1
Egalitarianism, Free Will, and Ultimate Injustice Saul Smilansky Egalitarianism is a major contemporary position on issues of distributive justice and related public policy. Its two major strands are the Rawlsian, following John Rawls’ monumental A Theory of Justice (1971), and what can be called “choice-egalitarianism”.1 Both strands are deeply connected to the free will problem, in a way that is not sufficiently recognized. The free will problem can be broadly relevant to the egalitarian project on a number of levels:
a. Establishing the egalitarian premise for ethical discussion. b. Deciding upon the proper egalitarian practices of distributive justice. c. Evaluating the egalitarian arrangements. d. Attempting to convince people of egalitarianism and implementing the egalitarian conception of distributive justice.
Most of the philosophical work of both Rawls and the choice-egalitarians has been connected to the free will problem on level (b), and this will be the ground for our discussion. The present paper is concerned with (c). In a nutshell, I claim that both Rawls and the choice-egalitarians have failed to recognize a deep sense of injustice, which I call Ultimate Injustice. This form of injustice, which follows from the implications of the free will problem, appears despite the very different directions that were taken by Rawls and by the choice-egalitarians, both in general and with respect to the free will issue. For the sake of this discussion we shall assume a strong sense of egalitarian ethical presumption on (a), without explicating it or defending it. In other words, we shall assume that the general egalitarian project of trying to formulate an egalitarian conception of justice, a project shared by Rawls and the choice-egalitarians, is valid. One could begin with strong anti-egalitarian assumptions about justice that would preclude debate about
2
most of the issues of the present paper, but arguing on this basic matter (e.g., against Nozick) would take us beyond the scope of present discussion. Part I of this paper briefly presents the Rawlsian and choice-egalitarian positions, in a way that emphasizes their free will-related aspects. Part II makes a quick survey of the free will problem and of the major alternatives on it, emphasizing the relation to distributive justice. In the process I briefly outline my own way of understanding this problem, which I call Fundamental Dualism (see Smilansky 2000: Ch.6). Part III presents my arguments for the claim that both the Rawlsian and the choice-egalitarian social orders are deeply unjust in terms of Ultimate Injustice. Part IV briefly reflects upon some theoretical and practical implications that may follow if I am right.
I. How Rawls and the Choice-Egalitarians Have Dealt With the Free Will Problem
I shall assume here general familiarity with the Rawlsian project. Rawls’ position involves the free will issue in three ways. First, Rawls begins by assuming what amounts to a hard determinist position, a position that causes him to reject desert as a basis for an egalitarian conception of distributive justice. In light of free will considerations, neither one’s native endowments, nor the character that allows one to cultivate one’s abilities, are inherently deserved, he argues: ‘The notion of desert seems not to apply to such cases’ (Rawls 1971: 104). Secondly, and as a result, Rawls constructs distributive justice in a free will and desert bypassing way (we can call this “Rawls’ bypass strategy”). Because inequality of treatment cannot be justified by desert based upon free choice, he thinks that we need to turn elsewhere for justice. Justice, according to Rawls, is fairness, and fairness follows from a certain recognizably fair decision procedure, namely the ‘veil of ignorance’ (Rawls 1971: 136ff). The result of this thought experiment that interests us here is the ‘difference principle’. The economic and social station of those worse off in society is not justified in the light of their agency-based desert in itself but because there is no alternative social
3
order where the worse off are better off. The talented need to be tempted into working hard by incentives which amount to unequal benefits, and those at the bottom will gain from the inequality. Since the worse off must acknowledge this, they must acquiesce that the social order is fair and just. Once the proper social order is established, there seems to be no conceptual room for criticism. Thirdly, once Rawls sets up his order in a seemingly free will-free way (except for the merely negative hard determinist rejection of “pre-institutional desert”), his book is thoroughly infused with the assumption of free will and moral responsibility. People are expected to fulfill their obligations, to take responsibility for their actions, and indeed to try to form their expectations, in a way that makes sense only if non-hard determinist positions on the free will problem are assumed. Rawls makes the strategic decision to throw “metaphysics” out of the door, but then quietly lets it in through the window, where it rightly (and perhaps inevitably) becomes dominant.2 The choice-egalitarians go in a very different direction. Instead of attempting to bypass the free will problem by avoiding the traditional paradigmatic connection between free will, desert and justice, as Rawls does, they (at least implicitly) return to the paradigm. Given a baseline of equality, the inequality needs to be justified (on the way the notion of the baseline operates here, see Smilansky 1996a; Smilansky 1996b). Otherwise it is morally arbitrary. As Cohen writes ‘a large part of the fundamental egalitarian aim is to extinguish the influence of brute luck on distribution’ (1989: 931). Such justification can be found with people’s free choices: inequalities can be justified so that, for instance, those who decide to work harder can justly be compensated for their efforts and contributions. As Cohen says, ‘since effects of genuine choices contrast with brute luck, genuine choice excuses otherwise unacceptable inequalities’ (1989: 931). There is a question how we are to interpret the choice-egalitarian notion of ‘free choices’ or ‘genuine choices’. I think that it is necessary to understand choiceegalitarianism as compatibilist (the terminology will be clarified shortly). As I have argued before (Smilansky 1997: 154-5), only this interpretation can make sense of the confident
4
way in which Cohen talks about choice, namely, claiming that it is a matter of degree and that its existence is obvious. Moreover, the other alternative, metaphysical or libertarian free will, is highly contentious and, as many believe even incoherent (see bellow). To pin the hopes of egalitarianism on libertarian free will would be suicidal.
