Animals and the ethics of war: a call for an inclusive just-war theory

The JW tradition, which scholars commonly trace to Augustine and Aquinas, is the dominant Western approach for ethics in war. It’s also been crucial in the development of IHL. Key early JW thinkers, such as Hugo Grotius and Emer de Vattel, were (also) jurists. Central sources of IHL – including the Hague Conventions, Geneva Conventions, and UN Charter – use concepts shaped by early JW thinkers, such as just cause, last resort, necessity, and proportionality. (We introduce these below, but note here that we use the terms as understood in JWT, rather than IHL.)

IHL has moved ahead of JWT when it comes to animal inclusion (compare §§4.1–2 and §5.1) and scholars of IHL note the relative quietude of philosophers. Karsten Nowrot, for example, welcomes the emergence of political-philosophical exploration of animals, offering a range of reasons why IHL is ‘a particularly noteworthy legal regime from the perspective of a political theory of animal rights’.¹⁵ Our exploration, then, provides a much-needed complement to burgeoning work on animals in IHL.

In brief, according to JWT, violence is just if it meets certain criteria. Though these criteria have developed over time, there’s remarkable agreement on the main points. The key distinction is between jus ad bellum, just recourse to war, and jus in bello, justice in war. Classically, theorists apply the former primarily to states’ conduct, holding that combatants need only concern themselves with the latter. This view has come under strain, both because of critique from ‘revisionists’¹⁶ and due to developments in modern warfare (e.g. the deployment of drones to target individuals outside of warzones). Scholars have also argued for two additions to this framework: Jus post bellum (justice after war),¹⁷ and jus ad vim (justice in force-short-of-war).¹⁸

The six requirements of jus ad bellum are: just cause; right intention; legitimate authority; reasonable chance of success; last resort; and proportionality.¹⁹ In contemporary JWT, just causes are self-defence, defence of a third party (including, e.g. NATO’s collective defence principle), and, to an extent, ‘humanitarian intervention’²⁰ (war on behalf of third parties, e.g. to prevent genocide). Right intention means that, even where a cause is just, the intention may be wrong, which would make a war unjust (e.g. the just cause may be a pretext for conquest). Legitimate authority requires an appropriate agent to declare war. Example agents include heads of states or legitimate independence movements, or the UN Security Council. Reasonable chance of success means that wars very unlikely to succeed cannot be just – even if the cause is just, the harm caused would be needless. Last resort requires that those resorting to war seriously consider non- and less-violent alternatives. If a political community bypasses less harmful means, harm caused by the war is unnecessary. Finally, proportionality dictates that harm caused by a war mustn’t outweigh the harm it seeks to avoid.

A war that doesn’t meet all requirements is not, according to the standard interpretation, ad bellum just. The criteria also remain relevant when war is underway – war aims and intentions can shift, and developments can reduce or increase the chances of success.

Jus in bello concerns the behaviour of combatants in war.²¹ Scholars typically identify three requirements: discrimination (or the principle of distinction); proportionality; and necessity.²² Discrimination requires that combatants differentiate between legitimate and illegitimate targets. Legitimate targets are military targets (including enemy combatants or military infrastructure). In other words, combatants may never target non-combatants or non-combatant property.²³ Proportionality, as with the ad bellum version, requires that the harm caused by an action does not outweigh the harm the action aims to prevent. Thus, an activity causing considerable harm to non-combatants, while only peripherally contributing to mission success, isn’t proportionate (e.g. soldiers can’t bomb a populated village to drive out one enemy combatant). Finally, necessity requires that a mission causes only necessary harm – if a less harmful route to the same goal is available, then combatants should pursue it. For instance, if combatants can choose between killing and incapacitating an enemy, they should opt for the latter.

Whilst the JW tradition has been instrumental in shaping contemporary ethics of war, there are other influences. For example, soldiers frequently subscribe to so-called military (martial) virtues.²⁴ Pacifist approaches have a long history as well. Pacifism, however, despite popular perceptions, doesn’t necessarily rule out violence. Varieties of contingent or ‘minimal’ pacifism may allow for some violence, or even some wars.²⁵ And while Jainism requires nonviolence or ahimsa of its monks and nuns, lay Jains may kill in self-defence under certain circumstances.²⁶

Although there’s much we admire in these approaches, we here focus on JWT. Why? As noted, it has played an important role in developing the concepts core to IHL today. More importantly, wars are, regrettably, going to happen. JW theorists tend to concede that states will wage war and that violence can be legitimate, but aim to reduce the occurrence of unjust wars and unjust behaviour in war. In the words of Michael Walzer, war may be hell, but even ‘in hell, it is possible to be more or less humane, to fight with or without restraint’ – JW theorists address ‘how this can be so’.²⁷ We’re interested in considering the ways war can be more humane, perhaps by making JWT a little less human. While there’d be no war in a perfect world, wars are happening here-and-now, and JWT (in concert with IHL) is the framework best suited to theorising how wars can be less destructive.

In other words, while seeking to end war altogether is an important goal, this is little help to the small country invaded by its more powerful neighbour, the soldier in the field who’s under fire, or the civilian caught in crossfire. JWT helps us conceptualise rules and practices that can alleviate the harm of conflict until war is no more.

Although war has always involved animals – for transport, protection, communication, and more – and the natural environment (think of ‘scorched earth’ tactics), philosophers writing on war have devoted little attention to animals. JW theorists may discuss the importance of protecting the environment in general (including animals?), but rarely consider the impacts of war on animals themselves. This is surprising. Animal rights is no longer a fringe topic in academia. Philosophers may teach animal ethics and JWT in the same course, but they are unlikely to explore them together. Indeed, there are scholars – such as Jeff McMahan – who have made significant contributions to both animal ethics and JWT, yet haven’t put them in conversation.

