Assisting the Assisters

The recent withdrawal of Western states from Afghanistan saw widespread endorsement of the view that these states owe especially stringent duties of rescue to Afghans who assisted their armed forces during the 20 years of Western intervention and occupation.1 Although concern has been expressed for those who worked with Western states quite broadly, it is the treatment of those who directly assisted Western forces—for example, by working alongside troops as translators or interpreters—that has received the lion's share of attention and attracted the most fervent criticism.2 When I refer to “assisters” to Western forces, then, I am referring to Afghans who provided Western forces with this kind of direct support: those who stood “side by side with those on the frontline of combat.”3 The claim that assisters are owed especially stringent duties of rescue has largely focused on a particular type of a more stringent duty to rescue. This duty holds that when rescuers must choose between victims, they should prioritize saving particular victims—such as assisters—over other victims, even if those other victims face similar threats of harm and saving them is equally costly. Political leaders, commentators and military personnel, including many veterans, routinely identified evacuating assisters as a priority during the fall of Kabul, for example, despite evidence that other Afghans, such as judges, teachers, journalists and politicians, face similar threats of harm under the resurrected Taliban regime and evacuating members of those groups would have been no costlier.4 This prioritization of assisters is also reflected by states' practices. The United States has a longstanding Special Immigrant Visa (SIV) program specifically for translators and interpreters. A separate “Priority 2” scheme for those employed by or on behalf of the United States was introduced in August 2021. There is no scheme aimed at those who took on dangerous public roles, such as journalists, teachers or judges. Evacuation efforts in the period leading up to the withdrawal deadline focused on those eligible for the SIV program.5 Likewise, the UK's Afghan Relocation and Assistance Program was introduced in April 2021 to help Afghans “who risked their lives supporting British troops” by providing linguistic services for the military, where these “exposed meaningful enabling roles [...] made a material difference to the delivery of the UK mission in Afghanistan [and] exposed [them] to public recognition in performance of their role, leaving them now at risk.”6 By contrast, the Afghan Citizens Resettlement Scheme for those “who have assisted the UK efforts in Afghanistan and stood up for values such as democracy, women's rights, freedom of speech, and rule of law” was not announced until January 2022—that is, several months after the withdrawal of Western forces.7 Commentators frequently invoke the idea of a moral duty to aid assisters, although the implication that Western states have duties to aid only assisters seems like an uncharitable interpretation of the popular view.8 After all, most people believe that we have duties to aid others when we can do so without incurring undue risk to ourselves or imposing undue risk on others. The more charitable interpretation, and more plausible view, then, is that while Western states have at least prima facie duties to aid all Afghans threatened by the Taliban, they ought to prioritize helping assisters over other Afghans.9 The popular view also suggests that Western states should be prepared to incur greater risks for the sake of rescuing assisters compared to other refugees.10 The importance of rescuing assisters, specifically, was invoked in arguments to extend the evacuation period at the end of the war despite the increasingly high risks of lethal harm to combatants from militia groups (as evidenced by the bombing of Kabul airport) and the Taliban's threat of “consequences” for remaining combatants if the United Sates exceeded the agreed deadline for withdrawal.11 Widespread claims that Western forces should nonetheless remain in order to rescue assisters suggest that, according to the popular view, these forces can be expected to incur risks for the sake of assisters that they need not incur for the sake of other refugees. The popular view is typically defended by appeal to one or both of the following features, each of which is plausibly morally significant: (a) assisters risked their lives to assist Western forces and are now endangered as a result12 and (b) assisters were promised protection by Western states.13 Proponents of the popular view argue that these features confer gratitudinal and/or promissory duties on Western states with respect to Afghan assisters and that these duties in turn ground especially stringent duties of rescue to those assisters.14 Assister Obligations: Other things being equal, when a refugee rendered a belligerent's forces assistance during a war, belligerents are morally obliged to (i) incur greater risks to rescue that refugee than they are obliged to incur to rescue other refugees and (ii) prioritise rescuing that refugee over rescuing other refugees, if the refugee is endangered as a result of providing assistance and/or was promised assistance by the belligerent. Section