Global Duties in the Face of Uncertainty

Shortsightedness poses an obstacle to any moral theory that tries to extend the scope of justice beyond existing political communities. In the field of intergenerational justice, it manifests itself by a tendency to give less weight (or even no weight at all) to the interests of future generations than to those of present generations. This tendency toward ‘presentism’ is furthermore not only presented as a human psychological propensity (to attach more importance to immediate benefits and burdens at the expense of future, and possibly greater, benefits and burdens), but is sometimes also supported as an economically and morally desirable behaviour. Among the moral arguments put forward in favour of presentism we find, for instance, the idea that one has special obligations toward people with whom one stands in a special relationship (especially friends, family members and co-citizens) and that since these relationships tend to disappear as temporal distance grows too large between people, it is morally permitted, indeed required, to give priority to the interests of present generations. Another, perhaps more common argument is formulated in terms of uncertainties: given that it is difficult, if not impossible, to predict in a reliable way what the world will look like in the distant future (e.g. what kinds of problems future people will encounter, what choices they will make when faced with different possible courses of action, what their tastes and values will be, what technologies will be available), and hence, what the needs and interests of future generations will be, it seems justified to focus on what is more easily accessible to us such as the needs and interests of those who belong to present generations. ^[1]

Shortsightedness also haunts global justice theories, this time in the form of a tendency to attach less importance (or even no importance at all) to the interests of distant strangers than to those of co-citizens or co-residents. Again, this myopia and related indifference to the misfortunes that befall people at the other end of the planet has not always been regarded as a mere natural tendency to be more easily moved by what we see with our own eyes. Arguments have also been made to show that this tendency is morally worth pursuing and, among them, the most prominent once again appeal to the idea of special obligations. Unlike temporal distance however, spatial distance does not exclude the existence of important special relationships. The fact that friends or family members become separated by thousands of kilometres does, for instance, not entail that they no longer have special obligations toward each other — the main reasons being that these obligations are constitutive of what it means ‘to be parents’ or ‘to be friends.’ ^[2] It is therefore not surprising that the strategies most frequently used to combat spatial myopia (and which appear to have somewhat monopolised the debate on global distributive justice) amount to revealing the existence of special relationships across state borders. This holds for theories seeking to establish causal links between our conduct and the severe harms suffered by people half a world away. This also holds for theories seeking to demonstrate that certain aspects of existing international institutions — such as their impact on individual life prospects ^[3] or their cooperative nature ^[4] — justify extending our duties of egalitarian distributive justice to all of the world’s inhabitants. In both cases, the point is to make us realise that our moral duties toward the distant poor are stronger and more demanding than we have hitherto assumed and to counteract in this way our ‘natural’ global indifference. Tackling that problem of indifference is all the more important that those who are in a position to address global harms — the so-called ‘global rich’ — belong for the most part to democratic countries and so to countries whose political leaders are expected to be responsive to citizens’ demands. If principles of global justice are to stand any chance of being implemented in this kind of setting, it seems indeed most appropriate to focus the attention on the way ordinary citizens view their own global responsibilities.

Surprisingly, however, relatively little attention has been paid to the other above-mentioned source of shortsightedness, namely the potential existence of uncertainties. Most global justice theorists seem to have taken for granted that once the problem of indifference is overcome, the problem of motivation will also largely be solved. Yet, as this paper aims to show, empirical and normative uncertainties can generate a feeling of powerlessness and demotivate even those who are sincerely willing to promote global justice imperatives. More specifically, this paper will identify four kinds of uncertainties that could potentially have an impact on the nature, the content and the very existence of global duties: first, uncertainties regarding the causes of global injustices; second, uncertainties regarding the consequences of global justice initiatives; third, uncertainties pertaining to the ‘imperfect’ character of certain global duties; and fourth, uncertainties regarding the conduct of other global actors. It will discuss each of these uncertainties in turn, with particular attention to their normative implications, to their distinctively ‘global’ source, and to the possibility of their being addressed. It will conclude with some reflections on how the normative issues raised by uncertainties related to spatial distance compare to those raised by uncertainties related to temporal distance.

