Introduction

In 1922, Franz Kafka wrote a short story about a philosopher who asks probing questions about dog food. So obsessive is this philosopher that he becomes detached from his fellows, begins fasting, and eventually collapses. Kafka’s philosopher is, as it happens, canine – the story is ‘Investigations of a Dog’ (2018). Perhaps, then, the philosopher’s obsession is understandable. But should we – decidedly non-canine – philosophers care about animals’ food?

I became interested in the animals’ diets because they raised ethical dilemmas that (in my view) existing approaches to animal ethics and the philosophy of food hadn’t satisfactorily answered. But, in Just Fodder: The Ethics of Feeding Animals – as the thoughtful commentaries from Gary O’Brien, Angie Pepper, Clare Palmer, and Leon Borgdorf show – I used these dilemmas to develop a novel approach to animal ethics.

In brief, we owe all sentient animals certain negative obligations – specifically, they have negative rights – because (and to the extent that) they possess certain normatively significant interests. But we owe sentient animals positive obligations when (and to the extent that) we have certain normatively salient relationships with them. This framework consciously combines utilitarian, rights-based, and feminist approaches to animal ethics. Crucially, it can make sense of the idea that the obligations that we have – individually and collectively – to (and concerning) animals ‘close’ to us are different to the obligations we have to (and concerning) animals ‘distant’ from us. Perhaps some of the details of my relational account go awry. And I thank O’Brien, Pepper, Palmer, and Borgdorf for pushing at some of the more surprising claims I make. But I maintain that developing a relational complement to animal rights theory is worthwhile.

In this short response, I explore the comments of my respondents around five themes.

Extensionism

Methodologically, my work is extensionist. I take existing philosophical frameworks and concepts and extend them to include animals. Borgdorf holds that this method is ‘inherently non-ideal’ (Borgdorf 2024). I disagree. If we should adopt a given normative theory, extensionism can be a method of ideal theory. The animal ethicist needn’t commit to redesigning ethics from the ground up, in ideal or non-ideal theory.

In this vein, Borgdorf worries about my acceptance of ‘a certain property ideology’ in my discussion of ‘animal neighbours’: ‘the question emerges whether property rights should be extended to nonhuman animals’ and whether ‘the whole notion of property in itself can be questioned’ (Borgdorf 2024). I am happy to have both conversations.Footnote 1 But my questions about animal neighbours are meaningful even if we are sceptical about (animal) property rights.

Borgdorf calls for designating spaces as property of urban animals, encouraging animals to participate in the design of shared spaces, and offering animals the opportunity to choose the kind of relationships that they want to have with people. These are undoubtedly worthwhile conversations. In Just Fodder, I argue that we must include companion animals in political decision-making. But I accept that which animals count as ‘companion’ animals for this purpose is an open question, and, more broadly, we need further reflection on which animals’ views we should include in political decision-making.

Borgdorf suggests that we might complement my position – which appeals to relatively abstract and universal reasons to justify claims about what we owe to animals – with ‘internal’ reasons, which ‘emerge through reflection and guide moral actions based on the desire to be consistent with one’s moral beliefs’. These are ‘intuitive rather than based on reason’ (Borgdorf 2024).Footnote 2 But I believe that appeal to internal reasons has its limits.

My proposal to protect field animals impacted by arable agriculture by shifting to vertical agriculture ‘does not solve the ethical problem here and now’, as it is a ‘long-term solution’ (Borgdorf 2024). But Borgdorf is quick to dismiss other counters to ‘burger veganism’,Footnote 3 which is the idea that harm to field animals justifies meat-eating. Instead, Borgdorf argues that

…one needs to develop and habituate mechanisms of desensitisation to be able to consciously harm or kill cows or other sentient beings and continue performing this behaviour. In the killing of animals in fields during crop harvest, this [is] much less the case. There is little habituation of an overriding of one’s empathic intuitions taking place. Accordingly, the conscious killing of a cow both requires and builds a more vicious character than the accidental killing of “animal thieves”. (Borgdorf 2024)

But this isn’t a compelling reason to kill this animal rather than that. Granted, we plausibly do not need to desensitize ourselves to mouse-crushing as much as we do to cow-slaughtering.Footnote 4 But that doesn’t justify crushing mice rather than slaughtering cows. (Perhaps it could explain attitudes, but that’s not my primary concern.)

