King’s Research Portal DOI: 10.1177/1474885116664191 Document Version Peer reviewed version Link to publication record in King's Research Portal Citation for published version (APA): Tebble, A. J. (2020). On the circumstances of justice. EUROPEAN JOURNAL OF POLITICAL THEORY, 19(1), 3-25. https://doi.org/10.1177/1474885116664191 Citing this paper Please note that where the full-text provided on King's Research Portal is the Author Accepted Manuscript or Post-Print version this may differ from the final Published version. If citing, it is advised that you check and use the publisher's definitive version for pagination, volume/issue, and date of publication details. And where the final published version is provided on the Research Portal, if citing you are again advised to check the publisher's website for any subsequent corrections. General rights Copyright and moral rights for the publications made accessible in the Research Portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognize and abide by the legal requirements associated with these rights. •Users may download and print one copy of any publication from the Research Portal for the purpose of private study or research. •You may not further distribute the material or use it for any profit-making activity or commercial gain •You may freely distribute the URL identifying the publication in the Research Portal Take down policy If you believe that this document breaches copyright please contact librarypure@kcl.ac.uk providing details, and we will remove access to the work immediately and investigate your claim. Download date: 06. Apr. 2021 https://doi.org/10.1177/1474885116664191 https://kclpure.kcl.ac.uk/portal/en/publications/on-the-circumstances-of-justice(093564e6-7f4a-4c44-9508-1fa5fb03a368).html https://doi.org/10.1177/1474885116664191 1 On the Circumstances of Justice This is the final accepted version of the article published by SAGE Publishing in The European Journal of Political Theory on September 5, 2016 as doi:10.1177/1474885116664191 An epistemic account of the circumstances of justice allows one to make three important claims about the Humean and Rawlsian ‘standard account’ of those circumstances. First, and contrary to Hume, the possibility and necessity of justice are rooted not in limited beneficence or confined generosity, but in the epistemic insight that the knowledge relevant to deciding what to do with the fruits of social cooperation is for a variety of reasons uncentralisable. Second, and regardless of whether Rawlsian ethical disagreement is more persuasive as a circumstance of justice than Humean confined generosity, it does not explain the possibility and necessity of justice, for the uncentralisability of social knowledge would be decisive even under conditions of unanimity. Finally, the epistemic account not only shows what the circumstances of justice are but, contra Cohen’s critique of the standard account, also provides at least some guidance as to what justice itself may be. Adam James Tebble, Department of Political Economy, King’s College Londoni 1. Introduction The idea of the circumstances of justice, or those facts that explain both the possibility and necessity of justice, has a long pedigree in political philosophy and makes an appearance in many of the works familiar to the student of the discipline. Most notable here is an account of those circumstances often attributed to David Hume which, as Peter Vanderschraaf (2006:344, n.7) and others have commented, is almost unanimously accepted by contemporary political philosophers. To be sure, there is some debate as to whether Hume should be considered the undisputed father of the notion of the circumstances of justice. As Vanderschraaf (2006:344, n.7) has also noted, aspects of Hume’s account are foreshadowed in both Plato and Aristotle, as well as in St. Thomas Aquinas and Hobbes, 2 although unlike the Scot none of these thinkers made self-conscious reference to the idea of the circumstances of justice per se in their respective discussions. Interesting and insightful as these earlier contributions are, it will not be our objective to discuss them at length here. Rather, we will be concerned to question the persuasiveness of the modern, or standard, account of the circumstances of justice, understood as that attributed to Hume and subsequently developed by John Rawls.ii To this end I will defend a substantial revision of the standard account as part of a wider argument that questions whether Hume’s and Rawls’s respective characterisations of the circumstances of justice do the explanatory work expected of them. In the section that follows, and in acknowledging the significant differences with regard to theoretical aims and objectives that exist between them, I will set out both the account of the circumstances of justice attributed to Hume and Rawls’s development of that account, via his discussion of confined generosity and ethical disagreement as circumstances of justice in A Theory of Justice. Subsequent to this, I will lay the groundwork for my revision of the standard account by examining G. A. Cohen and Michael Sandel’s discussions of Rawls on ethical disagreement. In the case of Cohen, we will see that disagreement militates against the significance of a circumstance that both Hume and Rawls include in their accounts: moderate scarcity. Moreover, in the case of Sandel, and whilst failing to dislodge it as a circumstance of justice, we will see that there emerges a significant problem with Rawlsian disagreement that necessitates a major revision of the standard account. Building upon this, I will defend an alternative account of the circumstances of justice that emerges from a consideration of what authors working in the epistemic tradition of liberalism call society’s knowledge problem. Central here will be the claim that the standard account errs in positing either confined generosity or the propensity to disagree as explanations of the possibility and necessity of justice. Rather, it is the altogether different need to coordinate human knowledge with regard to the ends of life and the means 3 to pursue them that accounts for the possibility and necessity of justice. I will conclude my argument by considering how, if at all, our epistemic revision of the standard account may respond to some possible objections. Among these will be Cohen’s critique of the standard account as containing hidden ideological commitments. In responding to Cohen, I will claim that the circumstances of justice do not only offer guidance with respect to the conceptual question of the conditions that make justice possible and necessary, but also with respect to the question of what justice substantively requires. 2. The standard account Hume, moderate scarcity and confined generosity Traditionally understood, the circumstances of justice are those facts that explain both the possibility and necessity of justice. In the modern era what is called the standard account of those circumstances is associated with Hume and Rawls, although as we will presently see that their respective aims when discussing them are quite different It is commonly held that Hume (1751[1995]:188, 1740[1978]:486-9, 494-5) identifies two circumstances that explain the possibility, or origin, of rules of justice governing property and of (typically contractual) promise-keeping, and of how the necessity, or obligatory nature, of the artificial virtue of obeying those rules may be made intelligible to us: moderate scarcity of resources and our limited beneficence, or confined generosity.iii With respect to moderate scarcity, the possibility of rules of justice and the necessity of obeying them is established counterfactually. If we lived in a world of material abundance - let us call this world Cornucopia - there would be no need for such rules, or for a requirement for us to act justly even when our immediate interests suggest we should not, as each would have more than enough with which to make do (Hume, 1751[1995]:183-4). Hume makes this point about the relationship of moderate scarcity to justice still clearer by inviting us to assume the other extreme and imagine ourselves in a state of aggravated material scarcity. Here, and in contrast to it being unnecessary in a world of abundance, there would be no possibility of our acting justly for this virtue would be 4 ‘suspended’ and supplanted by the imperatives of ‘necessity and self-preservation’ (Hume, 1751[1995]:186). To be sure, and as both Jonathan Harrison and Cohen (2008:332, 334) point out, Hume may be understood to read the scope of justice somewhat narrowly in this respect, for his claim about abundance only shows that there would be no need for rules governing property, the keeping of promises and the virtue of our following them.iv These, however, do not exhaust all the rules that we would typically consider to be rules of justice. If, for example, under conditions of material abundance people ‘were sometimes malevolent’ society would still ‘need some rules of justice to protect each of its members from the others’, a point which we will see presently is amplified by the notion of Rawlsian disagreement (Harrison, 1981:264). Hume adopts a similar approach to establish the importance of our confined generosity, or limited beneficence as a second circumstance of justice. Were we wholly beneficent and miraculously unconfined in our generosity, he claims, we would have no need for