1. Uncertainties regarding the causes of global injustices

The first kind of uncertainty that global justice theorists must face has to do with the causes of the present predicament of poor countries. Its normative impact is particularly significant for theories that conceive of global duties primarily as duties not to unduly harm others and to compensate for the harm done. For if the causal nexus that relates a specific conduct to a given harmful situation cannot be clearly established, then these theories will also have difficulties identifying who can be said to have a special duty to address this situation and what exactly they ought to compensate for.

The most striking illustration is probably provided by theories of ‘historical justice.’ To the claim that the situation in which many poor countries currently find themselves is for a large part the result of unjust actions committed by rich countries in the past (such as colonisation, genocide or slavery) and that the latter have therefore today a special duty to help poor countries cope with the economic, social and perhaps even political problems they encounter, it could indeed be objected that it is difficult, or even impossible, to establish causal chains between the harms suffered by the global poor and the conduct of the global rich. One reason for this is that, whereas a state can to some extent be seen as preserving its identity across time, the present members of rich countries can impossibly be considered the cause of injustices perpetrated at a time they were not even born. On what grounds could they then have a special duty to rectify these injustices? One way of circumventing this difficulty is to shift the attention from past injustices to their present harmful effects. So it could be argued that although the present members of rich countries have not themselves engaged in colonisation or slavery, they are still benefiting — and present members of poor countries still suffering — from the continuing effects of these injustices. But the difficulty with this view is that, in addition to diluting the ‘causal’ element that was supposed to activate global duties of justice — it is one thing to cause an injustice, but another to benefit from an injustice one has not caused — it leaves it unclear to what extent present harms or benefits can be traced back to injustices perpetrated several centuries ago. More radically, one could argue that it is always uncertain what the world would have looked like today had the past been different, ^[5] and on this basis, question the very idea that rich countries might have duties of compensation toward poor countries. In order to determine whether rich countries have caused harm to poor countries, we need to describe the counterfactual situation to which the actual situation is to be compared and relative to which it can be said to cause harm to the poor countries. Yet, it might be argued, we will never be in a position to know what the effect would have been of births, events or actions that never took place. As Mathias Risse puts it, “[t]he point is not that a threshold of reasonable certainty cannot be met, but that we must plead complete ignorance.” ^[6] This also holds for such morally reprehensible acts as slavery and colonisation: even if we acknowledge that these acts were profoundly wrong and should never have taken place, it remains impossible to know whether poor countries would have been better off today had these acts never taken place, and hence, to know whether these acts have on balance harmed or benefited the poor. Yet, from a historical justice perspective, so long as these uncertainties persist, the proper content or even the very existence of global duties of justice will also remain uncertain.

Now, theories of historical justice are admittedly not the best example to address the question that interests us here. Even though they may represent a certain conception of global justice — a conception to which even right-libertarians could adhere — the uncertainties they face have more to do with temporal than with spatial distance, and seem therefore to be more a question of intergenerational justice than of global justice. It is however possible to ‘purify’ this example and to develop a similar rectificatory argument but this time concentrated on the present global situation. We find this kind of argument in the work of one of the most prominent global justice advocates, namely, Thomas Pogge. Very much in the same spirit as theories of historical justice, he places the emphasis on our basic negative duty not to cause undue harm to others and on associated positive obligations to compensate for the harm we have done. But he departs from them in basing duties of global justice not on past, but on present injustices. His argument can be divided into the following three steps: first, the normative thesis that we all have a basic negative duty not to harm others by cooperating, without compensating protection and reform efforts, in imposing on them an unjust institutional order; ^[7] second, the empirical thesis that the present citizens of the rich countries are currently cooperating in imposing an unjust global institutional order on others; and third, the conclusion that, under existing empirical circumstances, the present citizens of the rich countries have positive obligations to help the global poor and would violate their negative duty not harm others if they failed to fulfil these obligations. ^[8]