Appeals to ‘vicious character’ could offer justifications. But I doubt that (much of) the wrongness of killing cows lies in the impact that it has on our characters – or even the impact it has on the people we might treat badly because of worsened characters. Instead, I suggest it’s down to facts about the cows. It would be surprising to find out that animals’ interests matter primarily because of the impact they have on our characters – though, to be clear, I take it that this is not Borgdorf’s view.

Instead, Borgdorf is saying that these concerns about character are a factor in determining the relative wrongness of killing cows versus killing mice. I partially agree – for example, character is related to intentionality, which can (I believe) make a big moral difference. But affording too much weight to character can lead us to put great weight on morally irrelevant distinctions. Borgdorf points to the size of the cow and the bloodiness of cow slaughter as barriers that the would-be cow killer needs to overcome. But it seems morally arbitrary to have a higher line for killing an animal just because she is larger or has more blood than another. I’m considerably larger than, and presumably contain much more blood than, my mother. But I doubt we should be convinced that this is a good reason to kill her rather than me.

Agency

Pepper contends that I downplay the importance of animals’ agency. I confess that, unlike Pepper, I doubt that many animals have an interest in ‘self-determination’ that is ‘equally weighty’ (Pepper 2024) to their interests in not being killed or not suffering. It is, I believe, much worse for a pig to be killed or to be made to suffer than it is for the pig to not be able to determine details of her life. Nonetheless, it’s certainly possible that some animals have some interest in exercising agency.

Pepper rightly observes that many ethical puzzles with which I grapple wouldn’t arise if we embraced abolitionism. I can’t resolve the abolitionist/non-abolitionist divide here, but I note three things. First, Pepper is right that my critique of abolitionism in Just Fodder is far from conclusive. Second, Pepper is right that the harms perpetuated by (or in the name of) companions offer a compelling case for abolitionism independently of concern for companions themselves. Third, Pepper is right that abolitionism overcomes many problems with which I grapple – ‘conflicts largely manufactured by’ our continued relationships with other animals (Pepper 2024). If animals have rights but we should reject abolitionism, there are difficult questions to answer. Just Fodder is trying to answer them.

Let me put abolitionism aside. Pepper worries about my claim that guardians aren’t necessarily obliged to provide companions with what they prefer to eat. Pepper instead argues that ‘companion animals have (or could have if appropriately supported) a right to determine for themselves what they eat, when they eat, where they eat, and with whom they eat’ (Pepper 2024, emphasis Pepper’s). In this domain, animals’ wills must ‘be recognised as normatively authoritative … the agent’s will ought to be regarded as decisive, such that it is generally impermissible for others to interfere in the agent’s choices or actions’ (Healey and Pepper 2021, 1226, emphasis Healey and Pepper’s).

This has its limits. We can, Pepper says, override companion choices that aren’t ‘decent’ (Pepper 2024), including choices bad for them or (presumably) others. And surely there are cases in which respect for choice is infeasible. For example, there’s little room for choice if I only have access to one ‘decent’ pet food.

Even beyond these cases, I contend that companion choice shouldn’t always be decisive. In Just Fodder, I conceptualize companion animals as dependent members of families. And families can collectively choose – including on behalf of dependent members – to eat one way rather than another. Some families sit at the table to eat, some sit in front of the TV. Some eat restricted diets, others eat lavishly. Some eat their main meal at noon, some eat it in the evening. That’s OK. This does limit the self-determination of family members. But that’s part of being in a family.

I’m sympathetic to three (interlinked) counters Pepper could offer.Footnote 5 First, if families limit choice in this domain, perhaps they must allow animals to self-determine in other domains. Second, as family members, companion animals should influence family norms. Third – the core of the abolitionist worry – there must be some capacity for companion animals to leave families.

