ATTACHMENT, AUTONOMY AND 1 Justice, Attachment and Natural Resources 1 Chris Armstrong, University of Southampton Forthcoming in Journal of Political Philosophy, 2013. Introduction This paper investigates the significance, from the point of view of egalitarian justice, of patterns of attachment to natural resources. Section I establishes what is at stake in arguments about attachment as a source of special claims over resources. Section II shows why, by contrast to the view associated with Charles Beitz, for instance, attachment over natural resources should be taken seriously by egalitarians. But even if egalitarians should take attachment seriously, it is not clear just how and whether they could accommodate important attachments whilst also holding firm to their egalitarian commitments. Section III, however, provides an account of just how egalitarians can accommodate good attachment-based claims within their accounts of justice, and also spells out the implications of that accommodation for theories of global justice. Section IV considers but rejects an alternative account of how we should accommodate attachments, and concludes by arguing that egalitarianism contains far greater resources for taking important attachments seriously than we might otherwise suppose. 1 Many thanks to Andrew Mason, Annie Stilz, Kit Wellman and Lea Ypi for comments, as well as the participants in workshops on benefiting from injustice at the University of Uppsala, and on territorial rights at the University of Frankfurt. Work on this paper has been facilitated by a British Academy Mid-career Fellowship on global justice and natural resources. 2 I. Theorising Attachment On the assumption that they are simply ‘there,’ with no-one responsible for creating them, theorists of justice have often suggested that we all have symmetrical general claims on the world’s natural resources, 2 of either a sufficientarian or egalitarian character. 3 But an adequate account of natural resource justice will also recognise and respond to special claims, which come in two varieties. Special claims based on improvement seek to ground certain rights over a natural resource on the way in which an agent has increased its value. Even if no-one has created a particular natural resource, acting on it so as to render it more valuable plausibly grounds a special claim to at least a portion of the income from it. 4 But though it has received much greater attention, it is not only improvement that is capable of generating special claims over natural resources. Indeed the claim that it is only productive use of land or resources which can generate good claims has an ignoble history, inasmuch as it has been used to justify dispossessing indigenous peoples from land or resources which they have inhabited or interacted with, but not used ‘productively.’ 5 Assuming that an account which only considered special claims based on improvement would be an impoverished one, this paper will consider a second kind of special claim, in this case premised upon ‘attachment.’ Attachment- 2 I assume, for the purposes of this paper, the conventional definition of natural resources under international law, which depicts them as non-human-made goods taken from the ‘natural wealth’ of the world. As such they are distinct from man-made products but also from the geographical sites from which they can be taken. Key examples include non-fugacious resources such as unrefined minerals, metals and wild plants, and fugacious (free-flowing) resources such as air, water, wild animals and various kinds of natural energy. Section IV considers an alternative view according to which these things should only count as natural resources when they are used in certain ways. 3 See e.g. Charles Beitz, Political Theory and International Relations (Princeton: Princeton University Press, 1979); Thomas Pogge, World Poverty and Human Rights (Cambridge: Polity, 2002); Mathias Risse, On Global Justice (Princeton: Princeton University Press, 2012). 4 See e.g. Hillel Steiner, An Essay on Rights (Oxford: Blackwell, 1994). 5 See e.g. James Tully, An Approach to Political Philosophy: Locke in Contexts (Cambridge: Cambridge University Press, 1993), pp. 137-76. 3 based claims seek to ground rights over resources on the close relationship which some agents have formed with specific resources, developing life-plans which depend upon secure access to them. For this project to be worthwhile, we do not need to assume that claims from attachment and improvement will always be entirely separate. It may often be the case in practice that the same agent has both improved, and become attached to, a specific natural resource, thereby generating overlapping special claims over it. But this will not always be the case. A poor miner might work hard to extract coal from a coal-seam – thereby generating a special claim based on improvement – but feel no attachment to that coal, and indeed disdain his career as a miner, and wish to change his job at the earliest opportunity. On the other hand someone might become attached to a resource without improving it. A Hindu might place great store in the expectation that on his death his ashes will be scattered on the Ganges; a Christian might place great store by her ability to baptise her children in the water of the Jordan. Neither will ‘improve’ either river from an economic point of view. If there are cases where improvement-based claims are weak or non-existent but significant attachment to a resource nevertheless exists, we need to know whether separate moral claims can derive from such attachment, and what force those claims might have. If we can identify such claims, it will thereby become more obvious, for instance, why the dispossession of indigenous peoples was, and is, objectionable. If the conventional account which focuses solely on economic improvement – and is often associated with Locke 6 - appears unduly restrictive in its implications for judging resource claims, one prominent alternative – the conventional Kantian account of ‘provisional right’ – appears under-determined. On Kant’s account, we can obtain a provisional claim over an object simply by declaring our will to control it, at 6 It is, in fact, controversial whether Locke actually sought to ground property rights in improvement. For a sceptical discussion, see A. John Simmons, ‘Makers’ Rights,’ in Justification and Legitimacy (Cambridge: Cambridge University Press, 2000). 4 the same time as declaring our willingness to submit that claim to the adjudication of a civil (and, ultimately, cosmopolitan) political authority. That account does not place any further conditions on what constitutes a valid claim over an object. Lea Ypi has recently suggested that the Kantian account can afford to be ‘ecumenical with regard to how groups of people end up occupying specific geographical areas’ – or, perhaps, controlling particular objects; Kantians can accept improvement-based criteria, or attachment-based ones, or indeed any criteria so long as the claims arising under the various accounts share the crucial commitment to binding authority. 