You’re living alone on a tropical island paradise. The tasty and varied local plant life provides all the nutrition you need (let’s suppose). You befriend a local family of pigs, often playing with them or watching these smart beings play with each other. Then one day you decide to strangle one of the pigs to death, cook it and eat it: not out of necessity but just to see how it tastes, or perhaps because it seemed slightly easier than foraging that day. Was your action immoral?
This post explores the big philosophical questions of animal welfare (and indeed of morality more broadly). In particular, it asks whether there is a philosophical framework that should make us care about farmed animal ‘welfare’ but not about slaughtering those animals for food. That combination of views is, I think, now quite dominant: mixing 1) a Darwinian appreciation that there is no fundamental difference in kind between humans and other animals; 2) the idea that morality exists: some actions are wrong, in general including killing humans for food; and 3) that it is reasonable to routinely slaughter (some) other animals for food. But – despite how ‘obvious’ all of these views seem to most people – the philosophical underpinnings for this combination are not obvious at all.
While in practice it’s rare for people to have internally consistent philosophies, below I try to classify belief systems into five categories, and I hope you will consider which one(s) you favour. (Let me know if you can articulate others.)
- Option 1: we have strong deontological obligations to non-human animals
- Option 2: we have no moral obligations to non-human animals
- Option 3: we have a mix of deontological obligations to non-human animals, in which non-productive harm is strictly prohibited but harm in the production of food and certain other goods is permitted
- Option 4: prior existence utilitarianism
- Option 5: total view utilitarianism
meat bulk of this blog is about utilitarian ethics (options 4 and 5), in which our actions are only judged in terms of their consequences for welfare (and in which the welfare of non-human animals is generally counted too). But there are of course many other world views out there, so first we look at the non-consequentialist and/or speciesist approaches to animal welfare.
Option 1: we have strong deontological obligations to non-human animals
I’ll begin with philosophies that give a straightforward answer to the question of whether it is ok to rear and slaughter animals for food. The ‘animal rights’ view is clear: there are some actions, such as killing, that we should never do; and these prohibitions extend to at least some non-human animals.
Gary Francione writes that “For the most part, when I refer to animal rights, I am really referring to one right: the right not to be treated as the property of humans.” Personally, as a utilitarian I believe that it might be worthwhile in some circumstances to e.g. rear, own and kill mammals for medical research. But, as rights go, the prohibition on ownership is attractive. The idea that sentient beings should not own other sentient beings is admirably clear and fits so well with modern notions of freedom and inalienable rights. (Given the harms that have so often stemmed from treating beings as property, it seems a prohibition that rule utilitarians could easily get on board with.)
In these views, the notion of ‘high-welfare’ animal husbandry and ‘humane’ slaughter is laughably misguided (in the same way that you might view high-standards dog milk production, or high-welfare human slavery). Tom Regan neatly presents this view in saying “the truth of animal rights requires empty cages, not larger cages.”
Option 2: we have no moral obligations to non-human animals
On the one hand this second option sounds (to me) quite an extreme position that few people would ascribe to. But on the other hand it is implied by many philosophies. I’d include here, at the extreme, egoistic and nihilistic theories that would deny we have any moral obligations to other humans, let alone non-humans. One could also simply (unreasonably) assume that only humans are conscious and therefore other animals are all – like mushrooms – incapable of suffering, despite being our distant cousins.
Looking at the contractualist philosophies, which struggle to extend any ethical consideration to non-human animals, Kant said that “so far as animals are concerned, we have no direct duties. Animals are not self-conscious and are there merely as a means to an end. That end is man. […] Our duties toward animals are merely indirect duties toward humanity.” Similarly, “[John] Rawls has little to say about animals in A Theory of Justice. He writes that, at least until arguments to the contrary can be devised, human ‘conduct toward animals is not regulated by’ the principles of justice, because only ‘moral persons’ are ‘entitled to equal justice’. This moral personhood is distinguished by two features. First, moral persons ‘are capable of having … a conception of their good (as expressed by a rational plan of life); and second they are capable of having … a sense of justice, a normally effective desire to apply and to act upon the principles of justice, at least to a certain minimum degree’. So, only those who can understand what it is to be just, and those who are able to claim it for themselves and respect the rights of others, are entitled to be beneficiaries of justice. … In other words, whatever else can be hidden behind the veil of ignorance, species membership is not one of them.” (Though this last claim is debated, in the same commentary.)
These views imply, on a basic level, that it doesn’t matter what we do to non-human animals. Torturing a puppy or eating an octopus alive are not immoral acts, as those animals lie beyond the bounds of morality. “Suppose, for example, you torture a dog. You have not, in Kant’s view, violated any obligations you owe to the dog. You have done the dog no wrong. A dog is not the sort of thing that can be wronged.” So, while these theories certainly allow the consumption of animal-based products, they do not require us to care about animal welfare at the same time. They do not explain why “we are not allowed to kick farmed animals, while we are allowed to kill them”.