II develops the idea that when someone is endangered as a result of assisting a rescuer, the rescuer can thereby incur a gratitudinal duty to suffer ordinarily supererogatory risks for the assister's sake. Section III argues that this duty can, in turn, ground a duty to prioritize saving the assister over other victims in some rescue cases. However, Section IV argues that gratitudinal duties fail to vindicate Assister Obligations and, hence, the popular view. Section V argues that although promises can ground especially stringent duties to rescue, promissory duties also fail to vindicate Assister Obligations. Section IV argues that, given the plausible assumption that the invasion of Afghanistan was an unjustified war, Western states' duties to Afghan assisters are not best explained by gratitudinal or promissory duties, nor by adverting to special relationships between assisters and combatants. Rather, as Section VII argues, unjustified belligerents' duties to assisters are best explained by the belligerents' duties not to harm. Since these duties also ground especially stringent duties of rescue to many non-assisting refugees, it is false that Western states owe more stringent duties to Afghan assisters than to other Afghan refugees. Section VIII concludes that we must reject the popular view of Western states' duties to Afghan assisters. Rescue: Adam is being chased by a shark. Bob can save Adam's life only at a moderate risk of having his arm bitten off. Saving Adam is supererogatory. Bob nevertheless attempts the rescue. As Bob successfully hauls Adam aboard his boat, Bob falls into the water. Adam can save Bob's life only at a moderate risk of having his arm bitten off. This duty to save Bob is one aspect of what seems like a broader liability on Adam's part to suffer costs for Bob's sake that he need not suffer for others. To say that Adam is liable to these costs is to say that requiring Adam to incur them, or imposing them upon him, does not wrong him. For example, imagine that Bob can save himself from the shark only by throwing a grenade that risks collaterally harming Adam. It seems permissible for Bob to impose greater risks of collateral harm on Adam than he may impose on an unrelated thirdparty. (Again, it is the asymmetry that matters here, rather than precisely how much more risk Bob may impose on Adam.) Correlatively, it would wrong Bob if Adam were to try to prevent Bob's throwing the grenade in order to avoid bearing this cost. One possible explanation of Adam's more stringent duty in Rescue is that Adam will be made no worse off if he loses his arm in the course of saving Bob than he would have been had Bob not saved him from the shark.18 But this fact cannot explain Adam's having a more stringent duty to help Bob in particular, since it would also obtain if Adam were to enact a similarly risky rescue of a stranger. And yet, plausibly, Adam does not have a duty to incur an ordinarily supererogatory risk of losing his arm in order to save a stranger's life. Friends of the popular view suggest a more promising explanation—namely, that Adam owes Bob a duty of gratitude that manifests as a duty to suffer extra risks for Bob's sake.19 On some accounts, gratitude is owed only when benefactors do something supererogatory—that is, only if they assist us in a way that goes beyond what is morally required of them.20 I suspect that many proponents of the popular view believe that Afghan assisters typically did act supererogatorily in assisting Western forces. The work itself was often dangerous—many assisters were seriously injured while working—and the risk of reprisals was significant even during the occupation. And, as above, Afghan assisters are now endangered as a result of providing this assistance. However, it's plausible that we can also owe gratitude for at least some obligatory actions, such as those that require the agent to incur a significant cost, or a cost approaching the threshold of what can be demanded by the relevant duty.21 If we also owe gratitudinal duties to at least some of those who discharge costly duties to assist us, then the number of assisters owed more stringent duties of rescue will be correspondingly greater. It is widely held that while beneficiaries can owe duties of gratitude, fulfillment of these duties cannot be demanded by the benefactor.22 However, those defending the popular view by appealing to gratitudinal duties certainly seem to believe that such duties may be demanded. My aim is to show that even if gratitudinal duties do support especially stringent duties to rescue assisters, this will not vindicate the popular view of Western states' duties to Afghan assisters. Hence, I will simply grant that duties of gratitude can be demanded in the way the popular view supposes. Rescue is a no conflict case of rescue—that is, a case in which the prospective rescuer can save all those in peril. Hence, while our verdict in this case supports the first clause of Assister Obligations—that rescuers can be obliged to incur greater risks for the sake of assisters—it does not obviously support the second clause that rescuers must prioritize saving assisters over saving other refugees. To assess the plausibility of that claim, we need to consider conflict cases of rescue—that is, cases in which a