This is tricky territory. More thought is required.

Predation

Pepper and Palmer challenge my claims about our occasional responsibility for animals’ predatory actions. If, as I argue, wildlife rehabilitation centres (WRCs) should not release predators they’ve rescued, why (Pepper asks) is it permissible for us to feed predators in our gardens? In both cases, we assist predators, meaning prey potentially have a complaint against us.

Could WRC personnel be morally responsible for harm when (or to a degree that) householders aren’t?

In Just Fodder, I don’t offer a full account of moral responsibility. But I do point towards elements a full account would contain. First, whether an (individual or collective) agent A is responsible for situation x is a matter of degree. Second, A’s responsibility for x depends upon a range of factors including, but not limited to: the extent of A’s moral agency; the extent of A’s causal responsibility for x; the extent to which A intended x; and the foreseeability of x for A.Footnote 6 Third, A’s responsibility for x bears on whether x involves injustice.

This means that there’s no clear line distinguishing when we’re responsible for predation and when we aren’t. The assistance we provide to a wolf can be anything from throwing her food scraps to years of involved care in a WRC. But a satisfactory account of moral responsibility will, I contend, be able to recognize differences between various cases.Footnote 7

My vision of moral responsibility is not, I think, as expansive as Palmer worries. WRC personnel are responsible for a released wolf’s actions to a degree, but this doesn’t mean that they’re responsible for everything this wolf (and her offspring) do. A plausible account of moral responsibility must allow that moral responsibility comes in degrees – and, at some point, the degree of moral responsibility will be sufficiently small that we shouldn’t worry about it. De minimis non curat lex – the law does not concern itself with trifles.

Palmer further observes that, by my own logic, I should oppose the release of prey, lest we become responsible for their deaths in predators’ claws. I’m not convinced the cases are analogous. Palmer notes that in rehabilitating a doe, we enter a relationship of hospitality with her. But one of the interesting things about hospitality – as I explore in Just Fodder – is that it does not carry with it an assumption of lasting care. I have responsibility for Palmer if I invite her to my office or (especially) home for tea. But after our goodbyes, my hospitality-linked obligations towards her quickly dissolve. In rehabilitating a doe who’d otherwise die of exposure, we do increase her risk of dying of predation – but only in the same way that the doctor who resuscitates me after a heart attack increases my chance of dying in a car crash. I suggest that, in releasing the doe, we’re not imposing risk, but ceasing assistance. These are, I think, morally quite different activities.

Withdrawing assistance can sometimes be wrongful, of course. Just as we can’t ‘release’ our companions to fend for themselves against predators, so we shouldn’t release any wild animals who have become relevantly ‘companion-like’. But our relationships with animals in WRCs will not usually be like this – for instance, relatively few of these animals will develop (echoing Palmer) ‘internal dependencies’, which partially account for our special obligations towards companions.

Another challenge Palmer offers is that rehabilitating a wolf who humans have harmed and then releasing her might be analogous to permissibly moving risk around (like, e.g., translocating a predator) rather than imposing risk. I am sympathetic to this perspective. As noted in Just Fodder, I’ve already changed my mind on this at least once. I concede that the source of the harm that led to an animal arriving at a WRC may make a great deal more difference than I allowed in the book.

Both Palmer and Pepper worry that I downplay the impact of permanent captivity on predators. Pepper worries that this creates a rights conflict between predator and prey, and (rightly) says that we should not assume that prey interests will win out.Footnote 8

As Martha Nussbaum (2023) has recently argued, we should seek to avoid such conflicts altogether. Rather than condemning captivity, this invites us to reflect on ways that ‘captivity’ could be compatible with the flourishing of predators. What kinds of community will suit them, us, and (crucially) other animals? It’s feasible that we could design an institution where ‘captive’ owls flourish while eating only respectfully produced food – but I do worry that mice are never going to flourish in owl talons. Rather than trying to decide whose interests ‘win’, we should be trying to imagine ways of living that don’t necessitate these conflicts.

