Keywords

1 Kant’s ‘Indirect’ Account of Duties Regarding Animals

The basic view Kant endorses on our moral relations to animals is that we do not have any duties towards them (MM 6:442.08–11). At best, we have duties ‘regarding’ animals (MM 6:443.23–25)—that is, duties that pertain to how we should treat animals. But these are not duties towards the animals themselves. This denial of moral status to animals has made Kant uniquely unpopular among animal ethicists. In response to objections from animal ethicists, Kantians’ main line of argument has been that Kant’s views are not as repugnant as they—admittedly—sound at first. They point out that Kant recognises an indirect duty to cultivate our capacity for sympathy (MM 6:443.10–16). So we should treat animals with sympathy, even though this is not a duty towards the animals. Enthusiastic readers go so far as to claim that “Kant’s protection of animals is quite strong” (Sensen 2011, 134), that his views have “strong implications for our treatment of animals” (Denis 2000, 406), or even that “Kant endorses many of the most important policies of animal rights and animal welfare philosophies” (Altman 2011, 31). Have animal ethicists just failed to understand that a duty of sympathy cultivation is an adequate replacement for moral status?

The answer, I argue in this chapter, is no. Even on a charitable reading, Kant’s view on animal ethics suffers from severe shortcomings—severe enough for us to start looking into developing a Kantian alternative. For animal ethicists, this chapter can double as a critical overview of the debate about Kant’s animal ethic and an explanation why they should be interested in new Kantian proposals. Hence, even though this chapter will present objections against a certain part of Kant’s philosophy, its intent is altogether constructive. The goal is to further motivate the creation of an alternative Kantian animal ethic that succeeds where Kant’s fails. This paves the way for the main discussion of Kantianism for Animals in Part II.

My argument will take the following steps: In the present section, I will briefly recap the main features of Kant’s account of animal ethics. Following this recap, I will discuss the two major lines of objection against Kant which have repeatedly appeared in the literature. The first is that there is something awry with the structure of Kant’s account (Sect. 3.2). The second is that Kant’s view simply demands too little (Sect. 3.3). Though the way in which these traditional objections have been presented in the literature is often a bit rough and ready, I will argue that there is something true in each of them. Before concluding, I will add a third and less traditional line of objection: Kant’s view of animal ethics is not helpful in the way it is supposed to be (Sect. 3.4). Kantian moral philosophy is meant to reinforce trust in ordinary moral feelings, but Kant has us second-guessing the ordinary feelings of obligation towards animals that he acknowledges we have. Overall, the upshot of this chapter is that Kant’s view faces powerful objections that should move us to consider how we can develop an alternative.

Let me begin with the basics: Kant’s central view in animal ethics, as stated in §§16–17 of the Doctrine of Virtue (MM 6:442.03–443.25.) and the Collins lecture notes (Collins 27:458–460; Collins 27:413), is that there are no duties towards animals, but that there is a duty towards self to cultivate sympathy and gratitude in our interactions with animals. In the Doctrine of Virtue, Kant places his reflections in an “Episodic Section” (MM 6:442.03) towards the end of his discussion of our duties towards self.

The main topic of the “Episodic Section” is the “amphiboly in moral concepts of reflection, to take what is a human being’s duty to himself for a duty to others” (MM 6:442.04–06). By an ‘amphiboly’, Kant means a confusion about the form of our duty, but not about its content. We are right to think that we have a duty to set the end of treating animals with sympathy and gratitude (the duty’s content), but we are mistaken about the beings towards whom we have this duty (the duty’s form). Presumably, an ‘amphiboly’ is not just a momentary confusion we can shake off as soon as we become aware of it. Rather, an ‘amphiboly’ is a confusion stable upon reflection. Our duty to cultivate sympathy does not cease to appear to us like a duty towards others, even after we are made aware that it can only be a duty towards self.Footnote 1

The idea is old—almost as old as the European philosophical tradition itself—that our duty to treat animals with sympathy is not a duty towards the animal, but towards human beings or God. Views to this effect were put forward by Thomas Aquinas, Maimonides, Clement of Alexandria, the Stoics, and according to Plutarch, even by Pythagoras (Sorabji 1993, 129, 173). There is nothing at all original about Kant’s view in this regard.

Kant however makes two amendments to the old idea: First, that the duty to treat animals with sympathy is directed towards self and not towards other human beings or God. We have a duty towards ourselves to cultivate our capacity for sympathy and not harm it, because sympathy for others can make it easier for us to fulfil our duties towards them (MM 6:443.14–15). Recall from Chap. 2 that our duties to self demand that we promote our own moral perfection (MM 6:385.32; MM 6:386.18–387.23; see Sect. 2.4). An important part of this is that we keep ourselves in good moral shape, that we do our best to remain capable of observing our duties. Thus, even though others benefit from our sympathy, our duty to cultivate the capacity for sympathy is a duty to self.

It is worth mentioning that Kant’s argument for sympathy cultivation therefore does not hinge on the empirical claim that cruelty to animals leads to cruelty to human beings, contrary to the standard portrayal of Kant’s view among animal ethicists.Footnote 2 First of all, all kinds of duty-violations against other human beings, not just cruelty, can result if we fail to cultivate one of morality’s motivational aides. For instance, a lack of sympathy might make us more egotistical, more miserly, or more arrogant. More importantly, however, Kant is not so much concerned with the actual consequences of sympathy cultivation. We always ought to keep ourselves in good moral condition no matter what, in keeping with the obligatory end of our own moral perfection (Kain 2018, 226). So our duty of sympathy cultivation obtains no matter whether we will ever actually need to rely on sympathy to observe our duties towards others.

Secondly, Kant adds to the old idea the notion of an ‘amphiboly’ between duties of different types—the idea that our duties to self deceptively appear like duties towards animals. This helps bring together the denial of duties towards animals with the perspective of ordinary moral agents who do take themselves to have such duties. Whereas Aquinas and others would simply have to contradict these agents’ moral experience, Kant at least tries to explain that experience (though he still cannot account for it on its own terms).

