Does morality require sameness?: a response and question to Jennifer Frey

Abstract

In a previous paper in this journal, Jennifer Frey presented three arguments against New-Kantian approaches. This paper briefly reiterates these arguments and shows why New-Kantian positions do not succumb to them. Most noteworthy, such positions are formal and not substantive. They care little about the question whether people pursue the same goods and instead stress the role of procedure in explicating rationality and consent in explicating the good. By stressing this distinction between formal and substantive approaches, this paper also provides a hint to the contentious topic of how Kantians can deal with cultural diversity, historicity, and plurality in ethics. It finishes with some questions to the author of the previous paper: do non-formal approaches imply that peace can only exist within similarity?

Keywords:
Ethics; Consent; Formal approaches; New-Kantianism; Kantian animal ethics

In a previous issue in this journal, Jennifer Frey (2018) presented three arguments proving that New-Kantian approaches in ethics are fundamentally flawed. First of all she argued Kantians are unfaithful to the grammar of the good, secondly that they look for universality in the domain of practice where none is to be found, and finally that there is a contradiction between the demand for universality and the diverse practical capabilities of rational agents. In a series of interesting publications, Frey pioneers a more promising alternative inspired by Aristotle, Thomas Aquinas, Foot, and Anscombe (Frey 2019a; 2019b; 2020).

The problems addressed in Frey's paper are of interest to anyone pondering the topic of universality in ethics. In this paper however, it will be argued that Frey's arguments fail to hit their mark, because she fails to distinguish between substantive and formal approaches in ethics. It will argue that New-Kantian approaches are formal approaches, since they stress procedure and consent when explicating the nature of ethics. Paragraph 2 and 3 will show that this implies New-Kantian don't require substantive universality or identically shared goods, desire, duties, or values in ethics. Paragraph 5 argues that this undercuts the problems raised by Frey. The intermediary paragraph 4 makes some general remarks on the related problem of the nature of animals in Kantian ethics.

The topic of formalism in ethics is relevant beyond the boundaries of this response to the paper of Frey.

Kantians are often criticized for being unable to deal with the plurality, contingency, and historicity which pervades our practical lives. This paper argues the opposite is the case: the formalism of Kantian approaches makes them very well adapted to accept contingency and plurality. This is why the conclusion will raise the question how the previous author could ideate a peace between people with different substantive values without to some extent becoming a Kantian. However, the paper will start by presenting Frey’s argument. This will also further clarify what is at stake here.

1. Frey’s critique of New-Kantianism

1.1

In the background of the worries expressed in the paper by Frey, seems to stand the general conviction that to a very large degree, Aristotle’s conception of the nature of value is correct. According to this conception, 'good' is a bound or relational term¹. That is, things are ‘good’ because they are ‘good for’ certain creatures given what these creatures are². This boundness to the nature of creatures naturally leads to the acceptance of a kind of pluralism. After all, if creatures are different in relevant ways, the boundedness of the good implies their goods will be different as well³. Thus we find Frey writing:

Aristotle holds out the possibility that there may be different rational animals with different characteristic capacities and ends, and therefore different desires and goods. (2018, p. 184)

The good for a human beings differs from the good for a fish, because human beings differ from fish. Frey thinks recognizing this boundedness will save us from what she calls the ‘Platonic Error’. She writes: “[The Kantian enterprise] attempts to find a single form of the good or the normative, a single mold in which all value or normativity can ultimately be said to fit. Let us call this the Platonic error about goodness” (2018, p. 177).

1.2

Once this notion of relational goods is on the table, it becomes easy to follow Frey’s arguments on what contemporary Kantians are getting wrong.

First of all, Frey claims Kantians can only accept “an excessively narrow and implausible account of good as a predicate” (p. 176). Kantians ground morality on rationality. But whatever this grounding might look like in detail, it is evident that a large portion of our normative talk does not concern rational creatures. For example, we happily say that rain is good for plants. To stress rationality seems to disregard the many other ways in which the predicate ‘good’ is used. Since every ethical position should be able to account for this aspect of normative language use, Frey can state that New-Kantians must be ‘unfaithful to the grammar of the good’ (p. 176). Aristotle's ethics with its notion of bound goods provides a clear and more convincing alternative (p. 177).

Secondly, Frey writes that “Kantian autonomy does not respect the difference between our theoretical and our practical cognition of the good” (p. 176). This might be surprising to many New-Kantian who are accustomed to strictly separating practical from theoretical philosophy, but on this point Frey is again in agreement with Aristotle⁴. She writes:

Aristotle’s thought [...] seems to be that if a fish attained sophia or theoretical wisdom - say, the truths about logic, mathematica, and other things eternal and unchanging - it would be the same as what a man could attain. The fact that 2+2=4 is not a truth relative to human beings, but perhaps the fact that it is good for a man to have friends is. (p. 183)⁵

In the domain of theory, we have universal and eternal truths and all rational creatures will ultimately agree on what these truths are. In the domain of practice however, this universality is not to be expected. So if Kantians demand universality in ethics, this means they demand universality in a domain where it is really not proper to do so⁶. Again, Frey mentions Aristotle as a more reasonable alternative.

Finally, Frey argues that “the universality requirement and the efficacy requirement stand in irreconcilable tension with one another, and that the autonomy of practical reason does not sit well with its purported productivity” (p. 176). The idea here is that any set of ethical principles should be realizable (e.g. what she calls the realizability criterion⁷). Yet what is realizable is clearly dependent on a huge set of contingent and species relative factors. She writes:

The thought here is relatively simple. What practical reason judges you ought to do, and what the will can realize is constrained by what you can do, given the kind of creature you are. (p. 182)

Therefore, it makes no sense to look for anything that all rational creatures should universally pursue (i.e. what she calls the Kantian ‘universality requirement’) since all creatures are different. This is a fact recognized by both Aristotle and Thomas Aquinas, but, she claims, not by Kantians (pp. 180, 182).

