1. Sangiovanni
Andrea Sangiovanni’s comments all focus on the first part of Kant’s Political Legacy (Caranti Reference Caranti2017). After a synthetic (and correct) résumé of my main foundational argument, Sangiovanni presents four major criticisms, three of which relate to my idea that autonomy is something we should investigate if we want to explain why we owe things to humans simply by virtue of their being human. Following Kant I argue that we are autonomous creatures and that our autonomy bestows on us an inherent worth and, as a consequence, dignity. The reason is that autonomy makes us capable of agency fully unrelated to our sensuous inclinations to obey a principle we perceive as morally mandatory. Kant says that autonomy places us above the natural world. Less emphatically, we could say that autonomy makes us creatures who systematically escape a description that would otherwise make us merely intelligent pursuers of our own well-being and preoccupied only with the satisfaction of sensuous desires.
But here is Sangiovanni’s first problem: if autonomy is an inherent worth that confers on us this status, there appears to be a gap between my having that value and my having rights. Ironically, this gap would be similar to the one I spotted in Griffin’s unjustified move from caring about agency to having a right to the conditions that make agency possible. This time the gap is supposed to be generated by the non-relational structure of inherent worth – a value that exists independently of whether someone can claim it and addressees of that claim can be identified – and the clearly relational structure of rights (rights are always claimed by someone against someone else). To explain his point Sangiovanni gives the example of great, extremely beautiful, hence valuable paintings, which as such have intrinsic worth. In this sense, he thinks, great paintings can be said to have dignity. Certainly, they bestow on us duties: we are to preserve them and pass them to future generations. But it would be bizarre to say that they have corresponding rights against us, because they cannot in their own name make claims on us. In the same way, even if autonomy were the high value I claim it to be, it would be unable to generate rights against specific addressees. Autonomy may be able to protect people generically against torture, but is unable to protect me against you torturing me.
I do not know if I correctly understand Sangiovanni here. To begin with, even granting that paintings have dignity (an expression not accidentally rather odd), this dignity is quite different from the one Kant attributes to humans. As Sangiovanni knows, the value of the painting is that of a commodity. In fact, it has a price on the market, like anything that is valuable to us because it satisfies some desire or serves some interest. The painting would be worth nothing if we did not take aesthetic pleasure from seeing it. Let us now recall the passage from the Metaphysics of Morals (MS, 6: 434)Footnote
1
in which Kant basically says that anything has a price in the world, except persons, because persons display autonomy. It seems that the ‘dignity’ of a great painting does not fall in the same category as that of persons. This should already caution against the impression that what I mean by autonomy can be likened to what we all mean by the value of a great painting or of any other commodity. Quite simply, a great painting has a commodity-like value, not a dignity. In fact, I have never heard of any painting in the history of art, even those of the highest monetary value in the world, being attributed this quality.
Secondly, even granting that the value of a painting is sufficiently similar to the value of human dignity as Kant and I construe it, I confess that I fail to see the relational problem Sangiovanni indicates. My dignity gives me a claim against everybody to be treated with respect. And in general the dignity possessed by all human beings gives all of them a right against everybody to be treated with respect. Obviously how respect gets translated into specific rights is a problem, but not one that has to do with the supposed lack of relational structure of dignity. Similarly, the dignity of a painting gives it a right against everybody to be treated in a certain manner. Of course it is odd that a thing has a right, but this does not have to do with a lack of relational structure of its value. It has rather to do with the fact that saying that we have duties towards a painting is a way of saying that we have duties towards other people that either own or have a right to see it, now or in the future, independently of ownership. In short, dignity, as Kant and I understand it, seems to have a fairly clear relational structure: All human beings have a right against all human beings to be treated with respect. And thus, when it comes to you torturing me, it only takes a simple syllogism to arrive where Sangiovanni wants: I have a right against you not to be tortured.
Sangiovanni’s second criticism is easier to understand (for me at least). He protests that I bet everything, so to speak, on a mere capacity to act morally as opposed to the realization of that capacity. Dignity for me has to do with the capacity to follow the moral law, not with how many times we have turned that capacity into concrete actions.
Before replying to this very reasonable concern, I would like to react to the way in which Sangiovanni frames his critique. To reinforce his point, he says: ‘After all, human beings can use their capacity to act morally to act immorally.’ This may be simply an unfortunate way of making the point, but as Sangiovanni knows, in chapter 1 I go to some length to clearly distinguish between what Kant at times calls practical freedom, or spontaneity, that is, the capacity to set any ends whatsoever, and autonomy, the more specific and higher capacity to act in complete independence from inclinations, hence to act from duty. So on my reading, when human beings act immorally they are not using the same capacity that they need to act morally. Of course, Sangiovanni may protest. If practical freedom is the capacity to set any ends whatsoever, this includes moral ends. Hence it seems that practical freedom encompasses autonomy, as opposed to being an altogether different capacity. Well, here the way we read the evolution of Kant’s theory of freedom becomes relevant. And what I am going to say will also apply to a similar criticism from Guyer further on. On my reading, which is inspired by Allison (Reference Allison1990), while up to the first edition of the first Critique Kant thought that practical freedom was sufficient for the sake of morality, the Groundwork (1784) marks a great novelty, that is, the introduction of autonomy as a new capacity, altogether different from practical freedom, and considered as necessary and sufficient for the sake of morality. Morality no longer rests on the generic capacity to set ends in a way that is not fully determined by sensuous motives, but on a capacity to set ends in complete independence from them. And this obviously leads to the conclusion, which I should probably have spelt out better in the book, that practical freedom can not set moral ends. And this explains why Kant now says that we have a faculty that distinguishes us from animals and yet it does not make us persons (MS, 6: 434). That faculty is precisely practical freedom, now redefined as able to set any ends freely, but never in complete independence of sensuous impulses, that is, as unable to set moral ends.