Again, this brings us into complicated territory. But unless we’re going to ignore the metaphorical – but reasonable – complaints of prey animals, it’s territory that we must broach.

Interventionism

I first approached the issue of wild animal suffering before beginning Just Fodder (see, e.g., Milburn 2015). Then, I saw claims about wild animal suffering as arguments against animal rights. But that framing is old-fashioned now – animal ethicists have offered sophisticated defences of intervening in nature. I remain sceptical about intervention, but I take the case seriously.

In Just Fodder, I argue that we may only protect gazelles if we can do so while respecting the rights of lions. Borgdorf challenges this by asking why ‘the negative duty towards the lion’ outweighs ‘the positive duty towards the gazelle’ (Borgdorf 2024). For me, the answer is straightforward. The lion has negative rights based on her interests. On the other hand, we might have a supererogatory duty of beneficence towards the gazelle. A right trumps a (possible) supererogatory duty.

Borgdorf makes a more compelling challenge to my claim. Even though the lion is a rights-bearer, we might permissibly harm (innocent!) rights-bearers in defence of innocent third-parties. I didn’t give this possibility the attention it warrants in Just Fodder. I hope to address the violent defence of animals at some length in the future, so hope Borgdorf will forgive me for saying little of substance about it here.

O’Brien, too, questions my mostly hands-off approach to wild animals in a particularly ingenious way. He suggests that given the histories that humans have with wild animals, we are relevantly ‘entangled’ with them, and thus have positive obligations towards them, just as we have positive obligations towards companion animals.

I agree with O’Brien that (what I’ve called) ‘relational non-interventionism’ tends to become more interventionist in practice than in theory, given the messy realities of our entanglements with wild animals (see, e.g., Milburn 2021). But what’s striking (worrying?) about O’Brien’s case is that he argues that, given humans’ historical and pervasive influence on ecosystems, we are relevantly entangled with (almost?) all wild animals.

O’Brien observes that I hold that the relationships we have with animals can be individual or ‘collective, political ones, grounded on our membership of morally significant groups such as states, or, perhaps, the human species’ (O'Brien 2023, 2). I confess to some scepticism about the idea that the human species is a group or agent in the morally salient sense. Humans qua species members are what Iris Young calls an ‘aggregate’ – humans qua species members are a collective ‘classification … according to some attribute’ (2011, 43). Aggregates carry with them neither an assumption of community/agency, unlike formal ‘associations’ (Young 2011, 44), nor shared interests, like Young’s ‘groups’. So perhaps they shouldn’t carry weight in our decision making.

This, though, calls into question my claims in Just Fodder about the significance of domestication. If (say) I reject O’Brien’s comments on prehistoric humans’ ‘overkilling’ as ethically irrelevant today, on what grounds could I claim that prehistoric humans’ domestication of animals is ethically significant?

I respond that the initial ‘act’ of domestication is not what morally matters. Instead, the ongoing domestication process – including the concomitant shaping of animals – is what matters. Separate an ancient domestication from the ongoing process, and it starts to seem ethically unimportant. As O’Brien notes, ‘[t]here seems to be little connection between us and our distant ancestors, and it sounds harsh that we would inherit the guilt for actions they performed’ (2023, 3). I agree, but when it comes to ‘inheriting’ moral responsibility, a black-and-white, all-or-nothing approach is surely not the most desirable.

O’Brien is surely right that ‘[i]f we think that animals have a right to their habitats’ – and, speaking generally, I do – then the way we have taken land from animals ‘constitutes a rights violation on a massive scale’ (2023, 6). Thus, O’Brien suggests, we ‘have some kind of duty to pay [animals] reparations’ (2023, 7).