The neglect of animals in the ethics of war isn’t only a matter of JWT’s anthropocentrism. Despite some calling for ‘a just war theory for animals’²⁸ or ‘an animal-inclusive just-war theory’,²⁹ animal ethicists have, on the whole, offered few reflections on war. Anecdotally, we’ve encountered reluctance to address warfare among animal ethicists because of nervousness about endorsing violence for animals’ sake – one philosopher writing about the violent defence of animals chose to publish pseudonymously in the Journal of Controversial Ideas.³⁰ Sometimes academics institutionalise this reluctance. The Journal of Animal Ethics ‘will not publish material that justifies or advocates . . . violence’,³¹ nor will the Oxford Centre for Animal Ethics ‘appoint Fellows who advocate violence’.³² JW theorists, then, need not apply.

After all, JWT endorses violence in principle – some (conceivable) wars are just. But one can be a JW theorist and deny that war is justified in practice; perhaps no real-world war meets the requirements. Similarly, an inclusive JWT need not endorse many (if any) wars – an inclusive JWT could, for example, conclude that harm to animals (in principle) provides a just cause, but violence on behalf of animals cannot meet the other ad bellum requirements (compare §3.1). Of course, such a theory would still have practical relevance. As JWT addresses much more than ad bellum justification, an inclusive JWT would address much more than waging war for animals’ sake.

Thus, to note that animal suffering could be a just cause is not the same as advocating violence on animals’ behalf. Indeed, JWT helps make sense of common sentiments about violence for animals – that while perpetrators’ cause and intentions are noble, violence is an overreaction or counterproductive (again, see §3.1). Such arguments implicitly appeal to the distinct, but jointly necessary, conditions of just cause, right intention, proportionality, and reasonable chance of success.

Though unusual in noting it explicitly, Cécile Fabre is typical of JW theorists, as her ‘account of the just war is human-centric’; it doesn’t ‘include non-human animals in [the] global community’ – this may, she accepts, be ‘arbitrary’.³³ We agree with her. Like many animal ethicists and animal studies scholars, we contend that excluding animals simply because they are animals is an example of ‘speciesism’, and we hold that such discrimination is arbitrary. Better than ethical theories resting on arbitrary distinctions, we contend, are ethical theories that treat like cases alike. This does not mean, of course, that we are committed to saying that bombing a rabbit warren is morally equivalent to bombing a village. It isn’t, because the inhabitants of a rabbit warren have different interests to the inhabitants of a village. But it does mean that, in our view, a JWT that doesn’t discount the interests of rabbits simply because they belong to rabbits will be more defensible than a JWT that does. To borrow a term familiar to animal ethicists, we believe that an inclusive JWT will entail the equal consideration of interests – at least when it comes to comparisons across species lines.³⁴

Thus far, we’ve implicitly contrasted an inclusive JWT with a JWT that excludes animals’ interests altogether. But we must acknowledge that there are two other (sets of) approaches that those incorporating animals into JWT could take. Though we cannot offer a full response to these possibilities, it is important to acknowledge them.

First, even an anthropocentric (thus non-inclusive) JWT could take account of animals indirectly (compare the discussions of IHL in §§4.1–2 and §5.1). Many animals might warrant protection qua human property, meaning damage to them constitutes morally considerable harm to humans. Or else animals may be parts of nature, or human culture – and nature or cultures may be worth preserving for humans’ sake. A conservative, anthropocentric JWT could thus offer some protection to some animals, but it would not be for their sake, and would likely exclude many animals. The distinctive feature of an inclusive JWT, we suggest, is that it takes seriously the widespread (though not universal) belief that animals matter for their own sake; that it’s sometimes wrong to do things contrary to animals’ interests because it’s contrary to their interests.

Second, one could acknowledge the limitations of the indirect approach, but stop short of embracing the equal consideration of interests in the inclusive JWT. This ‘hierarchical’ JWT theory would allow that animals’ interests count for something but suggest that they count for less than human interests simply because they are animal interests. So, for example, even if a chimp’s interest is as strong from the perspective of the chimp as a human’s interest is from the perspective of the human, the human interest will ‘win out’. Indeed, even if the chimp’s interest is stronger (from the chimp’s perspective) than the human’s interest (from the human’s perspective), the human’s interest may ‘win out’ – but that will depend on details of this hypothetical theory that remain un-worked-out.³⁵ (We reject this view – despite its intuitive appeal and its ability to offer animals protection – as it still relies upon arbitrary distinctions, thus falling foul of ‘speciesism’.³⁶)

But this description of an inclusive JWT theory – and a comparison to slightly less inclusive approaches – doesn’t offer a clear picture of what the approach in practice. Although, as we’ll see, scholars have linked animal ethics and JWT in a range of contexts, a systematic inclusive JWT doesn’t emerge from the literature. What would such an account look like? It would answer, or provide tools to answer, many of the questions we canvass below. Inter alia, it’d tell us how to include animals in considerations of jus ad bellum. For example, it’d tell us whether the eradication of animals could itself be a just cause – and if that idea even makes sense. It’d tell us how to factor animals into questions of fighting justly, and thus how to include animals’ interests in calculations of proportionality and necessity. It’d tell us under what circumstances animals might be legitimate military targets, and thus whether the targeting of animals could be discriminate.