But let us go back to Sangiovanni’s central point. Why bow to a sheer capacity and not focus on people’s actual level of morality? Does that not beg the whole question of explaining what makes us worthy of human rights protection? My answer is articulated in three parts. To begin with, when we deal with human rights we are trying to make sense of rights that are supposed to protect even the worst of us. Unless Sangiovanni wants to say that people like Saddam Hussein, Gaddafi, Mobutu did not deserve the protection of human rights, it seems we had better focus on a capacity as opposed to actual behaviour.
Secondly, Kant’s story – at least on my reading – is that we have a capacity that sets us above the sensible world. Now if we all have this capacity, and this fact sets us above the sensible world, I think we have moved a step in the direction of showing why people deserve the protection of human rights. I do not see myself as begging the question. Proving that much, of course, is still insufficient to show that the argument is sound. Should one in fact accept Kant’s linkage between a capacity and dignity, as an article of faith? This leads me to the final part of my answer.
Since empirical confirmation is ex hypothesi not to be sought, I can only refer to what I said in the book. On my reading, the strongest argument Kant has for the ‘reality of freedom’ is the immediate consciousness each of us has of the fact that we can do in all circumstances what morality requires. This is a classic, complex and still debated issue in Kant studies, turning on the interpretation of the ‘fact of reason’. I will therefore ask to be excused if I do not defend in detail my reading of this concept crucial for understanding the entire edifice of Kant’s moral thought in this limited space. But the gist is rather simple. As the example of the person threatened to offer false testimony against an innocent man shows, we know that even in the worst circumstances we could choose the moral path no matter what the consequences are. Acts of moral heroism exist, history is full of them. And the reason why we connect to them, wish we could be as courageous or the like, is that we identify with these heroes. We do not say that they were extra-terrestrial entities whose functioning we can ignore. Each of us in foro interno knows that he or she could have been that hero.
Sangiovanni’s next worry is that we may be autonomous, but not all to the same degree. As he puts it, ‘why should we believe that individuals’ capacities to act in accordance with the moral law are equal? Some individuals’ capacity to act morally are very high – on average and in normal conditions, they display great moral strength, resoluteness, resilience and courage – whereas others’ capacities are very low. So, if dignity resides in the (awesome) capacity to act morally, then why should not those with greater moral capacities have greater dignity, and therefore higher status?’ A conclusion that would obviously be in stark contrast with what Buchanan (Reference Buchanan2010) called the egalitarianism of human rights. Moreover, Sangiovanni warns against a possible easy way out here. If I say that the difference individuals display in moral behaviour is insufficient ground to deny that we all have the same capacity, but simply that we exercise it more or less often, or more or less well, then ‘we seem to have departed very far from a common-sense view of what a capacity is, and to have won the argument at the price of a highly controversial metaphysical view’ (p. 254). Let us see.
To begin with, it is important to realize that on my reading autonomy is ‘simply’ the capacity to overcome all sensuous impulses if that is necessary to act morally. I do not want habits, ‘second natures’, developed in an Aristotelian way through exercise, to be part of autonomy. It is simply not what I mean. For being entitled to the status that confers human rights, on my theory, you simply need to pass a minimal threshold. This threshold is precisely that you can overcome all your inclinations. Now, is it ‘self-evidently false as a matter of fact’ that all human beings have that ‘minimal’ capacity? I do not think so, and refer back to what I said above about the immediate consciousness that each of us has of possessing this capacity. I do not claim that this immediate consciousness is a rock-solid basis for proving that the opposite view is ‘self-evidently false’. After all, short of the possibility of appealing to evidence, and given that we are not talking about an a priori truth, could there be such an argument? I am happy with the result that, construed in the minimal way I just described, autonomy is not at all far away from personal experience and common sense. We find it any time we reflect on what we could do if forced into the shoes of the poor man threatened to give false testimony.
The point can perhaps be further illustrated by a thought experiment. Let us imagine that we have to judge a woman who has stolen property but claims to be affected by kleptomania. It seems that we only have two options: either we believe the story, consider her insane, hence not responsible for the crime, or we do not and, independently of all the possible mitigating circumstances, we find her guilty. We do not say that her autonomy (and therefore her accountability) was depleted. We do not punish her because she did not develop past moral fortitude. We either say that she was free/autonomous, hence she is responsible, or we decide that she was sick, hence not responsible because not an agent at all. And, if we decide that she is guilty, we do not make the gravity of punishment dependent on how autonomous she was but on how pressing were the sensible motives that suggested that kind of behaviour (she was hungry, needed the stolen property for the well-being of her kids and the like). It seems to me that this is the way we usually reason about people and their responsibility, and I propose that a minimal, non-scalar conception of autonomy is an integral part of our judgements. If that is the case, perhaps my account (and Kant’s) is not so removed from common sense or helplessly metaphysical.