The conclusion Frey draws from these arguments is a very strong one: “Although I treat them in concert, any one of these objections would suffice to bring Kantian autonomy into serious doubt” (p. 176). So do these arguments hit their mark? Against which kinds of theories would they be effective?

1.3

The arguments Frey presents are clearly damning for any position which presupposes there should be a single list of well-described goods or duties that all rational creatures should pursue. For instance, if we need to presuppose that duties are always realizable by the creatures bound to fulfill them, and these creatures have radically different capacities, this seems to falsify the notion that all creatures can share the same list of duties. We might call such an approach, which looks for specific shared goods and duties, a universal substantive list approach (for this term, compare Korsgaard 2009, p. 47). And such an approach would not only be falsified by the arguments given above, but also by the general fact of diversity and contingency which Frey stresses at different points in her paper (for instance pp. 161, 177, 182, 184, 186, 188).

But at this point misunderstanding looms. Many New-Kantian approaches are neither substantive universalists nor substantive pluralists. They don't think universal morality requires there to be concrete shared goods. Neither do they presuppose agents must necessarily have different goods. Instead New-Kantian approaches like those of Korsgaard and O'Neill tend to make the specific plans which agents might act upon largely irrelevant to our explication of the nature of ethics. To put it succinctly: substantive sameness in values is not a necessary condition for the kind of universalism involved in New-Kantian ethics. Since such approaches are not interested in issues of substance, we might call these approaches formal^{8 9}.

More concretely, this formalism stems from at least two characteristics of many New-Kantian approaches. The first of these is the way these approaches stress procedure as the keypoint for being rational. The second is the shift towards putting consent at the center of ethical theory. The aim of the rest of the next two paragraphs is to elaborate on these two characteristics and show why stressing procedure and stressing consent might move the focus away from questions of substantive sameness.

2. Testing approaches do not require substantive sameness

The first reason why New-Kantian theory does not require substantive sameness, is that they do not presuppose that practical reason can tell us what to do in concreto. For instance, Christine Korsgaard writes: “[...] the categorical imperative does not directly tell us to do, or not to do, this or that particular action.” (2018, p. 120).

2.1

Instead of looking at what actions are to be performed, New-Kantian analysis stresses the way that actions come about. For instance, in as far as New-Kantians believe that the morality of action is determined by its rationality, they model this rationality as a capacity to make actions dependent on a test procedure¹⁰. Korsgaard writes:

Morality, on Kant’s account, is not a certain set of considerations, identified by their content, but a way of deliberating: the categorical imperative [...] is part of the structure or logic of practical reason. Of course Kant tries to show that the thing that we usually regard as our substantive duties will be shown to be necessary by this way of reasoning, in order to make a plausible case for it. But what distinguishes the moral agent, on Kant’s account, is not first and foremost what he thinks about when he decides, but rather the way he deliberates when he makes his decisions. [...] Kant offers us what I think of as a ‘testing’ rather than a ‘weighing’ model of reasons. (2009, pp. 48, 51)

The creatures capable of moral actions are those who are capable of arriving at actions through rational reflection. These actions are the right ones, since they survived being tested. Rationality has less to do with the content of the plan we propose to act out, and more with the way a creature might arrive at an action. Conversely, rational creatures are exactly those creatures who are called upon to test whether their plans are permissible, and thus limit or constrain their actions by the result of such a test¹¹.

2.2

This stress on procedure breaks with a demand for substantive shared goods or duties. The notion that practical reason is a capacity for testing plans, implies that the source of these plans remains independent from reason itself. And indeed, as another famous New-Kantian puts it:

The Categorical Imperative provides a way of testing the moral acceptability of what we propose to do. It does not aim to generate plans of action for those who have none (O'Neill, 1998, p. 106)¹².

The maxims or plans which rational creatures will need to test are not themselves derived from reason or mere thinking. They appear outside of our commitment to reason.

We do not autonomously generate them by reason itself¹³.

This largely pushes all concerns of substantial agreement on what is good between different rational creatures to the background¹⁴. After all, even if we universally shared a certain testing procedure, we might still end up with different plans, since different agents pursue different substantive goals. For example, if two groups of friends both decide to do whatever the majority of them thinks is an acceptable way to spend this Friday night, this might still lead to one group going to the cinema and the other to a café. In both cases, this outcome is acceptable, given the procedure agreed upon, but in both cases this result is dependent on contingent and highly personal preferences.

If we stress procedure over content, the topic of substantial unity in values and shared life forms becomes largely irrelevant. There is no a priori reason to presuppose that actions which survive critical testing and scrutiny will necessarily share some further particular characteristics, given that the plans which are tested stem from different agents who might have very different backgrounds. So universality on the level of procedure neither implies nor requires universality on the level of substantive plans¹⁵.

Frey might be less at odds with Kantians then she presupposes, when she writes that: "What it is reasonable and good for a man to do might be totally different from what it is reasonable for a rational fish in some distant galaxy to do" (184). New-Kantians can simply agree. Such a rational alien fish would come up with radically different plans, given that its life is probably radically different from ours.

2.3

At this point it might be objected that stressing the nature of practical rationality as being procedural does not really help shake the demand for universal agreement on substantive values, at least if the test involved in this procedure would still somewhat arbitrarily enforce specific goods or values as the main criteria of restriction. We might for instance presuppose such a test would be consequentialist, and presuppose it really only checks whether our actions maximize some substantial good like drinking enough water. It would then evidently only be applicable and relevant to certain kinds of creatures, namely those for whom water intake is an important part of staying alive. For all we know, this does not necessarily involve all rational creatures¹⁶.