II. The Basics of the Free Will Problem
The free will problem in general, and a particular view of it in particular, lie at the heart of my analysis of justice, but this means that we shall have to go over the outline of the problem and of the pertinent view on it. Hence I ask the reader for his or her patience: all the free will-related work will emerge in the end as relevant for the discussion of egalitarianism and Ultimate Injustice. In compact form, the free will problem can be presented as the conjunction of two questions: a. The first question is whether there is libertarian free will, and it can be called the libertarian Coherence/Existence Question. Libertarians think that there is libertarian free will; everyone else disagrees. This question is metaphysical, or ontological, or possibly logical. b. The second question concerns the implications if there is no libertarian free will. It is traditionally called the Compatibility Question, namely, are moral responsibility and related notions such as desert and justice compatible with determinism (or with the absence of libertarian free will irrespective of determinism)? Compatibilism and hard determinism are the opponents on the Compatibility Question. This question, as we will understand it here, is ethical.
The traditional positions can be seen clearly from their answers to these questions: First, libertarian free will. We are all, more or less, familiar with the idea of libertarian free will. For our purposes it can be characterised roughly as the ability to control one's action and actually do otherwise in exactly the same situation, with internal
5
and external conditions held constant. People naturally assume that they have libertarian free will, and it has formed the basis of most of the ethical teaching of the Western religions and of major ethical systems such as Kant's. To help us intuitively hook on to it, it is the sort of freedom that determinism would preclude (although indeterminism would also not help and in fact the issue of determinism is not important). Libertarianism of course answers yes to the first question. Typically libertarians are incompatibilists, that is they think that if we did not have libertarian free will we would be in trouble, and there would not be, for example, moral responsibility. But luckily, we do have libertarian free will. In other words, the libertarian is demanding but optimistic. Compatibilism is, roughly, the position that the forms of free will most people clearly have to some degree, such as the ability to deliberate and do as they wish, suffice to meet the requirements of morality insofar as they are affected by the issue of free will. In particular, the compatibilist rejects the idea that some sort of ‘metaphysical’ or ‘libertarian’ notion of free will, such as would be negated by a completely deterministic ontology, is necessary in order to have moral responsibility. Hence, the term ‘compatibilism’: the compatibilist insists that free will, moral responsibility, and their concomitant notions are compatible with determinism (or with the absence of libertarian free will). For example, the compatibilist would claim that most people in the West choose a career with some measure of freedom, and are morally responsible for this choice, although it follows from their desires and beliefs. Lack of relevant freedom would result only from atypical causes eliminating or severely curtailing control (such as pathological compulsion or external coercion). It is important to stress that compatibilism is not utilitarian or consequentialist but maintains contact with the traditional paradigm requiring control for moral responsibility, and moral responsibility for blameworthiness and desert. Compatibilists maintain that the traditional paradigm can be sustained even in a deterministic world, and does not require libertarian free will. On the compatibilist level of deliberating, choosing, and acting, most people are basically free, such matters are within their control, and it is this that matters. The compatibilist is non-demanding and hence optimistic.