Yet, this present-oriented argument also appears to be weakened by the existence of uncertainties. How do we know whether – and if so, to what extent – the difficulties encountered by poor countries are the product of global rather than domestic factors? ^[9] The difficulty is not only that researchers deeply disagree on the relative significance of domestic and global factors for understanding the why and how of existing world poverty, and hence, disagree on whether one should opt for an ‘explanatory nationalism’ or for an ‘explanatory globalism,’ but also and more fundamentally, that it is extremely difficult if not impossible to disentangle global from local causes of poverty, and hence, to decide between these two options. Just as there is no way of establishing counterfactually that existing world poverty results from injustices perpetrated several centuries ago (rather than from temporally intermediate events or actions), so too there is no way of establishing counterfactually that existing world poverty results from the design of the global institutional order (rather than from local factors such as ethnic divisions, widespread corruption or a political unwillingness to reform domestic institutions). In order to establish that poor country X’s present predicament has been produced by the existing global institutional order, we should be able to compare X’s actual situation with what its situation would have been like without the existing global institutional order; the trouble, however, is that “[w]e have only this one world to work with” and that we cannot ignore the alleged global causal factor while keeping other factors constant. ^[10] Yet, if it remains uncertain whether or not the existing global institutional order harms the poor, it also remains uncertain whether or not those who contribute to its imposition would be harming the global poor by failing to take compensating action.

Pogge himself does not believe that empirical uncertainties as to the causes of present global injustices really affect his argument. According to him, all that needs to be shown is that present citizens of the rich countries cooperate in imposing a global institutional order that foreseeably gives rise to avoidable human rights deficits. In other words, in order to establish that we are harming the global poor, it suffices to know that we are currently taking part in the imposition of global institutions whilst being able to foresee that these institutions could be quite more respectful of human rights if properly reformed. The consequences of these reforms need, for its part, not be predicted with precision provided that the expected reduction of human rights deficits is highly probable in regard of empirical facts. ^[11]

In the same vein as Pogge, one might think that shifting the attention from what the rich ‘have done’ – except for their (often non-voluntary and inescapable) contribution to the imposition of a global institutional order — to what they ‘could do’ but ‘fail to do’ helps avoid a good many of the difficulties faced by more traditional causal approaches to global justice. Freed from the need to disentangle the role played by respectively ancient wrongs, recent and/or domestic factors, and existing global institutions in the production or perpetuation of world poverty, and hence also from the uncertainties surrounding these questions, we seem to be in a better position to establish the existence of global duties of justice, to specify their content, and to identify their bearers and recipients. Indeed, following Pogge, one could simply say that all those who participate in the existing global institutional order have duties of justice to try to improve this order toward a better protection of human rights, and that they owe this obligation to all those who are subject to this order and still suffer from human rights deficits. However, as we will see in the next section, things are more complicated than may seem at first glance. One reason for this is that even global justice theories that allocate duties primarily on the basis of agents’ capacities (such as their financial or technological resources) regardless of what they ‘have done’ will need information on the real and potential consequences of particular courses of action, and hence, on complex empirical facts concerning causation. ^[12]

2. Uncertainties regarding the consequences of global justice initiatives

What is striking when reading empirical studies dealing with world poverty relief is not only the number of initiatives that have not worked as expected, but also the uncertainties that remain regarding the kind of initiatives that could work. As some authors have pointed out, however well-intentioned they might be, many of the initiatives that have been undertaken in the context of international humanitarian and development aid over the last decades either have not led to a genuine, long-term improvement of standards of living in poor countries or have in some cases even led to a deterioration of these standards by, for instance, disrupting local systems of production, delaying the needed political reforms, or by encouraging the continuance of conflicts. ^[13] Still more worrying, there remain lots of disagreements among researchers, indeed experts, regarding both the kind of empirical data that are worth collecting (not to mention the difficulty of collecting them in certain local contexts) and the kind of methods that are appropriate to assess them. To this must be added the unavailability of public and reliable reports on the effects that particular aid initiatives have had: NGOs have, for instance, made few attempts to assess their work, and when they have, the results of these assessments have either remained unpublished or of little use due to their poor quality. ^[14] All of this indicates that, in addition to having to take note of the failure of a large number of international aid initiatives, we also have to acknowledge that we are for the time being unable to form a sound judgment on the kind of international aid initiatives that could be effective. There is an “overall uncertainty in the empirical literature about what aid really works.” ^[15]