rules of justice or for acting justly when personal interests and the limits of our affections suggest otherwise as we would always have the interests and needs of our fellows in mind. ‘[I]f everyone had the same affection and tender regard for everyone as for himself,’ Hume (1740[1978]:495) writes in the Treatise, ‘justice and injustice would be equally unknown among mankind’. Similarly, he (1740[1978]:494-5) claims that if we ‘[e]ncrease to a sufficient degree the benevolence of men ... you render justice useless, by supplying its place with much nobler virtues’. At the other extreme, and similarly to aggravated scarcity in the case of the first circumstance, were we wholly if not malevolently self-interested justice would be impossible, premised as it is upon virtuous behaviour when interests conflict, rather than upon a state of perpetual warfare like that to be found in a Hobbesian state of nature, or what Hume describes in the Enquiry (1751[1995]:187) as ‘the society of ruffians’. Rawls’s angels: from limited beneficence to ethical disagreement How, then, does Rawls fit into the standard account? As has already been noted, despite being associated with one another in the literature, there are significant differences between the approaches 5 of Hume and Rawls. First, the author of A Theory of Justice conceives of justice as something we may predicate of rules, institutions and schemes of social cooperation - or what he also calls ‘practices’(Rawls, 1999a:112. See also Rawls, 1999b:45-8; 1999c:190). Thus, in contrast to Hume, for whom we have seen justice is the artificial virtue of our following rules governing property and the keeping of promises, for Rawls justice is the virtue that makes these rules just. ‘Justice’, he (1999a:3) famously notes in A Theory of Justice, ‘is the first virtue of social institutions as truth is of systems of thought’. Moreover, and again in contrast to his predecessor, Rawls is not concerned with the circumstances of justice as either those conditions that explain the origin of rules of justice, or with how, despite its factitiousness, the virtue of following them may nonetheless be made intelligible to us. Rather he is concerned with the conditions which make it possible to characterise the rules that set the terms of social cooperation as just. We may say, then, that the Humean questions of the circumstances that explain the origin of rules of justice and which make the virtue of acting justly intelligible are different to Rawls’s question regarding the circumstances which make it possible and necessary to predicate justice of those rules. Rawls (1999a:109-10) also distinguishes himself from his Scottish predecessor by introducing a distinction between objective and subjective circumstances of justice, with the former being facts about the world and the latter facts about our ethical existence.v With respect to the first, objective, circumstance Rawls follows Hume in citing moderate material scarcity, although now as an explanation for why the terms of social cooperation can and need to be just, rather than of the origin of justice or of the virtue of our being just. ‘[A]lthough social cooperation is productive and mutually advantageous (one person’s or group’s gain need not be another’s loss)’, Rawls (1999e:256. See also Rawls, 1999d:234) notes, ‘natural resources and the state of technology are such that the fruits of joint efforts fall short of the claims that people make.’ Thus, just as in the bountiful conditions of Cornucopia there would be no need for rules of justice, so there would be no possibility of them under conditions of aggravated scarcity. ‘Natural and other resources’, he (1999a:110) writes, ‘are not so abundant that schemes of cooperation are superfluous,’ yet at the same time ‘nor are conditions so harsh that fruitful ventures must inevitably break down’.vi 6 In contrast to their agreement with regard to the first circumstance, in the case of the second, subjective, circumstance of justice, Rawls can be taken as disagreeing with Hume to the extent that it is supposed to show how it is possible and necessary to characterise the rules that set the terms of social cooperation as just. Our confined generosity, that is, does not explain the necessity of rules of justice. It is, rather, the notion of ethical disagreement that does this work. ‘[P]ersons and associations,’ Rawls claims in ‘Kantian Constructivism and Moral Theory’ (1999f:323. See also Rawls, 1999e:256, 1999d:234) ‘have contrary conceptions of the good as well as of how to realize them, and these differences set them at odds, and lead them to make conflicting claims on their institutions.’ Rawls (1999a:129) is happy, of course, to concede Hume’s point that ‘in an association of saints agreeing on a common ideal ... disputes about justice would not occur. Each would work selflessly for one end as determined by their common religion, and reference to this end ... would settle every question of right.’ The question, however, is whether such a community of ‘perfectly selfless angels’ has ever, or indeed, could ever exist, and matters are made hopelessly complicated in this regard when we pause to consider the situation of a modern pluralistic society (Vanderschraaf, 2006:3300. ‘[B]enevolence,’ Rawls (1999a:166) famously writes, ‘is at sea as long as its many loves are in opposition in the persons of its many objects.’ Similarly, he (1999a:112) comments insightfully that ‘[t]he spiritual ideals of saints and heroes can be as irreconcilably opposed as any other interests.’ Vanderschraaf (2006:330) makes this point lucidly by invoking, following C D Broad, the example of the survivors of a shipwreck. If all the survivors were such perfectly selfless angels, Vanderschraaf writes, ‘then each will be tempted to leave all of what little food they have for the others.’ Yet, doing this will result in all of them dying of hunger despite their express altruistic intentions. Because of this danger, the angelic survivors agree instead to divide the food so that each has something to eat. ‘What is necessary for justice,’ Vanderschraaf (2006:330) concludes, ‘is not moderate selfishness, but a moderate conflict of interests,’ regardless of whether the interests in question are ‘completely selfish, utterly selfless, or somewhere in between’.vii To be sure, it could be objected that, despite his own intentions, Rawls is talking past Hume and that the standard account is little more than a figment of the modern imagination. Indeed, given 7 that in seeking to explain the origin of rules of justice and the obligatory nature of just behaviour Hume is doing something altogether different to Rawls, what is called the standard account is more accurately described as Rawls’s development of his own reading of Hume, rather than of what Hume himself postulates.viii We will return to this matter in section 4, where we will see that Rawls’s are not the only grounds upon which we may reject the explanatory power of limited beneficence. For now, however, we may note on Rawls’s behalf that, were we concerned to set out the circumstances that show how the terms of social cooperation both can and need to be just, we ought not resort to Humean confined generosity to do so. Rather, the relevant subjective circumstance of justice is the fact of ethical disagreement about ends. Taken together, then, we see how the objective and subjective circumstances make justice possible and necessary insofar as it responds to the problem of the allocation of limited resources to people who, following Rawls’s development of Hume, disagree. In our imperfect world we require some way of adjudicating what should be done with the fruits of social cooperation and this is what justice achieves. 3. Ethical disagreement and the standard account Cohen, disagreement and moderate scarcity There are a diversity of responses to the standard account which, whilst certainly interesting in their own right, will not be engaged with in great detail here.ix Two responses that are relevant to our assessment of the explanatory adequacy of the standard account, however, are those of G. A. Cohen and Michael Sandel. As was suggested earlier, Cohen’s discussion of Rawlsian ethical disagreement is significant insofar as it may be read to prompt a further development of the standard account beyond that often considered to have been achieved by Rawls. Here, ethical disagreement is not just significant for the explanatory standing of Humean confined generosity, but also for the moderate material scarcity that both Hume and Rawls agree is indispensable to their respective accounts. Sandel’s discussion of Rawls’s view is also significant in this developmental respect insofar it enables us to clarify the conceptual terrain upon which an alternative epistemic account that calls into question the significance of either limited beneficence or ethical disagreement may be established. Finally, 8 Cohen’s discussion is significant in a second respect insofar as it highlights the issue of the relationship between the question of what the circumstances are in which justice is both possible and necessary and the normative question of our identifying what justice is. Leaving this discussion to the penultimate section of this paper, and discussion of Sandel’s contribution to the next section, let us commence with an analysis of Cohen’s discussion of the relationship between ethical disagreement and moderate resource