Kant’s account is the most prominent example of what animal ethicists, following Regan, call an ‘indirect duty view’ (Regan 2004, 174). In fact, Kant himself asserts that the duty to cultivate sympathy is an ‘indirect’ duty (MM 6:457.26). He also asserts that it ‘belongs indirectly’ to our duties, though ‘viewed directly’ it is a duty towards a human being (MM 6:443.23–25). One should mind, however, that the label ‘indirect’ has a different meaning in contemporary animal ethics than in Kant’s writings. To ward off confusion, let me separate the two meanings. Following Regan (Regan 2004, 150), the standard understanding in animal ethics is the following:

Indirect duty (Regan)

An indirect duty is a duty owed to someone else than its main beneficiary.

For example, imagine a babysitter who has promised to a parent to look after their child for the evening. The duty to make good on this promise is owed to the parent, though its straightforward beneficiary is the child.

Kant’s duty to cultivate sympathy and gratitude towards animals is ‘indirect’ in this sense.Footnote 3 This seems to be what Kant expresses when he uses ‘directly’ and ‘indirectly’ as adverbs (‘belongs indirectly’, ‘viewed directly’, MM 6:443.23–25). But it is not what he usually means when he uses the label as an adjective. His main use could rather be summarised as follows:

Indirect duty (Kant)

An indirect duty is an action or end which is not itself a duty, but which is an accidental means for fulfilling a duty.

Hence, in contrast to the vocabulary in today’s animal ethics, Kantian ‘indirect duties’ are not truly duties, but merely accidental means we ought to use to fulfil our duties (see Timmermann 2005, 140).Footnote 4 For example, consider Kant’s duty to promote one’s own happiness. There is no such duty in general, Kant insists (MM 6:388.26–28). But there is an ‘indirect’ duty to promote one’s own happiness, insofar as it is a contingent fact that being too unhappy will make it impossible for us human beings to observe our duties (G 4:399.03–07; MM 6:388.17–30). So even though there is no ‘direct’, or true, duty to promote one’s own happiness, there is an ‘indirect’ duty to secure a certain minimum of happiness that is necessary for autonomy.

Cultivating sympathy and gratitude is an ‘indirect’ duty in this Kantian sense too, since it is a means for being practically benevolent. Indeed, one could even call Kant’s duty doubly indirect, since the purpose of cultivating sympathy is to be able to act on sympathy. But acting on sympathy is itself only an indirect duty, as Kant points out (MM 6:456.20–27).Footnote 5 For beings like us, acting on sympathy is a means to the end of promoting the happiness of others, and this, finally, is a direct duty.

What does Kant’s account say we should do? As we have already seen, it is not all about sympathy. Kant also gives a prominent place to gratitude (MM 6:443.22–24; Collins 27:459).Footnote 6 We can read Kant more generally as asking us to cultivate any and all of our natural capacities insofar as they are serviceable to morality. Besides sympathy and gratitude, he also mentions the capacity for aesthetic appreciation (MM 6:443.02–09). This is a capacity to value things disinterestedly, which is a prerequisite of morality (MM 6:443.04–08). So we ought to cultivate this capacity in our interactions with our natural environment, be it alive or lifeless (see Chap. 10). However, when it comes to animal ethics, it is sympathy that does the most work for Kant.

Sympathy is taking pleasure in another’s joy and displeasure in another’s suffering (MM 6:456.20–23). It is an emotional response that takes the feelings of others for its object. While the response matches the valence of others’ feelings (positive if positive, negative if negative), it is qualitatively different. For instance, to have sympathy for someone with a toothache is not to have a toothache, nor is it to imagine having a toothache, nor to feel a pain similar to a toothache. It does not require that we put ourselves in their shoes, but only that we feel a certain unpleasant emotional response at the sight of their suffering. This makes it possible for us to sympathise with others whose experiences are qualitatively very unlike ours, so long as we can discern the valence of their feelings. However, sympathy is also clearly distinct from stronger feelings of community such as solidarity (in the sense of recognising a common plight) and from what Gruen (2015) calls ‘entangled empathy’ (which would be an ongoing cognitive and emotional process of reflecting on one’s relations with another).

When it comes to the cultivation of sympathy, we can think of it as the preservation and strengthening of two connections: first, the bond between the perception of joy and suffering in others and our own covalent sympathetic feelings; secondly, the bond between our own sympathetic feelings and benevolent action-intentions. Ideally, when we see someone suffering, we respond affectively with sympathy, and then respond practically with an action-intention to help. Of course, this does not yet ensure that we act in accordance with duty, let alone from duty. But at least, the inclinations of a sympathetic person are generally more in line with duty, making it easier to act from duty. Sympathy cannot guarantee morality, but it can remove some major obstacles. This is what Kant asserts when he writes that sympathy is “a natural predisposition that is very serviceable to morality in one’s relations with other people” (MM 6:443.14–15).

Our duty of cultivation admits of a negative and a positive dimension. In its negative dimension, our duty demands that we avoid weakening the bonds between the perception of suffering, our sympathetic response, and benevolent action-intentions. As Kant puts it:

violent and at the same time cruel treatment of animals is far more intimately opposed to a human being’s duty to himself, because his fellow feeling for their suffering is thereby dulled and a natural predisposition that is very serviceable to morality in one’s relations with other people is thereby weakened and little by little erased. (MM 6:443.11–16)

Hence, we violate our duty of sympathy cultivation in the most egregious way if we act for the sake of cruelty, that is, if we set out to derive pleasure from others’ suffering. But Kant’s account also plausibly prohibits desensitisation: numbing oneself to the perception of suffering. One could also emphasise the connection to our benevolent action-intentions: Kant can plausibly prohibit that we remain indifferent in light of our sympathetic responses. Once we respond affectively to the perception of suffering, we must act in accordance with our emotional response so as to retain it as a resource to aid moral motivation. Besides outright cruelty, Kant also takes the overworking of animals to be a violation of duty (MM 6:443.17–19), as well as all animal experiments “merely for the sake of speculation” if other methods are available (MM 6:443.19–21). In sum, anything we do which weakens the bond between our perception of suffering, covalent sympathetic feelings, and benevolent action-intentions is against our duty. Sympathy is a morally helpful machine that needs maintaining.