To answer these questions, however, an inclusive JWT must also address more foundational questions. We have already pointed towards the equal consideration of interests as an important principle for a genuinely inclusive JWT, but any account that includes animals – whether indirectly, hierarchically, or as equals – must offer a clear assessment of the value of animals,³⁷ and the weight (if any) that commanders must afford to animal interests in decision-making. This includes relational questions, on which the inclusive approach we have so-far sketched is neutral; for instance, should commanders afford greater weight to the interests of ‘their’ animals over the interests of ‘enemy’ (or ‘neutral’) animals? And it needs to offer a clear account of which interests animals have, and how strong those interests are. (Note that an interest’s strength and the weight decisionmakers should afford an interest are distinct questions.)

Proponents of an inclusive JWT must also ask whether JWT is suitable for the protection of animals. It has, after all, developed with only interhuman harms in mind. Even if scholars can expand JWT to include animals, perhaps they should change its criteria – perhaps (boldly) they should add to or modify them, or perhaps (less boldly) they must accept that the criteria are suitable only for some of the problems around animals and war. Ultimately, JWT exists to address a particular set of problems. The most important questions of animal ethics may not be among them (compare §3.3).

All existing human communities harm animals on a massive scale. No extant human community has ceased farming animals for food, and, consequently, humans kill trillions³⁸ of sentient animals for food annually. An inclusive JWT must explore whether the death and suffering of these animals could, in principle, be a just cause for military intervention.

To distinguish it from ‘humanitarian intervention’, Alasdair Cochrane and Steve Cooke label military intervention on behalf of animals ‘humane intervention’.³⁹ ‘In reality’, Cochrane and Cooke say, if animals have ‘morally relevant . . . interests and rights’, the just cause criterion ‘is incredibly easy to satisfy in a range of cases’.⁴⁰ Thus, (coalitions of) states could (in principle) wage war against states that don’t respect animals’ rights, if (and only if) the other jus ad bellum criteria are met in particular cases. As we will see, however, Cochrane and Cooke argue that (in practice) they won’t be.

Belligerent animal-rights-respecting states are not a realistic threat for any real-world political community. There are, however, individuals and organisations who aim to live in accordance with animal rights. These might present a more immediate threat to those violating animals’ rights.⁴¹ For Mark Bernstein, these ‘animal liberationists should be viewed as being at war with animal oppressors’, precisely because this allows ‘a means of testing [liberationism’s] legitimacy’.⁴² Liberationists, he says, have a just cause: ‘the altruistic nature of the liberationist’s battle harmonizes perfectly with the sanction to intervene on the behalf of another who suffers injustices’.⁴³ He ultimately concludes – with the caveat that there may be no just wars – that ‘the liberationist’s war is at least as proper as any ordinary war carried out by any regime’.⁴⁴

Lisa Kemmerer, similarly, asks whether aggressive ‘warrior activists’, fighting on behalf of animals, are ‘social activists, terrorists, or . . . soldiers engaged in humanitarian intervention against brazen indifference’.⁴⁵ She too tests animal liberationists’ ‘war’ against the requirements of jus ad bellum. JWT reveals, she argues, that activists are combatants in ‘an asymmetric war of humanitarian intervention on behalf of a beleaguered and abused population’. Activists, she says, ‘are engaged in a counteroffensive on behalf of the exploited, and are, by all counts, engaged in just warfare’.⁴⁶

Cochrane and Cooke hold that, even though war on behalf of animals may be justifiable in theory, it is not justifiable in practice. This is for at least two reasons. First, death is a greater harm for humans than for animals, meaning that it will be hard (not impossible) for violence against humans on behalf of animals to be proportionate.⁴⁷ (This is a contentious claim. As noted in §2.2, one of the tasks for an inclusive JWT is working out which interests animals have, how strong those interests are, and how much weight decisionmakers should grant them.) Bernstein, however, imagines a less violent war. Liberationists ‘seek no person’s death or harm’, using ‘no torpedoes, rockets, bombs, or guns’.⁴⁸ But Kemmerer argues that the ‘logic’ of those advocating violence on animals’ behalf is ‘difficult to refute’ if one takes animals interests seriously.⁴⁹ However, she overlooks the contention that animals typically have a weaker interest in continued life than do humans. Her case is thus incomplete.

Cochrane and Cooke’s second concern is that there’s no international agency with authority to intervene on behalf of animals – and ‘it is impossible to conceive of any state or coalition of states who could realistically be regarded as having the moral legitimacy to authorise human interventions’, as ‘all states are themselves involved in their own massive violations of basic animal rights’.⁵⁰ Bernstein admits ‘there is no venue for formal declaration’ for war, but believes – questionably, we think – that ‘the requirement . . . is not so much unsatisfied as inapplicable’.⁵¹ Kemmerer’s justification is more developed. She argues that, in cases of asymmetric warfare, JWT permits ‘covert leadership on the disadvantaged side’,⁵² and that people may declare war upon a ‘poorly accountable, inept, or corrupt regime’ running their own state.⁵³ ‘The U.S. government’ – among others? – ‘might legitimately be considered corrupt and inept in the eyes of’ liberationists,⁵⁴ precisely because it has failed to act against animal oppression, instead siding with oppressors.

Evidently, an inclusive JWT has questions to answer on legitimate authorities. A critic of Bernstein and Kemmerer’s respective cases could contend that the lack of legitimate authorities delegitimises the war they propose – or that what they’re describing is not war in the technical sense, and that JWT is therefore inapplicable. (More on this shortly.)

