This article focuses on duties correlative to human rights. We can distinguish between two different kinds of duties: negative and positive ones. Negative duties are duties not to perform an action, while positive duties are duties to perform an action. I focus on the latter, and in particular on understanding who holds them. I introduce the thesis that all individuals hold positive duties correlative to human rights, which I call the Universal Scope Thesis, and defend a refinement of this thesis.
In section I, I outline the initial assumptions. In section II, I present the Universal Scope Thesis, provide an argument by analogy in its support, and outline three objections to the argument. I then argue that these objections do not warrant abandoning the Universal Scope Thesis, but rather require us to refine it. In section III, I lay out the notion of aggregative duties: duties that can only be adequately grasped when we focus on the aggregate effect of the actions and omissions of different agents. In section IV, I elaborate on positive aggregative duties correlative to human rights and lay out the implications of this framework for the Universal Scope Thesis. In section V, I explain how this framework allows us to meet the objections presented in section I. In the process of doing so, I specify the Universal Scope Thesis. Section VI assesses an alternative thesis on the scope of positive correlative duties, and section VII concludes.
I. INITIAL ASSUMPTIONS
I assume that human rights are moral rights, and that their correlative duties are strong requirements of justice. I endorse the influential account of rights developed by Raz, according to which ‘X has a right’ if and only if ‘X can have rights, and, other things being equal, an aspect of X's well-being (his interest) is a sufficient reason for holding some other person(s) to be under a duty’.Footnote 1 On this account, rights are sufficient reasons for their correlative duties.Footnote 2 I assume that the function of human rights is to protect certain especially important interests of the right-holder.Footnote 3 I take it that those plausibly include, among other things, interests in adequate health, nutrition, political participation, education, and shelter. In the rest of this article, I use the term ‘fundamental interests’ as a place-holder to refer to the interests protected by human rights.
Thomas Pogge draws a distinction between institutional and interactional accounts of human rights. The interactional conception holds that both individual and institutional agents have a direct responsibility not to violate human rights. The institutional conception holds that only coercive institutions can directly uphold or violate human rights, while individuals have a duty to support institutional conduct complying with these claims.Footnote 4 The institutional account entails that violent husbands, abusive parents and racist attackers are not violating their victims’ human rights.Footnote 5 I find this unacceptably counterintuitive: following, among others, Ashford and Caney, I thus endorse an interactional approach.Footnote 6
I assume that a group can itself be a duty-holder, over and above its members, only if it satisfies the conditions for agency. I am neutral with respect to the specific content of these conditions: on whether, for instance, members of a group need to commit jointly to acting together, or share a decision-making procedure, for a group to be an agent.Footnote 7 However, I assume that one such set of conditions needs to be satisfied for a group to be an agent and, in turn, bear duties. Hence, I assume that groups which are not agents cannot hold duties: since duties are reasons to act, it is not clear what a duty would do for something that is not an agent.Footnote 8 I call the groups which satisfy conditions for agency ‘collectives’ and the duties held by those agents ‘collective duties’. Collective duties are opposed to ‘individual duties’: duties held by individual agents (such as rational human beings or other intelligent creatures). For simplicity, in what follows I focus on individual agents and duties. The picture I sketch can be extended to include collectives and their duties.
Another important question concerns the claimability of rights. A right is claimable if and only if its correlative duties have specified holders and content. Onora O'Neill has famously argued that rights must be claimable.Footnote 9 This grounds an influential argument against the thesis that positive duties can be correlative to human rights. Following Tasioulas, I do not take claimability to be an existence condition for rights.Footnote 10 We need to distinguish claimability from feasibility. The feasibility of correlative duties is an existence condition for rights: the existence of a right requires the existence of correlative duties. In turn, feasibility is a relatively uncontroversial condition on the existence of a duty: as the renowned dictum tells us, ‘ought’ implies ‘can’ – one cannot have a duty to do something one is not able to do.Footnote 11 Feasibility, however, does not entail claimability: sometimes we can tell that a duty is feasible even though it lacks specified duty-holders and content. This could be, for instance, because several agents can fulfil the duty in different manners. But once the distinction is drawn, it is unclear why the fact that correlative duties lack a determinate duty-holder and content should be a problem for the existence of a right. It seems more plausible to think that the specification of the holder and content of the duties is a further matter, one that can be settled after we have established whether a right exists.Footnote 12
II. THE UNIVERSAL SCOPE THESIS
a. The Universal Scope Thesis and an argument by analogy
The Universal Scope Thesis (UST) holds that all individuals hold positive duties correlative to human rights, i.e. that all individuals hold positive duties to protect others’ fundamental interests. Note that the UST is stronger than a thesis simply stating that everyone holds positive duties to protect everyone's fundamental interests. These positive duties could be characterized in other terms: for instance, as duties of charity or beneficence. Describing them in terms of human rights entails they are duties of justice, and duties that provide strong reasons for action.