In the paragraph below one reason is presented why this test does not involve such a substantial good, namely because it involves stressing the notion of consent. Yet as a side note, even outside of this, many ethicists working in the Kantian vein simply do not involve substantive values into their testing procedures¹⁷. The deeper reason for this, is that many New-Kantians like Korsgaard and O’Neill are ethical constructivists. For them, picking some substantive value or well-described principle as definite, would be reversing the natural order of how moral terms are supposed to be used¹⁸. For instance, Korsgaard writes:

If your maxim does pass the test, then you are acting in a way that is acceptable to everyone. And if you are acting in a way that is acceptable to everyone, then no one has a legitimate complaint against you. If no one has a legitimate complaint, then you have done nothing wrong. Moral standards are just the standards of conduct that we can all agree that people should adhere to. But it is not, as you might have thought, that we can all agree to this conduct, because it is morally right. Rather it is morally right because we can all agree to it. (2018, p. 123)

So it is not really any substantial good or specific principle which makes something agreeably good. Instead, calling something morally good just means calling it justifiable to other rational agents¹⁹. Although this is not argued here, justifiability, much like consent, does not necessarily require substantive unity in values as a necessary condition²⁰.

3. Respecting consent does not require substantive agreement

A second characteristic which makes substantive agreement less relevant to New-Kantian approaches, is their focus on the second formula of Kant’s Categorical Imperative which stresses the notion of 'not using others merely as means'²¹.

3.1

In a recent overview paper, Kant scholar Pauline Kleingeld defended the thesis that Kant's second formula should be interpreted in terms of consent. She writes:

The distinction between using others merely as means and using others as means but not merely as means can now be stated as follows: you use others merely as means if and only if you use others as means to your ends without making your use of them conditional on their consent. (2018, p. 402)²²

When Kant’s Categorical Imperative suggests that we should never act out plans which involve treating others as ‘mere means’, New-Kantians argue that in concreto, this means that we should make all our action in relation to others conditional upon the consent which we get from them.

Not only does Kleingeld's paper argue that this formula stresses consent, but it also shows that most modern Kantians agree on taking consent to be the central point of Kant's second formula. For instance, some argue that Kant is asking the question whether rational consent to a proposed action is possible (Parfit 2011, pp. 220-1). Some others think Kant wonders whether certain actions are inconsistent with the possibility for giving consent (Korsgaard 1996, p. 139). Others propose that Kant is asking whether a proposed action would be acceptable to everyone else (O’Neill 1989, p. 113). And finally, some argue that Kant is looking to limit all actions involving others to the condition of the actual and situational consent of this other (Kleingeld, 2018, p. 402). This paper will not argue that this last reading is preferable, since it is not required to make the point below.

That consent does not require substantive agreement becomes particularly clear if we reflect on the conditions for getting and not getting consent.

3.2

To get consent, it seems clear we do not need a universal or shared stock of substantive goods or desires. For instance, if my action is conditional on consent, I can only take the fancy drinking cup from Diogenes the Cynic if he agrees I’m allowed to do so. But to give his consent, he does not need to agree with me that this drinking cup is valuable. What is good to him might be different from what is good for me. On a formal level, we both understand morally justifiable action to be action conditional upon consent, even if we then continue to substantially disagree on the question of what things are worth pursuing or having²³.

There is one reason to worry about diversity in cognitive capacities though. It is often argued that getting genuine consent will require something like mutual understanding²⁴. For instance, O’Neill argues that giving consent is a propositional attitude which requires some non-perfect degree of understanding of what exactly one is consenting to (2003, p. 5)²⁵. If you do not understand that I am actually proposing to cut your hair, it matters little that you agree to me doing so since your lack of understanding implies that you cannot genuinely consent to my action. Since mutual understanding is a complicated accomplishment, this suggests one sense in which Frey is exactly right when she worries that:

We cannot possibly know a priori what other rational creatures can realize through the use of their own powers, and thus what they can cognize as goods to be pursued and realized, and so we cannot possibly say that what we cognize as good is something they can cognize as good (2018, p. 185).

So on the topic of getting consent, issues of diversity in forms of life indeed pop up, in as far as these might lead to a breakdown of mutual understanding. Even if such breakdowns are hardly ever complete and are never to be used as an argument for flat out ignoring the topic of consent altogether.

3.3

But this still does not imply that shared ethical principles require substantive shared goods. Respecting consent also involves respecting lack of consent. And lack of consent does not require either mutual understanding or shared values or agreement on substantive goods. If you do not understand what I propose to do, you did not give your consent. And if you have different values from me, and think my action in relation to you is unacceptable, you did not give your consent. And if you do not give me consent to touch your hair, this should already be enough reason for me not to do so.

Your fleshed out substantive considerations why this is unacceptable to you will most likely involve your complex personal history and private life, or idiosyncratic feelings, or cultural customs in relation to the touching and meaning of hair. And yet all of this is entirely irrelevant. The fact I did not get consent should be enough. As for the rest, I can simply trust you to make up your own mind.

3.4

Of course actual cases are more dynamic and less binary than just getting or not getting consent. Navigating the topic of consent will most likely always involve careful conjecture, guessing, backtracking, reassurance, and much refraining from action. But substantive goods are not what agents who make action conditional upon consent need to agree upon.

4. Intermezzo: some appearing tendencies in Kantian animal ethics

Below I will argue that the two considerations above heavily suggest that Frey’s problems are not really worrying for New-Kantians. But some misunderstandings loom. At many points in her paper, Frey refers to animals, for instance, as we have seen, fish. This is very helpful in keeping the theme centered on the nature of value, and above I have been happy to follow suit. Yet Kantian discourse on animal ethics and its relation to rationality is peculiar, and their use of the word ‘rationaĺ’ is relevantly different from that of Frey. It might be good to make five short notes which elaborate on some developments in recent Kantian thinking about animals and which will help frame the final discussion of Frey's arguments.

