by Jeroen Bouterse
Without really looking into them, I have always felt sceptical of Kantian approaches to animal ethics. I never really trust them to play well with creatures who are different from us. Only recently, I cared to pick up a book to see what such an approach would actually look like in practice: Christine Korsgaard’s Fellow creatures (2018). An exciting and challenging reading experience, that not only made a very good case for Kantianism (of course), but also forced me to come to terms with some rather strange implications of my own views.
In very broad strokes, Korsgaard’s argument is as follows. We are a kind of creature that values things as important to us. We use those evaluations as reasons in our choices, and in doing that, we assign them importance in an absolute sense. This is why we ought to recognize that similar creatures are similarly sources of value (in a sense that what they assign importance to has a claim to being important not just ‘to them’, but absolutely). Now, other humans are similar to us in both respects – valuing things, and then using those values as reasons – while non-human animals are similar to us primarily in that they, too, assign importance to things. This is true even if they are not rational creatures and do not use their evaluative perspective on things as reasons. The argument, then, is that this animal capacity or activity is actually what we are recognizing as absolutely important when we make decisions based on reason, and that therefore our respect for our fellow creatures as sources of values stretches out not just to other humans, but to non-human animals as well.
All throughout the book, Korsgaard is careful to emphasize that subjects assign value; value is not simply ‘out there’ to be discovered. Simple-minded utilitarian that I am, I realized I slip into moral realist language and thinking rather easily. Among friends, you will sometimes hear me say that “suffering is bad”. Does this mean I believe that objective value is sprinkled over the planet, coinciding with sentient life or conscious experience?
Korsgaard’s first chapter is devoted to weeding out such pagan thoughts: all value, she says, is “tethered” (1.3), which is to say that everything that is important is important to some creature. Saying something is ‘good’, ‘bad’ or ‘important’ and omitting to whom is something of an error in philosophical grammar. But if nothing comes with its own measure of objective importance, then there is no answer to the question whether creature A is more important than creature B (1.4). There may, I suppose, still be an answer to the question whether it is more important to me or to us, but any such answer can be subject to philosophical criticism: do we have reasons for why humans are more important-to-us than animals, or is this just a result of our evolutionary and cultural biases and interests? In the last case, are we really doing ethics or are we just expressing contingent preferences?
In chapter 2, Korsgaard identifies animals as creatures that things can be good (or bad) for – in the sense of functionally good, but also as goals that we strive for. “An animal”, she says, “functions, in part, by making her own well-functioning, the things that are good for her in the functional sense, an end of action, a thing to go for, a final good.” (2.1.7) Korsgaard emphasizes that we pursue things that are good to our selves, which are conceptually distinct from what is evolutionarily functional for our genes or species; death may make genetical sense in some way, but it still isn’t good from the perspective of the creature (2.2.6). We share our capacity for having or pursuing a good with other animals (2.3.4), but that is not the full story.
Humans are also rational, which means that we can reflect on the reasons we have for our actions, and can use the results of this reflection as motives for further actions. In one sense, animals definitely do things for a reason as well, in the sense that they do things on purpose, serving goals that are not (or not always) just implicit evolutionary purposes, but that are their own goals. They duck, for instance, because they want to avoid the lion’s attention (even if their desire to do this and live on is of course adaptively functional, as it is with humans) (3.2.4). Humans, however, can also adjust their goals based on reasons; this makes them agents, owners of their actions, in a stronger sense than other animals are. Our actions express our selves in a different way, because we are creatures who evaluate not just the things around us but also our selves (3.3.3). In fact, we learn to understand rationally that other things and creatures do not just exist in their relation to us (3.5.2).
The cat and the mouse
Korsgaard believes that we can follow Kant this far, but still deny his conclusion that humans have higher moral standing. Her case against human superiority is the subject of the fourth chapter. Is our capacity for morality – in the reflective, reasons-motivated sense of the word – itself a morally good thing (4.2.3)? One reason to say that it is, is that a capacity for morality helps us act better. Korsgaard rejects this argument.