7 But Kant’s account is able to be ecumenical precisely because it is so undemanding. The emphasis on the need to generate political authority means that Kant’s account will grant claims just wherever the declared will to control an object exists, along with the willingness to commend that claim to final adjudication by a higher political authority, and independent of any relationship of attachment or improvement. The Kantian account has several significant limitations. First, whilst it might provide us with guidance on how to treat claims generated prior to the establishment of civil authority, it does not provide us with any guidance on how to deal with special claims arising once a civil authority actually exists. Once the civil condition exists, the account simply tells us that property rights are confirmed, but also possibly constrained, by the rightful civil authority – but we lack criteria for how they ought to be constrained. Second, the connection between the need to establish civil authority, on the one hand, and the (highly permissive) criterion supplied for judging valid claims, on the other, is rather underdetermined. If we assume that transition to a civil condition is morally necessary then it is certainly plausible to consider it a necessary condition of valid claims over objects that claimants express a willingness for those 7 Lea Ypi, ‘A Permissive Theory of Territorial Rights,’ European Journal of Philosophy (forthcoming 2013), at 16. 5 claims to be bound by some future civil authority. What is simply not obvious is why we should consider that willingness to be bound by civil authority a sufficient condition (or, to be more precise, a condition which, alongside the will to control an object, is jointly sufficient) for judging a claim to be valid. The focus on the importance of submission to a binding authority does not prevent us from qualifying such claims in additional ways. On this latter score the conventional Kantian account does not seem to provide a satisfying account of when a morally weighty claim has been made. The fact that I merely will to control something, regardless of any other features of my relationship with it, is a very undiscriminating basis on which to grant rights over resources. If I simply declare a wish to control the oil buried under the ground on the island where I live, ought we to place any weight on that claim? At the very least, we will presumably want to place some constraint on how much of my island’s oil I can control. But more importantly, why grant even prima facie claims to control resources with which I have had no relationship at all, other than willing to control them? 8 The best foundation for attachment-based claims on resources is likely to be one which focuses on the conditions for the pursuit of significant individual life-plans. Any plausible account of justice will consider it important that we are able to see ourselves as at least jointly directing our own lives, making our own plans for life which we in turn have some prospect of achieving. On Rawls’s account, that will require justified confidence in our ability to develop our talents and to exercise them, 8 Another interpretation of Kant’s position suggested to me by Annie Stilz would maintain that what is normatively significant is not the mere act of willing control over an object, but the actual achievement of bringing that object under one’s control. But it is not clear to me why the achievement of control in and of itself would make a difference in justifying claims. It strikes me that if control has normative significance, it will lie in the fact that as a consequence of controlling objects we might come to develop expectations of continued access – in short, to begin to orient our life-plans around those objects. 6 so that one’s ends fall within the realm of the reasonably attainable. 9 It will require, in addition, the basic liberties necessary for pursuing our various projects. And it will also have material preconditions: at the very least, we will need secure access to the objects of our basic rights. And an egalitarian account will specify more demanding material prerequisites, constraining tolerable material inequalities in a variety of ways. If we accept that, we have grounds for specifying general claims to various material resources, including natural resources. These will be general in the sense that we all possess such claims simply as human beings, and insofar as we are all capable of generating life-plans that ought to matter. But what of special claims? Special claims over natural resources will be particular in two ways: they will be claims which only some of us possess, because of our precise relationship with natural resources. And they will also be particular in the sense that they apply to specific natural resources. By this we mean not specific kinds, but specific instances or ‘tokens’ of natural resources. An attachment-based justification for special claims over natural resources will emphasise the way in which we sometimes form life-plans which depend upon continued secure access to specific resources. If they are to be successful, attachment-based claims will be claims which cannot be met, at least without irretrievable loss, merely by providing equivalent shares of other resources. Such arguments should also appeal to egalitarians. Take welfare egalitarianism, for instance. A welfare egalitarian will care about resources precisely insofar as they fuel human wellbeing, in their various ways; resources matter, derivatively in that sense, inasmuch as they are important to the success of the various projects that matter to us. But if so, he or she has good reason to pay heed to any information we 9 See e.g. John Rawls, A Theory of Justice (Harvard: Harvard University Press, 1971), p. 440 (on confidence in one’s abilities); Justice as Fairness: a Restatement (Harvard: Belknap, 2001), p. 60 (on the ‘social bases of self-respect’). 7 might have about how particular resources are important to us. At the very least, information about attachments ought to be seen as valuable data for those seeking to promote welfare. More strongly, the presence of attachment to resources produces a defeasible reason for granting the so attached continued access to them. Notably, contemporary Locke scholars have offered support for such a view, arguing for a move away from a focus on either improvement or ‘labour-mixing’ as the only basis for special claims over resources, and suggesting instead (or in addition) that special claims can be granted wherever an individual integrates an object into what Simmons calls her ‘purposive projects.’ 10 Given the centrality that objects can come to play in our life-plans, we have a prima facie reason not to reallocate those objects lightly. Some contemporary Kantians have also given support to this kind of account. In her (ostensibly Kantian) defence of rights of ‘occupancy’ over land, Anna Stilz suggests that there is a good prima facie claim for someone to continue to occupy land wherever that someone has incorporated access to that land into their life-plans, such that occupancy becomes ‘fundamental to the integrity of his structure of personal relationships, goals, and pursuits.’ 11 Individuals may have used land as part of their projects, of what gives their lives meaning and makes it valuable to them. In particular we ought to focus on those life plans which are ‘comprehensive’ rather than ‘peripheral’ – which ‘structure many of our goals and choices, and are fundamental to our sense of ourselves as self-determining authors of our lives.’ 