Finally, is this minimal conception of human autonomy compatible with my making room for animal autonomy, that is for different degrees of autonomy among species? I think it is. In the book I attributed a certain degree of moral agency to non-human primates. However, I never suggested that they reached the ability to silence all their natural impulses, that is that they reach the minimal threshold of autonomy. Sure, I talked about ‘animal autonomy’ but that was a rhetorical way of expressing the point that animals approximate, without reaching, the capacity of ‘pure’ agency. In fact, I attributed to them the capacity of self-sacrifice (how could one deny that to dogs, for example?), but this is different. Animals seem to sacrifice themselves because the empathic impulse to save their master or their offspring can be stronger than the instinct of survival. In contrast, humans are supposed to make (or to be able to make) a conscious and detached deliberation about their course of action. In animal morality it is just this conscious, free, reflexive endorsement of such an act that seems to be missing.
The fourth and last criticism of Sangiovanni has to do with the very possibility of having a truly Kantian theory of human rights. In particular, Sangiovanni cites and endorses the six features of human rights I list at the beginning of my book, but invites me to see that five of them are clearly incompatible with Kant’s injunction against unilateral imposition, that is, his (in)famous idea that once a state is formed, one may not question its legitimacy either from inside or outside. It would follow that my idea that there could be a Kantian theory of human rights (something like the very book I wrote) looks very dubious, because human rights are precisely supposed to enable internal and external criticism.
The irony here is that I agree with Sangiovanni. The five features of human rights in question are incompatible with Kant’s prohibition to resist the supreme authority. What I would like to emphasize, in a way that I must not have succeeded in doing in the book, is that I subscribe to Buchanan’s distinction between human rights (HR) and international human rights (IHR). HR are those rights that we may have all sort of philosophical or non-philosophical reasons to consider as human rights, independently of whether they are already part of international law. IHR are those rights already present in official legal documents. The two sets do not coincide and this is good because, as Buchanan emphasizes, this allows a dialectic between the theory of HR and the theory and practice of IHR. If we come up with a list of HR too distant from the practice of IHR, we may want to rethink the philosophical reasons that led us to that list. Conversely, if IHR are in sharp contrast with what we have strong reasons to believe should count as HR, we may want to change international law. Now, in the passage that Sangionvanni cites I was speaking about human rights in general (both HR and IHR). But I thought it was clear that most of what I say in the book concerns HR, not IHR. I am – in a very traditional manner – offering a philosophical thesis that tells us why certain things should count as rights we have merely by virtue of our humanity, leaving aside the question of the fit between this theory and IHR. In this sense I will concede to Sangiovanni that there cannot be a truly Kantian theory of IHR because too many features of them conflict with Kant’s prohibition against resistance. But I insist that there could (and should) be a theory of HR, like the one I tried to offer, from which, among other things, we may get guidance on how to change IHR, if the misfit between the two domains is too large.
In addition, it is important to remember that Kant qualifies his unconditional duty to obey. In the relevant passage in the Metaphysics of Morals he writes: ‘Obey the authority who has power over you (in whatever does not conflict with inner morality)’ (MS, 6: 371). One reasonable reading of the parenthetical clause is that that if the despot demands of me that I do something that conflicts with morality – say give false testimony against an innocent man – then the non-conditionality of obedience starts shaking. Whether this is meant as a full-fledged authorization to rebel or resist is open to debate. What interests me is to see that even the rigid Kant of ‘obey no matter what’ leaves some room for having an external, moral perspective from which legitimate authority may and should be evaluated, criticized and possibly resisted. This suggests that even the just conceded incompatibility between Kant’s theory and IHR may need qualifications.
2. Williams
Howard Williams esteems my work probably beyond its merits, and I am not going to fight hard to change his mind. Since he does not have much to quarrel with in my interpretation, and I have little to quarrel with in the general framework in which he nicely inserts it, I hope he will excuse me if my reply is rather short. Williams, however, does have three points of disagreement with my account. The first relates to my reading of the status of Kant’s thesis that there is a guarantee of perpetual peace. The second has to do with my ‘isolated view’ of the moral politician. The third concerns my failure to insert in fine enough detail Towards Perpetual Peace within Kant’s general theory of law.
Regarding the first point, Williams thinks that I am making the ambitions of the guarantee thesis too high. He agrees with me that Kant wants to get his hands dirty with some empirical considerations, but thinks that anything Kant says to convince us that perpetual peace is possible, or worse, necessary, is to be taken as nothing more than a counterargument against the sceptic. As such it is not meant to have any more empirical solidity than the argument presented by the opponent. In a similar vein, Williams argues that the (sole? main?) point of the guarantee thesis is to ‘clear space for full deployment of practical reason in politics’ (p. 267). Williams explains this by equating the role of the guarantee thesis with that of the highest good: ‘It [the guarantee thesis] is a future related way of conceiving our political actions that gives them moral coherence. Just as “the highest good is the necessary highest end of a morally determined will and is a true object of that will” ([KpV] 5: 115), so perpetual peace is a highest end for a morally motivated political leader’ (p. 267).