4.1

First of all, many Kantians are anthropocentric, in the sense that they assume there is a marked difference between the ways non-rational animals arrive at their actions and the way rational creatures do. The details of this distinction can be spelled out in different ways. For instance, Korsgaard’s preferred way seems to be to contrast creatures who intelligently pursue the ends prescribed by their instincts from creatures who can reflectively evaluate these ends themselves (2018, p. 40). This leaves the difference qualitative and not merely one of degree.

4.2

Secondly, Kantians can therefore also admit that animals are not subjects of morality. They are not bound to act morally and not to be blamed for acting immoral. As Korsgaard writes: “Do the animals have obligations to us? The idea is absurd. They are not reflective, they do not construct or endorse their identities. They don’t even know they have them.” (1996, p. 157). Animals have their own goods, which they instinctively pursue. We can recognize this good exists. Yet since they are incapable of moral considerations,we cannot just copy over these goods, since we have different capacities than they do²⁶. Thus there is a difference between the moral good and the different final goods that animals pursue (This is also why Korsgaard thinks that carnivorous lions do not prove that we should not be vegetarians. We might have capacities which they radically lack (Korsgaard 2006)).

4.3

Thirdly, some New-Kantians distinguish between two different kinds of autonomy, one of which we can attribute to all animals, and one which we can only attribute to rational ones. Korsgaard writes:

In one sense, to be autonomous or self-determined is to be governed by the principles of your own causality, principles that are definitive of your will. In another, deeper, sense to be autonomous or self-determined is to choose the principles that are definitive of your will. [...] Only responsible agents, human agents, are autonomous in the second and deeper sense (2009, p.108)

The first sense of autonomy implies there is a difference between the cat herself jumping on the mouse, or the cat passively being carried around by its mother. In this first case, she acts from an instinct within her, and in the second case her body is moved by a cause outside of her.

The second sense of autonomy implies the creature in question is capable of regulating its own behavior in a further way. On the account of Korsgaard, this further control is provided by the capacity for reflectively testing whether the desires and plans we might come up with make for good actions²⁷. Being capable of such tests implies partaking not merely in the having of a self with a specific nature, but also of actively deciding what this self should be.

Here it is important to note that if we hold such a testing approach of what rationality boils down to, there is no strict contradiction between these two kinds of autonomy. That is to say, all living creatures will produce desires and plans which are expressive of their nature. Rational creatures will additionally be capable of making up their minds in regards to these instincts. This second kind of autonomy does not provide us with some substantive goods which could contradict or replace the substantive goods nature, tradition, and upbringing have already provided for us.²⁸

4.4

Fourthly, the fact that animals are not subjects of morality does not prove that they are not objects of moral concern. This might be surprising, since Kant himself made the inference that, given that animals are incapable of moral reasoning, they deserve no moral respect. But as is shown by further conceptual distinction, this argument is simply not valid. Korsgaard writes:

A passive citizen or a foreigner can obligate an active citizen by appealing to a law whose authority the active citizen recognizes. An animal, of course, could not actively make such a claim, but if she falls under the protection of our laws, we can recognize that she has such a claim. The fact that non-human animals cannot participate in moral legislation is insufficient to establish that they cannot obligate us in this sense. (2018, p. 126).

Not all Kantians might agree on the notion that animals deserve moral respect, but it is clear that the idea itself is not incoherent within the Kantian framework.

4.5

Fifthly, since not all creatures are subjects to morality, New-Kantians do not need to claim that the demands of a universal ethics for rational creatures extends to all living beings. What exactly constitutes practical rationality is an extensive topic in its own right (perhaps it involves cognitive faculties like self-reflection, imagination, having a theory of mind, empathy, etc.²⁹), but New-Kantians claim that only and all rational creatures are bound to the Categorical Imperative.

When Frey mentions that some fish in some distant galaxy might be rational, this can lead to misunderstandings. New-Kantians would generally care more about the question whether this fish is self-reflectively capable of criticizing it’s first impulses, and they would care less about questions like how many vins it has, whether it’s species has three sexes or just two, or whether it’s aquatic environment averages over 55℃, etc.³⁰ Even though of course these latter considerations might matter a lot to the fish in question and form the substance of its life.

4.6

And finally, all of this does not contradict that we might owe non-rational animals different things than we owe rational animals. In relation to non-rational animals, we often allow ourselves a degree of paternalism which would be unthinkable in relation to adult human beings. For instance, I cannot unproblematically help a lady cross the street once she made it clear to me that she does not want me to aid her. Yet it seems it is perfectly fine for me to lock up my cat to take him to the vet, even while he makes it perfectly clear he does not consent to me doing so. To explore this difference is not the topic of this paper. Perhaps it is easier to know what animals intend to do, since human action is complex and precise. Perhaps we know what animals want before they do. Perhaps ignoring consent does not harm animals in the way it harms human beings. Or perhaps it is the other way around, and we only really know what is good for human beings, namely to make up their own mind, and conversely animals are largely obscure to us.

5. Evaluating Frey’s arguments

At the start of this paper, it was claimed that Frey’s arguments evidently falsify any approach which looks for universality on the level of shared substantive values and goods. But as we have now seen, the core tendencies of contemporary Kantian approaches do not lead to a commitment to such a universally shared set of substantive goods or values. Instead, these approaches model ethics on more formal considerations of testing procedure and respect for consent. Do the three arguments Frey presented harm such a formal position?³¹

5.1

First of all, Frey proposed Kantian are unable to use normative terms to refer to anything outside of rationality. On the one hand this is false, since Kantians can happily accept that plants and animals all have their own substantive goods which are good for them and which they pursue, just like different people pursue different notions of the good life. This is part of the first kind of autonomy distinguished in the previous paragraph.

On the other hand, Frey is correct. Or at least in as far as New-Kantians have a tendency to differentiatie strongly between the moral good on one side, and substantive final goods of people and animals on the other. Since animals (presumably) lack moral capacities, we can only talk about the substantive goods that are contained in their desires and needs. But in relation to rational creatures, we can also use a moral vocabulary which according to Kantians is tied to testing actions by the Categorical Imperative and respecting consent.

