Skip to content

Is Animal Liberation Speciesist?

Written by Joseph Moore

This year, Peter Singer published Animal Liberation Now, a significantly updated version of his 1975 animal rights classic. Both the original and revised text argue that humans should refrain from inflicting unnecessary suffering on non-human animals, especially the cruel practices still commonly employed in factory farming and animal experimentation. And as a step towards this collective action, Singer urges his readers to modify their individual purchasing practices by preferring cruelty-free products or, even better, committing to vegetarianism or veganism.

The bulk of the revisions in the new edition concern the empirical facts on the ground, both the positive changes in the treatment of non-human animals since the original printing as well as ongoing, legally sanctioned cruel practices. Unfortunately, the philosophically weakest part of Singer’s influential argument, which occurs in the first chapter, has received no additional support in this edition. This is his claim that ‘the capacity for suffering and enjoyment is a prerequisite for having interests, a condition that must be satisfied before we can properly speak of interests at all’. The supposed necessity of sentience for having interests is why Singer limits his ‘principle of equal consideration of interests’ to (some) animals and does not extend it to living things in other kingdoms—plants, fungi, bacteria, etc.—or other kinds of subjects. But this relatively undefended assertion was dubious in 1975 and is even more dubious now. Singer’s restriction of interests to sentient beings is just as arbitrary as the speciesism he decries.

What Singer calls ‘interests’ are now more commonly described as contributors to well-being or welfare or prudential value. By any name, the concept is that of what is (non-instrumentally) good for or beneficial to a given subject, i.e., what makes that subject better off. (To avoid begging questions, understand ‘subject’ in a syntactical sense, like the subject of a sentence, rather than a psychological sense involving conscious ‘subjectivity’.) There is now a rather large philosophical literature on well-being and the necessary and sufficient conditions for contributors to well-being. Singer’s hedonism, the view that only pleasure is fundamentally good for subjects and only pain fundamentally bad, is just one theory of well-being among many others.

At least one other theory of goodness-for-a-subject is almost as old and just as influential. According to Aristotle’s naturalist perfectionism, it is good for every kind of living thing (as well as for at least some collectives, such as human city-states) to fulfil its natural function based upon its distinctive capacities. Derivatively, it is also good for every kind of living thing to have internal states (e.g., bodily integrity) and external resources (e.g., sources of nutrition) that enable it to perform its natural function well. And, in fact, Aristotle’s general biological theory explains how we know that all living things have their own forms of well-being. Aristotle claims that it is distinctive of all living things—plants, fungi, animals and, if he were aware of them, single-celled organisms—to have capacities and needs for nutrition, growth, self-maintenance and reproduction. Further, it is distinctive of all animals to have capacities for sensation, perception and locomotion and to have corresponding sensory, desiderative, emotional and social needs. While what is distinctive of humans is (supposedly) our rationality, we still share in all of these common animal and ‘vegetative’ needs. We already accept that it is good for humans to meet these needs, to survive and reproduce and avoid serious pain and engage in social relationships. Well, it is just as good for other animals and other living things to meet these same, common needs. And notice how this is just like Singer’s argument that non-human animal pain matters morally: we know human pain matters and non-human animals can experience the same kind of pain, so theirs matters just as much.

Now, Aristotle’s views on the subject are important and influential enough in the history of philosophy that his theory of goodness or ‘interests’ should not have been dismissed without comment by a philosopher in 1975. But the philosophy and ‘positive’ psychology of well-being has exploded in the intervening decades. There are now many more views under discussion, and relatively few expert participants in these discussions nowadays are hedonists. There are plenty of other purely psychological views, as well as more ‘objective’ (not-purely-psychological) views. If we accepted any of these alternatives, we would thereby draw the line of interests and moral concern in different places from Singer’s line of sentience (i.e., capacity for pleasure and pain). On a perfectionist view like Aristotle’s, more living things than just sentient animals would have interests; on a more cognitively complex view from positive psychology (involving, e.g., a sense of meaning or purpose), only some but not all sentient animals would have interests. Here in 2023, Singer should not rest his book’s entire argument on a widely rejected assumption with no more defense than that Bentham held the view.

To be fair, rejecting Singer’s hedonistic premise does not undermine his practical moral conclusions. We can charitably revise his argument to ignore the claim that pleasure and avoidance of pain are the only morally relevant interests and instead assert that pleasure and pain avoidance at least are morally relevant interests. We already accept that we ought not inflict needless suffering on humans, non-human animals are also capable of suffering and so, on pain of arbitrariness, we ought not inflict needless suffering on non-human animals. Combined with all the factual reporting about conditions in factory farms and labs, this is a convincing enough argument that we, collectively, ought to stop such unnecessarily cruel practices.