On a ‘theoretical’ level, this is especially important when duties conflict: for example, in cases where duties to protect the fundamental interests of individuals not related to the duty-holder clash against duties to protect less important interests of – say – their family members or co-citizens. In those circumstances, the former duties are more likely to outweigh the latter if they are stronger duties correlative to human rights, than if they are weaker duties of charity or beneficence. Relatedly, characterizing these duties as correlative to human rights makes a difference on a ‘practical’ level: human rights have become a powerful language shared by a large variety of actors across the globe. They are able to mark the special urgency of the issues they refer to, and resonate with a variety of different individuals and institutions. Framing our obligations to protect very important interests, such as health and freedom from slavery, in terms of human rights can be a very effective way of communicating the importance of the duties to protect these interests, and, in turn, motivating agents to act.
The argument I put forward in support of the Universal Scope Thesis is an argument by analogy. It consists of three steps: first, I point out the similarities between positive and negative duties correlative to human rights; second, I argue that negative duties correlative to human rights are generally thought to have universal scope; third, I claim that, by analogy, this provides a reason to believe that positive duties correlative to human rights have universal scope.
Both positive and negative duties fulfil the same function. To begin with, they both protect fundamental interests. But the analogy holds also at a more fine-grained level. Consider, for instance, the interests protected by so-called welfare rights, such as health and education. These interests are protected by positive duties to provide goods and services, but also by negative duties not to deprive others of these goods and services. Similarly, consider the interests protected by so-called liberty rights: the interests in being free from pain and slavery or to participate in political life. These interests are protected by negative duties not to torture, enslave or violently prevent political participation, but also by positive duties to provide police forces, criminal courts or voting facilities.Footnote 13
The second step consists in pointing out that negative duties correlative to human rights are universal. For instance, it is uncontroversial that everyone has a duty not to torture or enslave anyone else. Each human being's right not to be tortured is correlative to everyone's duty not to torture; similarly, each human being's right not to be enslaved is correlative to everyone's duty not to enslave, and so on.
The third step consists in pointing out that if negative duties are universal, then, by analogy, we have reason to believe that positive duties are too. Positive and negative duties correlative to human rights play the same function. Accordingly, why should we think that they have different scope? If it is everyone's responsibility not to act in certain ways to protect fundamental interests, then plausibly the responsibility for acting to protect these values falls on everyone too.
The burden of proof is now shifted. It is the objector's task to find dissimilarities between positive and negative duties which explain the difference in scope. Note that positive and negative duties differ in many ways, but not all of those are relevant to the question of their scope. For instance, an important putative disanalogy which is not relevant to the issue addressed is the one concerning strength. Many think that acts and omissions have different significance: being causally responsible for a bad state of affairs is worse than letting the same state of affairs occur.Footnote 14 This is compatible with the thesis that everyone holds both positive and negative duties correlative to everyone's human rights: one can hold that there is a difference between acts and omissions and still argue that positive duties correlative to human rights are universal.Footnote 15 Hence, the difference in strength is not a relevant disanalogy for the purpose of this article.
In what follows, I consider three putative disanalogies which are relevant to the argument outlined: the first concerns the feasibility of the duties, the second concerns their demandingness, and the third concerns their relationship with rights.
b. The Feasibility Objection
The first objection concerns feasibility. On one side, the objection runs, it is always feasible to fulfil negative duties: it simply requires refraining from certain kinds of actions, and this is always an available option – an agent can always refrain from torturing, enslaving, and so on. On the other hand, it is not always feasible to fulfil positive duties. The fulfilment of positive duties demands the performance of certain actions, which in turn requires meeting specific material conditions. For instance, to provide food or health services one needs to have resources such as food and medicines, time to distribute them and the ability to reach recipients. But sometimes these conditions are not met.
Moreover, the objection goes, the problem is amplified by the fact that human rights are held by all human beings. This means a duty-holder has to act in order to protect everyone else's fundamental interests. For example, discharging positive duties correlative to human rights would require an agent to have enough food to feed everyone. But, as Onora O'Neill and Henry Shue have pointed out, only exceptionally rich people could provide some assistance to everyone, and surely no one could provide everyone with enough assistance.Footnote 16 But ‘ought’ implies ‘can’ – one cannot have a duty to do something one is not able to do. Since most people do not have the resources to fulfil the positive duties correlative to everyone's human rights, most people cannot hold these duties.
c. The Demandingness Objection
The second objection concerns demandingness. Negative and positive duties, the objection runs, have very different costs. Refraining from performing an action has little or no cost, while performing an action is generally extremely demanding: it requires facing costs with respect to time, material, intellectual and emotional resources, and so on.