I have very little to argue against the ‘practical’ reading which Williams defends and I think that it is perfectly compatible with my more theoretical one. I simply think that Kant wants to do something more than point to a necessary object we need to conceive of in order to make our practical (moral or political) agency coherent. He provides some essentially theoretical, broadly empirical reasons to believe that progress in human affairs is more likely than any other outcome. And he does so by referring to natural mechanisms that, combined with background assumptions that Kant does not make explicit, but evidently assumes, are supposed to back up this view. If the guarantee thesis were merely to play the role Kant reserves for the highest good in our moral life, then it would be hard to explain why he talks in Towards Perpetual Peace of a visible (hence rather empirical) ‘mechanical course’ of nature that exhibits ‘purposiveness … letting concord arise by means of the discord between human beings even against their will’ (ZeF, 8: 360). And why he makes an effort to detail how concord is to spring from discord, by mentioning unsocial sociability, the unintentional benefits of linguistic and religious differences, the equally unintended peaceful significance of greed (trade as an unconscious force for peace) and so on. Williams says that the guarantee thesis is ‘is neither wholly an empirical prediction nor merely a source of reassurance for the right-minded’ (p. 267). I agree. I would simply add that if it is to be more than a reassurance for the righteous, then some quasi-empirical claims have to be given (and in fact are). One of the main goals I had in interpreting this controversial part of Kant’s political thought was precisely to spell out as clearly as possible these quasi-empirical claims.
Regarding the second point, Williams agrees with me that the life of the moral politician is more difficult than that, already not easy, of the moral person, mainly because the former has a responsibility to foresee the consequences of his or her decision, while the moral person is
relatively free from that responsibility. This Weberian contrast between an ethics of responsibility and an ethics of principles was mainly meant to underscore that in order to predict the consequences of moral political actions a good deal of (good) empirical science and, most importantly, of phronesis is necessary. This led me to conclude in a rather pessimistic tone (at times even my optimism wavers) that it is very rare to encounter moral politicians. But Williams reassures me that there is no need to construe the moral politician as an isolated hero expected to be near-omniscient. A moral politician, quite simply, may ask for help from counsellors and experts. This is part of the reason why Kant cared about the freedom of the pen. Experts have to be able to enrich the public debates, even if the final decision remains in the hands of politicians. I think Williams is right. Even when I try to be a pessimist, I do not succeed.
Regarding the last point, I am also inclined to concede the point, if this means that I may have not sufficiently emphasized the parts of the Metaphysics of Morals relevant for the peace project, or the continuity between Towards Perpetual Peace and the later work. But I reject the criticism if this is meant to charge me with having paid scant attention to the role that the law is supposed to play in Kant’s project. While I resist its full reduction to an institutional recipe, because I think that the morality of common people and leaders remains an essential part of the story, nonetheless I do see Kant as arguing for a priority of the law over morality in the progress towards our cherished end. As he puts it in a passage in Towards Perpetual Peace, the importance of which is difficult to exaggerate: ‘it is not the case that a good state constitution is to be expected from inner morality; on the contrary, the good moral education of a people is to be expected from a good state constitution’ (ZeF, 8: 366). My interpretation not only of the 1795 essay, but also of Idea and of On the Common Saying, remains faithful to this principle (Caranti Reference Caranti2017: 121–2, 126, 138).
3. Guyer
Paul Guyer honours me with a careful reading of my work, aptly combined with clear statements on the points in which our views differ. I will try to do justice to each of them.
The first minor disagreement has to do with the permissions Kant infers from the innate right to external freedom. Guyer thinks that these permissions are three, not four as I suggest. The authorizations to be one’s own master, to be beyond reproach and to communicate one’s thought freely would be nothing but articulations of the right to formal equality. So neither do they add anything to formal equality nor does formal equality add anything to them. Now, from a textual point of view, it is true that Kant introduces the three authorizations through a mithin (hence), which may be taken to suggest that they ‘follow’ what has just come before, that is, the right to equality. And I am ready to concede that equality may stand to these authorizations as a kind stands to its species. What I am not convinced of is that innate equality can be ‘exhausted’ by the three authorizations. The notion of (formal) innate equality seems to entail more. It entails, among other things, that no discrimination of sex, race, religion is allowed, that we are equal before the law, that no noble rank is recognized and so on. This seems to go way beyond the three specific authorizations that Kant cites. Indirect evidence of this comes from the 1784 Naturrecht Feyerabend where, as Guyer knows well (Guyer Reference Guyer2016), the authorizations Kant infers are neither three nor four, but five (Kant adds now the right to acquire property). This suggests that the list of rights or authorizations one may derive from the innate right to freedom (and equality) is open-ended, something hardly compatible with the notion of equality being exhausted with three (or four, or perhaps any) of them.
More important is our disagreement about the role that the notion of ‘humanity’ plays as basis for the innate right to (external) freedom. Guyer alerts me to the fact that Kant’s official definition of humanity (‘the capacity to set oneself an end – any end whatsoever’) does not seem to square with the role that I cut out for it in the argument. Having the capacity to set any end is what enables us to set moral as well as immoral ends. Hence in humanity, no intrinsic worth is to be found.