It assumes that moral standards apply to the actions of animals, even though the animals themselves cannot see that. But they do not. It makes no more sense to say that animals act wrongly when they do these things than it does to say that tornadoes and volcanoes act wrongly when they kill people at random. (4.2.5)
[…] I argued, with Aristotle, that we can regard animals as functional objects. When we do, we see their actions as having a function: to promote the animal’s good, in the way characteristic of that animal. If animal action has a function, it is subject to an internal standard. [… D]istasteful as it seems to us, cats who toy with mice, birds who push their siblings out of the nest, and lions who kill other lions’ cubs are doing just what they ought to do. (4.2.6)
I wish Korsgaard had said something about human cruelty at this point. Are homo sapiens who commit atrocities in ways that have some kind of internal rationale also doing “just what they ought to do”? If not, why not? Evidently, they failed to be swayed by rational reflection upon their own motives. Perhaps they didn’t know better morally speaking, and it served them well otherwise. Just like with animals, we project moral norms onto people who don’t thereby gain the capacity to act on it. Does this make it nonsensical to prefer a person to develop their capacity to act morally? Is that always an intellectual error? If we are not allowed to prefer Socrates over a cat, are we not allowed to prefer Socrates over Genghis Khan either?
In that case, in what sense do we actually value our own rationality at all? Korsgaard ends the previous chapter with a call to heed our rational capacities: “if we act as if the animals were put into the world for our use, it is a failure of our rationality, and with it our humanity”, she says there (3.5.2). It may be a failure of our humanity, but it is not a failure of our animality; and isn’t that just as well, then? Isn’t it one or the other: either we assign additional value to rationality, or we don’t? If we don’t, then what is the normative force of the appeal to it? And if we do, then are we not following Kant at least one step further, in establishing a hierarchy of human and non-human animals (even if we may still turn out to have direct duties to non-human animals)?
A similar objection applies when Korsgaard considers the possibility that humans may be ‘better off’ than the other animals. Quoting John Stuart Mill in his claim that it is better to be Socrates dissatisfied than a pig satisfied, because humans have the capacity for higher pleasures, Korsgaard replies that it does not really make sense to such a thing unless we can specify ‘for whom’ it would be better. Consider reading poetry. Is the pig at present missing out on this valuable pleasure? “Poetry is not good for a pig”, Korsgaard remarks (4.4.1); what pigs like is rooting around in straw. The capacity to enjoy this is missing from our life just as the capacity to enjoy poetry is missing from that of the pig. In neither case is anybody missing out on ‘higher’ pleasures.
I think we have to accept this argument when it comes to goods that are not intertwined with our capacity for rationality. Poetry, however, is intertwined with our capacity for reflection. (Lots of what Mill would consider higher pleasures are.) By Korsgaard’s own lights, that capacity is not just another animal capacity like having good night vision or being able to jump large distances. Tellingly, Korsgaard believes that a human person who does not enjoy poetry or does not think about the good of other people would be ‘better off’ if he did: he has the capacity to do these things, and to the extent that he really doesn’t, he is not a fully functioning human being, but someone with a defective nature (4.4.5, 5.2.3). Arguments like this set off an essentialism alarm with me, and I think they entangle the concept of rationality needlessly with our biological species.
‘Judgments’ pertaining to rationality, and to morality based on rationality, are about what someone would choose to do in a given situation if they were rational. They only make sense if we also believe that the rational thing is the better thing to do. It would be neat, then, if we could get cats to reflect on their instinctive tendencies to torture mice. If they were able to act on those reflections, that would turn them into better creatures. I do see, I promise, how that is an order of magnitude more speculative than getting Genghis Khan to ponder the categorical imperative and to change his bloodthirsty ways; but that is just the difference between the extremely unlikely and the vanishingly unlikely (or between the biologically impossible and the anachronistic). Those probability values are incidental, contingent facts, and I simply don’t see why they change the moral situation. The mouse is a fellow creature to the cat, even if it doesn’t realize it.
The wolf and the lamb
For Korsgaard, every being that “has a good” is a fellow creature. In the argument she gives for this, our rational capacities still play a crucial role. When we move from seeing things as good (or bad) for ourselves to seeing them as good (or bad) ‘absolutely’, we use reasons. Those reasons only work if we take “ourselves to be ends in ourselves” (8.5.5), which is why we have to believe that we are, indeed, ends in ourselves. The aspect of ourselves that we are respecting as such, however, is not our rational capacity, but simply our capacity to see things as good (or bad) for ourselves. That is an ‘animal’ capacity, not a specifically human one: non-human animals, too, clearly pursue things that are good for them. This means our respect for valuing subjects – which is inevitably assumed if we are to be moral creatures at all –has to include animals. “Beneficence requires respect for someone’s animal nature, not merely for his rational nature”, Korsgaard says (8.5.4).
It is easier for me to concur with this conclusion than to follow the argument leading to it. I see a lot of convergence, if not conceptually then at least in sensibility, between Korsgaard’s Kantianism and my own (much less well thought-out) utilitarianism. Sentient beings have moral standing as such; the lives and interests of every creature matter. This seems to me the main message of the book, and I think (though I would understand if Korsgaard objected) that it can be seen and appreciated independently of the constructivist argument that leads to it.