Once again, the difference is amplified by the fact that human rights are universal. The claim that negative duties correlative to human rights are universal entails that everyone ought not to perform certain actions against everyone. But this, the objection runs, has little cost for the duty-holder. The case of positive duties is very different. If positive duties are universal, everyone ought to perform certain actions for everyone: we have to give everyone food, education, and so on.Footnote 17 Even if it were possible for one to fulfil the positive duties correlative to everyone's human rights, the objection goes, it would still be excessively demanding for them to do so.
d. The Necessity Objection
The third objection concerns the difference between positive and negative duties with respect to their relationship to the correlative rights. Negative duties, the objection runs, need to be universal for human rights to be respected. But the same is not true for positive duties: in this case, it is not necessary that everyone holds the duty, but only that someone does.Footnote 18
By way of illustration, consider the right not to be tortured, and two of its correlative duties: the negative duty not to torture and the positive duty to support police forces protecting potential victims. The negative duty needs to be held by everyone, for the right to be upheld: only if everyone refrains from torturing can the right-holder's freedom from torture be protected. For positive duties, however, it is only necessary that someone fulfils this obligation. For instance, it is sufficient that co-nationals support police forces protecting potential victims: it is not necessary that everyone does so.
III. AGGREGATIVE DUTIES
These objections should not lead us to abandon the Universal Scope Thesis. Rather, they should push us to improve our understanding of the positive duties correlative to human rights and provide a refined version of the UST. I take up the first task in this part and the following one (sections III and IV) and will move on to the second task in section V.
Our common-sense understanding of responsibility tends to focus on the outcomes deriving from individual actions and omissions, taken in isolation. However, this does not give us the full picture. Some duties can only be adequately grasped when we zoom out and focus on the aggregate effect of the actions (and omissions) of different agents. When I say that some duties can only be ‘adequately grasped’ when we focus on aggregate effects, I mean that their strength, content or existence can only be assessed when we focus on these effects. I call the duties that so depend on aggregate effects ‘aggregative duties’. By contrast, I call the duties whose content, strength and existence only depend on the effect of the action (and omission) of single agents ‘single duties’.
Aggregative duties have featured in several discussions in both moral and political philosophy.Footnote 19 However, they have not been, so to speak, ‘labelled’ as a distinct category of duties. Because of this, and because they might be easily confused with other types of duties arising in collective contexts, it is worth briefly clarifying their nature.
First, the distinction between aggregative and single duties is different from the distinction between collective and individual duties. Collective duties are duties attributed to a collective itself, over and above its members; by contrast, aggregative duties are attributed to each one of the agents contributing, or failing to contribute, to the relevant outcome. Aggregative duties can be attributed to both individual and collective agents (though, as I said above, I here focus on individual agents). Moreover, the existence conditions for aggregative duties are different from the existence conditions for collective duties. As I pointed out in section I, it is plausible to assume that a group can itself be a duty-holder, over and above its members, only if it satisfies the conditions for agency. This is not so for aggregative duties. Rather, aggregative duties obtain whenever the aggregate effect of the actions (and omissions) of different agents is necessary to determine the strength, content or existence of their duties.
To best explain the notion of aggregative duties, let me present a couple of examples. Imagine the following scenario: a lake in a remote location hosts the last breeding population of an endangered species of fish. One day, this lake is discovered by some tourists; the following day, the lake is visited by several more people. However, the tourists are careless, and each one of them drops some litter in the lake. Each piece of litter would individually be enough to make the fish sick, but not to kill them. Yet, the overall amount of litter dropped in the lake is enough to kill the fish, and the fish die.
In the above example, focusing on single agents and actions (be it individual or collective) will not be enough to appreciate adequately the content and strength of the negative duties violated. We might be able to say that the individual agents have violated a duty – that is, a duty not to harm the fish. But we would be unable to characterize the duty as a weightier duty not to contribute to killing the fish. We might then think that there is a collective duty not to kill the fish. But it is not plausible to hold the group of tourists to be a collective agent, since the tourists have no joint commitment, nor do they share a decision-making procedure. So the duty not to kill the fish cannot be held by a collective agent. To account adequately for the content and strength of the duties violated in the above scenario we need to switch focus from single to aggregative duties. Since the actions of the individual agents aggregately result in the death of the fish, we can say that the individual agents have violated an aggregative negative duty not to contribute to killing the fish.Footnote 20
The above example focuses on negative duties, but we can also find aggregative positive duties. Once again, this is best illustrated with an example.Footnote 21 Imagine a man is drowning in the sea; on the nearby beach, ten people are sunbathing. The man could be easily saved by any two people, if one of them pushed the boat in the sea and the other fetched the oars in the shed nearby, and they then rowed to pick the man up. Imagine also that each of the ten people on the beach can easily perform either task. However, none of the people sunbathing does anything to save the drowning man, and the man dies. Here too, focusing on single agents and actions (be it individual or collectives) will not be enough to characterize the duties adequately. On one side, no individual agent can save the man by herself. Since, as I said, one cannot have a duty to do something one is not able to do, no individual agent can have a duty to save the man. On the other side, the conditions for collective agency do not obtain (since the group has no shared decision-making procedure, nor joint commitments), so no collective agent can hold a duty to save the man. To account for the existence of the duties violated in this scenario adequately, we need to switch focus from single to aggregative duties. Since the individuals can aggregately bring about the rescue, and the omissions of the individual agents aggregately fail to prevent the death of the man, we can say that the individual agents have violated a positive duty to contribute to saving the man.