Kant and others do propose some reasons why torturing animals might actually be bad, but only based on how those actions affect humans rather than accepting the animals themselves suffer and that that is bad. Peter Carruthers writes that “animals must fail to be accorded direct rights, through failing to qualify as rational agents. While contractualism allows that we do have duties towards animals, these only arise indirectly – on the one hand, out of respect for the feelings of animal lovers, and on the other hand, through the good or bad qualities of character that animals may evoke in us.” My view is that these are pretty weak attempts to avoid the unattractive conclusion that torturing animals is fine. I also suspect that most people think that “animal welfare matters because it matters for animals”, as Tatjana Višak puts it, so the Kantian approach would not be palatable to most people today.
Option 3: we have a mix of deontological obligations to non-human animals, in which non-productive harm is strictly prohibited but harm in the production of food and certain other goods is permitted
The two options above suggest that either it’s always bad to harm animals or that it’s never bad. But is there a set of deontological beliefs that says it’s bad except for purposes such as food production?
As with all deontological views, we might ask: “where do rules come from?” and “what exactly are the rules?”. For example, religions may specify that animal suffering is to be avoided where possible but also that producing and eating meat is permissible. If you are not religious, such prescriptions are not very useful for figuring out ethics. But even for the religious, how we should aspire to behave is not always obvious. In the Bible, for example: Genesis 1:26 gives man “dominion” over animals; but Genesis 1:29 suggests man should be vegan; then Genesis 9:3 gives Noah and co. the right to eat “every moving thing”; then Leviticus 11:7 forbids the eating of pigs, for example; but then the New Testament seems to repeal those same rules. Should Christians and Jews conclude that ideally we would all be vegan, based on Genesis Chapter 1 (and Isaiah Chapter 11)? How much concern, if any, should they have about animal welfare? None of this seems clear. So even for those who have religious texts to guide them, both the spirit and letter of the moral code are often unclear.
For those without religious texts, determining what actions are right or wrong is clearly even more challenging in some ways. We can construct a rule such as “causing unnecessary animal suffering is bad, but causing necessary animal suffering is permitted” – and indeed this is very similar to the Animal Welfare Act 2006 in British law. But what counts as ‘necessary’ suffering? In a rich society where varied plant-based nutrition (and B12 supplementation) is readily available, is it ‘necessary’ to artificially inseminate cows and then take away their calves in the name of milk production? Would it still be ‘necessary’ to transport pigs to slaughterhouses and lower them into gas chambers if we could grow pork in vats?
We could construct a rule whereby it is generally forbidden to harm animals but it is permitted if it is instrumental to productive activity – e.g. for food, leather, wool and medical research. But the ultimate moral basis for this is unclear if we’re not taking a utilitarian, contractualistic or mainstream religious approach. It would therefore not be clear why food production should be exempted but, say, cock fighting, fox tossing or fur coat production should not. One may also wonder what this moral approach would say about meat that is wasted rather than eaten, or about animals that have to be incinerated or turned into biodiesel rather than food.
I would welcome any pointers towards common (or uncommon) moral theories that give us strict duties not to harm animals but also exemptions for those harms that are most useful to humans. My suspicion is that any such theory would at heart really just be a representation of norms, i.e. activities that are currently normal are unquestionably permitted, while activities that are currently not normal are frowned upon; or, that they are really just utilitarianism, whereby causing harm is certainly permitted if it causes a greater amount of good.
If you want a moral theory that really cares about animal welfare but also allows animals to be used and harmed for human ends, it is perhaps best to focus on utilitarianism and its varieties.
Option 4: (Wide) Prior Existence Utilitarianism
In contrast to the ‘deontological’ viewpoints above, utilitarianism aims to maximise well-being – including minimising suffering – by any means. Actions (e.g. killing or lying) are not right or wrong in themselves, but only judged based on their consequences. So utilitarianism is perhaps a more promising avenue for providing a framework that cares about animals but can also justify them being killed if it is for a good reason. (This is the starting point for Tatjana Višak’s book, Killing Happy Animals, on which this post draws heavily.) Note that there are versions of utilitarianism that aim to maximise average well-being, or that prioritise increasing utility for those with the lowest welfare; but for now let’s focus on theories that, in abstract, would ‘simply’ add up the net well-being of all sentient beings and try to maximise that value.
The ‘total view’ of utilitarianism, discussed below, takes this adding up and maximising quite literally. It pushes us to fill the universe with as much happiness (and/or preference-satisfaction) as possible, and argues that a world with 20 billion happy people would be twice as good as a world with 10 billion happy people, all else equal. This approach has its attractions and possible disadvantages.