rescuer must choose between victims. Stranger Rescue: Adam can save either Bella or Clare, but not both, from being eaten by a shark. Both are strangers to Adam. Adam can save Bella at a tiny risk of the loss of his finger. Saving Bella therefore falls within the scope of Adam's usual duty to rescue. He can save Clare only at a moderate risk of the loss of his arm. Saving Clare is therefore supererogatory. Here, then, the fact that Adam can save Bella at an obligatory risk means that Adam has a duty to save someone. But it doesn't mean that he has a duty to save Bella. Rather, Adam's all-things-considered duty is a duty to offer Bella at least an equal chance of rescue. Adam's prima facie duty to risk losing a finger for Bella's sake is compatible with his nonetheless permissibly failing to incur that risk for Bella's sake. Busy Rescue: As Rescue, except that Clare, a stranger, has also fallen off her boat. Adam can save either Bob or Clare, but not both, from the shark only at a moderate risk of having his arm bitten off. I argued above that, by the lights of the gratitudinal account, saving Bob is not supererogatory for Adam in Rescue: rather, Adam is required to incur this risk for Bob's sake. It is hard to see how Clare's presence could affect the degree of risk that Adam has a duty to incur for Bob's sake, since that seems to depend on facts about the relationship between Adam and Bob. Given this, it looks as if Adam has at least a prima facie duty in Busy Rescue to incur the risk of the rescue for Bob's sake but not for Clare's. In this respect, the case looks like Stranger Rescue. If, however, Adam is required to save Bob in Busy Rescue—and not just give him an equal chance of rescue—then the duty he owes Bob cannot be of a kind with the duty that he owes Bella in Stranger Rescue. And this seems right. The ground of any duty to save in Busy Rescue is Bob's previous assistance of Adam. It would be odd if Adam's gratitudinal duty to Bob made it obligatory for Adam to rescue someone at an ordinarily supererogatory risk, but not obligatory to rescue Bob. Saving Clare is not an appropriate means of discharging a gratitudinal duty to Bob. It would also be odd if Adam were obliged, or even permitted, to offer Bob and Clare equal chances of rescue. Adam's gratitudinal duty to Bob can hardly give Clare a claim to an equal chance of rescue. And if Adam has a duty to save Bob in Rescue, it's unclear how Clare's presence could permit Adam to decrease Bob's chances of rescue in Busy Rescue, given that Clare has no claim to be saved. If this reasoning is sound, then it reveals one type of case in which rescuers are required to prioritize saving assisters over other victims—namely, when saving the other victims is supererogatory. When Clare has no prima facie claim to be saved, Adam must save Bob.24 On this view, it can be impermissible to save Clare in cases where doing so would normally be supererogatory (and therefore permissible). Hence, even though failing to save Clare does not violate any claim of hers, a prohibition on saving her nonetheless makes her worse off. I do not think that Clare has a complaint against being made so worse off. We are sometimes permitted to do things that make others worse off by rendering it impermissible to save them. Nevertheless, reflecting on Clare's position does, I think, give us reason to prefer the account of gratitudinal duties that I have developed here—namely, that these are duties to do more for the subject of that duty than one must ordinarily do for other people—to the alternative view that gratitudinal duties require us to generally give more weight to the subject's interests. Easy Rescue: As Busy Rescue, except that Adam can pull either Bob or Clare, but not both, aboard his boat only at a tiny risk of having his finger bitten off. By contrast, Adam's duty to save in Easy Rescue is explained by his usual duty to rescue, not by his having a duty to incur extra risks for Bob's sake. Adam owes both Bob and Clare prima facie duties of rescue independently of Bob's earlier assistance and, thus, owes them equal chances of rescue. The account of gratitudinal duties developed here does not support prioritizing assisters when the rescuer can save either of two victims at an obligatory risk. It therefore allows us to recognize gratitudinal duties, granting them genuine weight, but without undermining other people's claims against us. This coheres with our broader intuitions about how we may discharge our gratitudinal duties. For example, no matter how keen Amy is to give her benefactor Bree a thank-you gift, she may not do so by stealing from Ciara. Ciara's claim that Amy not steal from her simply blocks Amy's choosing to discharge her duty in this way. But Amy is permitted (and perhaps required) to refrain from buying Ciara a lavish gift that she had intended to buy her in order to buy Bree a thank-you gift instead. Doing so is permissible, even though it makes Ciara worse off, because, and insofar as, Ciara has no claim to the lavish gift. The alternative view implies that, in light of his gratitudinal duty to Bob, Adam no longer owes it to Clare to give her an equal chance of rescue, since he must weigh Bob's interests more heavily. We should resist the idea that Bob's earlier saving of Adam can undercut