scarcity. In addition to what Rawls has to say about its impact upon the standing of confined generosity, for Cohen ethical disagreement also shows why moderate resource scarcity cannot qualify as a circumstance of justice. Adopting a similar line of argument to Harrison in his discussion of Hume on moderate scarcity and malevolence, Cohen claims that Rawls’s shift from confined generosity to ethical disagreement with regard to the subjective circumstance of justice also reveals the difficulty presented by the objective circumstance of moderate scarcity to be moot. The problem here is that the example of material abundance, to which appeal is made counterfactually to establish moderate scarcity as a circumstance of justice, fails in this regard (Cohen, 2008:331, n. 88). Even under the bountiful conditions in Cornucopia, and contrary to the claim that it shows how justice would be unnecessary, there would still be conflicts of interest and hence a need for rules of justice, once it is recognised that, in addition to being capable of having what Cohen (2008:333) calls ‘spiteful motivations’, people also have different conceptions of the good that may require that we do different things with resources. This is not to say, of course, that Rawls either fails to see or ignores this point. ‘Although moderate scarcity may possibly be overcome or largely mitigated,’ he (Rawls, 1999f:326, 329) writes, ‘justice as fairness assumes that deep and pervasive differences of religious, philosophical, and ethical doctrine remain.’ The important point in this connection, however, is whether from the standpoint of a persuasive explanation of the possibility and necessity of justice, the overcoming of moderate scarcity matters. It appears that Rawls's point about mitigation is that it shows moderate scarcity to be merely a less important circumstance of justice than disagreement. But, we may concur with Cohen that it does more than this. The mitigation of 9 moderate scarcity having no impact upon ethical disagreement shows that moderate scarcity is of no explanatory import at all. Indeed, the problem of moderate scarcity’s explanatory irrelevance is deeper than Cohen recognizes. For not only may different conceptions of the good require that we do different things with resources. They may require that we refrain from doing things that other conceptions enjoin. It is plausible, for instance, that in Cornucopia one conception of the good - let us call it conception A - requires that we build place of worship x, and is indifferent to the construction of place of worship y that is required by conception B. It is also the case, however, that in Cornucopia conception B requires that only place of worship y be built, to the exclusion of x, and that another conception, conception C, is doctrinally committed to forbidding the construction of any places of worship at all. In such circumstances, it is evident that no amount of material abundance in Cornucopia would obviate the need for rules of justice to decide such conflicts peacefully. It follows, therefore, that if we would still need justice even under conditions of abundance it cannot be the case that the less demanding assumption of moderate material scarcity explains our need for it. Rawlsian disagreement, therefore, has an effect upon the standard account beyond that which both Rawls and Cohen expect. Not only is it decisive with respect to the explanatory power of confined generosity, the fact of disagreement about the ends of life is also decisive with respect to moderate material scarcity. To be sure moderate scarcity does make more vivid Rawls’s deeper, ethical explanation for the necessity of justice and of why it needs to assume its ‘peculiar’ and ‘special’ role as the first virtue of social institutions (Rawls, 1999d:234, 1999e:256). Making an explanation more vivid, however, is not the same thing as giving that explanation, and insofar as it incorporates moderate scarcity into its explanation of the possibility and necessity of justice Rawls’s development of the standard account is therefore to be rejected. Sandel, disagreement and the inverse proportionality thesis Having seen how the explanatory standing of both moderate scarcity and of confined generosity as circumstances of justice is undermined by ethical disagreement, it remains to consider whether this 10 Rawlsian reformulation of the second, subjective, circumstance of justice provides an explanation that is satisfactory. Doing so is doubly significant for we will presently see that it is out of this discussion that the need to articulate an alternative conception of the circumstances of justice becomes apparent. A useful starting point here is Sandel’s (1982:28-47, 168-72) critique of the underlying epistemic commitments of Rawls’s account and of how it undermines Rawls’s claim that justice is the first virtue of social institutions. As we have just seen, and as Sandel also tells us, it is the separateness of persons, where our diverse conceptions of the good and associated wants and desires are different that explains the possibility and necessity of justice. Yet, Sandel also wants to claim that it is not just the fact that our ends differ and that we consequently disagree either about whom we should act benevolently towards, or what doing so requires, that accounts for the need for justice. The consequences of Rawls ‘taking seriously the distinction between persons,’ Sandel (1982:171) claims, ‘are not directly moral but more decisively epistemological’ insofar as they highlight how our conceptions are largely inscrutable and opaque to others. ‘What the bounds between persons confine,’ he (1982:171) claims, ‘is less the reach of our sentiments ... than the reach of our understanding, of our cognitive access to others’. To be sure, in the case of more intimate relationships, such as those to be found within the family and amongst friends, one may have at least some knowledge of the conceptions, needs, wants and desires of others (although, as we will see presently, St. Thomas Aquinas gives us reason to be sceptical about even this much). This notwithstanding, it is what Sandel (1982:172) calls this ‘epistemic deficit’ between subjects that accounts in Rawls for the possibility, necessity and the ultimate priority or ‘pre-eminence’, of justice. ‘[W]here for Hume we need justice because we do not love each other well enough,’ he (1982:172) writes, ‘for Rawls we need justice because we cannot know each other well enough for even love to serve alone.’ Sandel, of course, is at pains to reject this view that he attributes to Rawls when claiming that considerations of fraternity and of community demonstrate that the circumstance of ethical disagreement, upon which the primacy of the virtue of justice depends, need not hold. Given, that is, a high level of fraternal feeling and benevolence of the type to be found in friendship, within the 11 family, or in community, there would be no conflict over moderately scarce resources. Moreover, ethical disagreement is only significant as a circumstance of justice if we additionally assume that we are largely inscrutable or opaque to one another. But why should we accept the Rawlsian picture of opacity, asks Sandel? A way of making Sandel’s claim clear here is to ask whether the society of angels that Rawls describes as one where there would be no need for justice could ever obtain. Clearly Rawls and other liberals think not. Yet, for Sandel, and even if the benevolence and familiarity that a society of angels could supply were not a realistic option, it may be supplied by the bonds of community. Indeed, when the constitutive bonds of community are taken into consideration, it is not at all clear that we are opaque to one another in the manner he claims Rawls believes gives justice its priority. ‘[I]n so far as our constitutive self-understandings comprehend a wider subject than the individual alone, whether a family or tribe or city or class or nation or people,’ writes Sandel (1982:172), to this extent they define a community in the constitutive sense. And what marks such a community is not merely a spirit of benevolence, or the prevalence of communitarian values, or even certain ‘shared final ends’ alone, but a common vocabulary of discourse and a background of implicit practices and understandings within which the opacity of the participants is reduced if never finally dissolved. Thus, for Sandel (1982:31) because ties of community overcome the epistemic deficit inherent in the notion of the separateness of persons, and upon which Rawls founds not just the possibility and necessity but the priority of justice, ‘justice is the first virtue of social institutions not absolutely, as truth is to theories, but only conditionally, as physical courage is to a war zone’. There is for Sandel, a relationship of ‘inverse proportionality’ (Badhwar, 1993:258) between relations of community and relations of justice, where the extent to which the former are extant is the extent to which the latter are surplus to requirements. We need not be detained by the question of whether Sandel’s reading of Rawls on the separateness of persons is a persuasive one. More important for the purposes of assessing Rawls’s 12 claim about disagreement is an assessment of whether Sandel’s claim about inverse proportionality is true. Answering this depends upon which of the two senses in which the inverse proportionality thesis may hold