IV. AGGREGATIVE DUTIES AND HUMAN RIGHTS
Recent literature has repeatedly pointed out the relevance of aggregative duties to questions of global justice. In her article ‘Negative Duties, Positive Duties and New Harms’, for instance, Judith Lichtenberg stresses how the aggregate effects of disparate actions are likely to harm others around the globe.Footnote 22
Aggregative duties have also influenced the recent literature on human rights. Elizabeth Ashford, in particular, has argued that (what I call) negative aggregative duties can be correlative to human rights.Footnote 23 Ashford outlines different types of harms that are the effect of agents' aggregative actions. I call these ‘aggregative harms’.
The first type of aggregative harm – additive harm – obtains in cases in which each agent's action, by itself, does not cause a serious harm to any particular victim, but the cumulative behaviour of a set of agents results in an extremely serious harm.Footnote 24 In order to explain how this notion relates to human rights, Ashford presents a version of a thought experiment originally outlined by Parfit.Footnote 25 We are presented with two scenarios. In both cases, there are a very large number of torturers and a very large number of prisoners. In the first case, each torturer causes excruciating pain to one of the prisoners. In the second scenario, each torturer causes only a very small incremental amount of pain to each of the prisoners; however, since there are very many torturers, each prisoner suffers excruciating pain overall. Moreover, in the second scenario the torturers are not intentionally coordinating to cause excruciating pain, yet they could reasonably foresee the result of their actions. Plausibly, the prisoners are subject to a human right violation in both cases, regardless of whether the harm is inflicted by a single agent, or rather brought about by the aggregate effect of the actions of different agents. Additive harms can thus bring about human rights violations, and causing such harms constitutes a violation of aggregative negative duties correlative to human rights.
The second type of harm – multiplicative harm – obtains when an agent's action magnifies the harm caused by other agents. Ashford presents once again two examples. In the first, a factory pollutes a river causing deadly illnesses. In the second, two factories pollute the same river. By themselves, both pollutants are harmful but not very dangerous; together, however, they are deadly. Ashford plausibly argues that if it is a human rights violation for the factory in the first scenario to bring about such harms, then the same is true for the factories in the second scenario. Multiplicative harms can thus bring about human rights violations, and causing such harms constitutes a violation of aggregative negative duties correlative to human rights.
While the notion of aggregative duties has been employed when discussing negative correlative duties, to my knowledge this framework has not been applied when discussing positive duties correlative to human rights. This is what I do in the rest of section IV. As we will see, extending this framework to positive duties allows us to identify different ways in which fundamental interests can be protected and clarify the normative implication of those opportunities.
The taxonomy outlined by Ashford provides a starting point for understanding the aggregative positive duties correlative to human rights. Inverting Ashford's taxonomy of harms, we can identify two types of aggregative benefits: additive and multiplicative benefits. The first type of benefit – additive benefit – obtains in cases in which each agent's action, by itself, does not cause a significant benefit for any particular beneficiary, but the cumulative behaviour of a set of agents results in a significant benefit. The second type of benefit – multiplicative benefit – obtains when the agent's action magnifies the benefit caused by other agents.Footnote 26
The two aggregative benefits above are the mirror images of the aggregative harms outlined by Ashford. But they do not exhaust the types of aggregative benefits and corresponding positive duties. Consider, for instance, the case of the drowning man described above, in which two people are needed to rescue the man. In this case, without both contributions, there would be no benefit at all: this is different from cases in which the action of one duty-holder adds to or magnifies the benefit caused by another duty-holder. We should thus introduce a third category, to include cases in which a number of different contributions are needed to bring the benefit about. Call these combination benefits.Footnote 27
With this taxonomy in hand, we can now consider the case of human rights and fundamental interests. Consider ways in which fundamental interests can be protected through additive benefits: good examples of these kinds of cases concern benefits which come from small individual donations, or consumption of fair trade products. While a single donation of, for instance, a few pounds might not cause a significant benefit for any particular beneficiary, the cumulative effect of many donations can be expected to do so.
Similarly, there are ways in which fundamental interests can be protected through multiplicative benefits. Consider, for instance, a scenario in which some agents provide health infrastructures and other agents improve road infrastructures, so that the latter action magnifies the benefit brought about by the former.
Finally, a good example of combination benefits is the spreading of knowledge: disseminating knowledge about human rights violations, for instance, can enable others to vote, consume or donate so as to protect fundamental interests. These are cases of combination benefits since different types of contributions are needed to bring about the benefit.
There are several reasons why focusing on aggregative positive duties is especially important for the protection of fundamental interests. First, many fundamental interests can only be protected through complex goods and services, such as schools and hospitals. The resources needed to provide such goods and services can often only be acquired through the contributions of multiple agents. In this respect, then, additive benefits are extremely relevant to the protection of fundamental interests.
Second, acquiring the needed resources and providing these goods and services are likely to require complex causal chains. A single agent might not be able to initiate and complete them. However, it is often possible for an agent to take up one of the steps in such chains. So, for instance, increasing foreign aid might require changing a government's budget priority, which, under democratic institutions, would require citizens' attention to the problem. Hence, an agent might not be able to directly affect government allocation of funds to foreign aid, but she can engage in political campaigning to raise awareness of the issue. This is an example of combination benefit since different types of contributions are needed to bring about the benefit.