The remaining philosophical problem for Singer, however, is that it is, pre-theoretically, very plausible that pleasure and pain avoidance do not exhaust well-being and interests. Indeed, to anyone not already in the grips of a psychologistic theory of well-being, it should be intuitive that things can be good or bad, better or worse for any kind of living thing and not just the relatively small number of those with mental states. It is straightforwardly good for a typical plant to have access to sunlight, water and soil nutrients. It is bad for a typical mushroom to be placed in a bright, hot, dry environment. It is good for a methanotrophic bacterium to have access to methane, bad for a typical bacterium to be bathed in pure alcohol. These non-animals have ‘interests’, in the sense of goodness-for-a-subject, even if they do not consciously ‘take an interest’ in them.

As a result, Singer has not given us a principled reason for caring about the interests of all animals but not other types of subjects with interests. He briefly considers the possibility that plants have interests of moral concern but still thinks it will depend entirely upon plants’ ability to feel pain. But plants and other living things do intuitively have interests independently of feeling pain, and this problematises Singer’s position. For he places great emphasis upon the idea that all species’ interests are equally deserving of moral consideration and that accepting anything less would constitute objectionable speciesism. But he is guilty of the same kind of arbitrariness by restricting ‘interests’ to sentient animals. Given a more plausible and inclusive theory of well-being, Singer’s argument commits him to the equal moral relevance of the interests of non-animal beings.

In this light, the ‘What about plants?’ objection that Singer dismisses takes on much greater force. It would require a great shift in our moral beliefs and our practices to treat the interests of plants, fungi and single-celled organisms as equally worthy of moral consideration as humans and other animals—a far greater shift than merely to give non-human animals equal consideration. If we wish to avoid the implication that chopping down a healthy tree is morally tantamount to a gruesome murder, we will need to provide some moral justification for privileging the interests of some kinds of living things, perhaps the interests of sentient animals over non-animals. It seems reasonable to look for some such justification.

But it is not clear that Singer, in particular, could accept this. In the first place, any such pattern of justification could turn out also to justify privileging humans over non-human animals, threatening his moral conclusions. But moreover, it is precisely this kind of privileging of interests that Singer denounces with comparisons to sexism, racism, etc. We could easily charge Singer’s animal-privileging position with ‘kingdomism’, which is just another form of ‘speciesism’. At the very least, Singer owes some principled explanation for drawing the line of moral consideration at sentience, as he clearly would like to. This could involve a sophisticated defense of hedonism about well-being against the objections that have been raised to it and the many alternatives on offer. Without any such argument or explanation in this text, however, Singer’s ‘zoocentric’ position appears to be one of the more arbitrary positions one could take here. It would be less arbitrary either to treat all living things’ interests equally (as counterintuitive as that might be) or to unabashedly privilege interests on some basis (perhaps in proportion to the similarity of other creatures’ complete interests to our own). Neither approach would be likely to vindicate animal cruelty, and so Singer may rest content to have preserved his main conclusion: the need for animal liberation. But a more consistent position might lend nuance to our more concrete interpretations of this goal or of our other moral obligations.

Share on

13 Comment on this post

  1. If you boil the water to some point it vaporizes.

    The same thing happens if you “boil” the rights. They vaporize. There will be no rights anymore. For nobody.

    Be aware. All the totalitarian theories had one characteristic in common. The human as an individual was not in the middle of their interest. These violent ideologies talked bout perfect society, perfect community. They talked about class war or race.
    What happen if in the middle of our interest there will not be human being but bacterium or mushroom? It is a little bit terrifying to guess.

    All this thinking about the rights of species are tempting. Because it is easier to imagine perfect world for all in the future. It is much easier than to think painfully about present and more painfully about past.

  2. Ian Douglas Rushlau

    Prof. Moore,

    Thank you this review of Singer’s claims.

    I’ll confess my biases at the outset- I believe consciousness is a more salient consideration than sentience in the identification of moral agents.

    I’ll reference some comments I made in another forum, in reply to an essay in Aeon written by David Borkenhagen covering some of the same ethical territory you are discussing. Borkenhagen presented a detailed description of the life cycle of the Octopus, which presents problematic scenarios for those who might wish to assign a degree of responsibility of moral care to humans in their relation to other animals. To briefly summarize- after mating, a male octopus dies within weeks (semelparity), because if they stuck around, there is a substantial likelihood they would consume their offspring. Octopi, it seems, are open to cannibalism. My attempt to grapple with this phenomenon within a human ethical frame is as follows:

    The rapid decay and death after breeding is chemically encoded for octopi, and we need to tread carefully when ascribing to these events any of our ethical categories. It’s tempting to see some sort of allusion to self-sacrifice for ‘those that come after us’ in semelparity, but since octopi are cannibalistic, the survival of the entire genus would be precarious if parents cohabitated with their offspring for long.