12 10 A. John Simmons, The Lockean Theory of Rights (Princeton: Princeton University Press, 1992), chapter 5. Note that Simmons considers this focus on purposive projects a development of the labour- mixing view. 11 Anna Stilz, ‘Nations, States and Territory,’ Ethics 121.3 (2009): 572-601, at 585. See also Christopher Heath Wellman, ‘Political Legitimacy and Territorial Rights,’ ms. 12 Stilz, ‘Occupancy Rights and the Wrong of Removal,’ ms. Whilst she does not explicitly invoke the kind of attachment-based account I am defending here, in motivating her defence of a Kantian account of permissive claims Lea Ypi also reaches out to examples whereby a claim ‘promotes a sufficiently important end of agents’ (Ypi, ‘A Permissive Theory,’ at 10), or where people unable to stake such claims have their ability to form and act on life-plans undermined. 8 To date, though, defences of this type of view have focused on control over land. Stilz’s account, for instance, focuses on the wrong of removing peoples from the land they inhabit, but does not apply the argument to natural resources. But the fact that natural resources are in principle removable from land means that we can intelligibly ask questions about whether rights over each should be allocated separately. Many natural resources (such as freshwater, fish, wild animals and possibly even oil) are not tied to a particular piece of land but are rather ‘fugacious.’ Even where resources are not fugacious, grounding claims to them purely in prior claims to land would leave us vulnerable to the challenge that if resource x could be removed without harming the integrity of the land (for instance, by some high-tech horizontal mining technology), any resource claims would then dissipate. To deliver a robust case we ought to investigate claims which are directly grounded in relationships with resources, rather than dependent on prior claims over land. There is, in any case, an increasing recognition that rights over natural resources cannot simply be extrapolated from the territorial right to govern a particular geographical area (or even from putative rights to occupy land). In that case, as David Miller puts it, rather than simply assuming that rights to exercise jurisdiction over territory, resources and membership need to be held together, ‘The relationship between the three sets of rights should be regarded as an open question.’ 13 This paper will investigate whether an attachment-based account can be extended to the case of natural resources, and on finding that it can, will set out some of the implications for the way in which we should integrate general and special claims within a broader account of resource justice. 13 David Miller, ‘Territorial Rights: Concept and Justification,’ Political Studies 60. 2 (2012): 252-68. A. John Simmons similarly observes that the ‘connection between territorial rights [over land] and rights over all resources in the relevant territory is not in any way (morally, physically, conceptually) necessary.’ Simmons, ‘States’ Resource Rights: Locating the Limits,’ online symposium on resource rights: http://territorynetwork.wordpress.com/ , at pp. 5-6. http://territorynetwork.wordpress.com/ 9 II. Attachment, Life-Plans and Natural Resources There are many cases in which individuals have formulated life-plans which are dependent upon secure access to particular natural resources. The most compelling instances are probably those in which members 14 of certain relatively small-scale groups have formed a close and enduring relationship with resources which have become key to central and enduring practices, and where that relationship is both sustainable, and non-substitutable in the way that it supports individual life-plans. Where access to specific resources is integral to one’s way of life and indeed one’s sense of self in this way, it is plausible to say that a fundamental interest in the preservation of one’s way of life generates a prima facie – but defeasible – claim to at least continued access to those resources. The resources in question must be specifically valuable, though, such that they are not straightforwardly substitutable as supports for life-plans. For a good claim to exist over these trees, it should not be the case that other trees would serve the purposes of a given individual without discernable shortfall. It must be the case that in carrying out his life-plans access to anything other than these trees would be, to him, a (possibly very poor) second best. 15 14 As explained later in this section, the argument I am making does not depend upon attributing value to ‘cultures’ or to collective identities; the argument from attachment can be persuasively made, I believe, solely with reference to important individual interests. 15 In an interesting recent paper, Margaret Moore has also argued that we should take seriously the diverse ways in which people value natural resources, beyond treating them as mere substitutable commodities. Moore, ‘Natural Resources, Territorial Right and Global Distributive Justice,’ Political Theory 40.1 92012): 84-107. However, Moore’s main concern is to show that a concern for collective self-determination points in the direction of state or communal control over resources, and hence is rather orthogonal to my project in this paper. I address arguments that seek to ground resource rights in the value of self-determination in Armstrong, ‘Against “Permanent Sovereignty” over Natural Resources,’ ms. 10 Some might still find attachment implausible, by itself, as an independent source of special claims. If no improvement has occurred, why then grant rights over resources? Why should I be able to ground a claim over a resource, simply as a result of developing an expectation that I will be able to access it, and then developing life- plans to match? We need to give serious attention to the suspicion that attachment to a resource is too weak a normative consideration to generate a claim to exercise rights over it. In his well-known discussion of the moral arbitrariness of resource distribution, for instance, Charles Beitz gives grounds for just such suspicion. He suggests that the argument from attachment to resources to rights over them was even more tenuous than the argument from the possession of talents to a right to the economic rewards attendant on such talents which Rawls had earlier done so much to unsettle. Specifically, he suggests that: ‘It would be inappropriate to take the sort of pride in the diamond deposits in one’s back yard that one takes in the ability to play the Appassionata…talents, in some sense, are what the self is; they help constitute personality. The resources under one’s feet, because they lack this natural connection with the self, seem more like contingent than necessary elements in the development of personality.’ 16 The reason that pride in the diamond deposits in one’s back yard is inappropriate, to be clear, is that pride is (something like) a justified satisfaction with one’s character or abilities, given that character and ability are elements of the self which we often deserve credit for cultivating. By contrast natural resources - such as the diamonds in 16 Beitz, Political Theory and International Relations, p. 139. 11 one’s back yard - ‘lack [a] natural connection with the self.’ 17 Or, to put it another way, when the Rector in George Eliot’s Middlemarch asserts that ‘It is a very good quality in a man to have a trout stream,’ we ought to take this as one of Eliot’s many jokes. 