The positive aspect of cultivation, which Kant also mentions, is that we ought to bring ourselves into situations in which we can indulge in sympathetic feelings and act on them. To use his example, we ought not to avoid poor neighbourhoods, but purposely walk through them, perceive the suffering of the poor, and then act on our sympathetic feelings (MM 6:457.29–35). With regard to animals, however, Kant puts more emphasis on the positive cultivation of gratitude, asserting that “gratitude for the long service of an old horse or dog (just as if they were members of the household [Hausgenossen]) belongs indirectly to a human being’s duty” (MM 6:443.22–24). So we are to indulge in feelings of gratitude and act on them, so as to further increase our capacity for gratitude.

As an aside, although the term ‘Hausgenossen’ translates literally to ‘members of the household’, the term used to refer to the lowest class of farm labourers, who did not own a house, land, or livestock, and who were explicitly excluded from using their communities’ common land (Otto 1900, 334). Thus, when Kant likens animals to ‘Hausgenossen’, he likens them to what is arguably the most disenfranchised class of human beings in the rural social structure of his day. Hausgenossen are labourers who are not considered to belong to the communities they serve and towards whom members of those communities have almost no duties. Nevertheless, of course one can be grateful to a Hausgenosse for their service, even though full members of the community do not strictly owe them gratitude for anything. Doing so marks us out as generally good cultivators of our own capacity for gratitude. So Kant’s comparison is not quite as heart-warming as it appears at first sight. It merely emphasises once more that our duty to treat animals with gratitude is not a duty towards, but only regarding them.

As a first upshot, animal ethicists should recognise that there is more to Kant’s account of animal ethics than just his denial of duties towards animals. In particular, Kant is far from claiming that we may treat animals in whatever way we please. This point is all the more important for those who take a critical stance towards Kant. If they bank on the initial repugnance of Kant’s denial of duties towards animals, Kantians will be right to point to the true intricacies of his view. Unfortunately, much of the debate between animal ethicists and Kantians follows this dialectic. In the coming two sections, let me explain the two main lines of argument that have been advanced against Kant by animal ethicists. In each case, Kantians have defended Kant by expounding his indirect-duty view. However, in each case, I claim that the defence has been overstated. As much as we should read Kant’s text charitably, we should acknowledge the true strength of the objections he faces. First, let me begin with structural criticisms: Is there anything inconsistent about Kant’s view?

2 Structural Problems of Kant’s Account

Many animal ethicists share the impression that there must be something structurally wrong with Kant’s ‘indirect duty’ account (e.g. Warren 1997, 51; Cavalieri 2001, 49; Zamir 2007, 28). Some of Kant’s readers suppose that his view must be a mere “auxiliary construction” (Wolf 1990, 34; Timmermann 2005, 132). Between the claim that we have no duties towards animals and the claim that treating them badly cultivates an immoral disposition, something just does not seem to fit. But what exactly does not fit? The literature offers several distinct answers. In this section, let me distinguish three structural objections and explain why each fails in its stated form. Afterwards, I will add a novel, more robust structural objection against Kant.

The first structural objection takes issue with Kant’s supposed view that there is nothing objectionable about animal cruelty ‘in itself’, but that it still cultivates an objectionable disposition. Call this the ‘alright in itself objection’. One of its earliest instances was put forward by Nozick:

Some say people should not [kill animals] because such acts brutalize them and make them more likely to take the lives of persons, solely for pleasure. These acts that are morally unobjectionable in themselves, they say, have an undesirable moral spillover. […] But why should there be such a spillover? If it is, in itself, perfectly all right to do anything at all to animals for any reason whatsoever, then provided a person realizes the clear line between animals and persons and keeps it in mind as he acts, why should killing animals tend to brutalize him and make him more likely to harm or kill persons? (Nozick 1974, 36)

Although Nozick did not explicitly address this objection to Kant, the passage is often understood as a point of criticism against Kant’s view (e.g. Franklin 2005, 38; Wolf 2012, 42FN). If Nozick is indeed talking about Kant, he diagnoses him with the following contradiction: On the one hand, Kant thinks that nothing at all is wrong with animal cruelty ‘in itself’, and on the other, he holds that there is so much wrong with it that its evil threatens to ‘spill over’ into the human domain.

But Nozick is mistaken.Footnote 7 First, as mentioned before in Sect. 3.1 above, it is decidedly not the case that Kant sees animal cruelty as merely instrumentally bad—as morally unobjectionable save for its bad consequences for other people. According to Kant’s account, we have a duty towards self to cultivate our natural capacities insofar as they are serviceable to morality. Sympathy is one of these capacities, and cruelty damages it. So anytime we act cruelly, we violate a duty towards self, no matter whether any other human beings ever come to harm from it. So there is no sense in which Kant claims that animal cruelty is ‘alright in itself’, to the contrary.

For the same reason, the image of a ‘spillover’ cannot accurately reflect Kant’s view of the evil of animal cruelty. It falsely suggests that Kant is concerned with an escalation of duty-violations of the same kind, as if his only worry was that an animal’s tormentor might one day graduate to tormenting human beings. Kant’s point is rather that we have a general duty towards self to cultivate our natural capacities insofar as they are serviceable to morality, and by being cruel, we violate this duty. Hence, the inconsistency Nozick points out is not Kant’s.Footnote 8

A second type of structural objection takes issue with Kant’s claim that moral restrictions apply to our treatment of animals, despite their status as ‘things’. Call this the ‘ordinary things objection’. Broadie and Pybus present it as follows:

[Kant’s] argument therefore is that if we use certain things, viz. animals, as means, we will be led to use human beings as means. If this argument were generalized, Kant would have to say that using things as means would lead us to treat rational beings as means. And Kant cannot avoid this generalization of his argument, since he cannot point to a morally relevant characteristic which differentiates animals from other things which, he would say, we can use as means. (Broadie and Pybus 1974, 382f.)

The claim at issue here is that Kant commits himself to this inconsistent set of propositions: We must not use animals as means, and any rule that applies to animals applies to all other ordinary things, and we may use some things as means.