This is a regrettably undertheorized question in animal ethics. Sue Donaldson and Will Kymlicka suggest that our focus should be on ‘forward-looking justice’ complemented by acknowledgement of past wrongs and ‘symbolic compensation’ (2011, 196). This is compatible with O’Brien’s suggestion that we offer reparations by ‘try[ing] to increase the welfare of wild animals suffering from various naturogenic harms’ (2023, 7). This is a compelling suggestion. If wild animals are owed reparations, rights-respecting actions to improve their welfare are good candidates.

O’Brien argues that ‘we should err on the side of taking too much, rather than too little, responsibility for our actions’ (2023, 5). This is compelling in the abstract, but I note that the potential costs of taking too much responsibility are comparatively low, while the potential costs of becoming too involved in the lives of wild animals is catastrophic. Wild animals have voted with their feet in living separately from us (Donaldson and Kymlicka 2011, 66), while ham-fisted interventions run the risk of exacerbating the harm they face. And if we start getting involved in the lives of animals unnecessarily, we then take on some responsibility for their plight.

To be blunt, sometimes we should mind our own business. Less bluntly, it’s plausible that we should err on the side of taking too much responsibility, but err on the side of taking too little action. Once we get involved, the cat is out of the bag, and it can be very hard to put her back. Or, as Palmer notes, once we start involving ourselves in the lives of wild animals, we’re damned if we do, and damned if we don’t (Palmer 2024, 8).

Pepper, too, worries that my account downplays duties of assistance owed to wild animals – in particular, she worries that the distinction I draw between animal friends and enemies is unsustainable. After all, the animals who come into our spaces don’t know whether they’re the animals we’ve invited by putting out seeds or one of the animals we’re trying to discourage by putting the seeds in feeders (Pepper 2024). Pepper writes that

If you want to feed some free-living animals, and potentially develop a relationship of dependency with them, then you have to accept the cost of being in a similar relationship with other animals that you may be less keen on. If that’s a cost that is too much to bear, then you should not start feeding any animals. Problem solved. (Pepper 2024)

This is certainly a position we could take.Footnote 9 Pepper thinks that ‘[b]y placing nuts and seeds in his garden [someone] is in effect welcoming all nonhuman neighbours with an interest in nuts and seeds’ (Pepper 2024). But pretty much anything we might do in our garden might attract animals. My flowers attract pollinators, and they attract bats. My trees attract birds. My compost heap attracts a whole ecosystem of invertebrates. Foxes dig in my beds. My decking – I think – shelters mice, while toads live under my shed. Slugs come to my lawn, attracting hedgehogs. And so on.

I’m personally open to taking on a minimal responsibility for these animals. But someone else might not be happy at the presence of these animals, and might have (unsuccessfully) tried to deter them. Can we really hold them responsible for these animals’ wellbeing?

If we can’t draw lines somehow, we become responsible for the wellbeing of basically all animals around us. The lines I’ve drawn arise partially through my engagement with hospitality, an apt idea from the philosophy of food. I welcome other proposals.

Environmentalism

Both Palmer and O’Brien point to how my account has some surprising consequences for many environmentalist practices and proposals.

Palmer observes that my approach entails that ‘only sentient animals matter in environmental policy’ (Palmer 2024, 8, emphasis Palmer’s), and thus that my position cannot account for many of the values that motivate environmentalism. I allow that there may be values beyond concern for sentient animals that should motivate environmental policy, but it’s undeniable that animal rights place roadblocks on certain (possible or actual) environmentalist policies. For example, disputes about the treatment of ‘invasive’ species are a familiar point of tension between animal advocates and environmentalists. I suggest that my approach is not going to be as hostile to some environmentalist claims as many approaches to animal ethics, including those that advocate wholescale interventionism. But I do think that my approach can offer an alternative and worthwhile view on certain environmental practices.