It’s also important to consider the ‘reasonable chance of success’ criterion in practice, and not merely in principle, as violence on animals’ behalf could be deeply counterproductive. Cochrane and Cooke’s contention ‘is not that humanitarian [or “humane”] interventions always work’ but simply ‘that in some circumstances, they could work’.⁵⁵ Kemmerer, however, rightly notes that animal activists sometimes oppose violence as counterproductive.⁵⁶ Animal activists ‘do not have a reasonable hope of success in any overt humanitarian counter-attack . . . at least not in the near future’, which could make their war unjust. She thinks warrior activists may ‘have a tremendous chance of success in the long run’, though.⁵⁷ An inclusive JWT needs to take these considerations seriously. Indeed, even the suggestion of violence may be counterproductive.

Parts of JWT are (for Bernstein) ‘inapplicable’ to the question of fighting for animals,⁵⁸ and, traditionally, JWT (for Kemmerer) ‘reserves “just war” for those already established in power’.⁵⁹ This raises the question of whether JWT is applicable to animal activism today. We needn’t invoke JWT to discuss moral permissions (or obligations) to protect animals; for example, several theorists explore violence in defence of animals using self-defence theory,⁶⁰ and in §3.2 we look towards jus ad vim as a complement to JWT.

An inclusive JWT, then, may hold back on making judgements about the actions of (real or hypothetical) animal activists – but it cannot avoid making judgements on whether harm to animals might serve as a just cause for future wars, and this may mean endorsing violence for animals. Crucially, however, just cause is only one of the requirements of jus ad bellum. Allowing that animal harm can be a just cause needn’t commit theorists to endorsing any actual wars.

Academics have used JWT when assessing state violence against poachers.⁶¹ For example, Amy Dickman and colleagues apply the jus ad bellum criteria to the use of lethal force by park rangers and soldiers against (suspected) poachers in countries such as Botswana.⁶²

As commonly framed in the literature, this is not a war for animals – Rosaleen Duffy’s phrase ‘a war to save biodiversity’⁶³ makes this clear. The concern is with protecting biodiversity, which includes animals as collectives (e.g. animal species), but not animals as individuals. (Concern for biodiversity needn’t translate into concern for animal suffering. In practice, it frequently means killing animals supposedly threatening biodiversity. Compare §3.3.) Alternatively, a war to save biodiversity may be (partially or wholly) about protecting valuable income streams, such as those offered by nature reserves.

Is this a literal war, in the technical sense with which JWT is primarily concerned? Dickman et al.’s use of JWT suggests that they think so, or at least that JWT offers a useful framework for assessing this state violence. We’ll later argue that jus ad vim may be more appropriate to analyse this ‘war’ than JWT as traditionally understood, allowing, therefore, that ‘wars’ to save biodiversity may not be literal wars, but rather a ‘force short of war’ that transcends the typical policing paradigm without reaching the level where jus ad bellum (or jus in bello) criteria are appropriately applied.⁶⁴ But let us accept Dickman et al.’s stipulation for now and assess their argument.

What could justify a war to save biodiversity? Dickman et al. claim the ‘instrumental value of the environment to human life, and the intrinsic value of individual megafauna’ is sufficient to meet JWT’s just cause threshold.⁶⁵ However, if a just cause must be defence of self, defence of allies, or humanitarian intervention, it isn’t clear how protecting the environment fits, if the environment is (merely) instrumentally valuable. Perhaps war to save ‘individual megafauna’ could be an example of Cochrane and Cooke’s ‘humane intervention’ – but tension between protecting biodiversity and protecting animals creates problems.⁶⁶ For example, what if megafauna themselves pose a threat to biodiversity?

Further, the focus on the ‘intrinsic value of individual megafauna’ seems arbitrary, as it’s not clear that only megafauna possess ‘intrinsic value’. On, for instance, Tom Regan’s rights view,⁶⁷ much-maligned, biodiversity-threatening animals (like rats⁶⁸) have ‘intrinsic value’, too. Consequently, a war to protect animals from poachers could, in principle, meet JWT’s just cause requirement, if Dickman et al. were prepared to embrace an inclusive JWT. But advocates of war to protect biodiversity would need to depart from prevalent approaches in scholarship on green militarisation to deploy this justification. Existing work in this area does not systematically consider animal interests, instead focussing on biodiversity and environmental protection. While biodiversity may have considerable value, it’s individual animals – not species, or ‘the environment’ – who have interests. Dickman et al. are thus treading a difficult line, sometimes appealing to (something like) an inclusive JWT, and sometimes appealing to something quite different – and this creates a tension, even if they can overcome the puzzle of JWT’s applicability.

There are, however, other questions an inclusive JWT must ask about ‘wars’ on poachers. Just cause is just one JWT requirement. In particular, we think it’s worth thinking about the ad bellum requirements of last resort and proportionality, as well as the in bello requirement of discrimination.

Poaching is complex, with many causes. For example, analyses link the growth of the bushmeat trade in Central and West Africa to increased logging and mining, while the Apartheid-era South African Defence Force, as well as all sides in the Mozambican civil war (compare §1), participated in poaching to gain funds. Indeed, Duffy suggests that, in sub-Saharan Africa, ‘anti-poaching . . . actually [relies] more fully on the production of regional stability than on militarized approaches to conservation’ due to the role of regional conflict in facilitating poaching.⁶⁹ Further, poaching tends to increase and fall depending on the circumstances – for example, disrupted transport networks meant ivory and pangolin scale poaching declined during the COVID-19 pandemic, but illegal hunting and fishing increased.⁷⁰ Illegal subsistence hunting is often occasional (e.g. increasing in the dry season),⁷¹ and can also rise in response to conservation efforts that alienate or uproot local populations.⁷²

This raises questions about who anti-poachers may legitimately target (an in bello concern), even if a war on poachers is, in principle, legitimate. Is any poacher a legitimate target, even if an occasional or subsistence poacher? Are poachers legitimate targets only while engaged in poaching?