When making these observations, Leif Wenar’s purpose is primarily to make global justice advocates aware of how difficult it is to predict the effects of particular aid initiatives. We cannot just say that “We would not be sacrificing anything significant if we were to continue to wear our old clothes, and give the money to famine relief. By doing so, we would be preventing another person from starving” ^[16] or that “Minor redesigns of a few critical features would suffice to avoid most of the severe poverty we are witnessing today.” ^[17] Donations of money and small institutional redesigns can prove completely inefficient or even harmful to their intended beneficiaries. Faced with the sad portrait he depicts of international aid, Wenar does not want to sound pessimistic however. Far from concluding that international aid cannot be efficient, he suggests the possibility of overcoming aforementioned challenges by intensifying independent research in order to gain comprehensive and systematic information on aid effectiveness, by subjecting aid agencies and programmes to more stringent standards of transparency and accountability, and by inquiring, as ordinary individuals, into the economic and political situation of potential recipient countries and into the working of particular aid agencies. ^[18]

As can be expected, Wenar’s concerns are particularly relevant to consequentialist theories of global justice. For from the moment the rightness of an action is made conditional on its outcome and the difficulty of foreseeing the outcome of an action is acknowledged (which is especially likely to be the case on a global scale), it also becomes difficult to determine what actions are to be considered duties. From a consequentialist perspective, the idea that an agent may be morally required to perform an action whose consequences are uncertain remains problematic.

These considerations may lead us to think that the existence of uncertainties is much less of a problem for deontological theories, which refuse to make the rightness of an action depend on its effects, whether actual or merely intended. As Immanuel Kant explicitly remarks:

[...] reason is not sufficiently enlightened to survey the series of predetermining causes that would allow it to predict confidently the happy or unhappy results of human actions […]. But it throws enough light everywhere for us to see what we have to do in order to remain on the path of duty. ^[19]

In his view, our moral duties can only be determined a priori, in a purely rational way, without reference to empirical considerations (except for what can be called the ‘general facts of human life’ such as the fact that men are both sentient and rational beings, that the earth on which they live has a spherical form, etc.); otherwise, so-called ‘duties’ would remain as uncertain as empirical circumstances and lose their unconditional, and hence, moral character. To appreciate the import of these claims, one need only recall the (in)famous essay On a Supposed Right to Lie from Philanthropy, in which Kant forcefully rejects the idea that one could ever be exempted from the moral duty not to lie to others, even when our interlocutor is a murderer asking for the whereabouts of our friend who is hidden in our house. But as we will see now, this does not mean that uncertainties do not reappear in Kant’s moral philosophy under another guise.

3. Uncertainties pertaining to the ‘imperfect’ character of certain global duties

Even if Kant insists on the exclusively rational foundation of all our moral duties, he also makes a difference between two domains of morals: right, which prescribes laws for actions, and ethics, which prescribes laws for maxims or subjective principles of actions. Right generates duties that are said to be ‘narrow’ or ‘perfect,’ that is, duties that are precisely and strictly determined and therefore raise no doubts as to the way they are to be fulfilled in particular cases. ^[20] This is one of the reasons these duties can be subject to external legislation and constraint. ^[21] Paradigmatic examples of ‘perfect’ duties include the duty not to coerce others and the duty not to deceive others. Ethics, by contrast, makes not the performance of certain actions but the adoption of certain ends into a duty. ^[22] Ethical demands can therefore neither be subject to an external constraint (their fulfilment can only result from a self-constraint) nor determine with precision in what way and to what extent an agent is to act. They are in this sense ‘wide’ or ‘imperfect’: they demand that certain ends be promoted, but they leave some “playroom” or “latitude” for free choice as to the specific way of promoting these ends. ^[23] Among these ‘imperfect’ duties, we find the duty of beneficence understood as the duty to promote the happiness of others, and which covers the duty to help others to meet their basic needs.

The significance of imperfect duties for global justice theory has been brought into relief by Onora O’Neill. ^[24] At the heart of her reflection is a deep discomfort with the increasingly widespread idea of ‘economic and social human rights’ understood as pre-institutional moral rights that would enable us to assess and to improve existing institutional rights and arrangements. O’Neill doubts that these alleged ‘human rights’ could indeed be considered genuine — that is, claimable and violable — rights rather than mere ‘manifesto rights,’ ^[25] and the reason she gives is that it is in the abstract entirely impossible to specify what their corresponding duties are. So long as these duties have not been allocated through concrete institutions, it is impossible to know who is to do what in order to provide for the economic and social needs of the world’s poor. This contrasts with universal (or human) liberty rights, which ‘naturally’ correlate with universal negative duties: we do not need institutions to know that we all ought to refrain from doing certain sorts of actions (such as killing or deceiving) to all others. ^[26] The same cannot be said of the duty to provide for the basic needs of the poor: given that no single agent has the capacity to provide for the basic needs of all the poor, this duty cannot be seen as a universal duty or as a duty owed to all others. It can only be regarded as a duty to provide for the needs of some unspecified others. Yet, as long as we do not know who has the duty to provide for the economic and social needs of a specific person, it also remains uncertain against whom this person could claim her economic and social rights and who could be accused of having violated her economic and social rights.