This means New-Kantians work on two different levels. For example, having fun is something good for all creatures who are capable of having fun. Furthermore, a creature might find it enjoyable to swing around large sticks. But if I am a rational creature, the fact that it is fun for me to violently swing a broomstick around, is irrelevant in relation to the fact that you did not consent to being hit by it.

5.2

Secondly, Frey argued that New-Kantian wrongfully expect universality in the domain of practice. Again, this is both true and false. As we’ve seen, Kantians don’t want universality, in the sense that they are not committed to the claim that all rational creatures should pursue the same set of goods. Rational creatures come up with different plans to test on the basis of contingent and personal considerations.

Yet on the other hand, New-Kantians would argue that any creature which is capable of rationality is bound to make their actions conditional upon a similar test involving similar constraints. All rational creatures, for instance, are universally called upon to make their action conditional upon the consent of the others involved.

As we've discussed, this does not require further substantive unity in values or goods. So in as far as there is universality in contemporary Kantians, it does not concern the content of specific maxims or plans, and the latter is not a necessary condition for universality.

5.3

Finally, Frey claimed that different agents have different capacities and thus cannot be bound to the same set of duties. Again, if we take these duties to be concrete and specific, New-Kantians can simply agree that this is true. It makes little sense for a cat to cook spaghetti, just as little as it does for me to eat mice. But as noted above, the universality demanded by New-Kantians is not like a substantive universality of shared goods or shared actions. At most, it requires a set of shared capacities involved in making our actions conditional to critical testing. And again, when these tests make action conditional upon consent, it matters little which substantive reason you might have for not giving consent. It also matters little whether I don't touch you with feathers, hands, paws, or tentacles.

Furthermore, there are of course creatures which are radically incapable of moral critique or of respecting consent. Such creatures will never be able to constrain their actions to only those which could survive testing against the Categorical Imperative. We might presuppose this is the case for snails, cockroaches, worms, goldfish, seagulls, mice, etc. But on this point, New-Kantians would simply stop calling such creatures ‘rational’. If instead we are to imagine a ‘rational fish in another galaxy’, as proposed by Frey (p. 184), this would mean imagining some creature who despite all differences is still rather like us. We should imagine them to be in the possession of some particular natural cognitive capacities³². Furthermore, New-Kantians are entirely in line with Frey’s realizability principle, and would probably conclude that creatures who are incapable of being moral are not bound to be moral at all. We don’t blame cats in the same way we blame people.

Conclusion

This paper argued that there are at least two reasons to differentiate between New-Kantian approaches and substantive list approaches in ethics. Two marked points of difference concern the way New-Kantians stress the role of testing procedures over issues of the content being tested, and stress respect for consent above dictating what particular actions are required.

After making some general notes on the current Kantian debate in animal ethics in the fourth paragraph, this paper proceeded to argue that the problems Frey raised do not really falsify such a formal testing approach. In doing so, the larger aim was to show why contemporary Kantian positions actually have little problem dealing with the contingency and diversity in our practical life. Their explication of ethics does not struggle with the fact of substantive plurality and cultural difference, since substantive considerations are largely irrelevant to it.

In arguing this, I do not claim that the historical Kant held such a formal position³³. Scholars might be entirely correct in fearing that New-Kantians take great liberty in interpreting Kant’s work (for instance, see Sensen (2009) and Kleingeld & Willaschek (2019)). Secondly, I also do not claim that Frey's arguments might not be effective against some New-Kantian approaches. But her worries are not problematic for the kinds of approaches we find in Korsgaard and O’Neill, and these authors clearly qualify as New-Kantians.

A final worry

As a final extension, however, I would like to raise a broader concern. In this paper, I argued substantive agreement in values and goods is not a requirement for the kind of ethics Kantians are pursuing. Rational creatures do not even need to share substantive goods to be committed to a communal ethics. Yet in her paper, Frey comes close to suggesting that substantive unity is actually a necessary condition for having a community. She writes:

Our lives may be irreconcilably different, our goods incommensurate. It may be impossible for different forms of rational animality to relate to one another in such a way as to share forms of life, let alone attempt to build a ‘Kingdom of Ends’ together (2018, p. 186).

The suggestion seems to be that creatures with different goods cannot share in a form of life. This raises the question to what extent these goods should be similar. Does she suggest there is to be no peace pact between creatures with different forms of life? Would she extend this to the claim that human communities are lifeless or disfunctional if they don’t share substantive values codified in culture, religion, or occupation?

Conversely, it seems unlikely to me that we can truly conceptualize a peaceful community with any kind of diversity, without invoking some minimal variety of formalism. To have pluriform communities seems to require that substantive differences are recognized to be irrelevant or acceptable. And this point seems to stand, even if we agree with Frey that analysis of ethics is better performed in the spirit of Foot and Anscombe, and not in that of Kant³⁴. Even on such a more Aristotelian approach, it might still be the case that we cannot satisfactorily deal with diversity without importing a large portion of the formal ethics which are, as this paper has argued, already explicated in contemporary Kantian philosophy.

