The Idea of Justice is much more than a collection of previously published papers. It is the pinnacle of a 40-year research project that began with important contributions in the field of social choice theory, a branch of studies that includes both political science and economics. Probably no one of the theses defended in the book are completely novel, but what is original and new is the goal (only partially achieved) to offer a consistent approach to the idea of justice, that differs from the main theories present in the philosophical debate. For this reason the book has an added value with respect to the sum of its parts.

The tradition of thought to which Sen wants to belong began in eighteenth-century France with Condorcet and Borda, then underwent much development in the twentieth century, thanks primarily to Arrow's impossibility theorem. Arrow's theory deals with the aggregation of preferences and individual values in collective or social choices, which are then used as guidelines for public choice. As it is well known, the messages emerging from that theory are primarily paradoxes and theorems of impossibility (one of which was formulated by Sen 40 years ago: ‘The impossibility of a Paretian liberal’); these bring to light the complexity of the transition from individual preferences to collective choices, the risks of manipulation, and the easiest solutions or shortcuts (the best known solution is a dictatorship, in which the preferences of a single person become those of the entire collective). But for Sen, social choice theory serves chiefly as an alternative paradigm and methodology to the dominant way of treating the issue of justice in classic authors of the past (Hobbes, Rousseau, Locke and Kant) and present (primarily John Rawls). Sen calls this line of thought ‘transcendental institutionalism’ (p. 5), an approach deeply related to the contractarian tradition, that tends to seek the ideal institutions and procedures to build a just society, usually through a social contract. A second characteristic of the transcendental approach is its focus on a single dimension (primary goods, resources, utility and so forth) to equalize in order to build a just or perfect society.

In fact, Sen writes in the Introduction the purpose of his book: ‘In contrast with most theories of justice, which concentrate on the ‘just society’, this book is an attempt to investigate realization-based comparisons that focus on the advancements or retreat of justice’ (p. 8). For him the most important question regarding justice is ‘how would justice be promoted?’, rather than ‘what would be perfectly just institutions?’ (p. 9); the latter is at the core of what he calls the ‘transcendental’ project.

The pars denstruens of Sen's book is tightly concentrated on the ‘transcendental’ vision of justice; according to Sen, the great effort that several major philosophers have put into finding the right institutions and procedures to construct the perfect society has made us forget the urgency of fighting ‘manifest injustices’ and of taking the time to improve an unjust situation so that it is less unjust, which for Sen is what the principal purpose of a theory of justice should be. An example that frequently recurs in the book is taken from the art world: to choose between a painting by Dalì and Picasso, it is not very useful to know that the perfect painting is Leonardo's Mona Lisa (p. 16). The central intellectual hero in Sen's ‘comparative’ theory of justice is Adam Smith of the Theory of Moral Sentiments (1759), particularly his theory of the ‘impartial observer’. In this comparative approach to justice, social choice theory meets his needs, since above all it is a theory that does not search for the optimal society, but rather compares different states of existence in the world and suggests how to choose one or the other (using criteria based on voting theory).

This theme of comparing imperfect situations that can be improved is so central in Sen's book that an apt title might be ‘The Idea of Injustice’, since the central theoretical and practical urgency that emerges through the entire work is intellectual engagement aimed at reducing manifest injustices.

Sen's theory of justice is primarily a critique of major theories of justice rather than a finished alternative theory (partly because Sen does not like this way of theorizing). Its central tenets are the following:

• • The central importance (along the lines of Mill) of discourse and public debate in all theories of justice, since democracy, in Bahegot's and Mill's words, as well as Sen's, is ‘government by discussion’. Sen's theory of human rights is also based primarily on public discourse (thus on civil society and its ‘voices’) and less on perfect obligations guaranteed by political and social institutions in view of those rights (Sen introduces economic and social rights into human rights, for which an obligation is never perfect).

• • The assertion that every serious theory of justice must be incomplete and open, as the incomplete spaces are filled in by public debate and the historical and cultural contexts that vary across people and cultures.