6
Hard determinism, despite its misleading name, is not only a position on determinism or on the existence of libertarian free will (although it is of course that as well). Hard determinism is the opponent of compatibilism, which is sometimes called “soft-determinism”, on the second question. In other words, hard determinism is a normative position according to which, given the absence of libertarian free will, moral responsibility and desert are impossible. Libertarian free will is required, but does not exist. The hard determinist is the pessimist in our cast of characters: she agrees with the libertarian that compatibilist free will is insufficient and that we require libertarian free will; hence, both are incompatibilists. But, like the compatibilist, the hard determinist believes that libertarian free will does not exist. Consider again the example of the person who chose his career freely according to the compatibilist: the hard determinist would claim that on the ultimate level the career-choice was not up to the person, who could not in the end form the sources of his motivation. These sources, the hard determinist will emphasise, are the basis for his “free” choice on the compatibilist level. In certain cases – such as when the man makes career choices which predictably lead to poverty – this absence of ultimate control is what matters, and eliminates moral responsibility. The hard determinist is demanding and pessimistic. We shall assume here that the answer to the first question is ‘no’, namely, that there is no libertarian free will.3 Hence we need to move on to the second, Compatibility Question, which asks about the implications of the absence of libertarian free will. In other words, we need to choose among the compatibilist and hard determinist interpretations of the ensuing situation. The free will problem, then, is about control: it issues from the core normative intuition that we must take human agency, control and its absence very seriously. This I call the Core Conception. It turns out, however, that the pertinent forms of control are fundamentally dualistic. On the one hand, we need to consider distinctions in local compatibilist control, if we are to respect persons. Questions about the existence of control, as well as about degrees of control, make sense and are morally and personally central. On
7
the compatibilist level we take the person as a “given”, and ask about his or her control in pedestrian ways: did he willingly do X? Was he coerced? Was he under some uncontrollable psychological compulsion? Most people most of the time do have compatibilist control over their actions, even if there is no libertarian free will (if, let us say, determinism applies to all human actions). The kleptomaniac or alcoholic are not in control of their pertinent actions in the way that, respectively, the common thief or occasional mild drinker are in control, irrespective of determinism. There are complex formulations of compatibilist control in the philosophical literature, and various borderline or problematic cases, but we need not enter into such matters here. On our level of discussion, which concerns the compatibilist perspective in itself, matters are sufficiently clear. But we can ask the question about control also on the ultimate level. Given that there is no libertarian free will, then asking about “ultimate control” lands us with the hard determinist conclusion, where ultimately there can be no control. Any person who we could agree was free on the compatibilist level (for example, one who could reflect on her options, decide to do what she wanted, and was not coerced) would be seen in a new light: under the ultimate perspective, the sources of her character and motivation would also be queried. And if we have no libertarian free will, then in the end we are just “given”, with our desires and beliefs, and any change in them is down to our earlier selves, which we ultimately cannot control. We are what we are, and from the ultimate perspective, with all our compatibilist choosing and doing, we operate as we were moulded. In my view, we have to take account of both valid perspectives on control, the compatibilist and the ultimate hard determinist, for each is part of the complex truth on the free will problem. Hence I propose a Fundamental Dualism encompassing both perspectives (Smilansky 2000: Ch.6). My claim is that the debate has been plagued by a too simplistic Assumption of Monism, according to which we must be either compatibilists or hard determinists. I argue that we need to take a dualistic approach concerning the implications of determinism: we have to be partly compatibilists and partly hard
8
determinists, and to try to integrate their insights while avoiding their inadequacies. We need to form a Community of Responsibility based upon distinctions made in terms of local compatibilist control: this is at once both morally imperative, and deeply unfair. Philosophically, respect for persons requires both the establishment and maintenance of a basically-compatibilist moral order, and the acknowledgement that this order is morally problematic, so that attempts must be made to mitigate its harshness. To be blind to any one of these two perspectives is to fail to understand the case on free will. For instance, a decision not to go to college because one does not like studying would at once be seen as free on the compatibilist level (one does what one wants), and as ultimately beyond one’s control (for in a deterministic world the sources of one’s motivation, which lie behind one’s compatibilistically-free decision, are not ultimately within one’s control). One must be allowed to live with the consequences of one’s decisions, when those decisions are free on the compatibilist level (i.e. when one’s decision is not coerced or grossly misinformed). But if those consequences are particularly grim, such as if they lead to poverty, the importance of ultimate hard determinist injustice would gain in importance. We need to try and limit injustice, but a great deal of Ultimate Injustice will follow from justified (compatibilist) social practice.
III. Rawls, the Choice-Egalitarians, and Ultimate Injustice
Both Rawls and the choice-egalitarians justify inequality, so that under their respective views a significant measure of inequality is not unjust. For Rawls, we recall, the justification bypasses the free will issue, and follows from the ‘difference principle’, while for the choice-egalitarians it is the very existence of free choice that (and only that) makes the inequality not unjust. However, both views fail to recognize Ultimate Injustice. Ultimate Injustice is the sort of injustice that, I claim, may follow when we do not take account of the absence of ultimate control. Given the egalitarian premise of the discussion, such injustice occurs whenever there is substantive inequality.