References

• Aristotle. Nicomachean Ethics. Translated by C. Lowe, Oxford University Press, 2002.
• Brandom, R. “Kantian Lessons about Mind, Meaning, and Rationality.” Philosophical Topics, vol. 34, no. 1, 2006, pp. 1-20.
• De Vogel, C. J. Theoria. Van Gorcum, 1967.
• de Waal, F. Primates and Philosophers: How Morality Evolved. Edited by Stephen Macedo and Josiah Ober, Princeton University Press, 2006.
• Frey, J. “Against Autonomy: Why Practical Reason Cannot Be Pure.” Manuscrito, vol. 41, no. 4, 2018, pp. 159-193. https://doi.org/10.1590/0100-6045.2018.v41n4.jf
  » https://doi.org/10.1590/0100-6045.2018.v41n4.jf
• Frey, J. “Anscombe on Practical Knowledge and the Good.” Ergo, vol. 6, no. 39, 2019, 10.3998/ergo.12405314.0006.039.
  » https://doi.org/10.3998/ergo.12405314.0006.039
• Frey, J. Neo-Aristotelian Ethical Naturalism. Cornwall, Cambridge University Press, 2019.
• Frey, J. “Revisiting Modern Moral Philosophy.” Royal Institute of Philosophy Supplement, no. 87, 2020, pp. 61-83. doi:10.1017/S1358246119000262.
  » https://doi.org/10.1017/S1358246119000262
• Gibbard, A. Wise choices, apt feelings : a theory of normative judgment. Harvard University Press, 1990.
• Hanna, R. “Kant's Theory of Judgment.” The Stanford Encyclopedia of Philosophy, 2022, ” The Stanford Encyclopedia of Philosophy, 2022, https://plato.stanford.edu/archives/spr2022/entries/kant-judgment Accessed 11 12 2021.
  » https://plato.stanford.edu/archives/spr2022/entries/kant-judgment
• Kant, I. Groundwork of the Metaphysics of Morals. Translated by Mary Gregor, Cambridge University Press, 1997.
• Kant, I. The Metaphysics of Morals. Translated by Mary Gregor, Cambridge University Press, 1996.
• Kant, I. Practical Philosophy. Edited by Allen W. Wood and Mary J. Gregor, translated by Mary J. Gregor, Cambridge University Press, 1999.
• Kant, I. Religion and Rational Theology. Edited by George di Giovanni and Allen W. Wood, translated by George di Giovanni and Allen W. Wood, Cambridge University Press, 2001.
• Kleingeld, P, and M. Willaschek. “Autonomy Without Paradox: Kant, Self-Legislation and the Moral Law.” Philosophers' Imprint, vol. 19, no. 6, 2019, pp. 1-18.
• Kleingeld, P . “How to Use Someone 'Merely as a Means.'” Kantian Review, vol. 25, no. 3, 2020, pp. 389-414.
• Korsgaard, C. The Sources of Normativity. Cambridge University Press, 1996.
• Korsgaard, C. “Morality and the Distinctiveness of Human Action.” Primates and Philosophers, edited by Stephen Macedo , Princeton University Press, 2006, pp. 98-119.
• Korsgaard, C. Self-Constitution: Agency, Identity, and Integrity. Oxford University Press, 2009.
• Korsgaard, C. “Kantian Ethics, Animals, and the Law.” Oxford Journal of Legal Studies, vol. 33, no. 4, 2013, pp. 629-648.
• Korsgaard, C. “The Relational Nature of the Good.” Oxford Studies in Metaethics, vol. 8, 2013, pp. 1-26.
• Korsgaard, C. “The Origin of the Good and our Animal nature.” 2015, http://www.people.fas.harvard.edu/~korsgaar/CMK.MA1.pdf
  » http://www.people.fas.harvard.edu/~korsgaar/CMK.MA1.pdf
• Korsgaard, C. Fellow Creatures: Our Obligations to the Other Animals. Oxford University Press, 2018.
• Meijer, E. Animal Languages: The Secret Conversations of the Living World. John Murray Press, 2019.
• Moran, R. Authority and Estrangement: An Essay on Self-Knowledge. Princeton University Press, 2001.
• O'Neill, O. “Consistency in Action.” Kant's Groundwork of The Metaphysics of Morals Critical Essays, edited by Paul Guyer, Rowman & Littlefield Publishers, 1998, pp. 103-132.
• O'Neill, O. Constructions of Reason: Explorations of Kant's Practical Philosophy . Cambridge University Press, 1989.
• O'Neill, O. “Some limits of informed consent.” Journal of Medical Ethics, vol. 29, 2003, pp. 4-7. 10.1136/jme.29.1.4.
  » https://doi.org/10.1136/jme.29.1.4
• Parfit, D. On what Matters. Edited by Samuel Scheffler, Oxford University Press, 2011.
• Popper, K. The Logic of Scientific Discovery. Routledge, 2002.
• Rawls, J. Lectures on the History of Moral Philosophy. Edited by Barbara Herman, Harvard University Press, 2000.
• Rorty, R, and Professor of Comparative Literature Richard Rorty. Philosophy and the mirror of nature. Princeton University Press, 1979.
• Scanlon, T. “Contractualism and Utilitarianism.” Ethical Theory: An Anthology, edited by Russ Shafer-Landau, Wiley, 2013, pp. 593-607.
• Scanlon, T. What We Owe to Each Other. Harvard University Press, 2000.
• Sensen, O. “Kant's Conception of Inner Value.” European Journal of Philosophy, vol. 19, no. 2, 2009, pp. 262-280.