    Do octopi possess consciousness? Clearly yes, with prodigious intellectual capabilities.

    Do they operate according to a moral philosophy? I’d pay cash money to know, but it’s unlikely we’ll decipher it if they do. What sort of moral philosophy might it be, which permits consumption of one’s own species when convenient, and pushes individuals to, essentially, commit suicide initiated by a single sexual encounter.

    If we grant other other animals moral status, according to moral standards employed (rather inconsistently, I’ll note) by Homo Sapiens, do we/ must we also assign to other animals moral responsibility for their actions? It would be absurd to charge an octopus with murder. They will dismember their kin and gorge on them, but it is morally suspect for me to grill one with olive oil and paprika?

    If granting an animal moral agency for the purpose of assigning them rights, but not applying standards of conduct that we apply to ourselves (also animals) seems an odd inconsistency.

    Such an approach- selective, circumscribed moral agency- is essentially the model we employ with children and the incompetent. (Again, full consciousness is implicated in moral agency.)

    Are we back to pastoralist paternalism in different dress?

    Are there moral considerations in how we treat animals, and our shared environment? Yes.

    Should we apply human ethical categories to other animals? Do we include plankton and protozoa? If not, why not? And if we don’t, are we simply making moral claims out of preference, a (literally, I think) warm fuzzy feeling for certain animals?

    That seems a shaky foundation for constructing an ethical framework upon.

    Best regards,


  3. Richard Yetter Chappell

    > “It is bad for a typical mushroom to be placed in a bright, hot, dry environment.”

    We might equally say that it’s bad for laptops and other sensitive electronics to be left out in the rain, or exposed to an electromagnetic pulse. But such talk hardly implies that laptops “have ‘interests’, in the sense of goodness-for-a-subject”, because it’s very implausible to regard laptops (or non-sentient plants) as welfare subjects at all. Rather, it’s just a vivid way of talking about what might cause damage to these objects, in a purely descriptive (non-normative) sense.

    A quick principled argument to support this: one can imaginatively inhabit the perspective of any genuine welfare subject. You can put yourself in the place of a pig, and assess how bad it would be to factory-farmed. You can’t put yourself in the place of a laptop, or a mushroom — again, because there is no perspective, no subject, there to inhabit.

    1. Thanks for your comment!

      You might say that about laptops, but it wouldn’t mean the same thing. When we say that things are good or bad for tools (e.g., computers, knives, etc.), we don’t literally mean that it’s good or bad for the tool’s sake. The knife itself is not better off for being sharp; the computer is not worse off for being fried. It is instead good for the tool-user, or for their purposes, for the tool to be functioning as the user intends. Subtract human interests in these physical objects and it’s unclear that we would even talk of ‘damage’. If I kicked a random stone and it broke in two, I wouldn’t say the stone was damaged, though a collector with an interest in keeping the stone intact might.

      Living things, on the other hand, can be benefited or harmed (e.g., ‘damaged’) in ways that are not simply relative to or dependent on the interests or purposes of other beings. A proof is that the interests of distinct living things can diverge. For example, what would be good for an especially human-lethal bacterium (zombie fungus, toxic/carnivorous plant, etc.) would clearly be bad for humans, unlike what’s good for my laptop. I agree, however, that talk of benefits and harms may be descriptive and non-normative. To know that direct sunlight is good for this houseplant, on its own, doesn’t begin to tell me whether or not I should move it into sunlight (or whether this would be good in some absolute sense).

      Unless we make the mistake of conflating being a syntactical subject (i.e., the bearer of a property) with psychological subjectivity, I don’t find much principle behind the assumption that ‘one can imaginatively inhabit the perspective of any genuine welfare subject’. What one individual or type of being can imagine is completely contingent, whereas we presumably think the conditions for having welfare are necessary. Whether a given creature has welfare shouldn’t depend on whether there is some (other) kind of creature with sufficient imagination. Moreover, some would challenge your assertion from the other direction. If Nagel is right, we have as much chance of imaginatively inhabiting the perspective of a mushroom as we do a bat or any other non-human animal: namely, no chance at all. In that case, your premise would establish that the pigs in factory farms aren’t welfare subjects either.

      1. Richard Yetter Chappell

        I don’t see how the diverging interests test distinguishes laptops from lifeforms. Imagine my laptop contains instructions for unleashing a lethal super-pandemic on the world. It would then be a good thing (for the world) for the laptop to be left in the rain, or to get fried by an EMP. But it’s still bad *for the laptop* to be left in the rain.