Third, globalization has increased the way in which agents can, and do, interact. What we buy and consume, the way we vote, how often we travel, where to, and so on: these actions can have effects at a global level. Much literature has focused on the harmful effects of our actions, and the negative duties deriving from those. But this is only part of the picture: the increased global interaction also provides us with new ways of benefiting others. We should also be aware of the positive effects of our potential actions, and the positive duties deriving from those.
Once we add aggregative duties to the picture, duties correlative to human rights can be understood as duties to act to contribute to the protection of fundamental interests. This allows us to refine the UST as follows:
Universal Scope Thesis: every agent has a duty correlative to human rights to act to contribute to the protection of everyone's fundamental interests.Footnote 28
Here, a clarification is in order. Duties can be specified at different levels of abstraction: for example, a duty to contribute to the prevention of life-threatening diseases is more abstract than a duty to provide malaria nets. The UST is a thesis about abstract duties: it tells us we have a duty to contribute, but it does not specify how. There are many ways in which one can contribute to bringing about this outcome. Hence, the UST about abstract duties is compatible with different allocations of concrete duties. Which allocation is to be preferred will depend on both empirical and moral facts. On one side, the relevant empirical facts will include the nature of the specific problems addressed and their solutions. On the other side, the relevant moral facts will depend on the principles that regulate the allocation of those duties. Those are likely to include considerations such as ability to effectively protect fundamental interest, cost, how well-off the potential duty-holders are, and so on.Footnote 29
The content of the moral and empirical facts that determine the allocation of concrete duties is outside the scope of this article. I thus won't include a detailed discussion of how duties should be allocated at the concrete level. Even setting aside the specifics, however, we can distinguish three general strategies. The first is to assign different duty-holders the responsibility for different right-holders. In this case, each duty-holder (or group of duty-holders) is responsible for a different group of right-holders, so that, together, the duty-holders will ensure that all the right-holders' fundamental interests are protected. Alternatively, it is also possible to assign different duty-holders the responsibility for different tasks. In this case, each duty-holder (or group of duty holders) performs actions that partly protect all right-holders, so that, together, the actions performed by all duty-holders will result in the full protection of everyone's fundamental interests. Finally, it is possible to endorse a mixed account, in which both types of distributions figure.
This line of reasoning points to a possible concern about the UST. The relevant moral and empirical facts might entail that the best way to discharge the UST is to assign different duty-holders the responsibility for different right-holders. For instance, the ability to aid others and the cost of doing so might strongly correlate with physical proximity, or the existence of shared political institutions. This could mean that, at a concrete level, each of us is responsible for aiding those near her, or with whom she shares political institutions. Accordingly, the UST might hold at an abstract level, but it would not make much of a difference in concrete terms. In fact, the UST would collapse into a different thesis, one we might call a ‘Limited Scope Thesis’ (LST).
Two points are worth stressing in reply to this worry. First, it is not far-fetched to imagine that, in at least some cases, all duty-holders would have a duty to act in a way that protects the fundamental interests of all right-holders. For instance, given the global nature of many health risks, everyone might be required to support an international institution preventing global pandemics. Moreover, even if the LST were the correct way to allocate concrete duties, it would still be important to hold the UST for abstract duties. First, even if the LST were the best way to discharge abstract duties given the current empirical facts, it might not be so in the future. Circumstances change: technological, political and economic factors, among other things, might change the nature of threats and possible ways of protecting fundamental interests. Even if the LST presently held true with respect to concrete duties, this might no longer be the case in the future.
Furthermore, endorsing the UST for abstract duties will be significant in situations of less than full compliance. If all agents have a correlative abstract duty to contribute to the protection of everyone's fundamental interests, when some agents fail to fulfil their duties, other agents have a duty to deal with the effects of non-compliance, to ensure everyone's fundamental interests are protected. This duty will be a correlative duty of justice, and thus extremely strong. On the other side, if the LST is endorsed at the abstract level, agents only have a correlative duty to protect the fundamental interests of a limited number of right-holders. Supporting the victims of non-compliance would not be required by a duty correlative to a human right. Offering such support might be required by another type of obligation (for instance, a duty of beneficence), but this is likely to be weaker than a duty correlative to a human right and more likely to be outweighed in case of conflict.