This is not an undisputed view within utilitarianism, however. In contrast to the ‘total view’, ‘prior existence utilitarianism’ seeks only to maximise the well-being of existing beings and (in the ‘wide’ prior existence view) also beings who will exist independently of whatever choice is being decided. It does not count the possible welfare of ‘contingent beings’. The distinction with the total view may seem minor but it is very important: not least for whether individuals and societies are obligated to try and maximise the number of children they produce, and – as we will return to soon below – whether we need to maximise how many sentient, happy non-human animals there are in the world.
Focusing for a moment on the details of the philosophy, Tatjana Višak does a good job of arguing that this view is a coherent one (she doesn’t claim to have settled whether it is the correct view). A key question is whether you can benefit a being by bringing it into existence: Višak’s view is that you logically cannot, as there is no being to benefit if it doesn’t yet exist.
(One major counterargument to the person view has been the ‘expected misery’ or ‘wretched child’ argument. This imagines the example of someone bringing a being into existence even though they know that being will live a life that isn’t worth living, e.g. a very sick child. Given that the the prior existence view says we aren’t obliged to create more happy beings, it conversely struggles to argue against creating irredeemably unhappy beings. Višak presents a counter-argument that involves assessing the ‘character’ of anyone who’d deliberately create an unhappy being (i.e. it would be bad to be the kind of person who thought that was a good idea). My own response to the expected misery argument would be to emphasise that utilitarianism would oblige people to relieve the new being’s suffering as soon as possible, or else to euthanise it. The thought experiment therefore tends to specify that such relief or euthanasia is impossible, but – in reality – we can’t know such things in advance. We might therefore distinguish between 1) someone deciding “I will create a suffering being, and then do nothing to try and relieve that suffering”, which is clearly an immoral stance even under person view utilitarianism; and 2) someone deciding “I will create a suffering being, and then strive to reduce that suffering, including by trying to euthanise that being if necessary”, which is deeply eccentric but perhaps not as wrong as the thought experiment tends to suggest. In practice, the problem of people conceiving unhappy children/AIs/animals, and then being obliged to try and resolve that unhappiness as soon as those beings exist, actually seems quite an unlikely, minor concern relative to the problems that other philosophies can pose when pushed to extremes.)
To get back to animal welfare, the prior existence view – which does not value the creation of more beings – does not seem favourable to animal agriculture. Višak is clear that if we focus on the welfare of existing beings then the pleasure you get from eating meat – relative to an alternative food – does not outweigh the utility that a (happy) pig gets from living another day or year. As Peter Singer says, “Since […] none of these practices cater for anything more than our pleasures of taste, our practice of rearing and killing other animals in order to eat them is a clear instance of the sacrifice of the most important interests of other beings in order to satisfy trivial interests of our own. To avoid speciesism we must stop this practice, and each of us has a moral obligation to cease supporting the practice.” To be clear, all forms of utilitarianism allow animals to be killed in some circumstances, for the greater good (and the calculations may differ if you are facing severe poverty and malnutrition), but bacon being slightly tastier than some alternative is simply not a sufficient cause.
One counter-argument here may be that if we could kill animals without pain or stress, then perhaps killing them does not actually harm them in a morally relevant way: that the animals in question (e.g. poultry, pigs, cattle, sheep) have no expectations of the future or no concept of death and so this is not a concern that needs to be incorporated in our utilitarian calculations. However, this is a very controversial view (not least because it raises questions about killing human babies, for example). Given that the perceived benefit of killing animals to eat them is generally so marginal – merely a question of taste or habit – I think you would need to be 100% certain that the animals in question have zero concern for their future (and that death would be 100% painless) to make it a welfare-enhancing trade-off, which does not seem a reasonable view at present.
Option 5: Total View Utilitarianism
In total view utilitarianism, as discussed above, welfare can be increased not just by helping existing beings but by creating additional ones. This also suggests that killing one animal but replacing it with another can perhaps be morally neutral. Superficially at least, total view utilitarianism is a plausible contender for a theory that cares about other animals but also allows (or even demands) animal agriculture. The argument here is the ‘logic of the larder‘: that eating pigs requires the existence of more pigs and, so long as these animals live lives worth living, this increases welfare. Another way of putting this may be that there is an unspoken contract between you and the animal that you eat: it provides you with food, and thanks to you (retrospectively) it gets to live a net-positive life rather than not existing at all. (I believe this is broadly what I thought when I ate meat and dairy.)