That world poverty cannot be adequately addressed from the perspective of economic and social human rights does not mean, however, that it must be relegated to the rank of secondary or optional concerns, nor that it could not become a matter of claimable rights. When stressing the normative uncertainties surrounding so-called economic and social human rights — that is, the impossibility of knowing in the abstract what their corresponding duties are — O’Neill shows that the human rights discourse, while being particularly well suited to “empowering the powerless” and to “giving voice to the voiceless,” ^[27] also has the tendency to divert the attention from potential duty-bearers. But according to her, there is no reason to believe that imperfect duties to help people in need are less important than perfect duties of right or of justice. Nor is there reason to believe that imperfect duties should not somehow be embodied in institutional structures, indeed give rise to specified institutional rights.

We find the same kind of reasoning in the context of humanitarian intervention. ^[28] So, it has become quite common to argue that the international community has an imperfect duty to protect victims of gross violations of human rights ― a moral duty whose importance may justify its taking precedence over sovereign states’ freedom not to intervene, but which in the abstract does not fall on any agent in particular ― and that this duty, far from being doomed to remain imperfect, essentially means that the members of the international community must first put in place the institutional arrangements allowing the allocation of responsibilities and the specification of their content. Accordingly, the imperfect duty to protect involves a duty to make this duty perfect, and hence, to transform what is initially a mere ‘manifesto right’ into a claimable and enforceable institutional right.

There are however several difficulties with this approach. To begin with, even if it is Kantian in spirit, it is unlikely that it can properly be attributed to Kant. Certainly, this is not an objection in itself. Still, it is worth noting that one of the reasons Kant can hardly be interpreted as supporting the idea that an imperfect duty chiefly demands that it be made perfect has again to do with considerations of uncertainty. So while Kant acknowledges that we have an imperfect duty to promote the happiness of others, he also continuously emphasises that, given that “people differ in their thinking about happiness and how each would have it constituted,” “no universally valid principle for laws can be given”: it is impossible for anyone to determine with certainty what would make others happy, and thus also to determine surely and universally what action would promote their happiness. ^[29] As Allen D. Rosen puts it, “Kant’s belief that the variability and indeterminacy of the concept of happiness render it unfit to serve as a basis for legislation.” ^[30] Trying to ‘perfect’ the imperfect duty of beneficence would amount to selecting and to imposing on others one particular conception of happiness, and therefore, to paternalistically negating others’ freedom to “seek [their] happiness in the way that seems good to [them].” ^[31] That is to say, those committed to perfecting the imperfect duty of beneficence always run the risk of behaving in an unjust way.

One may also wonder whether duties to ‘perfect’ imperfect duties are not themselves marked by important empirical and normative uncertainties, especially when their fulfilment is to proceed on a global scale. More specifically, one may wonder whether a duty to work toward the establishment of institutions in charge of relieving world poverty or of protecting the victims mass atrocity crimes can be seen as anything other than an imperfect duty given the uncertainties that remain regarding the kinds of institutions that could efficiently fulfil these tasks and the identity of those against whom victims could legitimately lodge their claims in case no one is acting. ^[32] If the point is not only to allocate global responsibilities, but to do so in a way that is both efficient and fair, then it is important to know what kind of institutional reforms could help us discharge our imperfect duties and what role specific institutions should play in helping us discharge these duties. Yet, some authors doubt that it is currently possible to reach that knowledge and the reason they give has to do with moral disagreements and uncertainties. ^[33] We simply do not know yet how global duties are to be allocated between domestic and global institutions because, global institutions and global justice concerns being of relatively recent origin and still evolving, the capacities of global institutions in matters of global justice, the proper goals of particular global institutions, and the proper scope of state sovereignty cannot yet be determined.