9
Inequality can be non-substantive: Alma chooses to work 8 hours a day and get full pay, Alex decides to work half-time for half pay, and both are equally well-off. In other words, substantively there is no inequality here. Substantive inequality exists whenever people are unequally well-off, some being less well-off than others (what it means to be well-off, and to be equally well-off, and how we determine how well-off people are, are of course difficult questions, but they are not our questions here). Both Rawlsian egalitarianism and the choice-egalitarians permit substantive inequality. For Rawls, one may have substantively less than others without injustice because this inequality is generated by the ‘difference principle’. For the choiceegalitarians the justification is, obviously, choice-based. It might seem as though there could not be substantive inequality for the latter, since why would anyone choose to be worse off? This, however, is a misunderstanding. Alma and Alex, we may assume, can choose at any time to be in each other’s situation: were Alma suddenly to prefer greater leisure, she could move to work 4 hours at half pay, and, similarly, if only Alex wanted he could double his working hours (and reduce his leisure) for Alma’s pay-bundle. Each simply gets the pay-plus-leisure mixture that he or she prefers. However, most choices permitted by the choice-egalitarians will not be similarly harmless. Choice-egalitarians permit one to decide not to acquire a profession, to spend one’s money on purchasing stocks, and, indeed, to risk it in gambling. Such practices inevitably generate substantive inequalities. The person who invested wisely will have more money than the person who invested foolishly, while the person who simply kept his money not invested will lie between those two. So, both egalitarian positions permit substantive inequalities. Why is this problematic? Rawls will be in the worst situation, in the light of the implications of the free will problem. Consider the people at the lower end of the socio-economic scale, the “worse off”. The economic and social station of those worse off in society is not justified for Rawls in the light of their agency-based desert in itself, we recall, but because there is no alternative social order where the worse off are better off. Since the worse off must
10
acknowledge this, they must acquiesce that the social order is fair and just. Even if we think that such an ideal Rawlsian order is indeed optimal in certain ways, and perhaps even justified overall, it seems to me that we can still say that there is something fundamentally unfair and unjust about such a situation. We have here a pragmatic compromise whose contours follow from the superior bargaining power of the talented, rather than a truly fair and just arrangement. If the talented do not deserve the fruits of their talents, then the fact that social contingencies (or their own greed!) make it preferable that they be motivated by the lure of unequal gains, does not make their superior social and economic position, however justified, fair or just (cf. Cohen 1995). The Rawlsian order has a number merits, but within the scope of our discussion it suffers from two grave faults:
a. Compatibilist Injustice: a Rawlsian order is often unjust in terms of compatibilist criteria for justice, namely, criteria that require justification in terms of every agent’s free choice, for the inequality. Rawls’ order is not based upon the Core Conception, but is generated in a different way. In it, the worse off will not be worse off because of their free choices; rather, we recall, they will be worse off because (for instance) they are less talented and social contingencies make it preferable for the worse off that the less talented like themselves– even though this situation is not related to choice and is morally arbitrary – will be less well off. Compatibilist Justice will not concern us here further.
b. Ultimate Injustice: a Rawlsian order is often also unjust in terms of the conclusions generated by the ultimate hard determinist perspective (which, we recall, is partly valid within the Fundamental Dualism, the correct attitude on the Compatibility Question). In other words, from the ultimate perspective, given initial egalitarian assumptions, any substantive inequality is unfair and unjust. The reason why it is unfair and unjust is that the only way in which, ultimately, inequality could be justified would be if it followed from libertarian free will.
11
However, libertarian free will is impossible. This means that any substantive inequality, whether justified for Rawlsian reasons or for compatibilist-choice reasons (as with the choice-egalitarians) is ultimately morally arbitrary.