This is not the most robust version of the ‘ordinary things objection’. One problem with it is that it ascribes to Kant a patently absurd view, namely that one must never use anything as a means (see Broadie and Pybus 1974, 382). If Kant were committed to this view, inconsistency would be the least of his worries. Another problem was pointed out early on by Regan (1976, 471; 2004, 180): Kant never asserts that we must not use animals as means. He is concerned with cruelty and ingratitude. There is nothing to suggest that Kant believes that using animals as means leads to using persons as means. And even if he did believe this, there is still an important difference between using persons as means and using them as mere means. So it is not clear that cultivating a disposition to use others as means would even be morally problematic to Kant—this might simply be a disposition to recognise others’ potential helpfulness, without denigrating their importance as moral patients.

However, these counter-arguments pick up on idiosyncratic weaknesses of Broadie and Pybus’s particular version of the ‘ordinary things objection’. Circumventing these weaknesses, we could still say that Kant gets tangled up in an inconsistency by claiming that no special moral rules apply to things, but special moral rules apply to animals, who are things.Footnote 9 But this line of objection would be mistaken too, for although Kant does assert that animals are things (G 4:428.20–21; Anth 7:127.08), he never declares them to be things just like all others (see Sect. 6.3 below). The point of Kant’s ‘indirect duty’ account is exactly that animals are things of a special sort, insofar as they can experience pleasure and suffering (as Herman also points out, Herman 2018, 177). This makes them the proper object of sympathy. The same cannot be said about unfeeling objects. So the ‘ordinary things objection’ fails.

A third type of structural objection argues that Kant cannot claim that cruelty damages a moral capacity if there are no duties towards animals which cruelty inherently violates. We could call this the ‘no moral damage objection’. As Zamir puts it:

the Kantian’s appeal to a notion of humanity that one distorts if one is cruel to animals seems to smuggle through the backdoor a tacit recognition that animals can be wronged (rather than merely ‘harmed’). Why else should cruel acts make for a flawed humanity if animals are morally neutral entities? (Zamir 2007, 28)

Zamir’s idea is that cruelty can only be damaging to our humanity if it is a moral transgression. After all, in the Kantian literature ‘humanity’ should be understood to designate a set of essentially moral capacities, including autonomy and the capacity to observe duties. If these moral capacities are to be damaged by cruelty, cruelty must be a type of moral transgression, a violation of some duty. But if there is a duty-violation in any act of cruelty towards animals, there must exist some duty we are violating, and this duty could be directed towards nobody but the animal. So in claiming that animal cruelty harms our humanity, Kant has tacitly assumed that there are duties towards animals.Footnote 10

Contrary to this objection, however, Kant never claims that animal cruelty damages our humanity. He does not assert that cruelty damages our autonomy or our capacity to recognise and act from duties. Animal cruelty is damaging to natural capacities that are serviceable to morality. It erodes the connection between the perception of joy and suffering, natural sympathetic responses, and benevolent action-intentions. This amounts to damaging a tool in humanity’s toolbox, not humanity itself. Hence, nothing in Kant’s picture commits him to the claim that there are duties towards animals. The ‘no moral damage objection’ misses its target.

In summary, previous objections of inconsistency against Kant fail. Of course, this does not show that Kant’s account is structurally sound. It merely shows that we must be precise about which inconsistency we mean to diagnose. I want to suggest that Kant has no structural problems at all when it comes to the question whether there actually exist any duties towards animals. He answers unambiguously in the negative, and no part of his account commits him to the affirmative. Things get tricky for Kant, however, if we refine the question: Are duties towards non-rational beings strictly unintelligible or are they intelligible but non-existent? As I will now argue, Kant is committed to both at the same time, and this is the true structural problem of his view. Call this the ‘intelligibility objection’.

The problem arises as follows: The starting point for Kant’s account of animal ethics is that duties towards animals are unintelligible. Such duties cannot exist, and as Broadie and Pybus are right to point out, “this is a logical ‘cannot’” (Broadie and Pybus 1974, 379). The reason why there can be no duties towards animals according to Kant is that directionality is a feature of our duties that can only obtain between rational beings (this difficulty will be the topic of Chap. 5). To say that a duty is directed ‘towards’ someone is already to say that their will necessitates us under a shared moral law (see MM 6:442.10–11). But then, duties can necessarily only obtain towards beings who share the moral law, who must be rational beings. And so, ‘duty towards a non-rational being’ is a contradiction in terms. This is essentially what Kant asserts when he says that “a human being has duties only to human beings (himself and others), since his duty to any subject is moral constraint by that subject’s will” (MM 6:442.08–11).

But now enters Kant’s innovative addition to traditional anti-cruelty doctrine, the ‘amphiboly’: What is truly a duty towards self appears like a duty towards someone else. This allows Kant to endorse that we have no duties towards animals while at the same time acknowledging that we all feel as though we had such duties. The trouble is that in order for anything to appear like a duty towards a non-rational being, such duties must at the very least be intelligible. By way of analogy, a square can appear like a trapezoid when viewed from an angle. But it cannot appear like a four-sided triangle. A part of a film set may appear like a wooden barn from a distance, but it cannot appear like a wooden barn made entirely of metal. And likewise, if we are to confuse any of our duties for duties towards a non-rational being, then duties towards non-rational beings must at the very least be intelligible, even if they do not happen to exist. This leaves us with an inconsistent pair of claims: Duties towards animals are unintelligible and must be explained away, but in the process of explaining them away as illusions or confusions, we have presupposed that they are intelligible. This is the true structural problem of Kant’s ‘indirect duty’ view.

A defender of Kant’s account could respond as follows: What is intelligible are duties towards furry, feathery, and scaly creatures, provided we think of them as rational beings. All that is strictly unintelligible are duties towards non-rational beings.Footnote 11 But this response forces us to claim that ordinary moral agents do not just mistake duties to self for duties to others, but that they also mistake non-rational beings for rational beings. This line of reasoning is not attractive. Ordinary moral agents might simply insist that they are not thinking of animals as rational beings whenever they feel obligated towards them. I certainly am not. So this counter to the ‘intelligibility objection’ increases Kant’s friction with ordinary moral phenomenology rather than decreasing it.

