sen, The Idea of Justice, 2009, p. 70). This might be correct in any particular case, though it seems like an entirely contingent matter. But it does not amount to a sound criticism of Rawls or of all contractualist ways of framing the relevant issues. After all, the very point of the Rawlsian model, with its ideally well-informed but crucially ignorant hypothetical representatives, is to emphasize the importance of impartiality. Rawls and related thinkers are also charged with being wrongly narrowly focused on institutions as the site of social justice. The thought here is that if we focus narrowly on our favoured institutions as manifestations of justice—for example, properly regulated relatively transparent markets, formal democratic electoral procedures, formal equality before the law— once these institutions are somehow concretely realized in the world, we may fail to take account of the effects that these may be having on actual person’s lives. As I noted above, Sen is to be commended for emphasizing, as fundamental, the realization of valued forms of functioning. But, again, this is not a particularly Against Sen Against Rawls on Justice 219 devastating criticism of Rawls. The latter attends to the basic structure of society as that over which the principles are properly authoritative precisely because in human social life, the nature of the overall institutional structure has thoroughgoing effects on the lives of actual people. It is for this reason that Rawls favours maximin as a theory of principle choice, rather than some other more familiar and less risk-averse strategy of choice under uncertainty.10 So it seems perverse to charge Rawls with neglecting this.11 The alleged populational plasticity problem is the final one that I will consider. A relevant point is made particularly clearly in Derek Parfit’s magisterial Reasons and Persons, viz., that choices of social policy in some society will affect the character and size of the future population of that same society (see Parfit, Reasons and Persons, 1984, pp. 355-77). Parfit was worried over the moral implications of this, which he calls the Non-Identity Problem. Sen thinks that a version of this problem arises in an acute form for Rawls and renders his account of justice incoherent. The thought is that the number of participants in the hypothetical assembly is indeterminate as it is affected by the decisions of that very assembly. Since the choice of principles will affect the basic structure and the character of the basic structure will affect the size and character of the population, the model is presumed to be unworkable (see Sen, The Idea of Justice, 2009, p. 145). It is admittedly difficult to know how to precisely take the interests of potential future people into account when thinking about the requirements of justice. After all, it is difficult to see how decisions we take now could be said either to help or to harm people who do not exist and whose very existence is causally down- stream from those decisions! But surely it is coherent to ask, of the members of a given presently living generation of citizens, how they could be represented to reason in a version of the Rawlsian OP. We are simply not required to assume that every possible future person be given a voice in the hypothetical procedure. In conclusion, Sen is certainly correct to think that comparisons of relative justice and injustice should concern those who care about actually making the world a better place. He is also on strong ground in insisting that aiming at a certain kind of theoretical perfection is quixotic and unnecessary for making useful comparative judgments. But these are not things that Rawls, the Rawslians, or social contract thinkers would generally deny. Nor does any feature of their general way of proceeding make them incapable of appreciating these points. There is one final point to register. I would like to resist cynicism about the prospects of concretely engaged political philosophy. Yet it appears highly unlikely that the genuine practical problems posed by the existence of serious injustices, and their going unaddressed, will be resolved by some general paradigm shift—perhaps moving away from the social contract model—in our theories of justice. What mainly keeps us from doing more to reduce injustice is not I think an over-utopian or obsessively perfectionist streak in our theories of justice. Instead, it is the drearily familiar litany of factors: ignorance, selfishness, confusion about what is, on reflection, important in life, failures of will, the distractions of living in a consumer society, the struggles of keeping life and limb together, the ideological state apparatuses, the dominance of our political institutions by the corrupt, the self-interested, and the incompetent, and 220 Indian Journal of Human Development so on. In pursuing the grand question publicly, we open up the reflective space to articulate and then to begin to redress these factors. That openness and the subsequent articulation of the demands of justice remains part of what we ought to want from the activity of theorizing about justice. NOTES 1. For a crisp recent statement of Sen’s sense of his own intellectual and personal relation to Rawls, see his “The Place of Capability in a Theory of Justice” 2010, pp. 239-253. 