The choice-egalitarians are here in a better state than Rawls. For, quite simply, their position does not suffer from Compatibilist Injustice. In their ideal moral order every substantively unequal situation among people follows from free choice, and, on the compatibilist level, free choice defeats moral arbitrariness. However, even choiceegalitarians are defeated by Ultimate Injustice: in their (compatibilistically-adequate) ideal social order a large measure of ultimate arbitrariness will prevail. People’s lives will follow in large measure from their compatibilistically free choices, but the content of these choices are not ultimately up to them. It might seem as though this ultimate sense is irrelevant: not only is it dubious (because after all we cannot have the libertarian-requiring sort of ultimate control), but why do we need it in the first place?4 In order to see the significance of Ultimate Injustice we need only turn to an example from criminal injustice. The choice-egalitarian who would deny the importance of Ultimate Injustice within distributive justice is like her parallel in criminal justice. True, it matters, and matters deeply, that the people in prison be there as a result of their free choices (i.e. justly on the compatibilist level), and not because they are, say, framed by the police for racist reasons. The latter case is (compatibilistically) unjust in a way in which the former is not. Nevertheless, this is not all we should want to say about criminal justice, once we become aware of the full import of the free will problem. The compatibilistically-guilty are in a deep sense also victims, whose punishment – even if it ought to occur – is deeply unjust. For, they choose as they are, and what they are is ultimately beyond their control. Hence if they chose a life of crime, then compatibilistically they ought to be punished, but this does not mean that all is well. They have not chosen to be such people who would (compatibilistically-freely) choose a life of crime. They are, in an important sense, victims of the constitutive forces that have made
12
them what they are. While people in general ought to be treated according to the way they behave, the sources of their motivation to behave as they behave are ultimately beyond their control, and we cannot remain blind to this. Similarly for distributive justice: in the choice-egalitarian’s just social order there will be large substantive inequalities that follow from people’s choices. We need to establish such a choice-tracking moral order. However, although those choices are morally significant on the compatibilist level this does not sum up all that is to be said on such cases. For, again, people choose as they are, and what they are is ultimately beyond their control. Put differently, my claim is that both Rawls and the choice-egalitarians are morally shallow: Rawls does not take proper account of either the compatibilist or the ultimate hard determinist implications of the absence of libertarian free will. The choice-egalitarians have only the second fault. On the other hand, since they explicitly see their view of distributive justice as based on free choice, rather than bypassing it, their limitation is all the more striking. Given that they recognize that control is crucial for justice, rather than bypassing the issue like Rawls did, one might expect greater concern with the ultimate absence of control. In any case, both positions do not adequately recognize the ultimate arbitrariness and its import for the issue of justice.5
IV. Some Theoretical and Practical Implications
We have concluded that both dominant strands of contemporary moderate egalitarianism, the Rawlsian and the choice-egalitarian, are gravely deficient in not recognizing Ultimate Injustice. What is the significance of this? First, we need to realize that even under ideal conditions, injustice is unavoidable. In the light of the prevalence of Ultimate Injustice, which follows from our best practices and with which we must live, I claim that we cannot deny that life is full of Unavoidable Injustice, which is just one of the ways in which the free will problem implies that life is
13
grossly unjust and inherently tragic. Quite simply, since we need to follow a large measure of Compatibilist Justice in social life, Ultimate Injustice will be unavoidable. This creates Unavoidable Injustice. It is not as though we can get out of this by opting for the ultimate side, abandoning any inequalities based upon compatibilistically-free choice, for then we shall fall into greater Compatibilist Injustice. As the choice-egalitarians have repeatedly argued, having other people pay for the costs of one’s free actions, in taking risks or acquiring expensive tastes, is also unfair, and avoiding to take seriously (compatibilist) choice will not lead to justice. Both the compatibilist and the ultimate hard determinist insights are crucial. Hence injustice is structural (see Smilansky 2000: sec. 11.1), part of the way things simply are. Both Rawls and the choice-egalitarians are shown to be optimistic and indeed complacent, thinking that if only we follow their different guidelines, we shall have distributive justice. This is not so: we shall have Ultimate Injustice, leading to structural Unavoidable Injustice. This of course does not mean that “anything goes”, that we should not try to do the best we can. Certainly matters can be better or worse even in terms of free will-related justice: suffice to think of the many countries where there is great inequality and distributive arrangements are based on corrupt political rule which has nothing to do with choice, effort or contribution. Compatibilist Justice matters greatly although we cannot avoid Ultimate Injustice, and so does the effort to mitigate the effects of Ultimate Injustice. Nevertheless it is important not to fall into the false optimism that both Rawls and the choice-egalitarians seem to share, whereby if only their egalitarian plans are pursued, distributive justice will prevail. Secondly, in a more constructive spirit, our conclusions imply that Ultimate Injustice needs to be recognized in our practical deliberations. Free will-related distributive justice is a balance between (compatibilist) choice-based just arrangements, namely a social order providing Compatibilist Justice, and the need to mitigate the results that will occur in terms of Ultimate Injustice. We need to establish and maintain a Community of Responsibility that follows compatibilist distinctions, but acknowledge that because of the ultimate perspective much injustice will result, and would need to be mitigated and
