by Amartya Sen
3
i n t r o d u c t i o n
need not await the determination of the relative priorities to be attached to these criteria. Arbitrary reduction of multiple and potentially conflicting principles to one solitary survivor, guillotining all the other evaluative criteria, is not, in fact, a prerequisite for getting useful and robust conclusions on what should be done. This applies as much to the theory of justice as it does to any other part of the discipline of practical reason.
r e a s o n i n g a n d j u s t i c e
The need for a theory of justice relates to the discipline of engagement in reasoning about a subject on which it is, as Burke noted, very difficult to speak. It is sometimes claimed that justice is not a matter of reasoning at all; it is one of being appropriately sensitive and having the right nose for injustice. It is easy to be tempted to think along these lines. When we find, for example, a raging famine, it seems natural to protest rather than reason elaborately about justice and injustice. And yet a calamity would be a case of injustice only if it could have been prevented, and particularly if those who could have undertaken preventive action had failed to try. Reasoning in some form cannot but be involved in moving from the observation of a tragedy to the diagnosis of injustice. Furthermore, cases of injustice may be much more complex and subtle than the assessment of an observable calamity. There could be different arguments suggesting disparate conclusions, and evaluations of justice may be anything but straightforward.
The avoidance of reasoned justification often comes not from indig-nant protesters but from placid guardians of order and justice. Reticence has appealed throughout history to those with a governing role, endowed with public authority, who are unsure of the grounds for action, or unwilling to scrutinize the basis of their policies. Lord Mansfield, the powerful English judge in the eighteenth century, famously advised a newly appointed colonial governor: ‘consider what you think justice requires and decide accordingly. But never give your reasons; for your judgement will probably be right, but your reasons will certainly be wrong.’2 This may well be a good advice for tactful 4
i n t r o d u c t i o n
governance, but it is surely no way of guaranteeing that the right things are done. Nor does it help to ensure that the people affected can see that justice is being done (which is, as will be discussed later, part of the discipline of making sustainable decisions regarding justice).
The requirements of a theory of justice include bringing reason into play in the diagnosis of justice and injustice. Over hundreds of years, writers on justice in different parts of the world have attempted to provide the intellectual basis for moving from a general sense of injustice to particular reasoned diagnoses of injustice, and from there to the analyses of ways of advancing justice. Traditions of reasoning about justice and injustice have long – and striking – histories across the world, from which illuminating suggestions on reasons of justice can be considered (as will be examined presently).
t h e e n l i g h t e n m e n t a n d
a
b a s i c
d i v e r g e n c e
Even though the subject of social justice has been discussed over the ages, the discipline received an especially strong boost during the European Enlightenment in the eighteenth and nineteenth centuries, encouraged by the political climate of change and also by the social and economic transformation taking place then in Europe and America. There are two basic, and divergent, lines of reasoning about justice among leading philosophers associated with the radical thought of that period. The distinction between the two approaches has received far less attention than, I believe, it richly deserves. I will begin with this dichotomy since that will help to locate the particular understanding of the theory of justice that I am trying to present in this work.
One approach, led by the work of Thomas Hobbes in the seventeenth century, and followed in different ways by such outstanding thinkers as Jean-Jacques Rousseau, concentrated on identifying just institutional arrangements for a society. This approach, which can be called ‘transcendental institutionalism’, has two distinct features.
First, it concentrates its attention on what it identifies as perfect justice, 5
i n t r o d u c t i o n
rather than on relative comparisons of justice and injustice. It tries only to identify social characteristics that cannot be transcended in terms of justice, and its focus is thus not on comparing feasible societies, all of which may fall short of the ideals of perfection. The inquiry is aimed at identifying the nature of ‘the just’, rather than finding some criteria for an alternative being ‘less unjust’ than another.
Second, in searching for perfection, transcendental institutionalism concentrates primarily on getting the institutions right, and it is not directly focused on the actual societies that would ultimately emerge.
The nature of the society that would result from any given set of institutions must, of course, depend also on non-institutional features, such as actual behaviours of people and their social interactions. In elaborating the likely consequences of the institutions, if and when a transcendental institutionalist theory goes into commenting on them, some specific behavioural assumptions are made that help the working of the chosen institutions.
Both these features relate to the ‘contractarian’ mode of thinking that Thomas Hobbes had initiated, and which was further pursued by John Locke, Jean-Jacques Rousseau and Immanuel Kant.3 A hypothetical ‘social contract’ that is assumed to be chosen is clearly concerned with an ideal alternative to the chaos that might otherwise characterize a society, and the contracts that were prominently discussed by the authors dealt primarily with the choice of institutions.