This criticism is similar to one of Sangiovanni’s we have already discussed, and I take the opportunity to expand my reply. To begin with, I recognize that Kant’s term for our capacity to set ends in complete independence from sensuous inclinations, that is, for autonomy, is personality, not humanity. Even in the passage from the doctrine of virtue I cite, it is the human being considered as a person that has dignity and is set above any price. It is also true, however, as I say in a footnote, that Kant is not always consistent in the use of these two terms and sometimes does use ‘humanity’ when he clearly means ‘personality’. In the Groundwork, for example, Kant is explicit that what confers worth on humans is our personality understood as ‘freedom and independence from the mechanism of the whole nature’ (G, 4: 87). But only a few lines later it is ‘humanity’ that we must consider as holy in us: ‘A human being is indeed unholy enough but the humanity in his person must be holy to him’, and of course, yet another few lines down, it is humanity (not personality) that is ‘an end in itself’. This terminological inaccuracy is understandable because not only human beings, but all rational beings are persons: ‘rational beings are called persons because their nature already marks them out as an end in itself’ (4: 428). Hence it is reasonable to assume that sometimes Kant uses the term ‘humanity’ as a short name for that subset of persons that are human. My suggestion is that it is precisely in this sense that he uses ‘humanity’ when he refers to the basis for the innate right to external freedom. Moreover, as I said in response to Sangiovanni, it is true that humanity in a sense encompasses personality because, as Guyer correctly notices, it is the capacity to set any ends, which of course includes moral ones. And yet, at least from 1784 on, autonomy is sharply distinguished from the freedom we have to set ends that are sensuously determined, and there is no doubt that Kant sees the former capacity, not the latter, as the seat of our worth.
To back up my textual reasons, in the book I offer few systematic reasons. If the basis in question were humanity, understood as what makes us superior to animals and yet having ‘ordinary value’ or merely ‘extrinsic value’ (MS, 6: 434), I do not see how Kant could claim that humans acquire any right whatsoever because of (kraft) it. Guyer thinks that the right in question arises not from humanity itself, but from the injunction to treat humanity never merely as a means, that is, from the categorical imperative. However, this only seems to push the question one level up. Why should we use humans never merely as means? I imagine Guyer thinks that Kant is allowed to borrow here the results of his foundation of morality in Groundwork and the second Critique. Well, possibly so. After all, I am a ‘dependentist’ like him in the reading of the relation between morality and law in Kant. On my interpretation, however, the mature Kant grounds morality itself on the fact that we are autonomous. Autonomy is the necessary and sufficient condition of morality and the ultimate basis on which the whole edifice of Kant’s juridical thought rests. So, even if Guyer is correct, this does not change the basic fact that it is autonomy that does the normative work.
This leads me to Guyer’s philosophical worry. The problem he sees is that the fundamental principle of morality cannot be respect for the capacity to be moral because this is stained with circularity. If I am supposed to respect those entities that are capable of being moral, I need an independent criterion that tells me what is moral, and then apply it to the appropriate individuals. My short answer here is that, as Guyer knows well, for Kant the content of the moral law, what is moral and what is not, is clear to any common human understanding (G, 4: 454). We have an immediate access to that criterion. Equally evident for us is the fact that we are persons, in the sense of being able to follow what we intuitively know is obligatory (KpV, 5: 87). Now, one can disagree with Kant that this content and our capacity to act on it are so evident. But I think this suffices to show that Kant’s argument, on my reconstruction, can be charged with intuitionism, not circularity. And I confess that, given Kant’s emphasis on how we discover the reality of freedom through the immediate consciousness of duty (the fact of reason, as I read it), I feel quite confident in attributing to Kant this bit of intuitionism.
What about Guyer’s interesting alternative to my foundation? He suggests that Kant derives the fundamental principle of morality from rationality itself, in particular from the principle of non-contradiction. As evidence, he cites some very early annotations on the theme of morality, handwritten on the margins of Kant’s own copy of On the Feeling of the Beautiful and Sublime (1764), according to which there is something not just ‘ugly’, but also ‘contradictory’ and ‘absurd’ in treating someone who has ‘a soul’ as if he or she did not have one. Kant seems to be suggesting that all we need to ground morality is to appeal to the basic principle of rationality, that is, to the principle of non-contradiction.
In proposing his alternative foundation, Guyer recognizes that it presupposes that it is evident to everyone that each human being has a ‘soul’ or a ‘free will’. But this is exactly the point that I am defending. My foundation says that morality rests on the recognition that we are all autonomous, that this entails awe and respect, and – here comes Guyer’s point – that we cannot, on pain of contradiction, treat people as if they did not have that capacity. Now Guyer may protest: No, the fundamental assumption is that we have a free will – the freedom to set any end – not that we are autonomous (the freedom to set moral ends). But now we are back to a point already made. I would reply that it is probably true that by ‘having a soul’ Kant meant in 1764 a ‘freedom of the will’ in the sense of what I call ‘practical freedom’. But the great progress Kant makes from the pre-Critical to his mature view of morality is precisely that he stops thinking, as he still did in 1781, that practical freedom, free will in the generic sense, is sufficient for the sake of morality. As I have already mentioned, from the Groundwork on (1784, so twenty years after that handwritten remark) Kant introduces the crucial idea that it is autonomy, not mere ‘practical freedom’, that is both a necessary and a sufficient condition for morality.
More importantly, leaving the evolution of Kant’s thought aside, if the ‘soul’ we are all supposed to see clearly were the sheer capacity to set ends in an intelligent, rational way, then there would be no contradiction in recognizing this ‘soul’ and treating people who have it merely as means. These people would be, as the mature Kant would say (MS, 6: 434), commodities, extremely precious ones, and yet still commodities. Hence, the most I feel confident to concede to Guyer here is that the principle of contradiction enters the picture once what really does the foundational work – autonomy – has already been secured.