one considers. The ethical version of the inverse proportionality thesis holds that ‘[i]n so far as mutual benevolence and enlarged affections could be cultivated more widely, the need for ‘the cautious, jealous virtue of justice’ would diminish in proportion, and mankind would be better for it’ (Sandel, 1982:169). By contrast, the epistemic version of the inverse proportionality thesis, the one that denies what Sandel (1982:172) calls the ‘epistemic deficit’, holds that ‘[i]n so far as justice depends for its pre-eminence on the separateness or boundedness or persons in the cognitive sense, its priority would diminish as that opacity faded and this community deepened’. As we have already seen, one reason to doubt the truth of the ethical reading of the inverse proportionality thesis is that benevolent people may still disagree about intermediate ends even if they do agree about ultimate ends. Rawls, for example, claims in establishing the explanatory importance of disagreement, that even benevolently or altruistically-motivated individuals may disagree about what to do with resources, a point which Simon Caney (1991:512, n.8) notes is also to be found in Kant. Similarly, for Aquinas even in an association of saints there may still be disagreement about what action is required and in this connection we may recall Rawls’s rejection of confined generosity. Of course, for Sandel friendship or community in its highest form obviates the need for justice because it involves the pursuit of goals known and shared by all. Yet, for Aquinas even under such circumstances, and even if this were an appropriate way to understand either friendship or community, justice would nevertheless be required. As Schwartz Porzecanski (2004:39. See also Schwartz Porzecanski, (2007:124-7) highlights in his reading of Aquinas, ‘[s]ince two or more courses of action may legitimately be seen as instantiating the common goal, reference to the common goal alone is not sufficient in solving disputes.’ Aquinas’s point is echoed by Badhwar (1993:9). Even if we assume shared final ends they would not aid us in making decisions about the intermediate ends that are subordinate to them. Thus, writing with respect to the example of a family who agree on final ends, but who at the same time have to make decisions about how they will be pursued she (Badhwar, 1993:10) concludes that ‘[r]eference to a common good will not ... settle every question of 13 right’. Similarly, Caney (1991:512) contends that the claim about benevolence presupposes that ‘there will only be need for reference to a set of rights if people are selfish.’ Yet, citing the examples of a highly benevolent religious community where there is dispute over the position of women with respect to office-holding, and the problem of altruists’ dilemmas, even under conditions of benevolence, individuals still need rights because they may ‘still disagree about who should have what’ (Caney, 1991:512).x Thus, contrary to Sandel’s claim about community, even if there is a goal that everyone knows and shares, it does not follow that there is no need to accommodate conflicting interests and aims. There is a more profound reason for rejecting Sandel’s inverse proportionality thesis in its ethical guise and thus side with Rawls on disagreement. Even if Sandel were right that relationships of benevolence obviate the need to demand that our rights be respected or that justice be meted, it does not follow for at least two reasons that we should not have such rights, especially when the reasons why people act benevolently may be morally troubling. Citing the contrasting examples of an altruistic office worker who performs all the undesirable tasks in full knowledge that he has as equal a right to more pleasurable duties as his colleagues, and of the housewife who alone does all the chores but who never questions whether this is fair or just, ‘[u]nanimity and benevolence,’ Caney (1991:512) writes, ‘are often generated in non-autonomous ways’. Susan Moller Okin makes a similar point in critiquing Sandel’s Humean claim that because love can do the work of justice within domestic settings, the family is not marked by the circumstances of justice. Citing the practice of couverture, she shows that even if relationships of love provide guidance for resolving many if not most conflicts of interest within such settings, it does not follow that they ought to resolve them all and that there is therefore no need for relationships of justice in them. ‘[E]ven if wives never had occasion to ask for their just share of the family property, due to the generosity and spontaneous affection of their husbands,’ she (1989:30-1) writes, ‘we would be unable to assess the families in which they lived from a moral point of view unless we knew whether, if they did ask for it, they would be considered entitled to it.’ 14 For Caney, such examples are instructive insofar as they make clear what Rawls has to say about disagreement and the priority of justice, and why Sandel’s discussion of community ties and the inverse proportionality thesis fails to grasp it. Rawls’s claim about the priority of justice does not commit one to the view that it must be invoked on all occasions, but rather that it needs to be available when we do need to call upon it. Even if relationships among families, friends and communities are usually benevolent, that is, it does not follow that justice is unnecessary. Thus, for Caney Sandel confuses the issue of whether we may on occasion waive our rights with that of whether we require them. Individuals, that is, may waive their rights when benevolence reigns, but that is not the same as their not requiring them, or that their background institutions need not be just. Personal relations need not always be marked by justice - indeed, it would be odd and oppressively dull if they were - but our institutions, such as a regime of rights and liberties upon which we may make claims upon one another about the fruits of social cooperation, including in usually benevolent intimate personal settings, do need to be marked by them. 4. The circumstances of justice: an epistemic view From the terms of social cooperation to the terms of social coordination Given our agreement with Rawls on the implications of ethical disagreement for confined generosity, and with Cohen on its implications for moderate scarcity, and granted that the ethical variant of Sandel’s inverse proportionality thesis fails to dislodge it, it looks as if ethical disagreement, and with it at least part of the standard account, remains unvanquished. Having vindicated Rawls’s position with regard to the ethical inverse proportionality thesis, therefore, we turn now to the question of the persuasiveness of Sandel’s argument in its epistemic guise. Doing so is important not just because it enables us to assess Sandel’s critique of Rawls on disagreement further but because, crucially for present purposes, it brings to light a problem that leads us to ultimately part company with Rawls and build an alternative, epistemic, conception of the circumstances of justice. 15 We may assess the epistemic variant of the inverse proportionality thesis in two distinct ways, the first of which is to assess the degree to which it holds true in the case of close, or what we may call face-to-face, social ties. One reason to reject Sandel’s argument as grounds for his claim that close social ties may do the work that Rawls claims justice ought is because even in the case of relationships within families and between friends the problem of interpersonal inscrutability or opaqueness, and of the epistemic deficit that is attendant to it, still remains. As was hinted at in our discussion of the ethical variant of the inverse proportionality thesis, Aquinas claims that justice is still needed between friends, for even friends do not have direct knowledge of one another’s motives or desires. ‘There is no unmediated access to the heart, not only because what occurs in the heart is hidden by the body - the imperfect vehicle of expression of the heart - but also because of the fact that “the will shuts up its secrets” [voluntas claudens sua secreta]’ (Schwartz Porzecanski, 2007:126). To assume that friendship is solvent of the epistemic opacity upon which the need for justice is founded would be to assume too much. As Aquinas (2014:I.qu.57a.4c) puts it in the Summa, ‘all that is in the will, and all things that depend only on the will, are known by God alone’. Moreover, and to the extent that mutual inscrutability may not be overcome by close social ties, Sandel is wrong to extend his argument to the case of inscrutability within communities where these too are solely marked by race-to-face relations. If we assume, reasonably, that a face-to-face community is no more and no less than a complex of face-to-face relationships, then Aquinas’s problem would also hold. It is at this juncture where we not only encounter a second set of reasons to reject Sandel’s epistemic variant of the inverse proportionality thesis, his critique of disagreement and with it the priority of justice. Most significantly, in assessing Sandel’s epistemic argument we may also uncover a different but more telling set of reasons to reject the significance of both Humean confined generosity and Rawlsian disagreement themselves as circumstances of justice. First, even if we had reason to doubt Aquinas’s view of the moral epistemology of