Another counter-argument to the ‘intelligibility objection’ could be that Kant never said that our feelings of obligation were always coherent. Something might vaguely appear to us like a duty we have, even though the thought of it turns out to be unintelligible under scrutiny. So we might, incoherently, have the impression that we have something like a duty towards a non-rational being. But this response again forces us to suppose that moral agents fall prey to yet another failing. Not only do they confuse duties of different types, but they fail to have coherent moral feelings altogether. Hence, this response fundamentally undermines trust in ordinary moral feelings rather than solidifying it. It hence runs counter to one of Kant’s main aims in moral philosophy (see Sect. 2.6 above). As far as I can see, there simply is no straightforward response that saves Kant from the intelligibility objection.

At the same time, note that the ‘intelligibility objection’ does not attack the very idea of an ‘amphiboly in moral concepts of reflection’. Duties of different kinds can be confused for each other, so long as they are all intelligible. Consider again the duty to cultivate our sympathy. This duty may appear to us like a duty towards other human beings, when it is really a duty towards self. No inconsistency arises. It only arises when we try to use the ‘amphiboly’ to explain away feelings of obligation towards beings towards whom, supposedly, there necessarily cannot be any duties.

To avoid the intelligibility objection, we however need to be careful in our use of the ‘amphiboly’. We must not use it to explain away feelings of obligation towards beings who do not qualify as finite rational beings—including not just animals, but also plants, lifeless nature, and God. In Kant’s view, there necessarily cannot be any duties towards any of these beings (MM 6:442.19–25). He explains away any feelings to the contrary in just the same way as he explains away feelings of obligation towards animals. But is it a big loss if we cannot use the ‘amphiboly’ to this end? I suggest that it is not. As for plants and lifeless nature, it seems doubtful in the first place whether ordinary agents have strong feelings of obligation ‘towards’ them. Kant’s own example is the obligation we feel not to destroy beautiful plants or delicate crystallisations (MM 6:443.08–09). Do ordinary agents really consider themselves obligated towards these things? I do not think I do.

Feelings of obligation towards God pose a more serious challenge, since many people might sincerely insist that they feel obligated directly towards God. However, feelings of faith admit of a different conceptualisation and explanation altogether. If we feel obligated ‘towards’ God, we likely do not think that God is on the receiving end of a moral relation, that he is a moral patient. We might take ourselves to ‘owe’ things to God, but not in the way we ‘owe’ things to other people. What is at issue here is rather the recognition of God as the authority of our obligations. God is not on the receiving end of duty. He is the ultimate authority to which the moral agent has to answer. Kant has more than enough resources to account for this recognition of divine authority. Indeed, Kant argues that we should think of all our duties as divine commands (MM 6:487.08–25; see also Anth 7:074.03–07; MM 6:152.33–37; MM 6:099:10–13), though certainly not as duties towards God (MM 6:486.02–03). And so, feelings of obligation towards God do not need to be explained away by appeal to the ‘amphiboly’. Ultimately, feelings of obligation towards animals pose the only serious challenge to Kant in ordinary moral phenomenology, and only here does he truly need the ‘amphiboly’.

Conversely, the idea of an ‘amphiboly’ would not have to lose its role in a Kantian framework which accommodates duties towards animals. Once we accept duties towards animals as intelligible, there can again be illusions. The duty to cultivate sympathy would still be a duty to self but may appear like a duty towards animals. But now, the statement of this confusion no longer leads into inconsistency.

To summarise, the alright in itself objection, the ordinary things objection, and the no moral damage objection all fail. But according to the intelligibility objection, Kant’s ‘indirect duty’ view suffers from a structural problem after all. So the ‘indirect duty’ account must go. On the one hand, this leaves Kant without an explanation for ordinary feelings of obligation towards animals, which is a pity. On the other, it gives Kantians an opportunity to rethink the conditions of interpersonal moral obligation. Is it really so important to Kant’s ethical system that duties ‘towards’ a being are thought of as duties arising from interpersonal constraint under a shared moral law? In Chap. 5, I will argue that it is not. Next up, however, let me consider the other main line of objection against Kant’s ‘indirect duty’ account in the literature.

3 Substantive Shortcomings of Kant’s Account

Besides those who object to the structure of Kant’s view, there are those who object to its substantive upshots. It simply does not demand enough to be plausible. Schopenhauer objected that according to Kant’s account, “one is only to have compassion on animals for the sake of practice, and they are as it were the pathological phantom on which to train one’s sympathy with men! […] I regard such tenets as odious and revolting” (Schopenhauer 1915, 94). In a similar vein, Hoff asserted: “Kant’s most serious shortcomings where animals are concerned are not logical but moral” (Hoff 1983, 67). Both Schopenhauer and Hoff react to Kant not just with theoretical disagreement, but with righteous anger. Their wrath is directed less at any particular argument of Kant’s, but rather at the moral offence of condoning indifference and exploitation.

Of course, an argument along these lines can succeed only if there is some prior moral standard by which we can condemn indifference and exploitation. For Schopenhauer, this standard comes with his own views he develops in the Basis of Morality (Schopenhauer 1915). But then, his criticism of Kant’s stance on animals is ultimately a proxy for more fundamental disagreements in moral philosophy. Hoff, on the other hand, makes it explicit that she does not base her ethical theorising on any particular system of moral philosophy, but on ordinary moral intuitions:

In the absence of good arguments in favor of a dogmatic and exclusive humanism, we may trust our moral intuitions. The well-being of an animal appears to be an intrinsically valuable state of affairs, and attempts to view it otherwise are unconvincing, unsatisfactory, and finally, perverse. (Hoff 1983, 68)

Hoff emphasises the intuition that we ought generally to be concerned with animal well-being, that it is wrong to be indifferent to it. The objection is that Kant’s view allows human beings to treat animals with indifference and that it therefore clashes with a widespread moral intuition. Call this the ‘indifference objection’.Footnote 12

Although the relation of Kant’s moral philosophy to ‘moral intuitions’ is complicated, the standard Kantian move has not been to claim blanket immunity against all appeals to intuition.Footnote 13 Rather, the main counter is that it does not matter for a duty’s material—what the duty prescribes—towards whom it is directed (Hayward 1994, 133; Denis 2000, 418; Herman 2018, 188; Ripstein and Tenenbaum 2020, 147). The same material can be prescribed by a duty to self and by a duty to others. And so long as there is some duty to treat animals well, the argument goes, it does not matter towards whom exactly it is directed. To further support the Kantian position, many emphasise the fact that Kant takes duties to self very seriously (Baranzke 2005, 340; Sensen 2011, 134; Svoboda 2012, 161; 2014, 316 f.; Camenzind 2018, 55). Duties to self are neither optional nor weaker than duties towards others. Therefore, defenders argue, Kant makes all the necessary moral demands with all the necessary urgency.