14
compensated for. A corresponding distributive order could operate within a market economy, but with particular sorts of emphasis within it and with correcting mechanisms added. A distributive order trying to accommodate the dualistic implications of the free will problem would hence allow a large measure of opportunity for people to become deserving according to their actions, permit them to benefit unequally as a result of their exercise of free choice, but also be sympathetic and provide a safety net if they failed. It would be broadly egalitarian in various ways: in emphasizing factors that are within people’s (compatibilist) control as the basis for justified inequality, in seeking to broaden access to opportunities for active agency and desert, and in the sense that however justified are the resulting free choice-based inequalities, they are also on one level unjust, so that people might merit compensation, because of the partial validity of the ultimate hard determinist perspective. Thirdly, our conclusions cast serious doubts about the possibility for the widespread acceptance of these positions. Neither Rawlsianism nor choice-egalitarianism are on the brink of taking over the minds and practices of Western democracies, but we can see how these positions can grow in attractiveness and come to be more influencial.6 However, once it is recognized that neither quite delivers on the promise of justice, and that both are indeed blind to Ultimate Injustice, such potential attractiveness may well decrease. This in turn leads to serious doubts about the idea of increasing knowledge about the free will problem. As we have seen, both egalitarian positions are based upon free will positions, even if those are not sufficiently or explicitly acknowledged. Rawls begins by espousing hard determinism, and choice-egalitarians are best understood as compatibilists. Rawls seems to think that he can bypass the free will problem and still get publicly recognized justice, and choice-egalitarians are free will optimists. However, once we see that both directions are not convincing, and that the picture is more complex and much darker, this casts a shadow on the advisability of public awareness of the implications of
15
the free will problem. I have developed this line of argument in detail elsewhere (Smilansky 2000: Part 2), and will not elaborate.
V. Conclusion
Justice theory, and in particular egalitarian justice theory, cannot be done without confronting the dreaded free will problem. The two dominating contemporary moderateegalitarian positions, those of Rawls and of the choice-egalitarians, are willy-nilly involved up to their necks in this issue. Rawls begins by assuming hard determinism, and then seeks to bypass the issue and allow inequality through the ‘difference principle’. The choiceegalitarians posit a free will-based view, albeit they have often been coy at admitting this. In any case, their position depends on an unfounded optimism about the simplicity of the issue and the robustness of their vital notion of free or genuine choice. Both Rawlsians and choice-egalitarians fail to note Ultimate Injustice, the injustice that follows from the best egalitarian practices because of the partly valid ultimate hard determinist interpretation. In one sense the Rawlsian suggestion does capture some of egalitarian justice, the sense in which the worse-off have no room for complaint, on the assumption that if the better off would not have more than the worse-off then the condition of the worse-off would become still worse. Hence, in certain constrained choosing conditions, the worse-off might choose this unequal social order. However, this hardly sums up people’s intuitions about justice and fairness, particularly those of egalitarians. For, we recall, a major intuitive fountain for egalitarianism was the thought that distinctions in what different people have must not be arbitrary. But in the ideal Rawlsian order many people will be worse off than others for what in a deep sense does seem morally arbitrary – namely, for factors (such as inborn talents) that are arbitrary for the egalitarian. It is perhaps a fact of life that if you do not pay the talented more than others, they will not be willing to, say, develop their talents and utilize them like others. The Rawlsian compromise addresses this
16
fact, but at the end of the day many people will still get much less than the talented – for reasons beyond their control. In one sense the choice-egalitarians do capture some of egalitarian justice, for if every inequality needs to be justified by free choice, then already much of the moral arbitrariness involved in people’s unequal states is prevented. Nevertheless, the sense of “free choice” the choice-egalitarians use is shallow. This shallow level is significant, it is part of the truth on free will, but it hardly sums up what egalitarians should want to say. Every free choice on the local compatibilist level that generates substantive inequality is also deeply unjust on the ultimate hard determinist level. Both of the dominant versions of moderate egalitarianism, then, permit a large measure of personal freedom and social inequality and seem to be able to live with a market economy, while capturing in different ways significant senses of egalitarian justice. However, they do not live up to the promise of delivering justice. Neither position takes proper account of the free will problem and of