relationships in face-to-face communities and agree with Sandel that there is no epistemic deficit in them, Sandel is nevertheless wrong to claim that the epistemic deficit does not exist in all forms of community and that justice is therefore ill-suited as a model for social cooperation. That is, even if we were to part with Aquinas 16 and accept Sandel’s claim that there is no epistemic deficit in face-to-face communal settings, he is wrong in the case of modern complex societies. Caney’s (1991:512) objection that ‘[m]odern societies are not analogous to such close-knit groups and thus the circumstances of justice do obtain’ becomes salient here, for it is not only the basis upon which he makes a claim about Sandel’s political naïveté; that it is unrealistic of Sandel to think that societies are not marked by conflict. Caney (1991:512) also claims that the examples of benevolent, conflict-free relationships Sandel cites are importantly limited with respect to modern societies insofar they are conspicuous by their absence relative to the sum total of all relationships to be found within them. To be sure, Caney’s point here is a different one to Aquinas’s with respect to the inscrutability or opaqueness of the soul. The problem is not that the innermost thoughts and motivations of even those closest to us are unknowable or knowable only to God, although the personal disappointments and betrayals that punctuate our lives may attest to this being the case. Rather, it is the inscrutability that accompanies social distancing in complex, or what Hume (1740[1978]:487) calls ‘large’ societies, that is decisive in this respect. Sandel’s friendship/family model is for Carey unpersuasive and cannot serve as an alternative form of social bonding and mutual knowing that exposes the contingency rather than the priority of justice not just because it naïvely assumes away the problem of conflict. It fails to fulfil this role because it is blind to the fact that, in the overwhelming majority of circumstances, we do not and cannot know either what our fellows want, need or require, the relative intensity with which they do so, nor which bundles of resources and production processes are most appropriate for meeting them. Given that the intimate face-to-face nature of the relationships to be found among friends, within the family and within community cannot therefore serve as a model of relations in large or complex societies, it would be ‘implausible,’ and not just naïve Caney (1991:512) suggests, ‘to think that political society could be conceived along these lines’. The persuasiveness of Caney’s response to Sandel notwithstanding, defenders of ethical disagreement should not take too much comfort from his argument. The point he makes about the epistemology of social scale is not only noteworthy because of what it rejects in Sandel’s discussion. It gains an additional significance in drawing our attention to an altogether different problem, 17 overlooked in the epistemic musings of Sandel, that undermines the explanatory power of ethical disagreement as a circumstance of justice and with it the persuasiveness of the standard account as developed by Rawls. In contrast to the standard account, what we may call the epistemic approach to the circumstances of justice that Caney’s observation anticipates is not concerned with the implications our ethical lives may have for how the subjective circumstance should be understood. It is concerned, rather, with the implications of the subjectivity of human knowledge and of social complexity. A useful starting point to build upon Caney’s observation and construct an alternative account of the circumstances of justice is the thought of one of the best known defenders of the epistemic variety of liberalism, Friedrich Hayek. Of course, despite being familiar with the work of both Hume and Rawls, Hayek never wrote about the circumstances of justice. Indeed, whilst certainly familiar with Hume’s discussion, he never used the term in his written work (Hayek, 1976a: 112–113, 1967b: 350). Nevertheless, Hayek’s approach gains entry into our discussion via his core insight that the knowledge of the circumstances relevant to the direction of cooperative efforts, including the division of its fruits - that is, the knowledge of the diversity of human needs and wants, their relative value and the most appropriate means of satisfying them - is never given in its entirety to any single individual. The first reason Hayek gives for this is because it is knowledge that is subjectively held by different people acting under different circumstances. We can imagine a state of affairs, for instance where Paul believes, as he has always done, that desired good x is best made with resource p, whilst his friend Peter has recently discovered that resource q is far better for this purpose. In contrast to both of them, however, Peter’s neighbours Penny and Patricia believe that resources shouldn’t be used for making something as devoid of value as x, but should be employed instead to satisfy the far more pressing need for good y. Precisely, then, because deciding what is of value and of how what is of value should be produced is dependent upon the beliefs of different individuals, for Hayek the question of how cooperative efforts are to be directed cannot be assumed to be resolvable upon the basis of objectively-given knowledge. 18 Hayek’s claim about the subjectivity of knowledge gains its full significance, however, when we recall the distinction between face-to-face and complex social relations from our discussion of Sandel. Under face-to-face conditions the subjectivity of knowledge presents no real problem for successful social cooperation, for each either already knows, or can without great cost come to know, the needs and wants of all the rest, how best to satisfy them and of how to act accordingly (Tebble, 2016: 28–29). For epistemic liberals, however, matters are quite different when we participate in cooperative schemes that extend beyond the ken of face-to-face social formations. In such complex societies the knowledge relevant to successful social cooperation is not only the subjectively held knowledge of different individuals, but the subjectively held knowledge of different individuals who, in the overwhelming majority of cases, are not in direct contact with one another, with the consequence that it is never given in the way that would make social cooperation as relatively costless as it is in face-to-face settings. To be sure, there is a sense in which this knowledge can be said to be given, as it is after all contained within society when viewed as a single collectivity. The important point, however, is that its dispersal among what we may call mutually ignorant individuals means that each has access to a different part of the sum, and therefore does not know what all the rest know, even though this would be relevant to ensuring that cooperative efforts are directed successfully (Tebble, 2016: 30–32). It is not only spatial distancing that accounts for the cognitive separateness and epistemic deficit prevalent in modern societies or, what theorists working in the epistemic tradition of liberalism call society’s knowledge problem. The question is not just that of who is in possession of the knowledge relevant to successful social cooperation, or of the distance between them, but also of the kind of knowledge this knowledge is that is decisive. Thus, related to its dispersal, and again showing why it defies centralization, are the facts that as subjective knowledge this knowledge is often circumstantial and fleeting. Unlike our scientific knowledge, for example, this knowledge is knowledge of specific circumstances that may only be of temporary duration, but which are nonetheless relevant to our decisions. Fourth, and perhaps most significantly for Hayek (1948a, 1948b, the subjectively-held knowledge of mutually-ignorant individuals is also often tacit in nature, 19 embodied in different traditions and practices whose action-guiding significance to the decisions of others is only ever communicated when acted upon by those who possess it. That is, it is knowledge that its possessors do not realize that they have, let alone that somebody else may know they have and could therefore make use of independently of them. Finally, recent work in the epistemic liberal tradition has shown how the knowledge problem is also significant because the circumstances to which agents must respond if their plans are to be successful are a continually changing result of their own past decisions and those of their fellows. In this connection we may note the contribution of complex adaptive systems theory and conceive the economy of a modern society as a complex adaptive system (Tebble, 2016:31-33). Of particular significance in this respect are two features that complex adaptive systems manifest. First, and as the above suggests, in contrast to complex physical systems that also display hallmark features of complexity such as self-organisation or, following Hayek, spontaneous order, complex adaptive systems are affected by the intentional agency of the elements that feature in them.xi Moreover, and related to this, complex adaptive systems are characterised by the phenomenon called parallel processing where individual agents respond to their circumstances simultaneously, although unbeknownst to one another. The result of these behaviours is that the