The significance of the difference is best shown by an analogy. Imagine the following two scenarios. In the first case Hans and Katie have two children, Tom and Rob, who, let's assume, have a right to develop their musical and sporting talents. This requires that they be driven to music and sport lessons. It is plausible to say that Hans and Katie both have an abstract duty to contribute to the protection of their children's interests. Assume that the concrete distribution of duties is such that Hans ought to drive Tom, and Katie ought to drive Rob. However, one day Katie stops fulfilling her duties. It then seems that Hans has a concrete duty to protect both Tom's and Rob's interests. The content of the concrete duty will depend on circumstances: if Hans can ensure Katie fulfils her duty, he might be required to do so; alternatively, he might be required to drive both children to both lessons, or to drive both to one lesson each. Importantly, he has a duty to deal with the effect of non-compliance, and this duty is correlative to the children's rights. Consider now a second scenario. Hans and Katie are not a couple and do not have children together – Hans is Tom's father and Katie is Rob's mother. Once again, the concrete distribution of duties is such that Hans ought to drive Tom and Katie ought to drive Rob, but this time Hans and Katie do not have abstract duties correlative to both children's rights. Imagine again Katie stops fulfilling her duty. In this second scenario Hans does not have a correlative duty to help Rob. He might have a weaker duty of beneficence to do so, but if the duty to help Rob conflicts with the duty to help Tom, the latter will take precedence over the former. Similarly, when we consider the scope of duties correlative to human rights, holding the UST for abstract duties and the LST for concrete ones will provide stronger protection for victims of non-compliance than endorsing the LST as a thesis about abstract duties.
V. REPLYING TO THE OBJECTIONS
a. Replying to the Feasibility Objection
We can now reconsider the objections presented in section II. Let us start with the Feasibility Objection. Recall that this objection holds that there is a disanalogy between positive and negative duties correlative to human rights, since it is feasible for everyone to uphold the negative duties, but not the positive ones. Once we include aggregative duties in the picture, we can see how this objection is misguided. The problem with the Feasibility Objection is that it unduly restricts the focus on single duties, ignoring aggregative ones. In so doing, it ignores an important way in which the protection of everyone's human rights can be achieved. While it is true that no one can provide services, help and benefits for all others by herself, many can contribute to the achievement of this outcome. Hence, the Feasibility Objection should not lead us to abandon the UST.
Yet, this objection still carries a valuable insight: even considering aggregative positive duties, we will still encounter agents who are unable to protect others’ fundamental interests in any way (for instance, people who live in severe destitution, or suffer severe disabilities). We thus need to refine the UST, with the addition of the following clause:
Universal Scope Thesis: every agent has a duty correlative to human rights to act to contribute to the protection of everyone's fundamental interests, except those for whom it is unfeasible to do so.
However, the point is different from the one pressed by the Feasibility Objection: the objection aimed to show that no one, or very few people, are able to fulfil all or some of the positive duties correlative to everyone's human rights. Extending our understanding of positive duties so as to include aggregative ones allows us to show that most people will be able to uphold such duties. The exclusionary clause is thus reversed: the point is not that some fabulously wealthy people might be able to uphold these duties, but rather that some specific agents might be unable to.
b. Replying to the Demandingness Objection
Recall that this objection holds that there is a disanalogy between positive and negative duties correlative to human rights, since it would be too costly for everyone to uphold the positive duties correlative to everyone's human rights, while it is not too costly to uphold the negative ones.
To reply to this objection, it helps to distinguish two issues: whether there are limits to the costs that moral requirements can place on us; and, if so, what those limits are. I assume that considerations of costs can constrain the existence and content of a duty, because they affect what is reasonable to require from a duty-holder.Footnote 30 For instance, it is reasonable to require that I provide others with assistance if I can do so at the cost of some non-essential good, but less so if I can only do so at the cost of my arm, or my own life.
With respect to the second question, it is hard to find a determinate threshold after which costs are ‘too high’.Footnote 31 The proposals for identifying this threshold vary hugely: Singer, for instance, takes it to be the point at which one would have to sacrifice something of significant moral importance.Footnote 32 Murphy sets the threshold at what we would be required to sacrifice in a situation of full compliance.Footnote 33 For the purposes of this article, I will not argue for any specific threshold, but I will assume that one such threshold does exist. I employ the term ‘unreasonably costly’ as a placeholder, and assume that duty-holders do not have a duty to act when the cost of doing so would be above the relevant threshold.
With these clarifications in hand, we can further refine the UST as follows:
Universal Scope Thesis: every agent has a duty correlative to human rights to act to contribute to the protection of everyone's fundamental interests, except those for whom it is unfeasible or unreasonably costly to do so.
Once this clarification is in place, we can address the demandingness objection and consider whether the objection holds once we include aggregative duties in the picture. As Lichtenberg points out, the opportunity cost of fulfilling negative aggregative duties can be extremely high: aggregative negative duties require us to change a very large variety of well entrenched behaviours, including, among others, what we buy, what we eat, and the means of transportation we use.Footnote 34 By contrast, once we spell out the different ways in which we can protect fundamental interests through aggregative benefits, there is no reason to think that positive duties should be especially costly. Aggregative positive duties require us to perform actions such as donating, voting, volunteering, buying fair trade, and so on, which need not be especially demanding. Once we take into account both the cost of omitting and performing the relevant actions, and the opportunity costs involved in upholding the duties, there is no reason to think that positive aggregative duties correlative to human rights should be more demanding than negative ones.