Korsgaard draws several important contrasts with utilitarianism, however. For one, she does not approve of its tendency to aggregate happiness as if it can be isolated from subjects. “Utilitarians”, she says, “regard the subjects of experience essentially as locations where pleasure and pain, which they see as good and bad experiences, happen, rather than as beings for whom these experiences are good or bad.” (9.2.3) I think this is indeed a potential problem, although the reasons that Korsgaard cites can often be factored in by the utilitarian as well. Peter Singer, for instance, has endorsed something of a ‘replaceability principle’ (where, to put it bluntly, it doesn’t matter whether a happy dog lives on or is replaced by a different happy dog) only for creatures that have no interest in continuing to live.[1] He and Korsgaard both believe that lives that are integrated (i.e. are more than a stream of isolated experiences) are not simply replaceable in that sense.
Another anti-utilitarian argument is the Aristotelian perspective Korsgaard formulates on what pleasure and pain actually are. Rather than being particular sensations, it is better to see them as “reflexive reactions to the things we experience” (9.3.2): namely, as welcome or unwelcome. Pleasure is (or is associated with) the unimpeded activity of a healthy faculty; pain is what we reject and want to escape. What does reading a good book have in common with the taste of chocolate? (9.3.1) Essentially, only the ‘form’ of experience: namely that they are both things we want to continue. This formal likeness relies on a subject that sees things in a “valenced” way. The argument shows that pleasure and pain are not independent natural facts to which we can reduce value; they are, like everything value-related, bound to subjects.
I think this is an extremely important point, at the same time a piercing critique of the foundations of utilitarianism and a great danger to the idea that we can make ethical decisions altogether. The utilitarian imposes some kind of commensurability of value onto the world, thereby allowing her at least in principle to adjudicate in case of a conflict of interests. Korsgaard is now saying that this is impossible, not just for pragmatic reasons, but in principle. It is good for the lion to catch the antelope; it is good for the antelope to escape. Respecting both subjects, we have to respect both their goods as absolute goods, and they are contradictory; there is no solution. (8.8.3)
A utopian answer might be to abolish predation from the natural world altogether. In the last three chapters of her book, Korsgaard shows sustained interest in this question, together with the question whether we should abolish the holding of pets and other domesticated animals. Her reason is that both these extreme conclusions might be seen to follow from her argument: if we cannot protect wild animals from harm, perhaps we should domesticate the entire natural world; but on the other hand, if we inevitably treat domestic animals as means to our end, maybe we should leave all animals alone (10.3.2).
None of this is realistic, and Korsgaard recognizes this (12.6.6). But what if we could indeed, without human clumsiness messing things up further, create a world where the wolf dwells with the lamb, and the lion eats straw like the ox? Even then, it seems that Korsgaard would refrain from intervening:
To whom would we owe that? As I pointed out […], we are not generally obliged to have children just because we could give them a good life. So why should we be obliged to create new species of animals just because they would be better off than the ones who would otherwise be here? (10.4.7)
My reply would still be: because suffering is real. Real enough, even after accounting for Korsgaard’s critique above that it is not an independent and measurable quantity but something that matters to subjects. After all, both the actual and the counterfactual world contain subjects who are capable of suffering; this provides an immediate disanalogy with our lack of duty to have children. The idea that it is (ex hypothesi) within our power to create a world without the intense suffering brought about by predation and other natural harms, but that we would refuse purely because we believe we have no duties with respect to the potential creatures in that world, strikes me as outlandish.
At this point, Korsgaard would probably ask me whether I really understood chapter 1 of her book, where she patiently explained that we have no access to intrinsic moral value; the idea that we would hierarchically order different possible worlds based on how much good or bad stuff is going on in them would probably strike her as outlandish. It is one reminder among multiple in this last part of the book, that a Kantian perspective on animal ethics, even while converging in many respects with my utilitarian intuitions, can suddenly completely part ways with them.
Luckily, these divergences seem to be mostly limited to utopian scenarios. On the easy questions – factory farming foremost among them – Korsgaard’s theory gives the same answers, and under only slightly imprecise descriptions, it often gives them for the same reason: animals with their interests are ends in themselves, and there is no reason to discount or ignore them. Since this was not obvious to Kant, and since my simplistic ‘a better world would be better’-utilitarianism is not available to the careful Kantian thinker, this is a point very much worth making.
[1] Peter Singer, ‘Killing humans and killing animals’, Inquiry 22 (1979) 145-156: 151-153.