its implications; primarily, neither recognizes Ultimate Injustice. There is no libertarian free will, either because the world is deterministic, or because even indeterminism cannot help us, the very idea of a worthwhile notion of libertarian free will being incoherent. What does this mean for justice? The implications are importantly dualistic, hence I have called it Fundamental Dualism. The Fundamental Dualism means that we must take account of both compatibilist and of ultimate hard determinist perspectives. On the one hand, we need to create a Community of Responsibility that tracks compatibilist distinctions in terms of control and its absence: for instance, we can make sense of the distinction between the lazy and the hard working, and need to follow this distinction, distinguishing it from the further distinction between what is up to people in some sense (e.g. making an effort) and what is in any case beyond their control (e.g. their race). We need to emphasize, for instance, the distinction between getting less pay because one chooses to work fewer hours as compared to getting less pay because of factors beyond one’s control. The former may be acceptable, for working fewer
17
hours is up to one in various significant ways. If, by contrast, the factor causing the inequality is not within one’s control, this is pro tanto problematic. Hence even in compatibilist terms the latter sort of inequality is not fair, in the way that getting less pay if one decides to work less may be. Not to work with these distinctions would be not to respect persons. Here the choice-egalitarians have it right. On the other hand, given that the sort of transcendence that libertarian free will is supposed to give us cannot exist, everything that people do, including their compatibilistically-free actions, is ultimately an unfolding of the given. Hence, for all the moral importance of following Compatibilist Justice, we cannot avoid the ultimate perspective that means that doing so is, in a crucial sense, unjust. This is Ultimate Injustice. Rawlsian and choice-based Compatibilist Justice seem to capture all of justice, in the free will context, only when we remain on a shallow level, and do not pursue the concern for control up to the ultimate level, where we end up with constitutive arbitrariness. And arbitrariness, we recall, is just what Rawls and the choice-egalitarians are out to eradicate. People who are poorer justly, both according to Rawlsian and to choice-egalitarian compatibilist lights, are nevertheless in an important moral sense victims of arbitrary forces beyond their control, and their fate is in one deep sense an injustice. By and large we need to follow Compatibilist Justice, and mitigate it when we can. But in a world such as ours, i.e. one without libertarian free will, even the best social orders will be deeply and inevitably unjust. This has various complex implications, which we have briefly noted but that need to be further discussed. In any case, if we are to do justice to an egalitarian conception of distributive justice, we need to go beyond Rawls and the choiceegalitarians.
Department of Philosophy University of Haifa, Israel.
[email protected]
18
Notes 1. Many broadly choice-egalitarian positions have been proposed. I will focus upon G.A. Cohen’s formulation, which seems to me to be the clearest in capturing the pertinent choice-related egalitarian intuitions (Cohen 1989). Other detailed alternative or explicatory treatments that would seem to be similar for our purposes would be, e.g,. Arneson (1989); Rakowski (1991); Roemer (1993). Yet other discussions (e.g. Dworkin, Sen) are more distant, and yet are likely to be affected by our claims. It is possible to doubt that the choice-egalitarians (or Rawls) are indeed egalitarians, but I will not enter into the pertinent debates on this point.
2. It is not quite clear how matters are to be interpreted here. One possibility is that Rawls became morally overwhelmed by the partial validity of the hard determinist insights, and by the ultimate level arbitrariness, and thus went too far. Then, when the Rawlsian mechanism got going, the tacit compatibilism naturally crept back in. Another possible interpretation is that Rawls is not such a dogmatic hard determinist in the first place, but that the decision to avoid as far as possible any talk about desert based upon people’s free and responsible actions is motivated by broadly pragmatic concerns: about the possible public contentiousness of “the metaphysics”, or about the difficulties of calculating desert (inherently, or in a way that would win wide acceptance). I do not have the space to consider the discussions of these issues in the literature.
3. The reason why libertarian free will is impossible is, in a nutshell, that the conditions required by an ethically satisfying sense of libertarian free will, which would give us anything beyond sophisticated formulations of compatibilism, are self-contradictory, and hence cannot be met. This is irrespective of determinism or causality. Attributing moral worth to a person for her action requires that it follow from what she is, morally. The action cannot count morally if it is produced by a random occurrence. We might think that
19
two different acts can follow equally from a person, but which one does, say, a decision to steal or not to steal, again cannot be random but needs to follow from what she is, morally. But what a person is, morally, cannot be under her control. We might think that such control is possible if she creates herself, but then it is the early self that creates a later self, leading to vicious infinite regress. The libertarian project was worthwhile attempting: it was supposed to allow a deep moral connection between a given act and the person, and yet not fall into being merely an unfolding of the arbitrarily given, whether determined or random. But it is not possible to find any way in which this can be done (for a more thorough presentation of this argument see e.g. Strawson 2002; Smilansky 2000: Ch.4).