Let us now move – I am afraid briefly – to Guyer’s two other criticisms of Part 1 of my book. The first is that I would need to do more than cite a few lines from sacred texts to bring home the point that adherents to any of the major religious traditions would – properly guided through Socratic questioning – recognize autonomy as a condition on the very possibility of having a significant religious life. Yes, as I already did with Sangiovanni, I plead guilty here. Let me just clarify that, contrary to what Guyer thinks, I was not expecting believers to show ‘habituation to establish a degree of detachment from [their] deepest comprehensive convictions in favour of a negotiation with other individuals who adhere to reasonable comprehensive doctrines radically different from the one[s they] uphold’ (Caranti Reference Caranti2017: 92). This is the expectation that proponents of global reason as the ultimate normative compass for human rights have to form. I argued that this expectation is unrealistic and expressed my preference for reflective equilibrium.
But Guyer’s main worry is that I cannot get a specific list of human rights from my refined Kantian account. He argues convincingly that I drew too little from Kantian resources in the move from the higher and fundamental innate rights to freedom and equality (possibly with the few ‘authorizations’ Kant infers from it) and left too much to each society’s freedom to come up with a specific list of basic rights. Guyer shows me how one could infer most, if not all, the rights that Kant lists in the Doctrine of Right from the innate right to external freedom. I have nothing to argue against Guyer’s expansion from this basic tenet. In fact, not only do I take it as fully compatible with my argument, but I also think of it as an important and welcome strengthening. I simply wish to clarify the following point. In the book I was mainly trying to show that Kant provides a powerful and non-parochial way of bringing home the fundamental principle that lies behind the human rights culture, that is, the idea that humans have rights by virtue of their humanity. This is why I did not focus enough on how to infer extra rights from the right to freedom and equality. But if there is a Kantian way to get from these two overarching and very generic rights of external freedom and equality closer to the specific and numerous human rights we are accustomed to, this is great news for my approach.
Guyer then moves to discuss the third part of my book, and he focuses on my defence of the guarantee thesis. The first concern, it seems, is that I do not sufficiently define the thesis I want to defend, because I oscillate between the idea that there is a guarantee of peace and the weaker thesis that peace is merely likely. Well, I do oscillate because I think, in a way that I should have explained better, that the two theses, properly qualified, are compatible, and I want to assert both. My position is that by pointing to some natural mechanisms Kant shows that the world has a propensity towards peace. A propensity, in the Popperian sense, is an objective tendency which ‘rigs’ a natural system towards a certain outcome. Now saying that a system has a propensity is fully compatible with saying that in certain specific instances the system will evolve (temporarily) in a direction opposite to that indicated by that propensity. So, given any time T1 in world history, we would have no certainty that the world at T2 will not progress, rather than regress. Nonetheless, we still have the certainty (guarantee) that the system will reach the condition of stable peace at Tx, if x is a point in the future remote enough to give the system the time necessary to adjust any possible involutions running against its natural propensity. Obviously, Kant thinks that we cannot know when this will happen. Now, when I say that peace is merely likely I am saying it is likely that the world will evolve in the near future towards a higher level of security. When I say that peace is guaranteed I am talking about a remote point in the future (unknown to us) when the process (whose mechanisms we do know) will be completed. In short, peace is merely likely if we are talking about any specific time before the system stabilizes itself. Peace is guaranteed if we enlarge our view toward an indefinite future point.
Perhaps the thesis can be further clarified (again in a Popperian key) by focusing on what would falsify it. If tomorrow a foolish democratic leader waged war against another democracy after 200 years of peace, this does not falsify my thesis. The thesis would be refuted if the world today were not less bellicose than it was 200 years ago, given reasonable standards of measurement. Now since the opinions of social scientists differ on this point, I confess I cannot offer non-controversial evidence. Maybe wars in absolute terms, that is in the world at large, are neither fewer nor less cruel than what they used to be. It is undeniable though (and evidence is abundant and clear) that within the part of the world where Kant’s three definitive articles have been implemented most, war is less likely than it was in the past, in fact, than it has ever been. I do not know if this is enough to ground the guarantee thesis. I hope it is enough to discourage the thought that it is just a simple-minded rosy prediction.
What I just said should also help in responding to Guyer’s specific point about radical evil. I do not need to deny that humans, being capable of subverting the path of morality even if it serves their best interests, could in specific circumstances act in a direction opposed to the one predicted by the thesis. I do not need to deny the possibility cited above of foolish leaders. For my thesis to be confirmed, these bits of agency that run both against morality and against the long-term best interests of agents have to become more and more rare. Now, I think Guyer will concede that people living in liberal democracies, at least those which are part of large international organizations such as the EU, and which enjoy a high degree of interdependence and permeability with neighbouring countries, are less exposed to wars waged ‘as a kind of amusement’ (ZeF, 8: 350) or as ‘whims’ than the people living in the nation states in Kant’s time. Sure, there are still today dictators of despotic states (Putin, for example) and the probability of having foolish democratic leaders is not zero (Trump is in power). But after the longest period of peace in European history, I would resist the idea that humanity is not less exposed to wars today than at the time when Kant wrote his ‘philosophical sketch’, at least that part of humanity that has, so to speak, ‘learned their lesson’.
I do not think that this comes even close to settling the question about whether we are in fact approximating perpetual peace. My goal in this part of the book was far more modest. I wanted to take Kant’s arguments seriously, try not to neutralize their theoretical potential, and show that they are not metaphysical and simple-mindedly optimistic, because they are meant to spot a tendency in an objective system, admittedly a rather large system: the world as it is and as it will be.