O’Brien discusses the deliberate introduction of animals as a morally salient entanglement between humans and wild animals, potentially grounding obligations to intervene on behalf of the introduced animals or on behalf of the non-introduced animals they impact (O’Brien 2023, 5–6). I broadly agree (cf. Milburn 2021, 18), but the challenge of a relational account is that, necessarily, we must look to the particularities of a situation to attribute (degrees of) responsibility. To use O’Brien’s own examples, perhaps the introduction of dingoes to Australia 8,500 years ago doesn’t carry much normative clout today – while the introduction of rabbits in 1859 may well do. But to look to the present, I confess that I am uneasy about (for instance) proposals to reintroduce predatory wolves to the United Kingdom, even while I am cautiously optimistic about the reintroduction of herbivorous beavers.

Palmer, meanwhile, raises the prospect of passive rewilding – something that, in Just Fodder, I am open to in many cases. In cases of passive rewilding, predation is foreseeably (and perhaps intentionally, insofar as thriving ecosystems are desired) increased. Should I thus oppose passive rewilding?

In Just Fodder, I argue that we can sometimes have a hand in creating an ecosystem, but still have little responsibility for the predation that takes place. The question Palmer poses is whether merely foreseeing that our actions will lead to an increase in predation (because of an increase of animal populationFootnote 10) makes us minimally responsible for said predation. I am not so sure, even if I do think active rewilding might make us more responsible for animals’ fates. Perhaps, in the case of active rewilding, we should ask whether the ecosystems actively created are animal-friendly.Footnote 11

To adapt one of O’Brien’s examples: Imagine possible government policies A, B, and C. A predictably leads to steady human population increases; B creates new towns with predictably high crime levels; and C creates new towns with predictably low crime levels. We would not hold a government favouring A responsible for increased crime, even though predictably increasing population will (all else equal) predictably increase the number of crimes committed. But we might hold a government that favours A + B rather than A + C responsible for increasing crime levels. Favouring passive rewilding is like favouring A; favouring active rewilding could be like favouring A + B or A + C.

In any case, we should be suspicious about any government that aims to create a thriving criminal underworld. It’s hard to imagine that as a goal compatible with human rights, whatever (cultural, aesthetic, economic…) merits a thriving criminal underworld might have. Similarly – I tentatively suggest – there might be something suspect about wanting to bring predation about, in the name of creating a thriving ecosystem. That’s not a goal that places respect for animal rights at its core, even if – I maintain (and hope) – there are many instances of wild animal predation that do not involve rights violations.

In short, what relational approaches means for rewilding, reintroductions, captive breeding, and other environmentalist policies is a worthwhile question for future research. My sense is that this could be enlightening for understanding the ethics of our relationships with ‘wild’ animals – or at least setting out possible approaches.

Conclusion

In Just Fodder, I aim to find a middle way through some possible extremes. For example, when it comes to domesticated animals, I aim to make sense of the idea that we can live with animals while still maintaining that these animals, and many of the animals who we might kill to feed them, have rights. And when it comes to wild animals, I aim to make sense of the idea that we aren’t obliged remodel nature to feed and protect animals, without endorsing the idea that it’s unproblematic for us to impose serious risks on them.

Balancing acts are always going to invite critique. This will include critique from advocates of the ‘extremes’ I seek to avoid. Abolitionists would see us do away with domesticated animals; welfarists won’t lose sleep over killing animals to make pet food. Interventionists will call for us to remodel nature; and environmentalists will welcome ‘nature’ red in tooth and claw, even when we’ve a hand in that bloodshed. Perhaps these critics are right. Perhaps, for instance (as Pepper presses) the animal rights position must ultimately tend towards abolitionistm. Or perhaps (as O’Brien presses) the pro-animal position must ultimately tend towards interventionism.

But these balancing acts also invite critique from people who might draw lines in different places or in different ways. Perhaps Palmer, with her criticism of my ‘moral over-extension’ (Palmer 2024, 1), will draw lines in different places. Or perhaps (echoing Borgdorf) someone opposed to ethical extensionism might seek a new method of line-drawing altogether.

Such criticisms are welcome. For me, philosophical publishing is about contributing to conversations. Pepper suggests that Just Fodder – and these reflections on it – are ‘only the start of the conversation’ (Pepper 2024). I hope she is right.