These considerations further suggest non-violent routes to resolving or reducing poaching, such as dismantling international trade networks, ensuring consistent employment, and engaging in less antagonistic conservation. ‘All out’ war on poaching would therefore be unlikely to meet the last resort requirement.

We can, perhaps, complement JWT analyses of the ‘War on Poaching’ with the jus ad vim paradigm. This framework allows us to think about the limited application of force outside of war. Jus ad vim is a response to the changing nature of warfare – including the rise of non-state actors (such as transnational terrorist organisations), and advancements in remote weapons technology. Daniel Brunstetter and Megan Braun argue that

While war used to be easily defined as a zone of combat where lethal force was justified (to be distinguished from a zone of peace, where it was not), the struggle against terrorism has created “in-between spaces” of moral uncertainty where force is used on a consistent and limited scale, but war is not declared.⁷³

They argue that the traditional jus ad bellum principles are inadequate for these new conflicts and their methods (e.g. targeted drone strikes), while the policing paradigms that govern (domestic) law enforcement clearly don’t apply either. They thus offer a new set of criteria, which are similar, though not identical, to the requirements of jus ad bellum, to judge when political communities may deploy violence outside of formally declared war.

Brunstetter and Braun apply their framework to international terrorism (and similar), but it’s potentially well-suited for militarised conservation. There are some similarities between international terrorist organisations and the international wildlife crime networks, in particular their transnational nature and their willingness to use violence, especially violence against innocents (including animals). Moreover, in at least some cases, these criminal networks pose a threat not just to animals, but also to states’ territorial control.⁷⁴

In jus ad vim, the just cause requirement offers a comparatively low bar, permitting (lethal) violence in cases where full-scale war would be impermissible, potentially meaning that authorities may target poachers outside of war. In other areas, though, jus ad vim is more restrictive than jus ad bellum. Brunstetter and Braun include a new requirement, the probability of escalation. If violence short of war could lead to more widespread conflict – consider evidence that militarised conservation alienates local communities⁷⁵ – that would be a significant mark against it. We note this framework probably cannot extend to the use of violence against all poachers. Subsistence hunters, at least where their hunting is occasional or opportunistic, likely do not contribute to the creation of what Brunstetter and Braun call ‘in-between spaces’, where states use force on a ‘consistent scale’.⁷⁶ Jus ad vim thus offers a promising complement to an inclusive JWT, allowing the assessment of violence on animals’ behalf outside of war as traditionally understood.

Though the idea initially sounds bizarre, some scholars have explored the prospects of wars against animals. Michael Morris analyses New Zealand’s ‘war’ against non-native animals,⁷⁷ ostensibly waged in defence of native wildlife.⁷⁸ Ultimately, Morris concludes, JWT demonstrates this war’s problems. Indeed, any such war is likely unjust. For example, the last resort requirement is difficult to satisfy in a ‘war on pests’. As Morris notes, contraceptive control could constitute a promising alternative to the use of lethal violence.⁷⁹ It’s true that contraceptive control may not always be possible given current technology, but this means that we should aim to develop new methods – as Morris suggests, conservationists have conducted several promising studies on mice and rats. Studies of the use of contraceptives to control grey squirrel populations in the UK are reportedly ‘making good progress’.⁸⁰

More fundamentally, it’s hard to imagine a successful war on pests outside the most contained ecosystems. As Morris notes, poisoning non-native predators can have unintended effects. For example, killing rats can lead to mouse population explosions. Mice also pose issues for native wildlife.⁸¹ Eliminating predators can also make things easier for subsequent predators – one study on rat control found that re-invasion meant more rats found 24–30 months after ‘treatment’ than before.⁸² This is further complicated by the fact that animals invasive in one place may be protected in others. Thus, a war on pests runs the risk of turning into an interminable ‘forever war’.

Though we find Morris’s analysis useful, a critic may charge him with equivocation. Specifically, a ‘war on pests’ might be war only metaphorically, much like the ‘War on Terror’ or the ‘War on Drugs’,⁸³ meaning that JWT does not apply. The violence directed at non-native animals is not unique in this regard. Indeed, there is much (sometimes war-like) violence against animals that JWT probably cannot usefully assess. It’s not our claim, for instance, that an inclusive JWT should replace existing theories of animal ethics⁸⁴ – only that it should complement them.

Tellingly, the political theorist Dinesh Wadiwel explores a ‘war against animals’ as a theoretical lens to characterise the violence that humans inflict upon animals.⁸⁵ But this isn’t to put it into the purview of JWT; Wadiwel doesn’t engage with JWT at all. Were he to do so, he would surely find the war against animals an unjust war,⁸⁶ for reasons echoing Morris’s case against the war on pests. But, this said, we doubt that Wadiwel’s war against animals is a war in the technical sense with which JWT is concerned, nor even the more general type of coercive violence (e.g. counter-terrorism) to which the JWT framework can meaningfully be applied.⁸⁷ While the analytical lens of ‘war against animals’ is influential in (critical) animal studies, one need not embrace the idea of a war against animals to meaningfully advocate an inclusive JWT (or vice versa).