18 The connection between personality and the possession of diamonds or trout streams is simply too weak, for Beitz, for concern with the former to justify rights over the latter. If we are at all tempted by the argument that talent is a morally arbitrary basis for distribution, we ought to apply that argument a fortiori to natural resources. But Beitz’s claim repays closer attention. He suggests that, by contrast to the kind of pride it is appropriate to take in playing the Appassionata, it would be inappropriate to take pride in one’s relationship with natural resources - given that one has, by hypothesis, done little or nothing to actively nurture them, and given that they are not, unlike talents, actually ‘constitutive of the self.’ Now that first clause is too quick, for of course a particular resource may well have been developed and improved, even in the autarkic world Beitz supposes in that part of his argument. But the second clause initially appears more secure. It does look to be a distinctive feature of talents that they are internal to the self, or constitutive of the self as Beitz puts it. However, consider again the Appassionata example. Imagine that Sophie has invested her whole being in playing that piece, and that now she can do it she thinks about herself in a different way – she appraises herself more positively as a person capable of cultivating her abilities and sensibilities, and she has a real sense of living within a musical tradition. Now that talent is always going to be vulnerable in certain ways (which is another way of saying that it has preconditions which are capable of being threatened). Sophie might injure her hands, and be unable to play any longer. 17 Ibid., p. 138. 18 George Eliot, Middlemarch (London: Penguin 1994 [1874]), p. 70. 12 Likewise if Sophie’s piano was taken from her – or if she was forbidden from playing any piano – she would be unable to exercise her talent. Whilst not directly deprived of her talent, she would be deprived of a crucial object for its exercise, and we would also expect her talent itself to suffer in time. If so, the loss to her sense of self, and her sense that she had skilfully brought her life-plans closer to fruition, could be just as acute as if she had injured her hands. Her control over her talent and her control over her piano – or, failing that, a substitute piano – are both crucial to her continuing to experience her sense of agency, and to the fulfilment of a project which is central to her life. Although there are some talents which do not obviously depend on control over external objects for their expression (like dancing, or singing, for example), there are many talents which are dependent for their exercise on secure access to objects such as musical instruments, sports equipment, vehicles, and so on. But if control over a piano – an external artefact, which she has done nothing to create or improve – is crucial to Sophie’s pursuit of her life-plans and even to her sense of self, then we have a reason to ensure she can continue to access it. And we can say the same thing, in many cases, about natural resources. Imagine that Ken, a fisherman, has oriented his life around fishing – developing his talents, his ability to work out where the fish are, what they want to eat at any given time, and so on. Here it seems that Ken’s sense of pride in the performance of fishing is no more misplaced than Sophie’s pride in her ability to play the Appassionata. But again Ken’s pursuit of his life-plan is vulnerable: we could remove his rod, of course (his equivalent of the piano), or we could take the fish from the river he fishes, or ban him from fishing it (or indeed from any river). If we did so, we would be directly interfering with Ken’s central life-plan. Indeed our removal of an object necessary to someone’s performance of his life-plans without a very compelling reason for doing will likely 13 be construed (often correctly) as evidence that we do not consider those life-plans valuable, and may even be interpreted as a sign of deep disrespect. Assuming that Ken’s life-plan is valuable to him, we have a strong reason not to do so. Are real-world cases available whereby individual identity is bound up in individuals’ right to securely interact with a specific resource, so that removing that right from them would present those individuals with a situation where they did not even believe themselves to be the same people any more? A good example might be the Saami people of Scandinavia: roughly sixty thousand members of an indigenous community which has sustained itself for more than a thousand years by herding reindeer. 19 The task of feeding and herding reindeer dictates the movements of Siidas – small groups composed of several families – across the seasons. It provides an economic foundation, but more than that it provides a shared life-plan, deeply wrapped up in a specific mode of existence in relation to the physical environment. What we have in this case is a situation where there is an intimate and profound connection between an individual’s identity and her ability to securely interact with a specific natural resource. Whilst all of us will have life-plans dependent on the availability of some natural resources (including, at the very least, the objects of our basic rights, such as clean air and drinking-water), some of our life-plans demand secure access to specific natural resources. We don’t understand ourselves as fishermen or as hunters if there are no fish to fish, or no prey to hunt. And we may not understand ourselves as Sioux or Tlingit fishermen or hunters unless certain particular fish or prey are available to us. To think that this represents a loss, from the point of view of justice, we do not need to think that cultures are in themselves valuable, or that their ‘essential characters’ ought to be preserved. The present account makes no 19 There is ongoing controversy – quite distinct from the question of whether animals can have rights - about whether humans can have rights over wild animals. If not, we can imagine examples where the attachment is to rocks or rivers instead. 14 such claim. We merely need to take seriously the claim that it matters to individuals that they are able to act on plans which are central to their lives, and that in order to act on such plans secure access to certain resources may be required. As a matter of legal reality, the various governments of the Scandinavian countries have recognised this in their individual enactments of Saami-specific rights over reindeer herds. 20 These do not typically extend to land ownership, but rather to rights to continue to herd reindeer. Similar accommodations have been made in Canada. 