overall economic context to which agents respond and with regard to which they must make decisions, in turn becomes the future context which they have helped to shape precisely as a result of their prior adaptive responses. In the complex economic adaptive order of a large-scale society, therefore, and regardless of whether one’s primary interests are one’s own or those of others, the circumstances that each must take into account in order to satisfy them are in a constant state of effervescent change brought about by the mutually affecting simultaneous adaptive responses of each to the activities of all the rest. It is not the case, therefore, that the knowledge problem is a one-off coordination problem that, at some point and with the requisite will and effort, may be solved. Rather, precisely because mutually ignorant agents simultaneously impact upon one another’s circumstances, and with this the chances of their respective plans being successful, the knowledge problem and associated need for mutual coordination are not 20 only ever-present with respect to a continually changing environment, but indeed intensify as the cooperative space expands to include more and more mutually impacting agents. The knowledge problem as a circumstance of justice These insights amount to an epistemic liberal variant of Sandel’s notion of the cognitive separateness of persons. Agents engaged in social cooperation are opaque to one another not only for the reasons that Sandel suggests are implicit in Rawls’s account, but because of the insuperable spatio-temporal distancing that exists between them. Thus, in addition to the inscrutability of the human soul, cognitive separateness and its attendant epistemic deficit arise from social complexity and it is in this sense that we are physically, and not only spiritually or ethically, distant from one another. Moreover, and again in contrast to the concerns of Sandel, it is the epistemic deficit manifested in society’s knowledge problem that assumes central importance with regard to the circumstances of justice. Similarly to the counterfactual reasoning that we have seen Hume and Rawls engage in with respect to the establishment of moderate scarcity, confined generosity and ethical disagreement, we may adopt a similar approach to establish the significance of the knowledge problem. In circumstances of what we may call aggravated solipsism - where each only ever has knowledge of their own needs and where knowledge of the preferences of others and the means to satisfy them is permanently absent - knowledge coordinating rules of justice would be impossible because coordination itself would be impossible. Similarly, were we to find ourselves bestowed with the power of omniscience, the rules of justice that enable the selfish, the moderately selfish and the wholly beneficent to coordinate their activities would, like coordination itself, be unnecessary. ‘[I]n a society of omniscient persons,’ Hayek (1982:II.39) claims, ‘there would be no room for a conception of justice: every action would have to be judged as a means to bringing about known effects, and omniscience would presumably include knowledge of the relative importance of the different effects.’ Most significantly, it is in this connection that we may claim that the knowledge problem deals a fatal blow to the explanatory potency of the two principal contributions to the standard account. Of course, in making such a claim it is important not to misrepresent Hume, for we have 21 seen that it is open to question whether he conceives of the purpose of his contribution in the way that Rawls and subsequent writers on the standard account think. It may well be the case, therefore, that we ought not resort to the notion of confined generosity to explain how the terms of social cooperation both can and need to be just, but this is not something that Hume is concerned to do in the first place. Yet, even if we accept that his purposes are quite different to Rawls’s, and in so doing go so far as to eschew all notion of there being a unified standard account, there is still an important epistemic case to be made against Hume’s view on its own terms. The origin of rules of justice requiring respect for property rights and the keeping of promises, that is, is not explained by the steady growth of small-scale societies gradually undermining the natural bonds of sympathy that would otherwise obviate the need for them, as Hume seeks to show in the Treatise (Clayton Hubin, 1979:4-5). Rather, our revision of the circumstances of justice explains the origin of such rules on the epistemic grounds that, without them, the gradual expansion of the terms of social cooperation would render the communication to each of the consequences of the actions of all the rest - so that evermore disparate plans may be realised in conformity with otherwise unknowable underlying conditions - increasingly fraught. Similarly, with respect to the virtue of our obeying rules of justice, even if there were complete altruism rather than confined generosity the knowledge problem would nevertheless necessitate just behaviour so that uncentralisable but nonetheless relevant knowledge of ever- changing resource needs and availabilities could be coordinated under complex conditions. It is not, therefore, as Hume seeks to show in the Enquiry, that without an account of moderate scarcity and confined generosity it would not be possible to explain why, despite being artificial, the virtue justice is nonetheless obligatory (Clayton Hubin, 1979:5). Rather, and regardless of whether we imagine ourselves to be wholly or partially beneficent, in Humean large societies that extend beyond the ken of the face-to-face group - where each knows of the needs of all the rest and of the appropriate means to satisfy them - we are obliged to act justly because we are never in a position to know all that we would need to know in order to attend adequately not only to our own needs, but to those of our overwhelmingly unseen fellows. Thus, as Harrison (1981:266) makes clear, even in a society where 22 everybody had ‘tenderness for every man’ there would still be a need for rules of justice (including rules governing exchange) and the need to clearly delineate proprietary boundaries to prevent ‘disorganisation’. ‘It follows,’ he (1981:266) concludes, ‘that Hume is mistaken when he remarks that if benevolence were strong enough to be a motive to obey rules of justice, such rules would be unnecessary.’ Indeed, as Hayek (1982:II.83) claims in his discussion of justice, under complex circumstances there is still ‘no possibility for the individual to know what he would have to do to secure a just remuneration of his fellows’, even if he were morally motivated to do so. In addition to the doubt cast upon Hume’s explanation, our epistemic revision of the standard account shows the weaknesses of Rawls’s account with regard to the justice of schemes of social cooperation. Even if there were Rawlsian unanimity about ends, the knowledge problem would nevertheless reveal the necessity of just schemes of social cooperation. ‘All the possible differences in men's moral attitudes amount to little,’ Hayek (1948c:14) writes, so far as their significance for social organization is concerned, compared with the fact that all man's mind can effectively comprehend are the facts of the narrow circle of which he is the center; that, whether he is completely selfish or the most perfect altruist, the human needs for which he can effectively care are an almost negligible fraction of the needs of all members of society. Even if we were to agree upon ends, therefore, we would still be faced with the problem of not knowing all that we would need to know in order to ensure that the decisions of each dovetailed with those of all the rest so as to make their realisation likely. Even if it obviates the need to include moderate scarcity and confined generosity in the standard account, then, ethical disagreement is not as decisive a circumstance of justice as the knowledge problem. It is not only, therefore, that were everybody to have ‘tenderness for every man’ there would still be a need for rules of justice so as to prevent disorganisation (Harrison, 1981:266). ‘[T]he possibility of justice,’ Hayek (1982:I.13) concludes, ‘rests on [the] necessary limitation of our factual knowledge, and that insight into the 23 nature of justice is therefore denied to all those constructivists who habitually argue on the assumption of omniscience.’ It is, then, the objective fact of our permanently limited knowledge - whereby each of us is only ever in possession of but a tiny fraction of all the knowledge necessary for rational economic decisions to be made - rather than the (moderate) scarcity of resources, our reluctance to agree, or the truncated nature of our affections that explains our need for justice. To be sure, the point at issue in this regard is not to be found in any divergence with either Hume or Rawls’s respective characterisations of the hearts of men. Rather, it is the fact that, unlike either limited beneficence or disagreement, the circumstance that explains the possibility and necessity of justice is objective. Regardless of whether we imagine ourselves to be wholly or only partially beneficent, or in agreement about ends or not, the facts of the subjectivity of knowledge and of social complexity mean that, in either case, we would still not be in any position to know all that we would need to know in order to attend to the needs of our overwhelmingly unseen fellows in complex societies. Rather than drawing philosophical attention to the need for setting the just terms of social cooperation, therefore, the knowledge problem highlights the need for setting the just terms of social coordination. 