[T]he difficulty is not just that there is considerable moral disagreement about the proper goals of global governance institutions and about the role these institutions should play in the pursuit of global justice; there is also moral uncertainty. ^[34]

Here again, as in Wenar, the proposed response is to focus on epistemic requirements. While the existence of normative uncertainties may justify temporarily limiting the demands of global justice that may reasonably be placed on global institutions, it also points to the need to put in place the epistemic-deliberative conditions that would allow us to cope with them. Making global institutions more transparent and more accountable is a first step in the right direction, but there is more to do than that. Given that the proper terms of accountability are themselves a subject of dispute and cannot be determined without gaining a clearer understanding of the proper goals of global institutions, there must also be a process of principled, informed deliberation, contestation and revision of these goals and terms of accountability, and this process must work in tandem with external actors controlling the reliability of the information, integrating it, and making it available in understandable form to the broad public. According to Buchanan & Keohane, this

[...] emphasis on the conditions for ongoing contestation and critical revision of the most basic features of the institutions captures the exceptional moral disagreement and uncertainty that characterize the circumstances of legitimacy for this type of institution. ^[35]

4. Uncertainties regarding the conduct of other global actors

The last kind of uncertainties that I wish to address may well be the most radical, since it is sometimes assumed to bring into question the very idea that there could be any moral duties in the international sphere. It is also the most often thematised one, especially by the realist tradition in international relations theory. A major argument within this tradition holds indeed that an agent cannot be asked to act in a moral way, and so to accept to constrain her freedom of action, so long as she does not have the assurance that others will behave similarly toward her, especially when her survival is at stake. Yet, it is argued, given that this assurance can only stem from agents’ submission to a state ― or more precisely, to common public coercive laws — and given that states are for the time being not subject to any world state, states cannot be considered as having any moral duty on the international scene, except a duty to look after the well-being of their people or to advance the national interest. ^[36]

This argument has often had Hobbesian overtones. So even though Hobbes discusses the international sphere only sporadically, the idea that this sphere represents a Hobbesian state of nature characterised by a “war of all against all” is quite widespread among realists in international relations theory. In the absence of a superior political authority, a state would always try to hold sway over other states and would by the same token never be immune from external attacks either. In this context of reciprocal threat and distrust, states are well advised to try to maximise their power, even if this implies committing acts of aggression. As Hobbes puts it, agents in a state of nature have a “Right to every thing; even to one anothers body.” ^[37] What is at stake is their very survival and any ‘warlike’ means may be used to ensure it.

Certainly, Hobbes acknowledges that it would be rational for states to renounce their natural right to everything and to observe certain moral rules in their mutual dealings. ^[38] But this holds only if other states also accept to renounce this right. No state can be bound to respect its agreements so long as a “just cause for fear” subsists, and for Hobbes, such a cause subsists so long as there is no assurance that other parties will also respect these agreements. ^[39] A state that would stubbornly keep acting morally toward other states, regardless of the way these states behave toward it, would made itself into a prey for others and no state can ever be obliged to do so. ^[40] Importantly, moral rules are worth pursuing only because and insofar as their observance contributes to one’s self-preservation.

Here again, one might think that, unlike prudential theories, deontological theories of Kantian inspiration are not affected by the existence of uncertainties as to the reciprocity of human behaviour. As Kant emphasises, “commands of morality” differ from “counsels of prudence” in that they are limited by no condition. ^[41] Whereas prudence makes the necessity of an action depend on what the agent in question regards as his happiness, morals represent actions as necessary in themselves, that is, as necessary regardless of their likely consequences or of other agents’ likely behaviour. Again taking the duty not to lie as an example, the fact that we know that our interlocutor has a strong disposition to tell lies in no way exempts us from the duty not to lie to her. The same holds a fortiori in cases where we are uncertain about her disposition.