Clare's claim in this way. This is not to say that we can never weaken other people's claims to be saved by us. For example, I argue elsewhere that Rescuer may contract with Alice to serve as Alice's bodyguard provided that Rescuer is not under a duty to rescue anyone else at the time of forming the contract. This contract can make it impermissible for Rescuer to offer equal chances of rescue when both Alice and Betty are imperiled even if she would otherwise be required to do so.25 But we have good reason to limit the ways in which we can so undermine our duties to others. The agreements that I describe, and their concomitant effects on others' claims, are justified by the moral importance of allowing individuals to pursue their own ends and projects. Allowing people some limited scope to make these kinds of agreements seems like an integral part of allowing agents to live autonomous lives. Prohibiting such agreements would thus pose a significant obstacle to living such lives. Gratitudinal duties are not like this: they are not, in general, duties that we take on voluntarily as part of helping us to pursue our autonomously-chosen projects and On the gratitudinal duties are and our scope for our own projects and This is such duties obtain only when are that the benefactor that the is by the an appropriate of the in an to someone under an to We then, Clare's claim to be saved by to the importance of allowing Adam some scope to pursue his own does Bob seem to have a claim that he be to Clare's claim in this doing so is not an integral part of Bob's being to live an autonomous life. By the account of gratitudinal duties that I have we avoid this conflict with Clare's In Easy Rescue, Adam is with two people he can save from similar harm at obligatory risk. The gratitudinal account I have developed does not undermine the that he must offer each an equal chance of rescue. Rescue: As Busy Rescue, except that saving Bob will a moderate risk of Adam's losing an arm and saving Clare will a tiny risk of Adam's losing a finger. may Adam simply save Bob. in Busy Rescue, where Adam can save Clare only at supererogatory risk and thus even a prima facie duty to save Adam can save Clare at obligatory risk. Bob's claim that Adam incur extra risks for his sake does not Clare's claim that Adam give her an equal chance of rescue when the risk of saving her falls within the scope of Adam's usual duty to rescue. The previous two argued that gratitudinal duties can plausibly require rescuers to (i) save assisters at ordinarily supererogatory risk to (ii) save assisters rather than other victims when saving the other victims is and give assisters an equal chance of rescue when saving them, but not the other victims, ordinarily supererogatory However, a duty to incur extra risks for an assister does not prioritizing assisters when the risk of saving other victims is as to fall under the scope of our duty to rescue. Assister Other things being equal, when a refugee rendered a belligerent's forces assistance during a war, that is morally obliged to (i) incur greater risks to rescue that refugee than they are obliged to incur to rescue other refugees and (ii) prioritise rescuing that refugee over rescuing other refugees, provided that the refugee is endangered as a result of providing supererogatory or costly obligatory assistance and/or was promised assistance by the and saving other refugees is supererogatory. is not the only possible ground of an especially stringent duty to rescue. As we saw in Section of the popular view often appeal to the fact that Western states promised to save Afghan assisters. It is plausible that promising to enact a supererogatory rescue is one way to incur a duty to enact such a rescue to the that one is not under a duty to rescue anyone else at the time of the in turn, can ground a duty to prioritize saving a particular over other victims similar it seems plausible that gratitudinal duties, can require rescuers to prioritize the even when saving other victims is risk to be ordinarily the of Alice's I argue that duty to save Alice even if she can save Betty at an obligatory risk, provided that Rescuer was under no duty to save Betty at the time she to save I take the duty to save to be in the that one a duty to save if evidence is that saving or that one is to save at cost to This is one does not fail in a duty to save from in the if one has no idea that she is in even if one of her one could save can also ground duties to rescue obligatory assisters who incur or no risk. that Bob a risk obligatory rescue of Adam and, by Adam doesn't owe Bob a duty of Adam nevertheless be and Bob that, should the need he will try to save him even at risk to It seems plausible that this could ground a duty for Adam to incur ordinarily supererogatory risks for Bob's sake. If Adam's could also make it upon him to prioritize saving Bob over other victims in some cases. duties therefore seem to do a of the popular view than gratitudinal duties, a to prioritize Afghan assisters even granting that Western states have duties to assist other Afghan refugees and especially stringent duties of rescue to a greater of obligatory assisters. promises can this on Western states were in a position to make promises to prioritize Afghan assisters over other refugees. I