        I agree it isn’t bad *for the sake of* the laptop, because it doesn’t have a “sake”. But exactly the same is true of non-sentient lifeforms. Direct sunlight isn’t good *for the sake of* the houseplant; the plant has no such sake. To say it is “good for” the plant is elliptical for a presumed function (e.g. to grow) in exactly the same way that not being left in the rain is “good for” the laptop to serve its presumed functions (e.g. of computation).

        The idea that non-sentient beings *literally* have interests, in a non-metaphorical sense, just seems bizarre to me — the same way that parallel claims about artifactual “interests” (correctly) seem bizarre to you.

        If you’re interested, I set out some independent reasons for thinking that welfare interests are not the same kind of thing as biological “interests” in the following old post:

        1. Thanks for another thought-provoking comment and post. I have an initial, direct reply and a longer dialectical observation.

          First, being good for some function is distinct from being good for (i.e., benefitting) a subject. Restricting ourselves to the latter, there is no true sense (you agree) in which rain is bad for the pandemic laptop that does not reduce, ultimately, to its being bad for some other subject—e.g., the user who wants to unleash a pandemic. But there are true senses in which things can be good or bad for non-sentient living things that do not reduce to being good or bad for other subjects: e.g., rain can be good for a plant even if that’s not good for any other thing. (Here I’m still talking about what you called a purely descriptive sense. Maybe you’d prefer ‘needs’ to ‘interests’.)

          This is not a mere difference of intuition or ‘seeming bizarre’. I’ve indicated a good reason for thinking non-sentient living things are more like sentient living things than like stones or computers in this respect. And it follows the same pattern as Singer’s reason for thinking that non-human animals feel pain and that this pain is bad for them: we know that we feel pain and that it’s bad for us; based on evolutionary, physiological and behavioural continuities with these other animals, we infer that they too feel similarly bad experiences. Well, we also know that, independently of pleasure or pain, it’s good for us, as individuals, to consume nutrients and maintain bodily processes and, as a species, to reproduce; based on evolutionary, physiological and behavioural continuities with all other living things, we infer that their consumption, self-maintenance and reproduction are good for them too. I still haven’t heard a refutation of this inference, but it would anyway be a Pyrrhic victory for a defender of non-human animal interests. Furthermore, we have opposite considerations undermining the attribution of similar interests to computers: even if we programmed a robot to charge itself every night and build other robots, we would know that this outward behaviour was explained by hardware and programming discontinuous with the biological underpinnings of our own behaviour.

          The dialectical observation: you seem to want to invest ‘interests’ with normative force. That’s fine. I make no claims about your ‘normative interests’. I am only talking about benefits or harms, not specifically morally or normatively relevant benefits or harms. You can say only the latter are real interests but we can still usefully think and talk about the former. I know I’m philosophically interested in normatively unspecified benefits and harms, even when considering only those of humans or sentient beings. We can ask and answer interesting, substantive questions about, e.g., their nature and their connection, if any, to morality. And I know I’m not alone in the philosophy of well-being as that concept is often defined, simply, in terms of what is non-instrumentally good for some x, without reference to morality, rationality, and so on. (There are, of course, some others, such as Darwall, who do define it in terms of reasons/oughts/etc. to desire.)

          As far as the dialectic with Singer goes, there’s no indication in either version of Animal Liberation that Singer wants to build substantive normativity into interests. Indeed, the fact that he posits the principle of equal consideration of interests as a substantive moral principle suggests to me the opposite: that he considers moral or normative relevance external to well-being rather than built into it. Maybe it would improve Singer’s position to accept your view but that would merely prove me right to criticise Singer for being silent where more detail and argument is needed.

          With all that in mind, you and I might only be disagreeing verbally. We agree that nothing literally benefits or harms a laptop, so we can set that aside. You assert that exactly the same is true of non-sentient lifeforms. I think it is literally true that non-sentient lifeforms have ‘sakes’ and can be benefited or harmed, but I’m only making that claim about the normatively ‘bare’ sense of benefits and harms, which you’re calling the sense of ‘biological “interests”‘. And in your post, you seem happy to say that humans and plants alike have those.

          On the other hand, I’m taking no stance on whether non-sentient beings have literal, normatively relevant interests. As per my original post, Singer’s principle of equal consideration of interests would commit him to any non-sentient beings’ interests (should they obtain) being morally relevant. If we interpret Singer’s principle using the ‘bare’ sense of ‘interests’, then the problems I described above will still arise, since we agree that they do have interests in that sense. If instead we interpret Singer’s principle using your sense of ‘normative interests’, then, in order to proceed, we do need to know whether non-sentient beings have those normatively relevant interests in the first place which can then be equally morally relevant according to Singer’s principle. But as I said, I have no intention of taking a stance on ‘normative interests’.