21 Although international law protecting indigenous rights is weak in many respects, it does clearly purport to defend access to resources where those resources are important to specific ways of life. Elaborating on the ‘right to culture’ – potentially, for the sake of our argument, a misleading term - defended in Article 27 of the Covenant of Civil and Political Rights, for instance, General Comment 23 suggests that: (c)ulture manifests itself in many forms, including a particular way of life associated with the use of land resources, especially in the case of indigenous peoples. [Thus cultural] right might include such traditional activities as fishing or hunting and the right to live in reserves protected by law. The enjoyment of these rights may require positive measures of protection and measures to ensure the effective participation of members of minority communities in decisions which affect them. 22 The reference to ‘culture’ here ought not to mislead us into believing that what is at stake are rights to use natural resources for exclusively ‘cultural’ or ‘symbolic’ 20 For some useful discussions of the recent history of these rights, see chapters 5, 12 and 14 of Svein Jentoft, Henry Minde and Ragnar Nilsen (eds) Indigenous Peoples: Resource Management and Global Rights (Delft: Eburon Academic, 2003). 21 S. James Anaya and Robert Williams, ‘The Protection of Indigenous Peoples’ Rights over Land and Natural Resources in the Inter-American Human Rights System,’ Harvard Human Rights Journal 14 (2001): 33-86, at 51. 22 Office of the High Commissioner for Human Rights, General Comment 23 (1994): The rights of minorities (Art. 27). See: http://www.unhchr.ch/tbs/doc.nsf/0/fb7fb12c2fb8bb21c12563ed004df111?Opendocument http://www.unhchr.ch/tbs/doc.nsf/0/fb7fb12c2fb8bb21c12563ed004df111?Opendocument 15 purposes. Rather the Covenant has been used to protect uses of a resource – whether ‘economic’ or ‘cultural’ or otherwise – which are essential to the way of life of a particular party. 23 The connection between specific resources and our life-plans and even identities does not always appear to be as tenuous as Beitz suggests, then. In many cases Beitz may be right that the connection between (individual or national) personality and the mineral deposits buried beneath one’s feet may be weak. Some natural resources will not even have been discovered, and others will be largely untouched, and will not feature in our projects to any significant degree. He may also be right that at the level of nation-states the connection will typically be weak. 24 But we should not rule out just any attachment-based claims. It does appear possible to identify particular natural resources which are hugely significant to particular people’s sense of agency, and to their ability to carry out projects to which they are deeply wedded. III. Responding to Claims from Attachment Attachment-based claims will be ill met by conventional means of undifferentiated redistribution – such as a principle of equal shares of total natural resource values - since what they typically demand in the real world are rights over specific resources. 25 What is at stake is the right to insist that these trees should not be cut down, or that 23 Anaya and Williams, ‘The Protection of Indigenous Peoples’ Rights,’ 52. 24 For a fuller defence of that claim, see Chris Armstrong, ‘Against “Permanent Sovereignty.”’ 25 Simmons claims that denying this ‘means denying the actual arguments made by Native American tribes for historical rights to particular lands and resources have any moral force at all, or any appeal beyond ungrounded emotionalism.’ Typically these claims are claims to the actual lands the Native Americans ‘lived, hunted, and worked on.’ Simmons, ‘Historical Rights and Fair Shares,’ in Justification and Legitimacy, at p. 241. 16 this agent should continue to be able to herd these reindeer. But on the other hand, attachment-based claims cannot be unconstrained by broader considerations of distributive justice: they surely cannot, for instance, ground unlimited rights over unlimited portions of natural resources. Just how stark a challenge, then, do they pose for conventional theories of natural resource justice? To make any progress in answering that question, we need to be clear about the precise content of attachment-based special claims. Specifically, we need to be clear about which rights over natural resources are being claimed, and whether those rights are necessarily exclusive – a move that is, regrettably, not often clearly made. As I suggest elsewhere, 26 there are a number of rights which an agent might enjoy over a given natural resource, including (at least) the right to access and non- subtractively use a resource, to withdraw and consume resource units, to manage a resource, to exclude others from its use, to alienate it, or to derive income from it. 27 Special claims over natural resources will be more demanding – and will appear less plausible, at least to some - if we assume that the only intelligible special claim is a claim to ‘full liberal ownership’ of a resource, comprising all six rights. 28 But this is certainly not the only intelligible claim, and indeed it probably does not reflect the spirit of most indigenous claims, for instance. If what matters is the ability to securely interact with a particular resource, ‘full liberal ownership’ is far from the only answer. If my life-plans centrally include an ability to walk in the woods near my home, that does not require that I own the woods, or that I am able to prevent others from walking in the woods, or that I can derive any or all of the income 26 Armstrong, ‘Against “Permanent Sovereignty.”’ 27 The list is adapted from Elinor Ostrom, ‘Private and Common Property Rights,’ in B. Bouckaert and G. De Geest (eds) Encyclopedia of Law and Economics, vol. II. Civil Law and Economics (Cheltenham: Edward Elgar, 2000), pp. 332-79, at p. 339, with the exception of the right to derive income which is my own addition. 28 On full liberal ownership, see Tony Honoré, ‘Ownership (1961),’ in his Making Law Bind (Oxford: Clarendon Press, 1987), pp. 161-192. 17 from the woods. It demands precisely that I can walk in the woods without disruption (it may also be dependent upon the woods being sustainably managed, whether by me or someone else). Likewise, my ability to live in a rainforest unmolested is compatible with granting outsiders the right to emit greenhouse gases, and hence sharing that rainforest’s absorptive capacity. Continued and unmolested access, in many such cases, is what is at stake in attachment-based claims. Such claims can therefore be granted without affecting anyone else’s claim to share the income or ecosystem benefits arising from the resources in question. In the case of the Saami the secure pursuit of central life-plans will require rights beyond mere access. Here the demand seems to be to rights of access (including, notably, a right to traverse Scandinavian national borders which is as old as those borders themselves 29 ), but also a share in management, some (limited and sustainable) rights of withdrawal, and the right to derive some income from selling meat, jewellery and clothing produced from reindeer. But the demand is not to ownership of reindeer, and even insofar as it extends it is not necessarily a claim to exclusive rights. Management rights over reindeer herds can plausibly be shared, access to the reindeer need not be exclusive so long as access by outsiders does not disrupt herding practices significantly, and so on. The income derived from selling meat, jewellery and clothing likewise offers a rather basic standard of living to the Saami herders, and is unlikely to trouble those with egalitarian or even sufficientarian sentiments. 30 Overall, the 29 Fae Korsmo, ‘Nordic Security and the Saami Minority: Territorial Rights in Northern Fennoscandia,’ Human Rights Quarterly 10 (1988): 509-24. 