Guyer finds the second part of my book the most convincing. I can therefore be brief in my response. First, I would like to clarify that I do not attribute to democratic peace scholars the view that majority rule is all that matters for the sake of peace – no serious theorist in that tradition thinks so, and certainly Doyle and Russett do not. In fact, what is now known as ‘democratic’ peace was originally called ‘liberal’ peace because crucial for Doyle, besides wide suffrage, were liberal institutions and values (Caranti Reference Caranti2017: 166–8). Moreover, Doyle’s and Russett’s ways of replying to generations of critics of the liberal peace thesis makes clear that the idea was (and has always been) that only when the three Kantian factors are combined is there ‘certainty’ of peace (191–4). My deepest criticisms of the democratic peace scholars are rather (a) that even if enriched by liberal elements, democracies should not be equated with republics – the crucial missing element being a republican ethos – (b) that they have misunderstood the nature and role of the international Federation, and (c) that they have reduced the significance of cosmopolitan right to the classical liberal idea that international trade promotes peace.
This leads me to Guyer’s final criticism. He thinks that all there is in the third definitive article is ‘an anti-colonialist restriction of any presumptive right of persons of one nation to claim and occupy other already occupied regions of the earth for their own benefit’ (p. 287). Sure, there is that element, as Kant’s harsh and merciless criticism of colonialism in the very discussion of the third definitive article proves beyond doubt. But is there not also a positive side? If all the third article is supposed to accomplish is to counter colonialism, why not say that frontiers should be impermeable? Kant is explicit that the way in which the right to visit promotes peace is first by allowing the creation of economic interdependence (‘nature … unites nations … by means of their mutual self-interest. For the spirit of commerce sooner or later takes hold of every people’, ZeF, 8: 368), and secondly because through economic and cultural interdependence people enter into a ‘universal community’ (8: 360), which Kant thought was so developed already in his times that ‘a violation of rights [we would say today human rights] in one part of the world is felt everywhere’ (8: 360). In short, cosmopolitan right counters colonialism and exploitation, but also enables commerce and the creation of a global moral conscience, which are the peace-promoting factors that the third definitive article adds to those of the first two. This positive side is not only compatible with anti-colonialism, but also complements it by showing how humanity is bound to progress towards a universal consensus (I call it a ‘global moral conscience’) on fundamental normative standards that we would call today human rights. And of course human rights are primarily meant to rule out things such as colonialism and aggression. Perhaps human rights, peace and progress in Kant’s political thought have to do with one another even more than I suspected when I wrote the book.
You have
Access
1. Sangiovanni
Andrea Sangiovanni’s comments all focus on the first part of Kant’s Political Legacy (Caranti Reference Caranti2017). After a synthetic (and correct) résumé of my main foundational argument, Sangiovanni presents four major criticisms, three of which relate to my idea that autonomy is something we should investigate if we want to explain why we owe things to humans simply by virtue of their being human. Following Kant I argue that we are autonomous creatures and that our autonomy bestows on us an inherent worth and, as a consequence, dignity. The reason is that autonomy makes us capable of agency fully unrelated to our sensuous inclinations to obey a principle we perceive as morally mandatory. Kant says that autonomy places us above the natural world. Less emphatically, we could say that autonomy makes us creatures who systematically escape a description that would otherwise make us merely intelligent pursuers of our own well-being and preoccupied only with the satisfaction of sensuous desires.
But here is Sangiovanni’s first problem: if autonomy is an inherent worth that confers on us this status, there appears to be a gap between my having that value and my having rights. Ironically, this gap would be similar to the one I spotted in Griffin’s unjustified move from caring about agency to having a right to the conditions that make agency possible. This time the gap is supposed to be generated by the non-relational structure of inherent worth – a value that exists independently of whether someone can claim it and addressees of that claim can be identified – and the clearly relational structure of rights (rights are always claimed by someone against someone else). To explain his point Sangiovanni gives the example of great, extremely beautiful, hence valuable paintings, which as such have intrinsic worth. In this sense, he thinks, great paintings can be said to have dignity. Certainly, they bestow on us duties: we are to preserve them and pass them to future generations. But it would be bizarre to say that they have corresponding rights against us, because they cannot in their own name make claims on us. In the same way, even if autonomy were the high value I claim it to be, it would be unable to generate rights against specific addressees. Autonomy may be able to protect people generically against torture, but is unable to protect me against you torturing me.
I do not know if I correctly understand Sangiovanni here. To begin with, even granting that paintings have dignity (an expression not accidentally rather odd), this dignity is quite different from the one Kant attributes to humans. As Sangiovanni knows, the value of the painting is that of a commodity. In fact, it has a price on the market, like anything that is valuable to us because it satisfies some desire or serves some interest. The painting would be worth nothing if we did not take aesthetic pleasure from seeing it. Let us now recall the passage from the Metaphysics of Morals (MS, 6: 434)Footnote 1 in which Kant basically says that anything has a price in the world, except persons, because persons display autonomy. It seems that the ‘dignity’ of a great painting does not fall in the same category as that of persons. This should already caution against the impression that what I mean by autonomy can be likened to what we all mean by the value of a great painting or of any other commodity. Quite simply, a great painting has a commodity-like value, not a dignity. In fact, I have never heard of any painting in the history of art, even those of the highest monetary value in the world, being attributed this quality.