that states can make promises of this promises are by the of others. If, for example, Adam promises Bob that he will steal Clare's for him, then the of what Adam has promised to do the usual moral of Adam does not incur any duty to steal Clare's since doing so would conflict with Clare's A to prioritize assisters with other to be saved. As above, individuals may sometimes agreements that weaken other people's claims against them. including are an part of allowing individuals a in which they can pursue their projects and thereby their own lives. states (and those as do not have projects to their autonomous that we have reason to the that Alice's Rescuer as her bodyguard and Adam's promising to save Bob cannot states' such And it is hard to see what else could states' promises to assist assisters, beyond what is demanded by their gratitudinal duties to those assisters, given that such promises would undermine other claims to assistance. to such promises would also wrong the own by those to incur costs that they have no duty to states have for promising people assistance beyond what they would owe them on the of can people to discharge their duties, for Adam to give Bob his in order to Bob to with his duty to save Adam's But such promises are not on the It is wrong for Bob to discharge his duty only on the that Adam give him his Adam is therefore under no duty to give Bob the And when such an with the claims of Adam is under a duty not to a to steal Clare's would wrong Given that, to any non-assisting Afghans had claims to assistance from Western it those other Afghans if Western states promises to assisters at the of those other claims to assistance. nonetheless be for such for example, other people will also fail in their duties if they see that promises are not But this explanation of belligerents ought to incur ordinarily supererogatory risks for the sake of assisters, at the of other refugees and imposing these costs on their does not on the moral of Rather, it on the moral importance of the ends that those promises are a means of As I argue in Section this does not vindicate the popular view of Western states' duties to Afghan assisters. if the have made these promises in order to people to assist them at supererogatory As I argue in Section insofar as assisters as a result of these the can incur especially stringent duties to rescue them. these duties are not promissory duties. Rather, they are in the duty not to harm. even if we grant that promises can sometimes ground more stringent duties to rescue, they do not support belligerents' having especially stringent duties to rescue assisters in States cannot make promises to beyond what they owe as a matter of gratitude, since such promises conflict with other claims to assistance. aimed at people to discharge their to aid moral Hence, promissory duties do no at the popular view. However, if what I have argued in II and III is then belligerents can owe assisters these duties of what they have After all, Adam makes Bob no promises in Rescue and Adam has a duty to incur supererogatory risks for Bob's sake. I above that Assister supports a duty for Western states to incur greater risks for the sake of rescuing Afghans who provided their forces with either supererogatory or costly obligatory assistance. This argues that this support is since it is false that Afghan assisters generally provided supererogatory or obligatory assistance to Western forces. The Western invasion of Afghanistan is widely as an unjustified war, including by many just war and, I many proponents of the popular view. people believe that Western states have especially stringent duties to assisters the war was On this view, the to rescue assisters the other of the For example, a Afghan argues that by them in an war for 20 if it our Afghan at the it will have us a British who in of being of the of our that the of Afghan [...] is, like the Afghan war the of a that we to an unjustified is a hence, providing such assistance is not supererogatory rather, either impermissible or justified only as a As I is no duty to prioritize those who assist even when saving other victims is supererogatory. Busy Adam's Adam to rescue Amy by throwing a grenade that will and some Bob Adam a so that he can a and the are Amy is saved. However, as Bob is his a him and an stranger, Adam can rescue either Bob or Clare only at the moderate risk of losing his arm. We also have reason to reject the idea of Adam's having a duty to save Bob. I argued in Section III that the fact that Clare is made worse off by Bob's having a claim to be saved doesn't show that Clare is But this holds only insofar as Bob permissibly in assisting Adam. It does wrong Clare if she is made worse off as a result of as she would be if Bob's assisting Adam to his made it impermissible for Adam to save If the fact that Bob's being imperiled is the result of his own seems to make it to saving Bob of if only Bob is threatened by the this case It is more plausible that Adam is required to save Bob in this as above, we cannot explain this by to Adam's Bob a gratitudinal or promissory Rather, if Adam owes Bob an

Afghan · Criminology · Political science · Refugee · Sociology · Asian Geopolitics and Ethnography · Law · Migration, Health and Trauma · Migration, Refugees, and Integration