One of the most important issues in jus in bello is so-called ‘collateral damage’: harm caused, unintentionally but foreseeably, to impermissible targets (chiefly civilians and their property). All core principles of jus in bello – discrimination, proportionality, and necessity – are relevant when considering collateral damage.

Discrimination doesn’t mean that militaries can never harm civilians – if interpreted this strictly, war would be impossible. Instead, combatants must ‘direct their operations only against military objectives’,⁸⁸ but they may still cause ‘collateral’ harm to entities that are not military objectives. However, combatants cannot cause widespread harm simply because it’s unintended; harm caused must be both proportionate and necessary. In other words, when choosing to pursue a particular course of action, they must first assess whether the likely harm is proportionate to the good they aim to achieve (if not, the harm’s disproportionate), and whether there are less harmful options available to them (if so, the harm’s unnecessary).⁸⁹

When combatants harm animals, it’s typically incidental or accidental harm.⁹⁰ Companion animals and animals living in zoos may suffer if attacks kill their caretakers, as noted in the Ukrainian context above. Wildlife are impacted by mines, bombs, and chemicals, as were Rwanda’s gorillas and Angola’s elephants during (and after) these countries’ civil wars; by the use of environmental destruction as a war tactic, as were the animals living in areas defoliated by American forces in Vietnam; or because their territories are used as a staging ground.⁹¹ Poaching and illegal hunting also tends to increase in the wake of conflicts.⁹² On an inclusive JWT, these harms would be accounted for in considerations of proportionality and necessity. Further, even if such harm to animals is non-intentional, it may be indiscriminate, as we will note below.

Insofar as animals are legally civilian ‘objects’, IHL already requires combatants to make proportionality calculations when their conduct may impact animals.⁹³ However, this presents a challenge as ‘it depends on the value attributed to the animals’⁹⁴ – this is, we note, a fundamentally philosophical question. If animals’ interests matter, Jérôme de Hemptinne argues, ‘their interests should no longer be automatically subordinated to those of humans’.⁹⁵ But this raises the difficult question of whether soldiers are obliged to put themselves at greater risk ‘to guarantee the protection of dogs and horses’ – and whether anything changes when it’s ‘pandas and white rhinoceroses’ threatened.⁹⁶ (To be clear, soldiers can sometimes be obliged to put themselves at greater risk to guarantee the protection of human civilians.⁹⁷) IHL can provide one set of answers – an inclusive JWT can provide another.

As important as these questions are, there’s a more fundamental challenge. If we take harm to animals seriously, might this render all war disproportionate? Some theorists might welcome this,⁹⁸ embracing the conclusion that taking animals seriously in the ethics of war means the end of warfare. On the other hand, we elsewhere argue that counting animals in war doesn’t prohibit fighting justly – it just changes what fighting justly looks like.⁹⁹

First, counting animals in war means changing officer training and the experts consulted in planning military activity.¹⁰⁰ For example, fighting justly already requires militaries to understand the cultural intricacies of human communities with which they interact. Perhaps, too, they should consult relevant experts to understand the animals impacted. Several navies have active guidance to mitigate the adverse effects of sonar on cetaceans.¹⁰¹ The Royal Australian Navy, for example, expresses a commitment to working with the relevant experts on its website:

Defence continues to support relevant research into marine mammal population biology and the effects of man-made noise on the ocean environment. . . . Assisting the [Australian Defence Force] to avoid any potential disturbance to marine animals during military training exercises is [Defence Science and Technology], who conducts research programs with experts at Australian universities and in collaboration with US scientists and the US Office of Naval Research.¹⁰²

Taking harm to animals into account should, then, lead to an expansion of the kinds of research consulted by militaries.

Second – we’ve argued, echoing Cochrane and Cooke – taking animals’ interests seriously does not entail that the death of an animal is morally equivalent to the death of a human. Many animals have interests in continuing to live that are weaker than those of (most) humans. Thus, the prospect of animal deaths needn’t be overridingly significant when commanders consider courses of action.¹⁰³

Third, the JWT requirement of discrimination already rules out many of the forms of warfare that are most impactful on animals.¹⁰⁴ For example, as chemical and nuclear warfare¹⁰⁵ or booby traps¹⁰⁶ are likely to devastate humans as well as animals, JWT already rules these out as indiscriminate.

To apply an inclusive JWT to in bello considerations in practice, we need more work charting the ways that militaries harm animals in practice and the steps they could take to limit said harm. The examples we explored in our ealier paper, by contrast, were stylised.¹⁰⁷ More engagement with real examples would allow an inclusive JWT to contribute to minimising suffering in the real world.

Much contemporary work on jus in bello concerns the distinction between combatants and non-combatants, and the distinction between ‘civilian’ and ‘military’ objects. This is partly because the principle of discrimination permits the deliberate targeting of combatants and military objects, but not non-combatants or civilian objects. It’s therefore crucial to be able to determine what does, and doesn’t, count as a military target. For an inclusive JWT, the question thus arises whether animals could ever be combatants or military objects, and thus targetable.

Militaries deploy animals in a range of contexts: in combat, as transport, as guards, for search and rescue, in training, as mascots, for therapeutic purposes, for ‘pest control’, or similar.¹⁰⁸ Some animals could be ‘animal soldiers’, and thus legitimate (‘military’) targets. Animal ethicists and JW theorists have written little about animal soldiers. Nonetheless, there are a range of avenues that an inclusive JWT could explore.