Of course, while aggregative positive duties need not be very costly, they can be – depending on how much we donate, how long we volunteer for, and so on, these actions can indeed be very demanding. However, this worry concerns the content, rather than the existence, of positive duties. Donating a high percentage of one's income, or volunteering most of one's spare time, are very costly activities. But this does not imply that there is no duty to donate or volunteer. Rather, it suggests that the ‘reasonable cost condition’ constrains the content, as well as the scope, of positive duties. So, for instance, we can say that there is a duty to donate a reasonable percentage of one's income, or volunteer a reasonable percentage of one's time. This is of course compatible with the UST. Moreover, there is no reason to think that this restriction would pertain only, or especially, to positive duties, since, as we have seen, aggregative negative duties can also be extremely demanding.
c. Replying to the Necessity Objection
The Necessity Objection holds that there is a disanalogy between negative and positive duties with respect to their relation to rights: while negative duties need to be held by everyone for a right to be respected, this is not so for positive duties; in the case of positive duties it is only necessary that someone holds such duties.
In reply to this objection, I want to stress two points. First of all, taking the largest possible set of duty-holders for abstract duties allows us to distribute costs more fairly. Once again, an example might help clarify this point. Let's go back to the case above: Hans and Katie have two children, Tom and Rob, who have a right to develop their musical and sporting talents, which requires that they be driven to music and sport lessons. Assume now each parent can drive both children to music and sport lessons, and that Hans and Katie are similar in all the relevant respects (for instance, they are equally able to drive the children, and it would be equally costly for them to do so). Consider two ways of thinking about Katie's and Hans's duties: according to the first, all parents have a duty to attend to the children's development; according to the second, this duty only falls on mothers. If we endorse the second thesis, then the duty to drive the children will fall on Katie. If we endorse the first, then, plausibly, one parent will drive one child while the other will drive the other child. The children's rights are respected in both scenarios; however, the second scenario is arguably fairer since it results in a more equitable allocation of burdens. Similarly, even though abstract positive duties correlative of human rights do not need to be held by everyone for the rights to be upheld, the UST results in a fairer allocation of burdens than a narrower allocation of duties.
Moreover, taking positive duties to have universal scope can itself extend the content of the rights. For a right to exist, it has to have correlative duties; in turn, for a duty to exist, it has to be feasible, since ‘ought’ implies ‘can’. Including everyone in the set of duty-holders allows us to consider the aggregative benefits deriving from everyone's contribution, and it thus places the lowest possible set of feasibility constraints on the existence of human rights. This, in turn, will extend the ability to protect fundamental interests and the content of the rights.
VI. COMPARING UNIVERSAL AND LIMITED SCOPE THESES
To highlight the significance of the UST it is helpful to compare it to a version of what I call the ‘Limited Scope Thesis’. Here I consider Wenar's ‘Least Cost Principle’.Footnote 35 In his article ‘Responsibility and Severe Poverty’ Wenar investigates who is morally responsible for alleviating severe poverty. He offers a general account of the allocation of responsibility to prevent threats to basic well-being, and then applies such an account to the case of severe poverty.Footnote 36
He holds that the responsibility to prevent threats to basic well-being is allocated according to the Least Cost Principle, according to which the responsibility for averting harms should be located in the agent who can do so at the least cost. He also adds that the principle can apply either directly or indirectly: in cases of one-off threats, the principle is applied directly to individuals; for instance, if a child is drowning in a pond, the responsibility to help that child falls on the person who could most easily save the child. But when predictable threats are at stake, the principle is applied through roles, which are created on the basis of general descriptions. For instance, the responsibility to care for a child is given to his parents, and the responsibility to avert car accidents is given to trailing drivers, etc. In order to deal with serious threats, Wenar argues, we construct systems of roles. For example, parents ought to feed their children, and officials ought to ensure parents are economically able to feed their children. The principle includes a stepping-back proviso: if the person with the original responsibility is unwilling or unable to fulfil his duty, the responsibility is allocated to the person besides him who can most easily bear the burden. Wenar's principle also includes an excessive burden proviso, according to which agents are only required to bear the cost if the cost is not excessively burdensome.Footnote 37 Following this framework he claims that the duties to avert the threat of severe poverty should be allocated in the first instance to the agent himself and, if the agent cannot take care of himself, they should ‘step back’, progressively, to his family, his community, his national government and, finally, the international community.
Wenar's account includes ‘systems of role’ and, in so doing, it includes some aggregative duties deriving from the interaction of sets of agents. However, it also leaves out all the other aggregative duties we can include when considering the interaction of all agents. As I said when replying to the Necessity Objection above, extending the picture to include all the aggregative duties has two advantages. First, it allows us to extend the content of human rights. For instance, Wenar holds that the duty to avert the threat of severe poverty is allocated in the first instance to the agent himself and, if the agent cannot take care of himself, it ‘steps back’, progressively, to his family, his community, his national government and, finally, the international community. But by focusing on these agents in turn, he does not include the benefits which might derive from the interaction of these categories of agents. Taking these benefits into account will extend the set of benefits included and, in turn, the number of fundamental interests that it is feasible to protect.