The overall result was to develop theories of justice that focused on transcendental identification of the ideal institutions.*
It is important, however, to note here that transcendental institutionalists in search of perfectly just institutions have sometimes also presented deeply illuminating analyses of moral or political imperatives regarding socially appropriate behaviour. This applies particu-
* Even though the social contract approach to justice initiated by Hobbes combines transcendentalism with institutionalism, it is worth noting that the two features need not necessarily be combined. We can, for example, have a transcendental theory that focuses on social realizations rather than on institutions (the search for the perfect utilitarian world with people blissfully happy would be a simple example of pursuing
‘realization-based transcendence’). Or we can focus on institutional assessments in comparative perspectives rather than undertaking a transcendental search for the perfect package of social institutions (preferring a greater – or indeed lesser – role for the free market would be an illustration of comparative institutionalism).
6
i n t r o d u c t i o n
larly to Immanuel Kant and John Rawls, both of whom have participated in transcendental institutional investigation, but have also provided far-reaching analyses of the requirements of behavioural norms. Even though they have focused on institutional choices, their analyses can be seen, more broadly, as ‘arrangement-focused’ approaches to justice, with the arrangements including right behaviour by all as well as right institutions.* There is, obviously, a radical contrast between an arrangement-focused conception of justice and a realization-focused understanding: the latter must, for example, concentrate on the actual behaviour of people, rather than presuming compliance by all with ideal behaviour.
In contrast with transcendental institutionalism, a number of other Enlightenment theorists took a variety of comparative approaches that were concerned with social realizations (resulting from actual institutions, actual behaviour and other influences). Different versions of such comparative thinking can be found, for example, in the works of Adam Smith, the Marquis de Condorcet, Jeremy Bentham, Mary Wollstonecraft, Karl Marx, John Stuart Mill, among a number of other leaders of innovative thought in the eighteenth and
nineteenth centuries. Even though these authors, with their very different ideas of the demands of justice, proposed quite distinct ways of making social comparisons, it can be said, at the risk of only a slight exaggeration, that they were all involved in comparisons of societies that already existed or could feasibly emerge, rather than confining their analyses to transcendental searches for a perfectly just society. Those focusing on realization-focused comparisons were often interested primarily in the removal of manifest injustice from the world that they saw.
The distance between the two approaches, transcendental institutionalism, on the one hand, and realization-focused comparison, on the other, is quite momentous. As it happens, it is the first tradition –
that of transcendental institutionalism – on which today’s mainstream political philosophy largely draws in its exploration of the theory of justice. The most powerful and momentous exposition of this
* As Rawls explains: ‘The other limitation on our discussion is that for the most part I examine the principles of justice that would regulate a well-ordered society. Everyone is presumed to act justly and to do his part in upholding just institutions.’ ( A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), pp. 7–8.) 7
i n t r o d u c t i o n
approach to justice can be found in the work of the leading political philosopher of our time, John Rawls (whose ideas and far-reaching contributions will be examined in Chapter 2 ‘Rawls and Beyond’).*
Indeed, Rawls’s ‘principles of justice’ in his A Theory of Justice are defined entirely in relation to perfectly just institutions, though he also investigates – very illuminatingly – the norms of right behaviour in political and moral contexts.†
Also a number of the other pre-eminent contemporary theorists of justice have, broadly speaking, taken the transcendental institutional route – I think here of Ronald Dworkin, David Gauthier, Robert Nozick, among others. Their theories, which have provided different, but respectively important, insights into the demands of a ‘just society’, share the common aim of identifying just rules and institutions, even though their identifications of these arrangements come in very different forms. The characterization of perfectly just institutions has become the central exercise in the modern theories of justice.
t h e p o i n t o f d e p a r t u r e
In contrast with most modern theories of justice, which concentrate on the ‘just society’, this book is an attempt to investigate realization-based comparisons that focus on the advancement or retreat of justice.
It is, in this respect, not in line with the strong and more philosophically celebrated tradition of transcendental institutionalism that emerged in the Enlightenment period (led by Hobbes and developed by Locke, Rousseau and Kant, among others), but more in the ‘other’
* He explained in A Theory of Justice (1971): ‘My aim is to present a conception of justice which generalizes and carries to a higher level of abstraction the familiar theory of the social contract as found, say, in Locke, Rousseau, and Kant’ (p. 10). See also his Political Liberalism (New York: Columbia University Press, 1993). The ‘contractarian’
routes of Rawls’s theory of justice had already been emphasized by him in his early –
pioneering – paper, ‘Justice as Fairness’, Philosophical Review, 67 (1958).
† In suggesting the need for what he calls a ‘reflective equilibrium’, Rawls builds into his social analysis the necessity to subject one’s values and priorities to critical scrutiny.
Also, as was briefly mentioned earlier, the ‘just institutions’ are identified in Rawlsian analysis with the assumption of compliance of actual conduct with the right behavioural rules.
8
i n t r o d u c t i o n
tradition that also took shape in about the same period or just after (pursued in various ways by Smith, Condorcet, Wollstonecraft, Bentham, Marx, Mill, among others). The fact that I share a point of departure with these diverse thinkers does not, of course, indicate that I agree with their substantive theories (that should be obvious enough, since they themselves differed so much from each other), and going beyond the shared point of departure, we have to look also at some points of eventual arrival.* The rest of the book will explore that journey.