There is something true in this Kantian response: The same end or action can indeed be prescribed either by a duty to self or by a duty to others. But this is not enough to shield Kant from the moral criticism of Schopenhauer and Hoff. For if we flatly deny that there are duties towards certain beings, this denial does restrict what our duties ‘regarding’ these beings can demand. The form and content of our duties are not completely decoupled.

To see the problem more clearly, consider this: If duties can only be directed towards human beings, there can only be duties ‘regarding’ animals insofar as those animals stand in some relation to human beings. The relevant relation for Kant is that of affecting the natural capacities of human beings insofar as they are serviceable to morality (which I will now call ‘natural-moral capacities’ for short). But unfortunately, what is good for animals is not always good for our natural-moral capacities. And what is bad for animals is not always bad for our capacities. We can visualise this using Table 3.1.

Table 3.1 What is good for animals versus for our capacities
Full size table

The limitation of Kant’s view is that he is bound to morally encourage behaviours in the left-hand column, be indifferent about behaviours in the middle column, and discourage behaviours in the right-hand column. This may not be so problematic as long as we focus on the cells in which good matches with good, and bad with bad—that is, the upper-left, middle, and lower-right cells. Acting on sympathy for animals, for instance, is usually good both for animals and for our natural-moral capacities. Acting cruelly, conversely, is bad in both respects. But Kant’s view becomes increasingly counterintuitive the more we consider behaviours that belong in the other cells of the table.

It would be convenient if there simply were no such behaviours—if the good of our capacities and the good of animals always aligned. But this is not the case. The most obvious counterexamples are behaviours which belong in the lower-middle cell: behaviours that are neutral to our natural-moral capacities but are clearly bad for animals. Consider that the suffering and death of billions of animals per annum today is not, for the most part, the result of cruelty. It is the result of what Noske called the “animal industrial complex” (Noske 1997, 22): a dynamic legal and economic system of animal use which effectively shields most of its participants from perceiving the suffering that their actions help to inflict. Participation in this system and failure to help dismantle it is bad for animals, but for most consumers, workers, and investors, it is utterly neutral to their moral capacities. Kant’s view of the morality of human-animal relations begins only where animal joy and suffering are perceived. It has nothing to say about those of us who contribute to animal suffering without perceiving that suffering. Notice that the point here is not about affective or practical indifference to suffering of which we are aware—Kant could condemn that. The point is that we never become aware of most of the suffering to which we contribute in the first place. Where there is no perception of suffering, there is no opportunity to cultivate sympathy.

Even if we narrow our view to those on the frontlines of inflicting animal suffering, such as slaughterhouse workers, Kant’s view has trouble casting any critical light on what they do. As Carruthers has pointed out, “almost any legitimate, non-trivial motive is sufficient to make [an] action separable from a generally cruel or insensitive disposition” (Carruthers 2002, 159; see also Fischer 2018, 253). Even factory farming and industrial slaughter, arguably the greatest moral atrocities committed against animals in our time, are not cruel, strictly speaking (Hsiao 2017). That is, people engage in these practices without taking any pleasure in the perception of suffering. We may even suppose that most engage in the professional infliction of animal suffering with a certain pro tanto regret. They may very well sympathise with animals but judge that other concerns override their beneficent action-intention in this case (such as earning a living, or providing other people with products they are presumed to need). The worker’s regret is evidence that their sympathy is not isolated from action. Were they to encounter human suffering, or indeed animal suffering outside the context of the animal industrial complex, they would still be inclined to help. It is only within a very specific economic context that they judge the intention to help to be insufficient to determine action. Hence, not only is the worker not being cruel, but they are not even desensitising themselves. Kant only has grounds to object once the worker’s regret starts to fade. This is a serious restriction of his view.

There are also examples of behaviours that are good for our capacities, but bad for animals. Suppose we sacrifice a bull to the Hellenic gods (or kill a sheep to serve to a benefactor who is guest of honour), cultivating our capacity for gratitude, for instance. And even if we turn to sympathy, there can be examples where cultivating our capacities is bad for animals. Suppose we mistake an animal’s suffering for pleasure due to an anthropomorphic misinterpretation. The facial expression of a hyperventilating cat can resemble an excited human smile. Imagine that I take a cat on a car ride. The cat enters a state of panic, but I keep confusing expressions of panic with expressions of wondrous enthusiasm. In my uninformed state I experience sympathetic pleasure for the cat. And I think to myself that going on more of these car rides together would be a great way to positively cultivate my capacity for sympathy. Kant has no grounds on which to object to me. This is a case in which promoting suffering cultivates sympathetic capacities in a way that is serviceable to our moral relations to human beings, in whom the respective expressions would in fact be expressions of joy!

There are also things which would be good for animals, but bad for our capacities. These things Kant must discourage, counterintuitively. For instance, it would be good for animals if we set out to confront their true suffering at the hands of human beings and then do something about it. But human beings inflict so much violence on so many animals that confronting even just a portion of it threatens to erode our natural-moral capacities. First, because our capacity to sympathise is limited, there is a point at which we develop ‘compassion fatigue’ (Figley 1995), an inability to sympathise any longer. In this way, confronting animal suffering might very well make us less sympathetic to human beings. So Kant must ask us to look away. Secondly, confronting the full scope of animal suffering may lead to a degree of misanthropy (Wuensch et al. 2002). In fact, it has been explicitly argued in the philosophical literature that misanthropy is a morally appropriate response to how people treat animals (Cooper 2018). But Kant emphatically discourages behaviours that lead to misanthropy (MM 6:450.22–29; MM 6:466.18–25) or which might diminish our respect for an individual person (MM 6:472.01–07). So Kant must, once again, discourage us from paying attention to most animal suffering, lest we lose respect for those who inflict it. In this case, what is good for animals is bad for our natural-moral capacities. But if something can be done about animal suffering, provided we confront it, it is counterintuitive that we should wilfully ignore it.