Before getting into the specifics of the logic of the larder, however, I would ask those attracted to it if they are really total view utilitarians. Are you happy with the ‘repugnant conclusion‘? Have you considered having as many children as possible, to maximise the net value you create, or supporting natalist public policies to encourage others to have more children? Do you see it as overwhelming important to start filling the apparently-empty cosmos with sentient (but possibly AI) life? Are you open to the idea – often explored in fiction – of rearing happy humans to be organ donors (or perhaps even food)? I think these are (mostly) all defensible positions, but most meat eaters – while paying lip service to total view utilitarianism when it comes to animal agriculture – would probably say these are crazy ideas.
The logic of the larder also immediately faces some big counter-arguments. Farm animals are not created in a vacuum: we are really talking about replacing wild land with farmed pasture and/or cropland (for feed), and therefore also affecting the number of wild animals. Taking this broader view, Matheny and Chen 2005 shows that farming cows or pigs (relative to soybean farming for equivalent protein production) reduces the number of mammals and birds in existence (see table)!
Indiscriminate animal eating therefore makes no sense on this logic: only poultry farming would be justifiable. On the numbers above, for example, getting 100g of protein from beef rather than soybeans is equivalent to reducing a non-specific animal’s life by 24 days. Furthermore, as the paper suggests, really following this approach means asking whether we should instead, for example, farm great numbers of mice – either for food or not – to maximise the number of happy beings: “by eating beans rather than chicken, and investing our saved money in raising colonies of mice, we could create on the order of 50 times as many life-years per dollar invested.” More fundamentally, I would also argue that it’s simply implausible that some forms of farming could ever be considered a reasonable way of increasing the number of happy beings in existence: if you started from a goal of increasing aggregate animal welfare, you would not invent the dairy industry, for example, given its unavoidable routine of artificial impregnation, separation and physically-demanding production.
All of this leads on, however, to the even thornier question of whether wild animals may in fact be living lives that aren’t worth living. What if those displaced wild mammals and birds are actually suffering? Or what if invertebrates are also sentient and suffering? Now the calculation may be that we want as few wild animals as possible, and replacing wilderness with fairly barren fields and sheds full of controlled animals is the utilitarian choice. I find the idea that the world is full of the silent screams of in-pain, starving, cold, terrified invertebrates horrifying but quite plausible. But even if we do accept this concern, we must again note that most meat eaters would say that destroying wilderness to try and reduce insect and worm suffering is an insane goal to work towards, let alone what drives them to eat meat.
Utilitarians must also weigh up other concerns. In particular, there is a clear climate and biodiversity case for reducing meat consumption (and increasing wild land cover). There is also a personal health case for eating less meat (possibly to the extent that the rational thing to do would be to stop eating meat even if you enjoy it and ignore all other arguments). And there’s also a pandemic risk-reduction argument for minimising animal farming and hunting. So, although I think we should worry about wild animal suffering, here in the 2020s I think it would be dangerously overconfident to argue on utilitarian grounds for the immediate destruction of more rainforests etc. and their replacement with animal farms. The committed total view utilitarian might also wonder whether meat’s high land and resource requirements (not to mention health impacts) are a constraint on human population growth, and whether perhaps the better outcome would be a world of more humans but fewer farm animals.
In contrast to earlier centuries, there is now a strong consensus (in many countries) that other mammals and birds are sentient beings and that higher animal welfare in farming is desirable: that we should care about the quality of their lives and deaths. But if we really respect and care about those creatures as sentient individuals (or ‘subjects of a life’) then, even if we manage to deliver happy lives and instant, painless deaths, I don’t think it is obvious that bringing those happy lives to an early end is morally justified.
It seems to me that to stand on relatively solid philosophical foundations, people (at least people living in a rich time and place) need to either:
- Be vegan, either on utilitarian or rights-based grounds, or at least agree that this is the right direction of travel even if, like a utilitarian who recognises in theory the case for giving what we can or donating a kidney, you fall short of the ideal in practice;
- Argue that non-humans have no moral standing, in which case we should not care about animal welfare at all; or
- Be a committed total view utilitarian, theoretically in favour of massive population growth, who believes that – on balance, and accounting for the suffering of wild animals (possibly including invertebrates) and environmental impacts – replacing wild land with (happy) farmed animals and their feed is a reasonably effective way to increase the expected amount of well-being in the universe.
To be clear, I respect the latter perspective and could come to the same conclusion if I see the balance of evidence (and farm animal welfare standards) change. But what I think is most interesting is that I suspect most people would not agree that they fall into any of those three boxes. Maybe there are simply ‘more philosophies in heaven and earth than are dreamed of in this blog’ – in which case, let me know. But it seems to me as though society’s assessment of other animals has changed significantly and left conventional norms (i.e. using animals for food) out of step with prevailing philosophies. Animal-loving meat-eaters either need some new philosophies, a well-evidenced bolstering of the ‘logic of the larder’ (not least through vastly improved farm animal welfare standards), or to throw in the towel.