Yet, quite surprisingly, Kant seems to adopt a different position when dealing with the rights of individuals in the state of nature since he claims that in this state “[n]o one is bound to refrain from encroaching on what another possesses if the other gives him no equal assurance that he will observe the same restraint toward him.” ^[42] What is even more disconcerting is that he too seems to ultimately ground this absence of obligations on prudential considerations. Like Hobbes, he links it to the malevolence of human nature and argues that, since each of us can observe in herself the natural tendency of all human beings to attack and try to dominate others as soon as they have an opportunity to do so, nobody can be obliged to wait until the facts provide a sad confirmation of this inclination before being authorised to defend herself. ^[43] Kant appeals to the same kind of considerations at the international level when he grants states in the state of nature a (provisional) right to wage war against states whose power becomes threatening (and not only against states that effectively aggressed them) on the ground that a superior power, even before having committed any act, already wrongs inferior powers by the very fact of its superiority. ^[44] Here again, what seems to underlie Kant’s recognition of a right to preventive war is, on the one hand, the idea that “men must be presumed evil until he gives the assurance of the contrary” ^[45] , and on the other hand, the idea that there can be no obligation to act in an imprudent way. This interpretation of the Kantian state of nature as devoid of duties to refrain from infringing on what belongs to others is further reinforced by Kant’s claim that, insofar as they intend to be and to remain in this condition, agents — whether they are individuals or states — “do each other no wrong” when they coerce or deceive each other for what holds for the one also holds for the other “as if by mutual consent.” ^[46] Thus, a state that breaks the agreement it concluded with its enemy cannot complain of being wronged if this enemy behaves in the same way when the opportunity presents itself. ^[47]

A major difference between Kant and Hobbes, however, is that Kant supports the existence of a moral — and not prudential — duty to leave the state of nature. As he emphasises, even though they cannot do wrong to each other, agents that refuse to leave the state of nature “do wrong in the highest degree” because they renounce the very concept of right. ^[48] If they cannot prevent their actions from having an impact on each other, then they have a moral duty to enter a rightful condition, that is, a condition in which their rights are determined and enforced through public coercive laws, and where remaining disputes about rights are settled before a court. ^[49]

But whether we adopt a moral or a prudential viewpoint, the same question arises: how to conceive of the existence of a duty to leave the state of nature when the uncertainty and the distrust that characterise this condition compromise the very existence of duties to others?

Kant’s answer to this question can found in his conception of jus in bello or “right during a war” ^[50] : while recognising how difficult it is to conceive of laws in a lawless condition, he also insists on the importance of not resorting to acts of war that would destroy trust between states. ^[51] These include: the use of assassins, poisoners or spies, breach of surrender, incitement to treason, violation of public contracts, etc. ^[52] Those who make use of such “dishonorable stratagems” do wrong in the highest degree (although they may do no wrong to each another) because they act on principles whose universalisation would make international peace impossible. It thus appears that certain duties are “of the strict kind” and apply whatever the circumstances in which one finds oneself, that is, “even in the midst of war” and even at the risk of one’s life. ^[53] In other words, while the malevolence of human nature and the lack of assurance that is peculiar to the state of nature may authorise us to use force against others, they can never authorise us to deceive others. The reason is that some trust must remain if states are ever to leave the state of nature, solve the problem of assurance and be secured in their rights.

5. Building knowledge and trust: some avenues for further reflection

This article has examined four kinds of uncertainties that cut across the field of global justice thinking. The first kind of uncertainties mainly affects causal approaches to global justice and revolves around the impossibility of establishing counterfactually that present global harms have been caused by injustices in the distant past (in backward-looking versions) or by the design of the existing global institutional order (in present-looking versions). One strategy to overcome these counterfactual difficulties is to shift the attention from the causes to the remedies of global injustices. But as we have seen, capacity-based approaches to global justice face a second kind of uncertainties, namely uncertainties regarding the real and possible consequences of alternative courses of action, and more specifically, uncertainties regarding the kinds of initiatives that could work in the context of international humanitarian and development aid. The third kind of uncertainties is more normative in nature and refers to the impossibility of allocating in the abstract the duties corresponding to so-called social and economic human rights, or alternatively, to the impossibility of determining at present what role particular global actors should play in the pursuit of global justice. Finally, the fourth kind of uncertainties stems from the anarchical character of the international sphere, and more particularly, from the idea that so long as there is no supranational political authority, states can have no assurance that other states will refrain from infringing on their fundamental interests and can therefore not be required to behave morally toward one another. ^[54]

The existence of these different kinds of uncertainties make the determination of our global duties difficult, if not impossible, and therefore risks demotivating even those who are sincerely committed to the cause of global justice. This is obviously the case for the third kind of uncertainties, which are essentially normative uncertainties about the allocation of duties among global actors, and for the fourth kind of uncertainties, which call into doubt the very existence of moral duties among global actors. But the same also holds for the first and second kinds of uncertainties which, although empirical in nature, also give rise to normative uncertainties regarding, among others, the bearers of global duties of rectification and the concrete measures that should be taken to address global injustices.