2. See John Rawls, A Theory of Justice, 1971. I assume that readers are familiar with Rawls' account of justice. In short, he seeks to model relevant impartiality in the reasoned choice of principles of social justice (viewed as governing the institutionally realized distribution of benefits and burdens attendant upon any society) by blocking the hypothetical procedure of choice from reliance on reasons which he thinks are arbitrary from the moral point of view. One’s representative cannot, for example, bargain in a self-interested way from one’s place in the social hierarchy to a set of principles, differentially favouring those in that place, since that sort of reasoning is blocked by what he calls the ‘veil of ignorance’. 3. It is not clear that this insight is the source of a powerful objection to Rawls. See Footnote 3 in John Rawls, The Law of Peoples, 2001 [1999], p. 13. 4. See in particular Amartya Sen, “What Do We Want from a Theory of Justice?”, 2006, pp. 215-38. See also Sen, “The Place of Capability in a Theory of Justice”, 2010, and his introduction in Adam Smith, The Theory of Moral Sentiments, 2009 [1790], pp. vi-xxvi. 5. See, for example, Rawls, A Theory of Justice, 1971, p. 566. “Now this sequence does not aim at the complete specification of conduct. Rather the idea is to approximate the boundaries, however vague, within which individuals and associations are at liberty to advance their aims and deliberative rationality has free play.” 6. The uncodifiability thesis holds that the deliverances of practical reason (whether judgments or actions) are, in general, not to be understood as consisting in the applications of a rule, every instance of which can be completely specified in advance. It is linked with what is sometimes called anti-theory in moral philosophy. For discussion of and motivation for the uncodifiability thesis, see John McDowell, “Virtue and Reason”, 1998 [1979]. 7. See Rawls, Justice as Fairness: A Restatement, 2001b, pp. 42-43. The idea that the principles would be chosen in lexical order reflects the concern that roughly we should not simply trade off economic gains at the expense of sacrificing the basic liberties. 8. I am not assuming that ‘better off than’ always names a relation strictly transitive at the level of general logic. See Stuart Rachels, “Counter-examples to the Transitivity of Better Than”, 1998, pp. 71-83. But I am assuming that here those worst off in S are both more worse off than the relevant those in S* and way more worse off than the relevant those in S**! 9. See Sen, The Idea of Justice, 2009, pp. 140-41. Rawls' original model and its global analogues are, of course, hypothetical, and it simply isn’t clear what sort of “exercise” Sen is labelling as “unrealistic”. If there is a sensible objection here, Sen has not succeeded in making it. Presumably, he is intending to side with Thomas Nagel in emphasizing the idea that at present, the political institutions that would allow for an immediate implementation of ambitious global social policies related to justice are simply not developed enough. Yet, it is not remotely clear that this constitutes a reason counting against the relevance, to global justice, of the Rawlsian apparatus. See Thomas Nagel, “The Problem of Global Justice”, 2005, pp. 113-47. 10. To put it simply, the maximin principle tells one to maximize the minimum or to choose the scenario with the best of the worst achievable outcomes across a range of feasible scenarios. It expresses a high degree of risk aversion. See Rawls, A Theory of Justice, 1971, pp. 152-53. 11. Compare Sen, The Idea of Justice, 2009, p. 85 and Rawls, The Law of Peoples, 2001a, Footnote 3, p. 13. Against Sen Against Rawls on Justice 221 REFERENCES Mackie, J.L. (1980). The Cement of the Universe, Oxford University Press, Oxford. McDowell, John (1998) [1979]. “Virtue and Reason”, Mind Value and Reality, Harvard University Press, New York. Nagel, Thomas (2005). “The Problem of Global Justice”, Philosophy and Public Affairs, Vol. 33, No. 2, pp. 113-47. Parfit, Derek (1984). Reasons and Persons, Clarendon Press, Oxford. Pogge, Thomas (1994). “An Egalitarian Law of Peoples”, Philosophy and Public Affairs, Vol. 23, No. 3, pp. 195-224. —— (2008). World Poverty and Human Rights, (Second edition), Polity Press, Cambridge. Rachels, Stuart (1998). “Counter-examples to the Transitivity of Better Than”, Australasian Journal of Philosophy, Vol. 76, No. 1, March, pp. 71-83. Rawls, John (1971). A Theory of Justice, Harvard University Press, Cambridge, MA. —— (2001a) [1999]. The Law of Peoples, Harvard University Press, Cambridge, MA. —— (2001b). Justice as Fairness: A Restatement, Edited by Erin Kelly, Harvard University Press, Cambridge, MA. Sen, Amartya (1980). “Equality of What?”, in S. McMurrin, Tanner Lectures on Human Values, Vol. I, Cambridge University Press, Cambridge. —— (2006). “What Do We Want from a Theory of Justice?”, The Journal of Philosophy, Vol. CIII, No. 5, May, pp. 215-38. —— (2009). The Idea of Justice, Harvard University Press, Cambridge, MA. —— (2009) [1790]. “Introduction” in Adam Smith, The Theory of Moral Sentiments, Penguin Books, New York, 2009 pp. vi-xxvi. —— (2010). “The Place of Capability in a Theory of Justice”, in Harry Brighouse and Ingrid Robeyns (eds) Measuring Justice, Cambridge University Press, New York, pp. 239-53. 222 Indian Journal of Human Development front 13-Evan Riley back