• 1
  Following Anscombe, we might call such a ‘bound good’ an ‘attributive term’. In another paper Frey writes: “An attributive adjective is one whose meaning is specified by the nature of that to which the substantive noun it modifies refers” (2019, p. 99).
• 2
  Frey calls it her ‘Hylomorphism’ (2018, 189).
• 3
  A part of the Nicomachean Ethics which is relevant here is this one: “Now if healthy and good are different for human beings and for fish, while white and straight are always the same, everyone will agree that what is scientifically known too, is always the same thing, whereas what is practically wise differs; for each kind of creature asserts that what is wise is what successfully considers the things relating to itself and will hand over decisions to that” (Aristotle, 2002 VI.7, 1141a20-27).
• 4
  On the nature of the objects of theoria, Frey is also in agreement with eminent scholars in antique philosophy like De Vogel (1967)
• 5
  Notice that this argument seems to hinge on accepting the Aristotelian idea of the boundedness of the good, as discussed in the previous section. Below, it will be shown that this Aristotelian boundedness is neither problematic nor unacceptable to a New-Kantian approach.
• 6
  In a more extensive quote, this connection between Aristotelian boundedness and the force of both this argument and the next is even clearer: “Practical thought does not aim at truth alone, but the realization of what is good. But again, what is truly good shifts from species to species, in large part because what can be desired, pursued, and realized shifts from species to species. The perspective of pure practical reason does not respect this fact, and for that reason it is not truly a practical perspective that concern what can be pursued and realized at all” (Frey 2018, p. 188)
• 7
  “[...] what constitutes the practical good (what Aristotle calls the prakton agathon) must be ends that are realizable by the very agents who cognize and desire them. Call this the realizability requirement on the practical good” (Frey 2018, p. 178)
• 8
  There is reason to believe the historical Kant himself was not much concerned with substantive ethics. As noted by Hanna, Kant seems to only believe in a procedural a priori and not in a content based a priori (Hanna 2022)
• 9
  A definition of formal versus substantive accounts in ethics is also presented in Scanlon’s What we Owe Each Other, where in the second section of the fourth chapter, he writes: “The strategy of formal explanations is to appeal to considerations that are as far as possible independent of the appeal of any particular ends. Kant’s theory is a leading example insofar as he undertakes to show that anyone who regards him- or herself as a rational agent is committed to recognizing the authority of the Categorical Imperative” (2000, p.150).
• 10
  The question whether Kant himself held such a testing approach, setting limits to the content of maxims which appear from elsewhere is a contentious issue. This is not the main focus of this paper. But some textual evidence can be provided. In the Groundwork, Kant writes “[T]he formula [of humanity] says: that the rational being, as an end by its nature and hence as an end in itself, must serve in every maxim as the limiting condition of all merely relative and elective ends”. (Kant, 1997, 4:436). Something like a limit appears again in the Metaphysics of Morals: “to be understood as the maxim of limiting our self-esteem by the dignity of humanity in another person, and so as respect in the practical sense” (Kant, 1999, 6:449). It also pops up in the second critique: “Hence the mere form of a law, which limits the matter, must at the same time be a ground for adding this matter to the will but not for presupposing it.” (Kant, 1999, 5: 34) and later in the same text: “Pure practical reason merely infringes upon self-love, inasmuch as it only restricts it, as natural and active in us even prior to the moral law, to the condition of agreement with this moral law” (Idem, 5: 73).
• 11
  See also the text by Kleingeld which is quoted in the next paragraph: “[Kant’s] idea, rather, seems to be that the condition [of respecting humanity] is to function as a general rider on one’s practical reasoning, imposing a general moral constraint on one’s use of others as means to one’s end.” (Kleingeld 2018, p. 401).
• 12
  I think there is a strong and underexplored parallel here with the philosophy of Popper, when the latter wrote: “The initial stage, the act of conceiving or inventing a theory, seems to me neither to call for logical analysis nor to be susceptible of it. The question how it happens that a new ideas occurs to a man [...] Accordingly, I shall distinguish sharply between the process of conceiving a new idea, and the methods and results of examining it logically.” (Popper 2002, pp. 7-8)
• 13
  We might argue Kant thought the specific ends we aim at are simply given by inclination. See the discussion by Christine Korsgaard (2013), in which she writes: “[...] as Kant observes, the ends that we choose are simply the objects of our inclination, and the objects of our inclinations are not considered just as such, absolutely or intrinsically valuable.” (p. 640)
• 14
  This is the option Frey seems to ignore when first writes: “the practical good, insofar as it is materially substantive, is created through willing” (2018, p. 174) and later continuous: “Practical principles, principles that regulate the formation of intentions, cannot be conceptually divorced from the conditions of their execution and remain practical in any meaningful sense” (p. 189). Arguably a lot depends on what exactly we should take ‘intention’, ‘principle’, or ‘conceptual divorce’ to mean. But in any case, on a testing approach, the idea that our principles should allow us to derive certain plans is false. The substance or matter of our actions is not determined by the supreme practical principle.
• 15
  This difference between contingent concrete material values and aims on the one hand, and the universal principles of rational action are also separated by Kenneth Westphal when he writes: “Debate about ‘values’, aims and aspirations as material premises in moral reasoning does not pertain to the most fundamental principles of justice, which are the most fundamental principles of morals, identified and justified by Kant’s Natural Law Constructivism.” (Westphal 2016: 141)
• 16
  For instance, the kangaroo rat (dipodomys phillipsii) is a species of nocturnal rodent native to the deserts of North America, who do not need to drink water at all. They extract enough fluid from the digestion of their food and their metabolism. New-Kantians would not be inclined to call such creatures rational though, as will be noted in the fourth paragraph of this paper.
• 17
  Below I, again, present the example of Christine Korsgaard.
• 18
  In the work of Kant, this is reflected in his attempt to conceptually make the moral law precede any objects of the will: "[...] it is [...] the moral law that first determines and makes possible the concept of the good, insofar as it deserves this name absolutely. [...[ It explains at once the occasioning ground of all the errors of philosophers with respect to the supreme principle of morals. For they sought an object of the will in order to make it into the matter and the ground of a law (which was thus to be the determining ground of the will not immediately but rather by means of that object referred to the feeling of pleasure or displeasure), whereas they should first have searched for a law that determines the will a priori and immediately, and only then determined the object conformable to the will" (Kant 1999, 5: 64)
• 19