5. Objections Rawls, knowledge and reasonable disagreement There are at least two objections that those keen to defend Rawls may raise at this point. The first is that, regardless of what it may tell us about the standing of disagreement, the epistemic account tells us nothing new. After all Rawls (1999a:110) does mention that, along with other shortcomings of ‘knowledge, thought and judgment’ the subjective circumstances of justice are characterised by the ‘incompleteness’ of our knowledge. Yet, given that we have argued that the knowledge problem with which epistemic liberals are concerned is an objective rather than subjective circumstance, it seems that the likelihood of Rawls meaning this in the same sense epistemic liberals do is remote to say the least. Moreover, and regardless of whether it is the same as the problem identified by epistemic liberals, it is also clear that for Rawls the incompleteness of knowledge is relevant only insofar as it 24 helps to explain what is for him the more important point of our ethical disagreement. ‘Some of these defects,’ he (1999a:110) writes, ‘spring from moral faults, from selfishness and negligence; but to a large degree, they are simply a part of man’s natural situation’. We do not need to deny the veracity of this claim about our natural situation, however, in order to sustain the argument about the knowledge problem as the primary circumstance of justice. If our claim about the knowledge relevant to successful social cooperation is of greater significance than the truncated nature of human affection or our propensity to disagree about the ends of life is true then the question of whether the incompleteness of knowledge is also rooted in humanity’s fallen moral standing is moot. Another important objection to the argument set out here is that it rests upon a mischaracterization of Rawls’s position. Particularly when discussing public reason and the burdens of judgment in later work, Rawls (1993[2005]:54-8, 66, 217, (2001[2003]:35-6, 83) is after all concerned to emphasize not disagreement but reasonable disagreement about justice. Here, and regardless of how impartial or altruistic people may be, they will continue to disagree in their religious, philosophical and moral judgments. Disagreements in these matters are therefore inevitable even among rational and reasonable people and this is known by the parties in the Original Position. It is because of this, moreover, that we are expected to forsake appeals to religious and other claims over which reasonable persons can be assumed to disagree when considering the requirements of justice, and instead appeal only to ideas that are held to be part of a liberal democratic political culture and as such may be assumed to be acceptable to all as both reasonable and rational (Rawls, 1993[2005]:217. Rawls’s notion of reasonable disagreement has of course attracted much critical attention in the literature and space will not permit us to enter into its merits here.xii What is significant for the assessment of the persuasiveness of our argument, however, is the place that reasonable disagreement occupies on the spectrum of possible positions that one may adopt regarding the subjective circumstance of justice. We will recall that the discussion thus far has been conducted within the context of an assessment of how differing degrees of disagreement impact upon the possibility and necessity of justice. Here we have claimed, with Rawls, that intractable disagreement renders justice impossible but that, as we have just seen and following Cohen, unanimity fails to make 25 justice unnecessary. Importantly, however, accepting this also shows why reasonable disagreement does not undermine the force of the epistemic argument. Precisely, because reasonable disagreement occupies a position somewhere between disagreement and unanimity - it is a stronger assumption than mere disagreement, but a weaker one than unanimity - it provides no reason to withdraw the claim about the significance of the knowledge problem for the second circumstance of justice. If it is accepted that reasonable agreement is merely a qualified form of ethical unanimity, and that unanimity does not undermine our claim about the knowledge problem, then reasonable disagreement is no more a circumstance of justice than its stronger, unqualified, relative. Cohen, ideology and the nature of justice In contrast to those above that defenders of Rawls may seek to raise, Cohen’s discussion leads to an altogether different and broader objection to the standard account of which the epistemic revision defended here is one version. Here the focus is not upon the alleged failure of the circumstances of justice to adequately ground justice but, as was mentioned in passing in section 3, upon the question of what justice is. That is, for Cohen the question of the circumstances that make justice both possible and necessary is entirely different to the question of what justice is. It is the elision of these two questions in discussion of the circumstances of justice, however, that in his view reveals the standard account to be ideologically loaded, insofar as it serves to obscure liberalism’s unadmitted individualist and capitalist commitments (Cohen, 2008:331). Hume’s discussion, Cohen explains, assumes a pre- given substantive account of justice itself as the artificial virtue of observing rules of property and of promising. But this is to put the matter back to front, for all that the circumstances tell us is that this kind of justice is either impossible or unnecessary under conditions of aggravated scarcity and extreme selfishness or material abundance and extreme beneficence respectively. Cohen makes his point via an analogy with the notion of craftsmanship. In order to know in which circumstances craftsmanship would be possible, he explains, we would first need to know what constituted craftsmanship. Cohen’s answer to this is that craftsmanship would be possible under conditions of limited abundance and limited talent. ‘But all of that’, he (2008:332) tells us, ‘follows from what 26 craftsmanship is and has no bearing on what it is.’ On this reading, therefore, the standard account of the circumstances of justice is more honestly described as the liberal account.xiii Beyond what it may tell us about t Hume’s account, much of the persuasiveness of Cohen’s argument depends upon how strongly we interpret the first part of his phrase ‘no bearing’. That the circumstances of justice have no bearing whatsoever upon the question of justice is an overstatement and, regardless of the unadmitted ethical commitments of Hume and Rawls, the epistemic view can give us some guidance as to why this is so. Firstly, the fact that no individual is in a position to authoritatively claim what an optimal outcome for society ought to be nor, even if this were known, what its achievement would practically require, suggests that justice must be concerned with offering an account of how the knowledge relevant to deciding these questions may be communicated across society. That is, justice must secure not a predetermined division of the fruits of social cooperation, but rather a procedure for the ongoing determination, under irrevocable conditions of moderate scarcity and the knowledge problem, of what that division should be. Against this view Cohen could claim that, similarly to the problem of defining what justice is in advance, the epistemic approach also fails because it too relies upon a controversial ethical assumption from the outset. That is, it is only because epistemic liberals already believe that private interests ought to be pursued that the knowledge problem, and the need to coordinate in spite of it, can be posited as significant. Such a response, however, is inadequate. As our discussion of face-to-face and complex coordination in the previous section suggested, even the successful pursuit of collective rather than private interests in a complex society would be confronted with the knowledge problem, because it would still be unclear what each would need to do, from within the confines of circumstances known only to them, to coordinate activities so that those interests could be satisfied (Tebble, 2016:28-33). The significance of the knowledge problem as a circumstance of justice and the consequent need for a procedural account of justice therefore, are not premised upon an unadmitted, undefended ethical individualism. They are premised upon the knowledge coordination problem presented by the subjectivity of knowledge and of social scale. Importantly, and to press the counter-objection still further, if this much is granted the epistemic account offers even more than the minimal guidance 27 above, for it not only suggests an account of procedural justice, but of pure procedural justice in the sense specified by Rawls. Here, and unlike a perfect conception such as the equal division of a cake guaranteed by the awarding of the allocational task to a person who may eat only the last portion, or an imperfect conception (such as a trial that seeks to discover criminal