30 I suspect, in fact, that the attachment-based argument for a right to derive income is less secure than the argument for, for instance, rights to access or manage resources. Specifically, I agree with Simon Green that if the ability to securely pursue one’s own life-plans is what is at stake, then a right to the ‘deployment of one’s talents’ as an artist (to use Green’s example) is not obviously intimately tied to, and does not obviously demand, the right to also sell one’s paintings on an open market. Green, ‘Competitive Equality of Opportunity: A Defense,’ Ethics 100 (1989): 5-32, at 18. Thus it might be that a Saami herder could enjoy her life as a herder without being able to sell reindeer goods, so long as she otherwise had the income to get by. That said, we might have other grounds for granting those 18 cluster of rights claimed appears here to pose no especial obstacle to those who would seek to distribute the benefits of the world’s resources equally. More pointedly, those who support welfarist (or partially welfarist) varieties of egalitarianism ought to treat our knowledge of the depth of attachment to reindeer or to walks in the woods as a valuable source of information about how equality can be achieved. In short, unpacking the precise content of the rights claims implicit in many special claims from attachment immediately renders the question of how they can be accommodated much more tractable, and reveals the potential compatibility of many special claims with protecting access for others, enjoying the benefits of shared management, using some share of income in the broader interests of justice, and so forth. We ought not, though, to imagine this move as a panacea. Special claims from attachment could conceivably be much more expansive than in the Saami or woods- walking cases. How should we think about the appropriate limits to special claims, then? One obvious constraint, it seems to me, is supplied by the injunction that no-one ought to be deprived, or deprive others, of natural resources (such as freshwater, or air) which are essential for meeting their basic human rights. Indeed, along with earlier thinkers such as Aquinas, Pufendorf, Hobbes and Locke, I find it hard to see how any right can be more fundamental and inalienable than the right to access those natural resources necessary to one’s survival. 31 That right may have important implications for broader discussions of control over resources, including, potentially, debates about international migration, but our goal is not to address those implications but merely to note the constraint it places on attachment-based special claims. If one party’s rights (such as improvement-based claims, or a belief that it would be more efficient, or infringe her liberty less if we allowed her to earn her own income rather than depending on handouts from others). 31 See also Risse, On Global Justice, chapters 5 and 6. 19 attachment-based claim to control a river has the inevitable consequence that others will imminently face dehydration, for instance, then it ought to be rejected. It might be objected here, of course, that we should understand the right to control or to securely access particular resource tokens as itself a basic human right. I do not believe that we should. It is plausible to suggest that as humans we all have an entitlement to the objects of our basic right to subsistence, for instance, and in some cases this will imply a (general) right to specific kinds of natural resources (such as freshwater, clean air, and sunlight). But it is not clear that the same can be said for specific instances or tokens of natural resources. No-one, I suspect, has a basic human right to continued access to or control over this tree, or that river. There may, to be sure, be good pragmatic reasons for allowing people to meet their basic rights by way of the specific resource tokens they already control and are attached to; if we believe that attachment is important in the way in which I am suggesting then there may be good principled reasons for preferring them to. But this does not take us very far, and in any case those reasons have to be seen as defeasible when the basic rights of others are at stake. Aside from a basic rights constraint, does egalitarianism place further constraints on the extent to which we should grant special claims from attachment? I believe that it does, but not in the way which is usually suggested. A very common view on the demands of equality with regards to natural resources suggests, roughly speaking, that the distribution of such resources is morally arbitrary, that it wrongly advantages and disadvantages people according to accidents of birth, and that we should all have, in fact, an equal entitlement to share in the benefits and burdens arising from the world’s natural resources. That view itself is often linked with an egalitarian constraint on appropriation which suggests that we ought to cease 20 appropriating resources once our shares reaches the maximum level which could be symmetrically held by everyone else. I have argued elsewhere that global egalitarians ought to reject such a view. 32 To see why we need to take a brief detour into the thorny question of the correct currency for equality. It may be that no single metric can be provided for egalitarianism or, to put it another way, that insofar as we care about advantage and disadvantage we are stuck with a conception of advantage which is irreducibly heterogeneous, incorporating a concern with the distribution of both resources and welfare. 33 Anyone persuaded of that point would immediately wonder why we ought to single out natural resources and hold that shares of the benefits and burdens arising from this good ought to be equal. But scepticism about insisting on equal shares of natural resources (or the benefits and burdens arising from them) does not depend upon prior scepticism about a single neat answer to the currency question. In fact, advocates of each of the three major currencies – equality of welfare, equality of capabilities, and equality of resources broadly construed – also ought to reject strictly equal shares of benefits. Why, for instance, insist on equal shares of benefits when those benefits drive welfare in unequal ways across persons? The welfarist should instead prefer a distribution in which unequal shares of resource benefits are granted, such that those unequal shares sustain equal welfare. 34 A similar conclusion will follow for the capability theorist: part of the reason for embracing the capability metric, after all, is that resources fuel capabilities in different ways, and that the 32 Chris Armstrong, ‘Natural Resources: the Demands of Equality,’ ms. 33 See G. A. Cohen, ‘Afterword to Chapters 1 and 2,’ in Michael Otsuka (ed) G. A. Cohen: On the Currency of Egalitarian Justice, and Other Essays in Political Philosophy (Princeton: Princeton University Press, 2011), pp. 61-72. 34 Michael Otsuka, ‘Self-Ownership and Equality: A Lockean Reconciliation,’ Philosophy & Public Affairs, 27 (1998): 65-92. 21 insistence on equal shares of resources (of whatever variety) is mistaken. 35 The resourcist, too, ought to be unsatisfied with a situation in which natural resources were equalized in isolation, and without regard to the impact of ‘internal resources’ or endowments – or, indeed, other social goods which ought to matter to the egalitarian, such as income, inherited wealth, healthcare or education. 