The listed behaviours have to do either with ignorance of animal suffering or with indifference to it. Kant’s account of animal ethics is all about how our treatment of animals affects our capacities. But what really affects our capacities is a package deal of certain perceptions of pleasure and suffering, our sympathetic responses, and our intentions to help or harm. As soon as one element of this package becomes unlinked, Kant’s argument ceases to be relevant. The intentions become unlinked from the package as soon as we harm animals out of innocuous, unrelated motives, as in the case of the sacrificed bull. Perceptions of suffering become unlinked as soon as we miss or misinterpret animal expressions, as in the case of the hyperventilating cat. Sympathetic responses become unlinked as soon as indulging in and acting out of sympathy harms our capacities rather than cultivating them, as in the cases of compassion fatigue and misanthropy. This is not a comprehensive list of all counterexamples to Kant’s view, but rather a template for generating as many new counterexamples as we like. And what such counterexamples bring out is not that Kant’s view has little loopholes here and there. They bring out the way in which animals truly do not matter, if they do not matter for their own sake. A degree of moral indifference towards animals—an objectionable degree!—is inevitable if we deny that we have duties towards them. This is, in effect, a refined version of Hoff’s ‘indifference objection’.

To this line of objection, one might respond with an even stronger reading of Kant’s position: If ignorance and indifference are the problem, attention is the solution. That is, one could argue that like sympathy, gratitude, and aesthetic appreciation, our capacity to pay close attention to others and their experiences is a natural capacity serviceable to morality. Along similar lines, Denis has suggested that we have an imperfect duty to keep ourselves informed about how our actions impact others (Denis 2000, 416), and Svoboda has argued that “[b]y ignoring the plights of animals whose suffering one could alleviate, […] one misses a chance to cultivate virtuous dispositions” (Svoboda 2012, 158).

But this line of argument only leads to more trouble. For what could it mean to cultivate our attention for others insofar as it is serviceable to morality? Surely, the kind of attention for others we ought to cultivate must help us recognise and observe our duties towards others. In the case of a human being, we ought to pay attention to their suffering and not look away because we may have a duty to relieve their suffering. And if we mistake another human being’s suffering for joy, we have failed to pay proper attention because we have a duty to relieve another’s actual suffering (as opposed to the mere appearance of suffering). But in the case of animals, there are no such duties to anchor our moral attention. Hence, there is no reason why cultivating our capacity for moral attention would require that we pay any thorough attention to animals’ joy and suffering. It suffices that we react in the right way to what we superficially perceive as their pleasure and suffering, whenever we happen to perceive it.

On the whole, then, we can grant that duties to self give rise to some prescriptions about how we ought to treat animals. But they do not give rise to just any prescriptions we like. There are severe restrictions to what Kant’s ‘indirect duty’ account can reasonably demand. These restrictions are too severe to survive scrutiny in the light of the moral intuitions that motivated Hoff’s (Hoff 1983) substantive criticism of Kant. The restrictions are also too severe to make Kant’s view a helpful resource for a critique of the animal industrial complex. Though the point has been overstated in the literature that Kant authorises total moral indifference regarding animals, Kant’s account does end up condoning by far too much indifference towards animals. In a refined version, the indifference objection is sound.

In the face of the indifference objection, one could argue that no matter whether our duties regarding animals demand very much, they at least give great weight to what they demand. It has been suggested in the literature that Kant considers perfect duties to self to be a particularly important kind of duty and that our duties regarding animals belong in this privileged class (Baranzke 2005, 340; Camenzind 2018, 55; Herman 2018, 188). However, even this well-intentioned attempt at painting Kant’s view in a positive light fails: First, there is no reason to think that all our duties regarding animals are of this specific kind. A duty to positively cultivate our sympathy, for instance, is straightforwardly imperfect (Svoboda 2012, 154). In general, any duty of any type whatsoever can be a duty ‘regarding’ animals if animals are affected by it, since that is all the label ‘regarding’ denotes. Secondly, whether a duty is perfect or imperfect does not indicate how important or stringent it is. Both types of duties have obligatory force, and according to the main text of the Metaphysics of Morals, neither allows for inclination-based exceptions (MM 6:390.10–12). So the suggestion is misleading that Kant assigns a “place of honour” to duties regarding animals (as Baranzke puts it, Baranzke 2005, 340). Their place is no more honourable than that of any other duties.

To be sure, Kant does take duties to self seriously. The Collins lecture notes even have him asserting that duties to self “take first rank and are the most important duties of all” (Collins 27:341). But this statement only asserts a philosophical primacy of duties to self (Timmermann 2006, 509), not a blanket normative primacy of duties to self over duties to others. Kant’s point, in a nutshell, is that we must comport ourselves in a manner appropriate to moral agents—to strive for moral goodness, remain good observers of duties, and acknowledge our moral equality with other rational beings (see Sect. 2.2). A person who does not do that at all would lack all moral worth and would also be unable to observe duties to others. By contrast, there is still hope for a person who observes duties to self but fails to observe duties to others (Collins 27:341). None of this implies that specific duties to self, such as cultivation duties, should take normative priority over other specific duties. We put the cart before the horse if we prioritise cultivating a capacity that can help us observe duties over the actual observance of our duties towards others. For instance, imagine that we can be crucially beneficent to a person by exploiting a piece of land, say by growing crops that help to save them from starvation. Clearly, we should not restrict our beneficence merely because the piece of land strikes us as sublime and presents a chance for us to cultivate our own capacity for aesthetic appreciation (see Chap. 10). The same is true, mutatis mutandis, for the exploitation of animals to the benefit of human beings. Whenever cultivating our capacity for sympathy would require that we do not make a certain contribution to another human being’s happiness (perhaps by complicating their way of earning a living, by failing to cater to their gustatory preferences, by posing an obstacle to their customs, and so on), Kant cannot prioritise sympathy cultivation. So our duties regarding animals have little relative weight.