From the foregoing, it appears that the responses that have been offered to the problem of uncertainties in the context of global justice are primarily of two kinds. First, gaining better empirical knowledge of the causal implication and of the capacities of global actors, and putting in place a framework for deliberation in order to reach a better appreciation of the proper goals and responsibilities of global actors. Second, maintaining or developing relations of trust between states with a view to replacing the law of the strongest with public coercive laws, thus making progress in terms of global justice. Underlying these responses is the idea that our global duties include a fundamental duty to put in place the conditions that would enable us to build (empirical and normative) knowledge and trust.

How do these responses fit within the framework of intergenerational justice? The trust issue undoubtedly has less relevance here. Certainly, there are more uncertainties as to the way future generations will behave within five hundred years than as to the way present global actors will behave within five months. Between temporally remote generations, there is also an obvious lack of reciprocity. Assuming that present generations observe their duties toward future generations: not only is there no guarantee that future generations will also observe their own intergenerational duties, but it is also impossible for these generations to reciprocate the benefits they received from generations in the distant past. This raises the question of whether the likelihood of non-moral conduct on the part of future generations can undermine the very existence of intergenerational duties in the present. Still, it is worth noting that at the heart of Hobbes’ and Kant’s arguments is not only a problem of reciprocity, but also and more fundamentally, a problem of survival: there can be no duty to act in a way that threatens one’s self-preservation. Yet this problem of survival cannot occur in the intergenerational context: we cannot argue that acting morally toward future generations would amount to making ourselves into prey for others and on this basis deny that we have moral duties to them, since the conduct of future generations can no longer have any impact on us.

By contrast, the knowledge issue is easily carried over to the intergenerational context. As we have seen, it is not possible to talk about the causes and the remedies for global injustices without taking into account their temporal dimension and the difficulties associated with it. A backward-looking perspective brings with it the difficulties related to the use of historical counterfactual baselines; a forward-looking perspective raises the question of the reliability of predictions. As indicated in the introduction, intergenerational justice theorists do not appear to deny the impact of uncertainties on what morality requires of present generations. They tend to agree that the less reliable predictions are, the more reasonable it is to disregard them. ^[55] But they also point out that predictions are often more reliable than is usually thought. ^[56] We can, for instance, have reliable knowledge of the perennial value of certain basic goods (such as unpolluted air or the absence of chronic diseases) and of the harms we cause to future generations. The latter is especially the case when one focuses on “regular patterns of behaviour” rather than on “discrete acts.” ^[57] Even if it is not always possible to predict the remote effects of a particular action, we can gain reliable information on the probable effects of our habits and ways of life (such as our daily car use). Therefore, the recommendation made by global justice theorists to increase our (normative and empirical) knowledge also seems appropriate in the context of intergenerational justice.

It would, however, be important to reflect further on the limitations of this recommendation. So, one may wonder to what extent it is possible for ordinary individuals, and even for experts, to gain the knowledge that would enable them to make informed decisions about the kinds of reforms that would promote the interests of the global destitute. It should not be assumed that this only holds for very specific, technical questions. As demonstrated by the controversies they raise in the public debate, even ‘basic’ questions we are confronted with almost every day — such as ‘Do we need more of less trade liberalisation?’ or ‘Do we need more or less Europe?’ — are obscured by a host of uncertainties. The point is not merely that decisions are made behind closed doors and/or by experts, nor that it would be too demanding for ordinary people to amass a large amount of information on the probable impact of different institutional reforms or patterns of behaviour. The point is also, and more worryingly, that most real world issues have become so complex and so multidisciplinary that expertise in one field no longer appears to be sufficient. Do we need to be omniscient in order to make the “right” choices? Are we not feeding the myth that omniscience is possible?