• • The necessity of an idea of impartiality in every theory of justice as equity (an idea of equity that Sen takes from Rawls and adapts as his own), an impartiality that Sen takes from Smith and his idea of an ‘impartial spectator’ – though Sen's ‘impartial observer’ is not Smith's, since Sen attributes a much broader role to the ideal observer than Smith's ‘man in the breast’. The ‘free’ use of Smith's impartiality can be considered a weak point of the book.

• • The necessity of a different idea of ‘neighbourliness’ that is impartial and not bound to one's ‘neighbour’; in this regard the pages devoted to a re-reading of the parable of the Good Samaritan are beautiful and evocative (about which we will have more to say).

• • The necessity of going beyond a contractualist approach based on mutual advantage in all choices of justice, since there are moral choices that go beyond personal advantage, in particular the choices that Sen calls ‘obligations of power’.

• • The development of freedom, in particular the freedom of agency, seen as the principal indicator of human development and quality of democracy.

• • Lastly – though we could continue with his theories of democracy, global justice, public goods and sustainable development – the necessity of surpassing the idea of happiness as the only thing worth valuing in life, since there are choices that have value even if they do not increase our happiness (understood as well-being).

In what follows, I will recount a few points that I consider central, not only to understand Sen's book, but also to grasp some of the core aspects of many current theoretical, cultural and civil debates.

A central example in the book features three children and a flute (pp. 12 ff.). The example is set forth by Sen to show how the ‘transcendental’ theories of justice do not work, because they focus only on one dimension of justice. Anne, Bob and Carla have but one flute, and a just criterion must be found to decide who should have the flute. Anne is the only one who knows how to play it; Bob is the poorest and it would be his only toy; Carla is richer than Bob, but she built the flute with her own hands. Sen points out that different theories of transcendental justice would lead to three different choices: a utilitarian hedonist theory would give it to Anne, economic egalitarians would give it to Bob, and a libertarian theory would give it to Carla. It is here that Sen declares the insufficiency of theories that use only one dimension to define just policies – including his own theory of capabilities formulated in the 1980s (by the way, in the book the shift from the first ‘transcendental’ Sen (Reference Sen1992) to the ‘new’ Sen is not emphasized).

Sen suggests that in these cases the ‘just’ criterion is that of public debate, which accepts the incompleteness of and conflicts among various criteria; despite these flaws, it arrives at a choice, perhaps using the aggregating techniques of social choice theory, on the basis of several assessments that cannot but be contingent; these assessments are subject to continuous review, critical scrutiny and even to being changed frequently over time and in different cultural and historical contexts.

As noted, what Sen proposes for each choice of justice is not a theory of justice per se, but criteria for choosing between situations that are more or less just, that Sen calls the realization-focused comparison approach (p. 7). To express this in standard economic parlance, for Sen ‘Pareto improvements’ are more important than ‘Pareto optima’ (although Sen would not like these terms), because the key issue of a theory of justice is changing situations from worse to better, even if both are non-optimal. To explain the difference between transcendental justice (based on the search for just or perfect institutions) and comparative justice (based on results and on lives that people actually lead), Sen adopts two words that used to indicate two different ideas of justice in ancient Indian cultures: Niti and Nyaya, which Sen relates to the Latin distinction between Justitia and Iustitium (pp. 20 ff). Niti is justice per se (Justitia), while Nyaya is justice practiced in a specific situation, thus never perfect and ever in contrast to another (Iustitium). For Sen, the official, dominant transcendental theory of justice is interested only in Niti, while his perspective (together with Smith, Bentham, Mill or Arrow) is more aligned with the Nyaya tradition.

Sen's references to Indian culture make this book absolutely original in the panorama of social theory; the examples are stirring, as they often take the form of epic and/or sacred episodes that are totally unknown to a Western reader. It should be noted in passing that one of the objectives of Sen's global research is to show that both of the concepts of rationality and democracy are present in the Indian world; they are not a legacy solely of the Greek–Roman–Jewish–Christian culture. Sen's discussion on impartiality, which for him is a prerequisite for any serious idea of justice, deserves additional comment. Sen refers to Smith because he sees Smith (along with Hume) as the only major ally for a non-transcendental and non-contractualist idea of justice, since Smith arrived at the idea of moral behaviour based primarily on impartiality rather than self-interest. For Sen (and not for Smith, we should note) impartiality should be open to people outside the immediate context of a given choice, an open and universal impartiality that gives each citizen of the world the right to invoke justice in situations in which he or she has no personal interest in intervening. Examples would be protesting for the survival of an animal species or for the rights of women in countries with a male-dominated culture. The theme of women is pervasive in this and in many others of Sen's publications (following the example set by J.S. Mill, one of of Sen's greatest sources of inspiration).