Once again, Kant’s critics may be wrong about various details, but they are right about the general issue: Kant’s account is too weak in its demands to match strong and common moral intuitions. The usual reaction to this insight is to reject Kant’s account of animal ethics, or indeed all of Kantianism. Of course, the converse reaction is possible too: One can cling to Kant and deny or downplay the moral intuitions with which his view conflicts. This is a general weakness of arguments from intuition, and objections to the substantive upshots of Kant’s view are no exception. Still, the discussion in this section has shown that Kant’s account is severely restricted in what it can demand. This should show to traditional Kantians that they are not in the comfortable position suggested by much of the Kantian literature. To animal ethicists committed to the moral importance of animals, it shows that Kant’s ‘indirect duty’ view is not a promising starting point for their philosophical case.

4 The Unhelpfulness of Kant’s Account

To the structural and material objections discussed so far, let me add a more practical objection: Kant’s account of animal ethics does not do what it is supposed to do. As I have explained in Sect. 2.6 above, Kant’s moral philosophy is best understood as offering a certain type of help to ordinary moral agents. It helps us restore and refine our moral outlook in the face of our natural inclinations and their accompanying rationalisations. It shows that our ordinary feelings of moral obligation have a ground in reason and reminds us of the stringency of our duties (Sticker 2017). But when it comes to our treatment of animals, Kant’s philosophy fails in exactly these regards.

The central move of Kant’s account of animal ethics is to ascribe a confusion to ordinary moral agents, namely the ‘amphiboly in moral concepts of reflection’. As mentioned before, Kant does not mean to say that ordinary agents are completely wrong about their duties in this area. They usually get the content of their duty right, he presumes: Do not be cruel or ungrateful to animals. Their only mistake is taking this to be a duty towards the animal rather than themselves. But then, what about agents who feel obligated to go above and beyond what Kant’s duty of cultivation can demand? What, for instance, about people—vegans, vegetarians, and conflicted meat eaters—who cannot shake off the feeling that there is something wrong about killing and eating animals, a practice Kant himself explicitly condones (MM 6:443.16–17)? It appears that Kant must after all correct them. Such strong and far-reaching feelings of moral obligation towards animals, as widespread as they may be, have no ground in reason according to his view. But this clearly undermines trust in ordinary moral feelings instead of reinforcing it. Indeed, Kant fails to give reassurance to those who take their duties towards animals most seriously, to the vegans, vegetarians, and animal advocates who actually act on what they view as their duties to animals. This is disappointing, given that animal advocates’ rationality is often questioned anyway (Wrenn et al. 2015). When it comes to animal ethics, Kant’s moral philosophy fails exactly those real-world agents who could perhaps use its rational vindication most direly.

Kant’s view is of just as little help when it comes to reminding us about the stringency of our duties, simply because cultivation duties do not demand very much and are unlikely to take precedent over other duties. As we have seen in Sect. 3.3, all that these duties straightforwardly demand of us is that we avoid adopting a practical-emotional stance that runs counter to capacities we should cultivate and that we adopt stances that cultivate these capacities. We should avoid cruelty for the sake of sympathy, ingratitude for the sake of gratitude, and lust for destruction for the sake of aesthetic appreciation. But this still allows most exploitation and mistreatment of animals, as long as it does not arise from a vicious stance. And where there is little to demand, Kant’s moral philosophy cannot do much to remind us of the stringency of our duties more generally.

Finally, Kant’s view fails to be helpful to contemporary ethicists interested in first-order normative questions. As we saw in Chap. 2, for the human domain Kant provides a whole taxonomy of duties. He distinguishes perfect from imperfect duties, direct from indirect ones, wide from narrow ones, and duties to self from duties to others. He provides an intricate account of the relation between happiness and morality, of inclination and duty, of free will and the moral law, and so on. And even if we zoom in on more specific moral questions, Kant has much to contribute, be it on lying, on beneficence, on friendship, on arrogance, envy, malicious glee, and so on. We find nothing even remotely comparable in Kant’s account of the moral relations between human beings and animals. The rich taxonomy of our various duties towards human beings contrasts with a single type of cultivation duty regarding animals. Rather than a full-fledged system, here we find only one key claim and only one, barely fleshed-out argument. Kant’s account of animal ethics fails to capture the complexity of our moral relations to animals as ordinary agents perceive them.

There are two important upshots here, one negative and one positive. The negative upshot is that Kant’s account of animal ethics is not just structurally unsound, and not just too weak in its demands, but that it also fails to serve the main purpose of Kant’s moral philosophy, viewed from a Kantian standpoint or the standpoint of a contemporary animal ethicist. On a sober look, this account is only a footnote to Kant’s philosophical system, and not one particularly worth defending. Kantians and animal ethicists are much better advised to investigate ways in which animals can be properly included in Kantian moral concern.

The positive upshot, however, is that a tweaked Kantian system holds great promise for animal ethics. Kantian moral philosophy has a point. It clarifies, systematises, and vindicates an ordinary moral outlook, particularly for the strict moralists in the room. It enables us to understand more about ourselves and do some rich and challenging philosophy along the way. More than anything, it is this positive promise of a Kantian account of animal ethics that drives the present project. But clearly, if we are to reap these benefits, we must first find a way to include animals in Kant’s ethical system.

To conclude, I have argued that critics of Kant’s account of animal ethics are mostly right. The account is inconsistent, its demands are too weak, and it fails to be helpful in the ways Kant’s moral philosophy is supposed to be. It just does not have a lot going for it. If we are more interested in animal ethics than defending a particular philosopher, we should abandon Kant’s view once and for all. But at the same time, a Kantian system that incorporates animals more thoroughly avoids these pitfalls and holds great promise.

Acknowledging that Kant’s ‘indirect duty’ account must go only reveals the bigger task ahead: We must develop some alternative. This will require that we get clearer on the reasons behind Kant’s initial denial of duties towards animals. One idea will likely spring to mind for many Kant readers: It must all be about the Categorical Imperative. Human beings are ends in themselves for Kant, but animals are not. And it is only ends in themselves which merit moral concern for their own sake. Perhaps surprisingly, I do not think that this is where the issue lies. Moral concern, including duties towards individuals, is at stake in an entirely different part of Kant’s ethical system. That is the topic of the next chapter.