36 So in the face of, inter alia, widespread and multifaceted inequalities in important social goods or simply in the face of unequal capacities to convert natural resources into, say, welfare or capabilities, equal shares of natural resource values becomes unattractive. This means, in turn, that what we could call a ‘narrow egalitarian proviso’ on natural resource appropriation also becomes unattractive. When we ask the question what quantity of natural resources (or benefits and burdens) individuals ought to be allowed to appropriate, we ought to recognise the utility of allowing agents to appropriate greater than equal shares of natural resource benefits insofar as this usefully offsets inequalities in other areas (including, but not necessarily limited to, cases of unequal capacity to convert resources into whatever broader category of advantage we care about). Distributing to each of us an equal portion of the benefits arising from the world’s natural resources in contexts of background inequality would be arbitrary in principle, and regressive in effect. The limits to appropriation ought not to be set by an unnecessary insistence on equal shares of these goods, but by equality across the range of goods relevant to distributive justice. Call this, if you will, a ‘broad egalitarian proviso.’ As with the recognition of the diverse set of rights which attachment claims might in practice target, this theoretical moves also establishes that egalitarians are 35 Amartya Sen, ‘Equality of what?’ Tanner lectures on human values 1 (1980): 195-220. 36 Ronald Dworkin, ‘What is Equality? Part 2: Equality of resources,’ Philosophy & Public Affairs (1981): 283-345. 22 able to be more accommodating towards attachment-based special claims than we might otherwise suppose. In cases where an agent has structured his or her life around secure access to particular resources equality can permit allowing him or her to control a greater than equal share of resources (or resource benefits), or indeed to block the use of a greater than equal share of resources where her central life-plans would be disrupted by their use (see below). 37 This again renders the case of indigenous claims to justice more tractable. Such claims often target secure access to large swathes of resources, but they are also, typically, accompanied by substantial inequalities in other spheres – by the general poverty and exclusion of indigenous peoples, for instance. There are limits to the degree to which we should be ‘integrationist’ about all of the goods relevant to egalitarian justice, 38 and if we have a commitment to equal citizenship then we ought not to be wholly content with a situation in which large shares of natural resources ‘compensated’ for political exclusion. But if indigenous peoples do indeed turn out to have a strong preference for greater than equal resource shares at the expense of, say, income, granting such attachment-based claims is compatible with equality. *** I have suggested so far in this section that quite substantial attachment-based claims to natural resources are compatible with egalitarian justice, and that egalitarians often have good reason to favour such claims as a means of pursuing equality. But this is 37 As Jeremy Waldron suggests in another context, any such claims cannot of course be granted once- and-for-all and without being subject to future revision; we will want to say that what counts as a just holding has to be seen as revisable, subject to for instance changes in population, in scarcity of resources, and so on. Waldron, ‘Superseding Historic Injustice,’ Ethics 103 (1992): 4-18, at 16. 38 For an excellent account of the basis of, and limits to, integrationism about egalitarian justice, see Simon Caney, ‘Just Emissions,’ Philosophy and Public Affairs 40.4 (2012): 255-300. 23 not to suggest, of course, that attachment-based claims will never conflict with equality. Indeed if that were the case my argument would be uninteresting. It is because conflicts can occur that we need to know exactly how far egalitarians can go in accommodating such claims. I have suggested that they can go much further than we might at first suspect, if they are prepared to carefully unpick the precise rights targeted by special claims rather than assuming that ‘ownership’ is the only option on the table, and if they are prepared to apply a broad rather than narrow egalitarian proviso to such claims. But there will still be cases which ask more than egalitarians can, in their capacity as egalitarians, accede to. Even if special claims from attachment do not directly target control over an expansive set of resources, the costs of meeting them could still potentially be very high. In many contentious cases in the real world a claim to exercise control over resource x is contentious not because x describes a large or valuable set of resources, but because control over resource x proves to be incompatible with exploitation of resource y by others. In legal practice, many indigenous land and resource claims appear to take precisely this form, being in essence defensive or ‘blocking’ claims against logging and oil-extraction activities, for instance, where those activities would seriously disrupt indigenous practices. 39 What is claimed, in such cases, is that secure access to these trees is incompatible with clearing the forests for agricultural use, or that the enjoyment of traditional fishing practices in these waters is incompatible with exploiting the oil shales of such a stretch of the coastline for petrochemical production. Whilst the claim to those trees or those fish is not in itself especially expansive – and, as such, not directly worrying to egalitarians – the opportunity costs of respecting such claims is high, because 39 Anaya and Williams, ‘The Protection of Indigenous Peoples’ Rights,’ passim. 24 respecting them will place a block on local economic development (development which might itself advance wider goals of social justice). We can imagine, then, a case where there is a definite global interest in the development of the resources contained in a region, 40 but where agent A would prefer such development not to take place because they are attached to a particular resource which would be destroyed if development occurred. Such blocking claims, I would suggest, should be treated in the same way as direct claims, and as such should also be subject to a broad egalitarian proviso. This has the implication that development of a set of natural resources can be blocked by indigenous communities where the benefits and burdens associated with those resources are not so great that no plausible direct claim could be granted over them; as I suggested above, in a context of countervailing disadvantage indigenous communities may be able to make quite large blocking claims. But it is hard to see on what grounds egalitarians could endorse blocking claims which violated even a broad egalitarian proviso. If the block on development has the result that outsiders inevitably become or remain worse off – across the whole range of goods relevant to distributive justice – than agent A, granting such claims would demand, in effect, that we treat the projects of agent A as more