4. If, as I claim, libertarian free will is incoherent, how then can the absence of ultimate control be important? This argument has recently been made against my position concerning Ultimate Injustice by James Lenman (2002). It is a version of the familiar “How can the incoherent be worth wanting?” move, and attempts to rule out any “ultimate worries” because of the very incoherence of libertarian free will, an incoherence that is transferred to those ultimate-level concerns (Lenman 2002: 68ff). I cannot take up here this issue in detail (see Smilansky 2000: 48-50), but do not think that it is disturbing. If libertarian free will is incoherent, then indeed in one sense we do not have a positive model of what libertarian-based justice would look like. But since my worry is negative, i.e., it is concerned with injustice, all I need is to point out the limitations of the compatibilist view about free will-related justice. The human condition under the notion that libertarian free will could exist should be contrasted with the true picture, where libertarian free will is impossible. In terms of the social, ethical and personal implications there is a huge difference. For instance, given that we shall continue to put people in prisons, the absence of libertarian free will means that this practice will have much shallower grounding, hence be much more unjust, than it would have been if justification based on the “ultimately guilty self” were possible. The same is true for the distributive injustice that follows from inequality, given strong egalitarian assumptions. The severe shallowness of desert and
20
value, and the grave injustice, that exist in a world without libertarian free will, are ethically and existentially momentous. Lenman also argues more directly against the idea of Ultimate Injustice (that is used in the present paper in the context of distributive justice). But as I show (Smilansky 2003, forthcoming), his argument is dependent on only one, broadly contractual, sense of justice, and this sense is not the one that we care about when we care (as we should) about the free will problem.
5. A defence of Rawls against previous presentations of my claims is Sung-Hak Kang (2002, forthcoming). I reply in Smilansky (2002, forthcoming).
6. Choice-egalitarianism seems to be a strong improvement over Rawls insofar as it avoids the liberal-egalitarian neglect of responsibility, which, as Samuel Scheffler has argued, has been socially and politically harmful for liberal-egalitarianism (Scheffler 1992). My argument in this paper for the need to recognize Ultimate Injustice has been philosophical, and I have not inquired after the practical effects of the recognition of this view. I have general worries about the possible effect of a thorough public awareness of the truth on the free will problem (see Smilansky 2000: Part II).
21
Bibliography Arneson, Richard (1989). ‘Equality and Equality of Opportunity For Welfare’, Philosophical Studies 56: 77-93.
Cohen, G.A. (1989). ‘On the Currency of Egalitarian Justice’, Ethics 99: 906-44.
Cohen, G.A. (1995). ‘The Pareto Argument for Inequality’, Social Philosophy and Policy 12: 160-185.
Kang, Sung-Hak (2002). ‘Free Will and Distributive Justice: A Reply To Smilansky’, Philosophia, forthcoming.
Lenman, James (2002). ‘On the Alleged Shallowness of Compatibilism: A Critical Study of Saul Smilansky: Free Will and Illusion’, Iyyun 51: 63-79.
Rakowski, Eric (1991). Equal Justice. Oxford: Clarendon.
Rawls, John (1971). A Theory of Justice. Cambridge: Belknap Press of Harvard University Press.
Roemer, John E. (1993). ‘A Pragmatic Theory of Responsibility For the Egalitarian Planner’, Philosophy and Public Affairs 22: 146-66.
Scheffler, Samuel (1992). ‘Responsibility, Reactive Attitudes, and Liberalism in Philosophy and Politics’, Philosophy and Public Affairs 21: 299-323.
22
Smilansky, Saul (1996a). 'Responsibility and Desert: Defending the Connection', Mind 105: 157-63.
Smilansky, Saul (1996b).'The Connection Between Responsibility and Desert: The Crucial Distinction', Mind 105: 385-6.
Smilansky, Saul (1997). 'Egalitarian Justice and the Importance of the Free Will Problem', Philosophia 25: 153-61.
Smilansky, Saul (2000). Free Will and Illusion. Oxford: Oxford University Press.
Smilansky, Saul (2002). ‘Free Will, Egalitarianism, and Rawls’, Philosophia, forthcoming. [This is a reply to Kang]
Smilansky, Saul (2003). ‘On Free Will and Ultimate Injustice’, Iyyun, forthcoming. [This is a reply to Lenman]
Strawson, Galen (2002). ‘The Bounds of Freedom’, in The Oxford Handbook of Free Will. Robert Kane, ed. New York: Oxford University Press.