The first is a firmly ‘abolitionist’ and/or pacifist line, saying that using animals in warfare is illegitimate, and must end.¹⁰⁹ Just as a purely pacifist approach offers little practical guidance to those facing difficult moral decisions in war, this approach offers no practical guidance on how to behave in a world in which animal soldiers do in fact feature on battlefields.

Second, then, we may turn to JWT, which already contains discussion of ‘innocent threats’, such as child soldiers.¹¹⁰ Animal soldiers, like child soldiers, may present threats to combatants, even though they are, themselves, innocent. There’s thus scope for exploring the expansion of claims about child soldiers to include animal soldiers. This approach is useful when it comes to how human soldiers may treat animal soldiers on the battlefield, but less useful on whether militaries may permissibly deploy animals on battlefields in the first place.

Third, we could look to the comparatively well-developed literature on animal soldiers in IHL,¹¹¹ in which scholars analyse which existing laws could protect animals, and what routes there might be to expand protection.

For example, Anne Peters and Jérôme de Hemptinne argue that we cannot easily recognise animals as ‘protected persons’ or ‘civilians’ under IHL.¹¹² However, there are other options. First, animals may be ‘objects’, meaning combatants may not target them unless they are weapons or legitimate military objectives. Second, animals (including domesticated animals) might have protection as components of the environment. Third, animals may be medical transport or equipment, meaning combatants may not impede them (qua medical transport/equipment) or attack them. Fourth, animals may be objects needed for civilian survival. Fifth, some animals (including members of endangered or endemic species) may be cultural heritage or property.¹¹³

In each case, the law affords limited and indirect protection. It does not offer protection to animals for their own sakes. (Compare the discussion of indirectly protecting animals in §2.2.) Recognising animals as persons under IHL might offer direct protection. It would be ‘conceptually possible to broaden the concept of “person” under IHL (and even “protected persons”) so as to encompass “animal persons”’. But, say Peters and de Hemptinne, these categorisations ‘are ill-adapted to the needs of animals’.¹¹⁴ Consequently, better we classify animals not as ‘combatants’ (and thus, when captured, ‘prisoners of war’), nor ‘civilians’, but ‘objects’, with Peters and de Hemptinne offering an ‘animal-friendly interpretation of IHL rules on objects’.¹¹⁵

While IHL is ahead of JWT on animal inclusion, Peters and de Hemptinne’s discussion highlights an important limitation of the IHL approach – international lawyers may find themselves constrained in important ways by existing (highly anthropocentric) legal frameworks, in a way that philosophers need not be.

Finally, we may turn to the recent literature in animal ethics on ‘social membership’.¹¹⁶ The idea is that animals, in addition to basic rights, warrant protection because of their membership in protected groups. Insofar as human soldiers warrant protections qua soldiers (compare the discussion of IHL above), the social membership model suggests that animal soldiers, too, warrant these rights. But ‘soldiers’ isn’t one of the groups that advocates of the social membership model have explored.

Other groups, however, may be relevant. The burgeoning literature on animal workers – and the rights to which they are entitled qua workers – sometimes looks to animal soldiers.¹¹⁷ This lens raises questions about some uses and treatment of military animals. If animal workers (qua workers) have rights to safe working environments, this limits the uses to which they can be put in combat zones. If animal workers (qua workers) have rights to retirement, this rules out killing working animals once ‘superfluous’. Infamously, during the Vietnam War, the US military classified dogs as ‘equipment’, killing them if no longer useful. Many were simply left behind at the end of the war.¹¹⁸

Advocates of the social membership model have also explored the idea of animal citizens.¹¹⁹ But how this interacts with the question about warfare is an open question. For instance, does the idea imply the permissibility of animal citizens ‘being sent to war, or even a duty on their part to serve [their nation’s] military interests’?¹²⁰ Could animal citizenship provide a lens through which to consider the legitimacy of animal soldiers on an inclusive JWT?

We observe that the social membership model speaks to a recurring issue. According to some authors in the ethics of war, states may legitimately favour their own civilians and soldiers when making proportionality calculations.¹²¹ More contentiously, some argue that it may be permissible for soldiers to give more weight to their own lives over the lives of enemy soldiers (or even enemy civilians) when calculating proportionality.¹²² If some animals are ‘members’, they may also be entitled to greater weight than those deemed ‘neutral’ or ‘enemy’.

Thus, despite the paucity of ethical analyses of the use of animals in war, there are many resources that an inclusive JWT could draw upon. This is an important future research area.

JWT historically distinguished jus ad bellum and jus in bello. Recently, theorists have proposed a third category – jus post bellum, or justice after war.¹²³ Jus post bellum addresses justice in the termination and aftermath of war. For example, it asks when and how hostilities should cease, who is responsible for post-war reconstruction, and who owes compensation/restitution to whom. As an indicative example of how just post bellum could incorporate animals, let us consider the final issue.