important than the projects of outsiders. Even if reasons can be produced for accepting such a result – whether reasons of justice or reason beyond justice – it is hard to see how egalitarians, as egalitarians, could endorse them. 40 This will not be true in many important cases, including the rainforest case briefly mentioned earlier. In such cases there may be a national interest in ‘developing’ the forests, but global interests are actually aligned in favour of preservation. Where this is the case, there may be an argument for outsiders sharing the opportunity costs of preservation with locals. See Chris Armstrong, ‘Sharing the Costs of Rainforest Protection,’ ms. 25 IV. Resources, Attachment and Justice This paper has suggested that if egalitarians care about people and their diverse interests, then a diverse range of egalitarian accounts also have reason to care about the way in which people form projects which often crucially rely on secure access to specific external objects such as natural resources. Although existing accounts of territorial rights over land do not automatically extend to ground rights over natural resources, the present enterprise shares much with the efforts of recent territorial rights theorists to take seriously the desire of many people to maintain access or ‘occupancy’ over particular sites – or resources - which have come to matter deeply to them and to their central life-projects. Far from helping ourselves to the conclusion that the actual distribution of natural resources is entirely arbitrary, and that there is accordingly nothing at all to be said, from a moral point of view, for current patterns of attachment to resources, we ought to care when such attachments persist and seek, within the proper constraints of an egalitarian account, to accommodate them. But although I have shown that egalitarians both can and should make much more room for attachment-based special claims than has often been recognised, it might be thought that my account still leaves some agents’ access to precious resources too vulnerable to the general claims of outsiders. Perhaps where attachment exists we should say something still stronger, to the effect that the resources to which an agent is attached should not actually be considered natural resources from the point of view of justice, and hence ought to be entirely immune from programs of global redistribution. In that way justice – egalitarian or otherwise – would cease to constrain attachment-based special claims at all. 26 Avery Kolers has recently defended such a view, 41 but I have suggested elsewhere that it is deeply mistaken. 42 On his account once we recognise the right kind of relationship between a community and a resource, questions about outsiders’ claims on that resource simply do not arise – no matter how strong the countervailing interests of outsiders happen to be. When we recognise these crucial relationships what we are doing, in effect, is accepting that communities invested in particular resources are entitled to say that they are not natural resources. But how do we determine when the requisite relationship exists? We should cease to consider something as a natural resource, from the point of view of justice, whenever the local community does not treat it as a mere substitutable goods to be traded as a commodity. 43 Whenever a good is not considered (by the community controlling it) to be non-substitutable either across types or across tokens – that is, when it cannot be substituted, without loss, with another resource of the same type, or indeed with a resource of a different type – external claims simply become irrelevant. Interestingly, special claims here do not gain force from the intensive use of resources (as, perhaps, in the case of improvement-based special claims), but from their non-intensive use. Where a resource is being used non-intensively – and not treated simply as an economic asset to be sold on open markets – this is likely to indicate a depth of attachment which overrides merely economic uses. 44 But non-intensive use looks likely to be a poor proxy for the existence of morally significant relationships between people and resources (it might, after all, 41 Avery Kolers, ‘Justice, Territory and Natural Resources,’ Political Studies 60.2 (2012): 269-86. 42 Chris Armstrong, ‘Resources, Rights and Global Justice: A Response to Kolers,’ Political Studies, forthcoming 2013. 43 Kolers, ‘Justice, Territory,’ at 279. 44 Margaret Moore also argues that ‘the capacity to “count” something as a resource is important to collective self-determination.’ ‘Natural Resources, Territorial Right,’ at 88. However, later in her paper she apparently repudiates the argument that local communities should get to decide whether something counts as a resource or not (at 103-4, n. 9). 27 simply mean that a given agent is utterly uninterested in a resource, or that they have such a great quantity of natural resources that they could not possibly use them all). It is also unclear how the fact that a given agent considers a natural resource not to be substitutable across types suffices to ground a claim to particular tokens of a resource (the fact that I have a project dependent on access to water in general does not suffice, surely, to justify a claim over the precise water I currently control, and certainly not all of it). But most significantly, Kolers’s account simply draws the wrong type of conclusion from the existence of attachment over some natural resources. It dictates that where there is a good claim based on non-intensive use, any broader questions about how to share access to a particular resource are simply voided. That resource (now a non-resource) becomes, in Kolers’s words, ‘immune to resource justice considerations.’ 45 But this is wholly implausible, especially in the face of possibly urgent claims from outsiders whose basic rights might be under threat. A more plausible account would seek to balance the different claims over resources, and endorse special claims only where there is a sufficiently weighty interest in retaining control over particular goods – which does not infringe on the basic rights of others - and not simply where we identify the mere existence of a certain kind of claim which immediately trumps other claims. Deciding on such cases is not easy, but I have suggested that we will find the task easier if we are prepared to explore more complex ways of sharing rights over resources, and if we are prepared to endorse a broad rather than narrow egalitarian constraint on any such claims. Indeed I hope to have shown that it is possible to take attachment-based special claims over natural resources seriously without simply disregarding the claims of outsiders. Recognising a resource’s usefulness to others is compatible with 45 Kolers, ‘Justice, Territory,’ at 282. 28 recognising – and, when appropriate, defending – its special significance to some people in particular. If so, we can provide reasons of justice to take attachment seriously, rather than immediately seeking to place attachment-based claims beyond justice. If we are prepared to advance reasons why we should take attachment seriously from the point of view of justice then, it seems to me, we can also enhance our understanding of just why colonialism and the forced dispossession of indigenous peoples was so wrong.