The idea of impartiality therefore needs a new idea of ‘neighbourliness’ that is unrelated to geographic proximity. It is in this context that Sen takes the parable of the Samaritan in the Gospel of Luke as a reference text to establish a ‘new’ idea of neighbourliness. Sen recalls that the Golden Rule (which is not just Christian) invites us to love not only God but our neighbours as well. The point however lies in the interpretation of the idea of ‘neighbour’, as neighbourliness is usually understood as geographical, ethnic, emotional or cultural nearness: we should love our neighbour, so we love those who are nearer – or more like – us rather than those that are more distant – or less like us. How should we then put the ethical imperative to love one's neighbour together with the necessity of impartiality and universality in a serious vision of justice, which must extend well beyond the geographic and ethnic confines of one's neighborhood? For Sen a deep fragility exists in thinking of people in terms of fixed communities formed by enclosing and exclusive boundaries. Sen finds a new idea of neighbourliness in the Samaritan in Luke's parable, which Jesus told in response to the question ‘who is my neighbour?’ In the end, the neighbour of the ‘man who fell among thieves’ was the Samaritan, and not the Levite or the priest, two people whose expressions of neighbourliness were bound to proximity, religion and ethnicity. For Jesus, the Samaritan became a neighbour because he took care of the man in need; the Samaritan who became a neighbour is a new sort of neighbour, because he went beyond all other conditions of neighbourliness. In fact, in order to grasp the rhetorical force of Jesus's argument, we have to remember that Samaritans were typically disliked or despised by Israelites: ‘Jesus does not, in this occasion, directly discuss the duty to help others – all others – in need, but rather raises a classificatory question regarding the definition of one's neighbor. He asks the lawyer with whom he is arguing: ‘Who was the wounded man's neighbor?’ The lawyer cannot avoid answering: ‘The man who helped him’. And that was, of course, Jesus's point exactly. The duty to neighbors is not confined only to those who live next door.’ (p. 172). And then Sen adds: ‘The Samaritan is linked to the wounded Israelite through the event itself: he found the stricken man so the need to help, provided that help and was now in a relationship with the injured person. It does not matter whether the Samaritan was moved by charity, or by a ‘sense of justice’, or by some deeper ‘sense of fairness in treating others as equals’. Once he finds himself, he is a new neighborhood. The neighborhood that is constructed by our relation with distant people is something that is pervasive relevance to the understanding of justice in general, particularly so in the contemporary world’ (p. 172).

One more remark on another central theme in Sen's work, the ‘obligations of power’ – linked to Sen's theory about happiness – that compel us to go beyond mutual advantage. The whole contractualist theory of justice, including Rawls's original view based on the ‘veil of ignorance’, is based on the idea of mutual advantage between subjects that engage in a social pact. For example, for Rawls the second ‘difference principle’ is basically a rule of individual rationality, since, once the veil of ignorance has fallen, I could be the worst-off in society, and thus it is rational for me to provide the maximum for the least well-off (measured in terms of primary goods). Sen does not deny that mutual advantage is a fundamental and foundational principle of all free societies, since an applied principle of reciprocity is at the base of every human society, more than altruism or selfishness, and particularly in a democracy comprised of free and equal people. However, he notes, although it is a necessary principle it is not sufficient for building a just society, because there are other ethical obligations that do not come from mutual advantage and contracts. In particular, referring to the Buddhist tradition, Sen identifies two instances in which there exists what he calls an ‘obligation of power’ (pp. 205 ff), which motivates people to act in such a way that one of the subjects – the more ‘powerful’ – derives no personal advantage; these are obligations to animals, and, indirectly, obligations of care of a mother for her child: ‘Buddha goes on to illustrate the point by an analogy with the responsibility of the mother towards her child, not because she has given birth to the child (that connection is not invoked in this particular argument – there is room for it elsewhere), but because she can do things to influence the child's life that the child itself cannot do. The mother's reason for helping the child, in this line of thinking, is not guided by the reward of cooperation, but precisely from her recognition that she can, asymmetrically, do things for the child that will make a huge difference to the child's life and which the child itself cannot do. The mother does not have to seek any mutual benefit – real or imagined – not seek any ‘as if’ contract to understand her obligation to the child’ (pp. 205–206). The conclusion of this discourse is particularly effective and evocative: ‘Mutual benefit, based on symmetry and reciprocity, is not the only foundation for thinking about reasonable behavior towards others. Having effective power and the obligations that can follow unidirectionally from it can also be an important basis for impartial reasoning, going well beyond the motivation for mutual benefits’ (p. 207).