culpability), there is no external criterion for what constitutes a just outcome, other than conformity with the procedure itself (Rawls, 1999a:73-78). Moreover, we may also be more specific than Cohen suggests with regard, if not to the question of whether the circumstances of justice aid us in identifying what procedures are appropriate, then at least with regard to which are not. Minimally, and if the epistemic circumstances of justice are taken seriously, tyranny and oligarchy do not look promising as procedures for making just decisions about the fruits of social cooperation on account of their assuming that the all knowledge relevant to doing so is accessible to either one or a few. In fact, however, our revision of the standard account is able to provide more than negative guidance, for there is good reason to believe that democratic theory may provide a plausible candidate theory of pure procedural justice in addition to the market proceduralism that epistemic liberals may favour.xiv 6. Conclusion An epistemic reading of the circumstances of justice makes clear the shortcomings of the standard account. First, with respect to Hume’s contribution, the epistemic reading has shown that it is not our limited beneficence that, along with moderate scarcity, explains either the origin of rules of justice or the virtue of obeying them. Indeed, to the extent that it suggests we disagree with Hume in this regard, the epistemic reading of the circumstances of justice concurs with Rawls that it is ethical disagreement that does more powerful explanatory work in this respect. Importantly, however, we have concurred with Cohen that the more important place that Rawslian disagreement occupies in the explanatory order also obviates the need to include moderate scarcity, as both Hume and Rawls do. Most significantly, however, and beyond any claims we may wish to make about the standing of moderate scarcity and confined generosity as circumstances of justice, our epistemic reading also shows that, contra Rawls, the fact of our propensity to disagree about the ends of life, reasonably or 28 otherwise, does not explain why justice is both possible and necessary. Rather, it is the knowledge problem resultant form the fact of the subjectivity of human knowledge and of modern complex conditions found in what Hume calls large societies that explains this need. Even if we were entirely altruistic and in unanimous agreement about what to with the fruits of social cooperation, we would still need rules of justice to offer us a means of making adequate decisions in this respect. Finally, it has been claimed that the epistemic revision of the standard account not only tells us why justice is both possible and necessary. Contra Cohen, it offers at least some guidance as to what justice is, to the extent that different candidate conceptions make the coordination of otherwise inaccessible knowledge in large societies possible. There are, of course, many such conceptions that could qualify as plausible theories of justice in this regard. Which is most suitable, and whether the epistemic circumstances of justice may tell us why, however, are questions which we must leave for another occasion. 29 Bibliography Aquinas (2014) Summa Theologiae. Oxford: Oxford University Press Badhwar NK (1993) The Circumstances of Justice: Pluralism, Community, and Friendship. Journal of Political Philosophy 1(3): 250-276 Broad CD (1930) Five Types of Ethical Theory. Oxford: Clarendon Press Caney S (1991) Sandel’s Critique of the Primacy of Justice: A Liberal Rejoinder. British Journal of Political Science 21(4): 511-521 Clayton Hubin D (1979) The Scope of Justice. Philosophy and Public Affairs 9: 3-24 Cohen GA (2008) Rescuing justice and equality. Harvard, Ma.: Harvard University Press Harrison J (1981) Hume’s Theory of Justice. Oxford: Clarendon Press Hart HLA (1994) The Concept of Law (2nd ed.). Oxford: Oxford University Press Hayek FA (1948a) Economics and Knowledge. In: Hayek FA Individualism and Economic Order, Chicago: University of Chicago Press, pp. 33-56 Hayek FA(1948b) The Use of Knowledge in Society. In: Hayek FA Individualism and Economic Order, pp. 77-91 Hayek FA (1948c) Individualism: true and false. In: Hayek, F. A., Individualism and Economic Order, pp. 1-32 Hayek FA (1960) The Constitution of Liberty. London: Routledge and Kegan Paul Hayek FA (1967a) The Legal and Political Philosophy of David Hume. In: Hayek, F. A. Studies in Philosophy, Politics and Economics. London: Routledge and Kegan Paul, pp. 106-121 Hayek FA (1967b) Freedom and coercion: some comments on a critique by Mr. Ronald Hamowy. In: Hayek FA Studies in Philosophy, Politics and Economics. London: Routledge and Kegan Paul, pp. pp. 348-350 Hayek FA (1982) Law, Legislation and Liberty. 3 vols. London: Routledge Hope S (2010) ‘The Circumstances of Justice’. Hume Studies 36(2): 125-148 Hume D (1740[1978]) A Treatise of Human Nature. Oxford; Oxford University Press 30 Hume D (1751[1995]) An Enquiry Concerning the Principles of Morals. Oxford: Oxford University Press Landemore H (2012) Democratic Reason: Politics, Collective Intelligence, and the Rule of the Many. Princeton: Princeton University Press Lister A (2005) Hume and Rawls on the Circumstances and Priority of Justice. History of Political Thought 26(4): 664-695 Lucas J (1989) On Justice. Oxford: Clarendon Press Okin SM (1989) Justice, Gender and the Family. New York: Basic Books Parfit D (1986) Reasons and Persons. Oxford: Oxford University Press Rawls J (1999a) A Theory of Justice (revised ed). Cambridge: Harvard University Press Rawls J (1999b) Justice as fairness. In: Rawls J Collected Papers. Harvard, Ma.: Harvard University Press, pp. 47-72 Rawls J (1999c) Justice as reciprocity. In: Rawls J Collected Papers. Harvard, Ma.: Harvard University Press, pp. 190-224 Rawls J (1999d) Reply to Alexander and Musgrave. In: Rawls J Collected Papers, Harvard, Ma.: Harvard University Press, pp. 232-253 Rawls J (1999e) A Kantian conception of equality. In: Rawls J Collected Papers. Harvard, Ma.: Harvard University Press, pp. 254-266 Rawls J (1999f) Kantian constructivism in moral theory. In: Rawls J Collected Papers. Harvard, Ma.: Harvard University Press, pp. 303-58 Rawls J (2001[2003]) Justice as Fairness: a restatement. Harvard, Belknap Press Rawls J (1993[2005]) Political Liberalism. New York: Columbia University Press Sandel M (1982) Liberalism and the Limits of Justice. Oxford: Oxford University Press Schwartz Porzecanski D (2004) Friendship and the Circumstances of Justice according to Aquinas, The Review of Politics 66(1): 35-54 Schwartz Porzecanski D (2007) Aquinas on friendship. Oxford: Oxford University Press 31 Sobel C (1972) The need for coercion. In: Pennock, J. and Chapman, J. (eds) NOMOS, vol. xiv: Coercion, Chicago, Atherton Press, pp. 148-177 Tebble, A. J., Epistemic liberalism: a defence, London, Routledge, 2016 Vanderschraaf, P. (2006) The circumstances of justice. Politics, Philosophy & Economics 5(3): 321- 351 Waldon, J., (1999) Law and Disagreement. Oxford: Oxford University Press i I would like to thank Robin Douglass and the two anonymous reviewers at EJPT for their extremely helpful comments on earlier drafts of this paper. ii H. L. A. Hart’s account of the five ‘elementary truths concerning human beings, their natural environment and aims’ that provide the basis of the ‘minimum content of natural law’(1994:1293), it has been claimed, bears much resemblance to the standard account. iii A third condition that Hume sets out - although it is not one that all commentators think is essential to his account - is that of our rough equality of moral powers. On this see Hume (1995:§152). iv See also Hope (2010:128-330) and Vanderschraaf (2006:329-30) who highlight the limited property rights and promise-keeping scope of Hume’s account. v For a discussion of this distinction insofar as it relates to Sandel’s critique of Rawls’ account of the circumstances of justice, see Caney (1991:515-16). vi Similarly, and along with a few other objective conditions, Rawls also follows Hume in claiming that we are roughly similar in our physical and mental powers. vii See also Badhwar (1993:7). viii On the relationship between Rawls and Hume on the circumstances of justice, see Lister (2005:664-695). ix With respect to Hume we may note the discussions of Broad and Vanderschraaf. Broad (2009:94-9) is concerned with his objection to Hume’s wider moral theory and not to the cogency of his account of the circumstances of justice per se. Arguing from a game theoretic perspective, Vanderschraaf (2006) claims that Hume’s three conditions are neither necessary nor sufficient for justice because moderate scarcity, confined generosity and rough equality of human powers also obtain in Hobbes’ state of nature where, crucially, there is no justice. 32 x On altruists’ dilemmas see Lucas (1989:49-50); Parfit (1986); Sobel (1972:148-177). xi For a discussion of complex systems in nature see Hayek (1982:I, 39-40). xii For an insightful discussion see Waldron (1999), chapter 7. xiii But see Hope (2010) who defends Hume against Cohen’s critique by distinguishing between the circumstances and the criteria of justice. xiv Hélène Landemore’s defence of democratic rule on epistemic grounds being the most recent example. On this see Landemore (2012). Badhwar NK (1993) The Circumstances of Justice: Pluralism, Community, and Friendship. Journal of Political Philosophy 1(3): 250-276