          I will note, however, that explicating Singer’s principle in this way makes it require us, really, to treat all normatively relevant interests equally. While this version of the principle is not entirely trivial–one might instead think all normative interests should receive some, but not equal, consideration—it does seem dialectically weaker and less compelling. It leaves room for Singer’s opponent (or a racist, sexist, etc.) to deny that the interests of non-human animals (people of color, non-males, etc.) are normatively relevant (i.e., that they are ‘normative interests’ and not just ‘bare’ interests), and thus block Singer’s conclusion, without denying his principle or the fact that these beings can suffer or have interests in the ‘bare’ sense. This interpretation seems merely to shift the point of contention, and Singer doesn’t suggest any way of proving that these prima facie interests are also normatively relevant besides the principle itself (which, it seems to me, is how he wanted to prove that they were normatively, viz. morally, relevant).

  4. Talking about ‘flow of thought’ in your essay
    brings us to the second key to writing an interesting
    essay, namely effective transition from point to point in the
    essay. The most uninteresting pieces of work presented in the name
    of essays tend to be disjointed jumbles of thoughts, which
    make for difficult reading. Ideally then, if what you come up with is to be termed as an interesting essay, it is upon you to ensure that by
    the time the reader (who might be a very important decision maker) gets to the end
    of the first paragraph, they have gotten interested
    in reading what you are to say in the next paragraph.

    The idea of getting the reader to desire to read what you have to say in the next paragraph brings us to yet another key to writing
    an interesting essay, namely that you have to ensure
    that you use EssayPro: A Comprehensive Guide to Login and Registration at Essaypro com tone that is engaging to the reader.
    For every point you identify as one of things you will need to talk about in your essay, you need to also ask yourself how you can present the same point in a
    way that is likely to personally engage the reader.
    This is perhaps the most important facet to writing an interesting essay; since if you can make your essay personally engaging to the reader,
    right there and then you get at least half the marks for having
    an ‘interesting essay.’ Your ability to personally engage the reader is dependant on your story telling skills, which are in turn dependant on how widely read you are.

    1. Fruit, I suppose, as that’s part of the plant’s life cycle.

      More seriously, that’s a potential problem if we adopt Singer’s principle of equal consideration of interests and also think that at least some interests (e.g., humans’) render it morally impermissible to eat the interested parties. I was only attempting to draw out an implication of Singer’s principle, not advocate it.

      To venture a guess, Singer might respond by appealing to his observation that equal consideration of interests doesn’t entail identical treatment. Maybe non-sentient living things don’t have the right kinds of interests to render it morally impermissible for humans to eat them (e.g., self-awareness, desires not to be eaten). In fairness, Singer doesn’t think it’s intrinsically morally wrong to eat animals either. He advocates veganism on pragmatic grounds to disincentivise ongoing cruel farming practices (among other reasons) but acknowledges that consumer- or government-mandated cruelty-free farming practices accomplish the same goal.

  5. The article makes no mention of recent research into the possible sentience of plants. These findings should somewhat recalibrate the debate.

    1. That’s because Singer briefly discusses this research in the new edition. He is ‘now more open to the possibility that plants can feel pain’ but still considers it unlikely. Moreover, his case for veganism, for example, doesn’t depend on the results of such research: even if plants are sentient and therefore worthy of moral consideration, a plant-based diet would harm fewer plants than a meat-including one. I don’t have any stance on plant sentience and I don’t disagree with this dialectical point of Singer’s.

      Instead, I have objected to Singer’s initial premise that sentience is a necessary condition for having interests (aka. well-being, goodness-for, benefits-to, prudential value, etc.). My whole point is that living things have interests beyond pleasure and avoidance of pain, so it’s also irrelevant to this particular argument of mine whether or not some members of any particular kingdom can literally experience pleasure and pain. Those empirical facts might bear on some debates and arguments but not this one.

      To adapt the line from Bentham which Singer quotes, the question is not just, Can they suffer? but, Can they be benefited or harmed? Interestingly enough, Singer comes close to conceding the point accidentally when he considers research that might suggest ‘that the plant has learned what does or does not benefit it’. He seems to be suspicious of plants literally learning but doesn’t obviously balk at the idea that they can be literally benefitted. And, after all, it would be quite natural to say, for example, ‘That tomato plant would benefit from more sunlight’.

Comments are closed.