Second, including all duty-holders allows us to distribute costs more fairly. For instance, consider one-off threats. Imagine a child is drowning and several people are on the beach. Each of them can perform the rescue at some cost. One of them is very strong, and would be able to individually perform the rescue at the least cost. According to the Least Cost Principle, the child should be saved by this person. But imagine further that the rescue can also be performed so that each beachgoer contributes to it, in such a way that each of them bears a very small burden. It seems that the latter option would lead to a fairer allocation of burdens. The same can be said in the case of predictable threats. The Least Cost Principle suggests that we pick the general description (or a limited set of general descriptions) under which it would be least costly to protect the value. However, we could also distribute duties in such a way that costs are distributed across all agents. Doing so is preferable because it leads to a fairer allocation of burdens.
Wenar's thesis also provides too simplified an account of the considerations that determine the allocation of concrete duties. While the article refers to the Least Cost Principle as a unique principle, in practice, he justifies his account by reference to three different considerations: cost, ability to protect fundamental interest, and how well-off duty-holders are. Considerations of cost underpin comments such as ‘Each competent adult is responsible for avoiding a great many threats to his own well-being because he is the agent who can do so at the least cost’.Footnote 38 Concern with ability to protect interests underpins comments like ‘we hold the trailing driver responsible for an accident [when he] is the person who can most easily keep the collision from occurring’.Footnote 39 Concern with how well-off duty-holders are underpins claims like ‘we think that the fairest and most efficient way to allocate responsibility . . . is to put the burden on the agent who can most easily bear that burden’.Footnote 40 However, distinguishing among these principles is important – because these principles can conflict.
This conflict directly bears on Wenar's thesis about responsibilities for world poverty. Recall that he claims that the duties to avert the threat of severe poverty are allocated in the first instance to the agent himself and, if the agent cannot take care of himself, they ‘step back’, progressively, to his family, his community, his national government and, finally, the international community. This claim might be plausible if we focus on costs. For instance, it might be plausible to say that these are the agents for whom it would be least costly to perform the action, because of physical and emotional closeness. However, it is not plausible to say that this allocation adequately accounts for duty-holders’ different levels of well-being. Locating responsibilities with these agents would be in line with this concern only if resources were equally distributed across families, communities and countries. But, needless to say, this is far from being the case: the current economic scenario is characterized by a high degree of inequality among households, regions and countries. Given this indubitable empirical fact, and the related disparities in power, it is also unlikely that the proposed allocation identifies agents that are best able to protect basic well-being. Setting aside the specifics of Wenar's thesis, this example highlights that the correct allocation of concrete duties does not rely on a unique principle, but rather requires identifying and weighting a variety of distinct moral considerations.
VII. CONCLUSION
In this article I defended a refinement of the thesis that positive duties correlative to human rights are universal, which I called Universal Scope Thesis. To do so, I put forward an argument by analogy. I argued that positive and negative duties correlative to human rights are analogous in so far as they both perform the same function; that negative duties correlative to human rights are generally thought to have universal scope; and that, by analogy, this provides a reason to believe that positive duties correlative to human rights have universal scope. I then outlined what I took to be the most pertinent disanalogies: one concerning feasibility, one concerning demandingness, and one concerning the relation between rights and duties. To address these worries, I focused on the nature of positive duties correlative to human rights and argued that an adequate understanding of these duties ought to include aggregative duties. With this framework in hand, I then addressed the objections. In the process of doing so, I also refined the Universal Scope Thesis.
APPENDIX
O'Neill worries that rights that can't be directed against some particular individual ‘may seem bitter mockery to the poor and needy’.Footnote 41 I disagree: framing the protection of fundamental interests in terms of human rights can have substantial implications, both in theory and in practice. I explain why in section II.
Tomalty objects to Tasioulas's reply on two grounds: first, that justifying the object of a right requires specifying duty-holders, so that ‘even claims that are held against all moral agents . . . presuppose that it is reasonable to require all moral agents to fulfil the relevant duty’.Footnote 42 Second, that without a specified duty-holder, rights do not play a normative role distinct from the one played by their grounding interests.Footnote 43 I do not find those replies compelling. I agree that it is necessary to establish whether it is ‘reasonable’ to allocate a duty to a certain agent – I spell out this intuition in terms of ‘feasibility’ and ‘demandingness’. However, when these requirements can be met by several agents, it is possible to determine whether they are met without specifying the identity of the duty-holders. Second, the normative role played by the right is to entail the existence of a correlative duty. This goes beyond the merely evaluative nature of interests, and thus plays a distinctive role in normative reasoning.
A concern related to claimability is that duties correlative to rights must be ‘perfect’. The distinction between perfect and imperfect duties has been drawn in several different ways.Footnote 44 On some of them, imperfect duties require duty-holders to exercise discretion in selecting their beneficiaries. Starting from this type of account, Hope has recently argued against theories that reject claimability as an existence condition of human rights, on the basis that they counterintuitively deny that failing to enact a correlative duty wrongs the right-holder.Footnote 45 However, denying that claimability is an existence condition of rights need not entail that correlative duties require discretion over beneficiaries. It can simply entail that it is possible to spell out correlative duties at a level of abstraction that omits this feature.Footnote 46 Hope's argument does not, therefore, challenge this article's assumptions.Footnote 47