To Sen, however, it is not sufficient to turn away from mutual advantage to establish an idea of justice as impartiality; rather, he wants to turn away from individual advantage, and even happiness, to which he dedicates the very important chapter 13 of the book. Sen is not the only contemporary thinker to criticize happiness as the basis of an ethical theory. Nozick, for example, was no less harshly critical in his last book The Examined Life, for reasons similar to those given by Sen. As might be expected from his general commitments regarding a theory of justice, Sen criticizes Bentham's utilitarian vision of happiness, that is the maximization of social and individual pleasure, as it is incompatible with his theory of capabilities (although, overall, he treats Bentham as an ally against transcendental institutionalism). In fact, much of Sen's early work was a critique of utilitarianism. Sen criticizes also – albeit not explicitly – the much broader idea of happiness that we find in Aristotle (eudaimonia) and in Mill. Sen tells us that even broadening the limits of the concept of happiness to include the idea of fulfilment or human flourishing, which for the Aristotelian tradition is the supreme and ultimate good (because everything is secondary with respect to happiness), happiness is not the only variable that counts in life, or the most important one. Why? The basic reason is simple. If, by reinterpreting Aristotle, happiness were to become all that has value in a person's life, then happiness would become a useless concept for social analysis, since knowing that all good and true things trace back to happiness leaves us without any criterion for specific choices, which frequently feature a conflict between several good situations. For Sen, then, one should keep happiness within the boundaries of common sense, including utilitarian happiness, considering it as but one dimension of life that in certain instances can give way to others. I personally have worked for several years around the idea of Aristotelian happiness (Bruni Reference Bruni2010), showing that many of the criticisms of the studies on happiness are actually applicable to Bentham's, rather than Aristotle's, view. I must acknowledge however that now I agree with Sen; while enriching and complicating the idea of happiness, it is more useful to limit happiness and compare it with other dimensions of the good life when analysing people's specific choices. As an example, consider a person (a woman, as Sen often does) who for religious or ideological reasons is convinced that being subject to her husband is part of her vocation, and she is content and happy in that state. Imagine that an impartial outsider to that situation freely suggests a path of awareness and freedom to her. If she begins a path of liberation and increases her capabilities, certainly in the early stages, and perhaps even in later stages, her happiness will decrease (if interpreted reductionistically), but her freedom will increase. Can we then say that the woman is happier, while possibly suffering spiritually and perhaps physically? I do not know; we can however say with certainty that she is more free, and thus prefer to trigger the process of her becoming aware, even though we know that there will be a cost in terms of happiness. We could certainly include freedom, in an Aristotelian sense, in the concept of happiness, but, again, in so doing the concept of happiness would become so broad as to be useless (somewhat as has happened with the concept of formal rationality, according to which all behaviour – even drinking petrol – can be described as rational).

The closing sentence of this chapter is also a good synthesis of the book: ‘one does not have to be a Gandhi (or a Martin Luther King, or a Nelson Mandela, or Aung San Suu Kyi) to understand that one's objectives and priorities could stretch well beyond the narrow limits of one's own personal well-being’ (p. 290).