Importance must be attached to the starting point, in particular the selection of some questions to be answered (for example, ‘how would justice be advanced?’), rather than others (for example, ‘what would be perfectly just institutions?’). This departure has the dual effect, first, of taking the comparative rather than the transcendental route, and second, of focusing on actual realizations in the societies involved, rather than only on institutions and rules. Given the present balance of emphases in contemporary political philosophy, this will require a radical change in the formulation of the theory of justice.
Why do we need such a dual departure? I begin with transcendentalism. I see two problems here. First, there may be no reasoned agreement at all, even under strict conditions of impartiality and open-minded scrutiny (for example, as identified by Rawls in his
‘original position’) on the nature of the ‘just society’: this is the issue of the feasibility of finding an agreed transcendental solution. Second, an exercise of practical reason that involves an actual choice demands a framework for comparison of justice for choosing among the feasible alternatives and not an identification of a possibly unavailable perfect situation that could not be transcended: this is the issue of the redundancy of the search for a transcendental solution. I shall presently discuss these problems with the transcendental focus (both feasibility and redundancy), but before that let me comment briefly on the institutional concentration involved in the approach of transcendental institutionalism.
* Also these authors use the word ‘justice’ in many different ways. As Adam Smith noted, the term ‘justice’ has ‘several different meanings’ ( The Theory of Moral Sentiments, 6th edn (London: T. Cadell, 1790), VII. ii. 1. 10 in the Clarendon Press edition (1976), p. 269). I shall examine Smith’s ideas on justice in the broadest sense.
9
i n t r o d u c t i o n
This second component of the departure concerns the need to focus on actual realizations and accomplishments, rather than only on the establishment of what are identified as the right institutions and rules.
The contrast here relates, as was mentioned earlier, to a general – and much broader – dichotomy between an arrangement-focused view of justice, and a realization-focused understanding of justice. The former line of thought proposes that justice should be conceptualized in terms of certain organizational arrangements – some institutions, some regulations, some behavioural rules – the active presence of which would indicate that justice is being done. The question to ask in this context is whether the analysis of justice must be so confined to getting the basic institutions and general rules right? Should we not also have to examine what emerges in the society, including the kind of lives that people can actually lead, given the institutions and rules, but also other influences, including actual behaviour, that would inescapably affect human lives?
I shall consider the arguments for the two respective departures in turn. I start with the problems of transcendental identification, beginning with the question of feasibility, and shall take up the issue of redundancy later.
f e a s i b i l i t y o f a u n i q u e
t r a n s c e n d e n t a l a g r e e m e n t There can be serious differences between competing principles of justice that survive critical scrutiny and can have claims to impartiality. This problem is serious enough, for example, for John Rawls’s assumption that there will be a unanimous choice of a unique set of
‘two principles of justice’ in a hypothetical situation of primordial equality (he calls it ‘the original position’), where people’s vested interests are not known to the people themselves. This presumes that there is basically only one kind of impartial argument, satisfying the demands of fairne
ss, shorn of vested interests. This, I would argue, may be a mistake.
There can be differences, for example, in the exact comparative weights to be given to distributional equality, on the one hand, and 10
i n t r o d u c t i o n
overall or aggregate enhancement, on the other. In his transcendental identification, John Rawls pinpoints one such formula (the lexicographic maximin rule, to be discussed in Chapter 2), among many that are available, without convincing arguments that would eliminate all other alternatives that might compete with Rawls’s very special formula for impartial attention.* There can be many other reasoned differences involving the particular formulae on which Rawls concentrates in his two principles of justice, without showing us why other alternatives would not continue to command attention in the impartial atmosphere of his original position.
If a diagnosis of perfectly just social arrangements is incurably problematic, then the entire strategy of transcendental institutionalism is deeply impaired, even if every conceivable alternative in the world were available. For example, the two principles of justice in John Rawls’s classic investigation of ‘justice as fairness’, which will be more fully discussed in Chapter 2, are precisely about perfectly just institutions in a world where all alternatives are available. However, what we do not know is whether the plurality of reasons for justice would allow one unique set of principles of justice to emerge in the original position. The elaborate exploration of Rawlsian social justice, which proceeds step by step from the identification and establishment of just institutions, would then get stuck at the very base.
In his later writings, Rawls makes some concessions to the recognition that ‘citizens will of course differ as to which conceptions of political justice they think most reasonable’. Indeed, he goes on to say in The Law of Peoples (1999):
The content of public reason is given by a family of political conceptions of justice, and not by a single one. There are many liberalisms and related views, and therefore many forms of public reason specified by a family of reasonable political conceptions. Of these, justice as fairness, whatever its merits, is but one.4