  This is what Scanlon famously claimed for his variety of Contractualism, when he wrote: “Part of contractualism's appeal rests on the view that, as Mackie puts it, it is puzzling how there could be such properties "in the world." By contrast, contractualism seeks to explain the justificatory status of moral properties, as well as their motivational force, in terms of the notion of reasonable agreement.” (2013, p. 601)
• 20
  A point for further consideration would be whether Scanlon in his surprising statement that his form of contractualism is ‘substantive’ (2000, p. 151). Both the focus on justifiability instead of the good, and the focus on wrongness instead of the right, heavily suggest that his theory of ‘what we owe each other’ leave it undetermined what specific actions and goods people should pursue.
• 21
  To briefly reiterate: “So act that you use humanity, whether in your own person or in the person of any other, always at the same time as an end, never merely as a means” (Kant 1997, p. 38).
• 22
  Kleingeld quotes the following passage from the Groundworks: “[H]e who has it in mind to make a false promise to others will see at once that he wants to make use of another human being merely as a means, without the other at the same time containing in himself the end. For he whom I want to use for my purposes by such a [false] promise cannot possibly consent to my way of behaving toward him and so himself contain the end of this action”. (4:429-30)
• 23
  Korsgaard writes: “The kind of value that Kant thinks attaches to ‘persons’ is one in response to which we are supposed to respect their choices [...]” (Korsgaard 2013, p. 632)
• 24
  Westphal seems to suggest that we should aim for ‘communication with everyone, not just our fellow partisans’ (Westphal 2016, p. 133). There might be a deeper disagreement between Westphal and this formalism I have just described. Perhaps this stems from his modal understanding of justification in ethics. At this point, I am not entirely sure yet.
• 25
  This is also why animals are difficult cases for ethics which stress consent, since they seem incapable of truly understanding what we intend to do with them. For instance, my cat cannot consent to being taken on a long drive to see a vet and cure his inflamed eye, since this is not communicable to him. Korsgaard, for instance, writes that: “animals cannot give their free, unforced, and informed consent to what we do to them.” (2018, p. 177) Although, see the brilliant overview book on communication and animals from (Meijer 2019)
• 26
  Though this article is not on Kant but on New-Kantians, something of this idea can be found in the first pages of Kant’s book on religion, where he notes that ‘natural creatures’ are the only place where ethically neutral disposition can be found: “A morally indifferent action (adiaphoron morale) would be one that merely follows upon the laws of nature, and hence stands in no relation at all to the moral law as law of freedom - for such an action is not a factum, I and with respect to it neither command, nor prohibition, nor yet permission (authorization according to law), intervenes or is necessary.” (Kant, 2001, 72)
• 27
  In her book Self-Constitution, the nature of action is explored exactly as our capacity to constitute ourselves as agents. For instance: “The intimate connection between person and action does not rest in the fact that action is caused by the most essential part of the person, but rather in the fact that the most essential part of the person is constituted by her actions” (2009, p. 100)
• 28
  This is where misunderstanding looms. Frey writes: “According to Kantians, we are autonomous when we are fully self-determined agents; we are fully self-determined agents when we act in conformity with the formal principle that constitutes an exercise of the capacity to will (or practical reason); in so determining ourselves through our conception of the good, we are self-consciously aware of ourselves as the efficacious cause of what is objectively good for all finite rational beings who bear the same capacity” (Frey 2019, p. 172). This final statement on ‘what is objectively good for all finite rational beings’ is ambiguous. Korsgaard would most likely say that if we understand this to be substantive goods (specific things we should desire or specific actions we should pursue), there isn’t a lot all finite rational beings will agree on. It will be good for those who need to drink water, to drink water. It will be good for those with sharp claws to sharpen these on pieces of wood. Rational beings don’t need to live exactly similar lives to be rational.
• 29
  Arguably, many Kantians like Korsgaard are committed to the idea that rationality exists and is roughly determinable in terms of cognitive capabilities. Different approaches are possible. For instance, the deflationary approach by Gibbard according to which rationality is simply the practical property of ‘making sense’ (1990) or the epistemological behaviorism of Rorty (1979) or the historic pragmatic approaches by Hegelians like Brandom (2006).
• 30
  Rawls in his commentary on Kant used to follow a similar process of thought as we are pursuing here, by letting self-reflection lead to autonomy, and autonomy to a testing procedure of more contingent desires. He wrote: “Thus, as ideally reasonable persons, we have the capacity to stand above and to assess our object-dependent desires. This gives us an elective power to determine from which of those desires, if any, we shall act” (Rawls, 2000, 152).
• 31
  For my use of the term ‘formal’, compare Korsgaard: “Roughly speaking - and I’m afraid all of the speaking will be a bit rough here - a substantive conception of morality identifies morality in terms of its content, while a formal conception of morality identifies it with a method of reasoning about practical issues. Most people, intuitively, think of morality in substantive terms” (2009, 49).
• 32
  Korsgaard writes: “Of course depending on their theories of rights and obligations, philosophers will disagree about which natural capacities are required. I myself think it is plausible to say that we cannot intelligibly assign any obligations to an entity which is incapable of reaching conclusions about what it ought to do and acting on those conclusions.” (2013, p. 633)
• 33
  Compare Kant in his Critique of Practical Reason: “What we are to call good must be an object of the faculty of desire in the judgment of every reasonable human being, and evil an object of aversion in the eyes of everyone”. (Kant, 199, 5: 61). This heavily suggests a more substantive approach. But we can also note Kantians like Korsgaard generally don’t interpret this passage this way: “What Kant means by this of course is not that everyone must care about the same things that I do, but rather, that if my caring about an end gives me a genuine reason for trying to make sure that I achieve it, then everyone else has a reason to value my achieving it as well, a reason not to interfere with my pursuit of it, and even a reason to help me to achieve it if I need such help.” (2018, 138)
• 34
  It is relevant to note that some Kantian constructivists, much like Frey, present themselves as heirs to the issues raised by Murdoch and Anscombe. Bagnoli writes: "In a way, Kantian constructivism is carrying this legacy [of Murdoch and Anscombe] forward, insofar as it addresses foundational issues about the objectivity of moral obligations within a more comprehensive theory of practical cognition and rational agency." (2013, 14)

• 35
  Article info CDD: 193

Publication Dates

• Publication in this collection
  19 Jan 2024
• Date of issue
  2023

History

• Received
  12 Dec 2022
• Reviewed
  15 Aug 2023
• Accepted
  13 Dec 2023