In animal ethics, questions about restitution to animals are underdeveloped.¹²⁴ Again, IHL scholarship has taken important steps in this direction, with Marina Lostal asking whether animals could count as ‘victims’ in the eyes of the International Criminal Court (ICC) – though she concludes that they couldn’t.¹²⁵

In JWT, compensation is an important tool for armies seeking to reduce the severity of collateral damage. Compensation cannot bring back a family member, but it can help lessen harm. It can also provide incentives for the military to reduce collateral damage.¹²⁶ Animals already feature prominently in compensation, and ‘have played a part in every single reparations proceeding’ at the ICC, ‘even when the criminal conduct at hand had nothing to do with them’.¹²⁷ Indeed, armed forces pay compensation to people for the loss of their animals without the ICC’s intervention – for example, the British paid £662 for the death of six donkeys after they ‘wandered on to a rifle range’.¹²⁸ In this case, they paid the donkeys’ owner, but many animals harmed in war are unowned.¹²⁹

But could armies compensate animals themselves? For example, might companion animals, dependent on humans for care, be victims of violence inflicted upon their human guardians? Might young, dependent animals be entitled to compensation if soldiers kill their parents? Lostal envisions that, if the ICC recognised animals ‘victims’, this ‘could lead to collective rehabilitation programmes in the form of reforestation, the construction of animal shelters and sanctuaries, the provision of veterinary services, or the awarding of resources to zoos and natural reserves affected by the crimes’.¹³⁰

Relatedly, militaries may owe compensation to their own animal soldiers. Recall the killing and abandonment of dogs by the US military. This was distressing for many of their human handlers,¹³¹ and surely harmed the dogs. The Army gifted some to the Vietnamese – who often didn’t see dog-handling as a ‘high prestige’ role, with disastrous results for the dogs¹³² – others they killed. Not only, we contend, was the Army obliged to take steps not to harm these dogs, but (echoing earlier ideas about animal workers) it may have owed them recompense for the work they had undertaken. At the very least, the US government may owe these dogs recognition – for example, memorialisation.

Memorials to animals killed (e.g. the UK’s Animals in War Memorial and Canada’s Animals in War Dedication) convey many different meanings.¹³³ Indeed, people can place different meanings on the same form of memorialisation. The British group Animal Aid introduced but then withdrew a Purple Poppy to wear alongside (or instead of) the Royal British Legion’s (red) Remembrance Poppy. Their ‘aim was to make it clear that animals used in warfare are indeed victims, not heroes. They do not give their lives; their lives are taken from them’. However, ‘too often the narrative promoted by the media’ about the Purple Poppy was ‘one of animals as the valiant servants of people in violent conflict’.¹³⁴

When war memorials frame animal soldiers as victims of war, this can be ‘suggestive of a political philosophy’ that the memorials’ erectors do ‘not believe’, argues Matthew McLennan. ‘[T]o truly honour nonhuman animals in war’, governments ‘would need to radically alter both [their] moral and political conception of nonhuman animals, as well as [their policies] of using some animals for military purposes to this day’.¹³⁵ Perhaps memorialising animals killed in war, until such time as militaries stop using animals in war, is an ‘atrocity added to an atrocity’ – as would be ‘a “children in war” monument that did not criticize the military exploitation of children’.¹³⁶ Real-world memorials to animals killed in war are thus complex – but do, apparently, offer acknowledgement of the significance of the harms faced by animals, and so a route into public consciousness for an inclusive JWT.

Tristam Derham and Freya Mathews argue that animals could be refugees.¹³⁷ The UN Refugee Convention adopts a narrow definition of who counts as a refugee, because it only recognises those fleeing persecution and war.¹³⁸ But some animals, such as African elephants, are victimised by human conflict and flee long distances for safety – they may even develop PTSD as a result. Even by the UN’s narrow definition, they could be refugees. One of the benefits of this reconceptualisation is rhetorical:

the inherent moral charge of the term ‘refugee’, at least in its original sense, ensures that the plight of a refugee makes a call upon anyone who can provide help. This moral call motivates international non-government organizations to assist human refugees and it might do the same for non-human refugees, mobilizing philanthropic organizations to assist animals on more morally complex and compelling grounds than on the grounds of conservation alone.¹³⁹

Even though legal reform is unlikely to be forthcoming, thinking of animals as refugees may provide impetus to protect animal victims of war.

Indeed, though it’s a different case to Derham and Mathews’s, animal refugeehood may be useful for thinking about the evacuations and forced migrations of companion animals with which we began this paper. Humans fleeing conflict zones frequently bring companions, and might refuse to leave without them. This suggests, for example, that evacuation plans as well as approaches to dealing with refugees should take animals’ interests seriously – for example, when thinking about the design of refugee camps. A parallel literature on the treatment of companion animals in disaster management provides insights on how this may be possible.¹⁴⁰

Taking animals seriously requires us to think about the impact that militaries have on animals outside of war. For example, recent research shows that the ecological impact of the US military is enormous – its estimated carbon footprint is larger than that of over 100 countries.¹⁴¹ This itself already has a significant impact on animals, who are vulnerable to climate change.¹⁴² But there are other ways in which militaries harm animals even outside the context of direct hostilities. Militaries routinely harm animals during research and development, either because they test on animals, because they use dead animals, or as a side-effect. Militaries purchase meat and other animal products on a large scale. (Much of this is routine, but it can impact public consciousness – for instance, there is periodic controversy over the British footguards’ bearskin hats.¹⁴³) And so on.

As such, urging militaries to take harm to animals seriously in war may require challenging – or dismantling – the foundations of the ‘military-animal industrial complex’.¹⁴⁴ This moves beyond an inclusive JWT.

Questions about shifts away from reliance on animals in economies, food systems, lifestyles, and more are familiar to animal ethicists. But putting these in conversation with the ethics of war will require reflection on specific elements of military culture. For example, violent or macho norms among soldiers threaten to limit uptake of animal-friendly lifestyles – at least while meat remains associated with masculinity. Further work might explore, for instance, the extent to which martial virtues might be able to accommodate animal-friendly values, ‘ecomasculinities’, and practical steps for a humane military – all issues which go beyond JWT.