Secondly, even granting that the value of a painting is sufficiently similar to the value of human dignity as Kant and I construe it, I confess that I fail to see the relational problem Sangiovanni indicates. My dignity gives me a claim against everybody to be treated with respect. And in general the dignity possessed by all human beings gives all of them a right against everybody to be treated with respect. Obviously how respect gets translated into specific rights is a problem, but not one that has to do with the supposed lack of relational structure of dignity. Similarly, the dignity of a painting gives it a right against everybody to be treated in a certain manner. Of course it is odd that a thing has a right, but this does not have to do with a lack of relational structure of its value. It has rather to do with the fact that saying that we have duties towards a painting is a way of saying that we have duties towards other people that either own or have a right to see it, now or in the future, independently of ownership. In short, dignity, as Kant and I understand it, seems to have a fairly clear relational structure: All human beings have a right against all human beings to be treated with respect. And thus, when it comes to you torturing me, it only takes a simple syllogism to arrive where Sangiovanni wants: I have a right against you not to be tortured.
Sangiovanni’s second criticism is easier to understand (for me at least). He protests that I bet everything, so to speak, on a mere capacity to act morally as opposed to the realization of that capacity. Dignity for me has to do with the capacity to follow the moral law, not with how many times we have turned that capacity into concrete actions.
Before replying to this very reasonable concern, I would like to react to the way in which Sangiovanni frames his critique. To reinforce his point, he says: ‘After all, human beings can use their capacity to act morally to act immorally.’ This may be simply an unfortunate way of making the point, but as Sangiovanni knows, in chapter 1 I go to some length to clearly distinguish between what Kant at times calls practical freedom, or spontaneity, that is, the capacity to set any ends whatsoever, and autonomy, the more specific and higher capacity to act in complete independence from inclinations, hence to act from duty. So on my reading, when human beings act immorally they are not using the same capacity that they need to act morally. Of course, Sangiovanni may protest. If practical freedom is the capacity to set any ends whatsoever, this includes moral ends. Hence it seems that practical freedom encompasses autonomy, as opposed to being an altogether different capacity. Well, here the way we read the evolution of Kant’s theory of freedom becomes relevant. And what I am going to say will also apply to a similar criticism from Guyer further on. On my reading, which is inspired by Allison (Reference Allison1990), while up to the first edition of the first Critique Kant thought that practical freedom was sufficient for the sake of morality, the Groundwork (1784) marks a great novelty, that is, the introduction of autonomy as a new capacity, altogether different from practical freedom, and considered as necessary and sufficient for the sake of morality. Morality no longer rests on the generic capacity to set ends in a way that is not fully determined by sensuous motives, but on a capacity to set ends in complete independence from them. And this obviously leads to the conclusion, which I should probably have spelt out better in the book, that practical freedom can not set moral ends. And this explains why Kant now says that we have a faculty that distinguishes us from animals and yet it does not make us persons (MS, 6: 434). That faculty is precisely practical freedom, now redefined as able to set any ends freely, but never in complete independence of sensuous impulses, that is, as unable to set moral ends.
But let us go back to Sangiovanni’s central point. Why bow to a sheer capacity and not focus on people’s actual level of morality? Does that not beg the whole question of explaining what makes us worthy of human rights protection? My answer is articulated in three parts. To begin with, when we deal with human rights we are trying to make sense of rights that are supposed to protect even the worst of us. Unless Sangiovanni wants to say that people like Saddam Hussein, Gaddafi, Mobutu did not deserve the protection of human rights, it seems we had better focus on a capacity as opposed to actual behaviour.
Secondly, Kant’s story – at least on my reading – is that we have a capacity that sets us above the sensible world. Now if we all have this capacity, and this fact sets us above the sensible world, I think we have moved a step in the direction of showing why people deserve the protection of human rights. I do not see myself as begging the question. Proving that much, of course, is still insufficient to show that the argument is sound. Should one in fact accept Kant’s linkage between a capacity and dignity, as an article of faith? This leads me to the final part of my answer.
Since empirical confirmation is ex hypothesi not to be sought, I can only refer to what I said in the book. On my reading, the strongest argument Kant has for the ‘reality of freedom’ is the immediate consciousness each of us has of the fact that we can do in all circumstances what morality requires. This is a classic, complex and still debated issue in Kant studies, turning on the interpretation of the ‘fact of reason’. I will therefore ask to be excused if I do not defend in detail my reading of this concept crucial for understanding the entire edifice of Kant’s moral thought in this limited space. But the gist is rather simple. As the example of the person threatened to offer false testimony against an innocent man shows, we know that even in the worst circumstances we could choose the moral path no matter what the consequences are. Acts of moral heroism exist, history is full of them. And the reason why we connect to them, wish we could be as courageous or the like, is that we identify with these heroes. We do not say that they were extra-terrestrial entities whose functioning we can ignore. Each of us in foro interno knows that he or she could have been that hero.
Sangiovanni’s next worry is that we may be autonomous, but not all to the same degree. As he puts it, ‘why should we believe that individuals’ capacities to act in accordance with the moral law are equal? Some individuals’ capacity to act morally are very high – on average and in normal conditions, they display great moral strength, resoluteness, resilience and courage – whereas others’ capacities are very low. So, if dignity resides in the (awesome) capacity to act morally, then why should not those with greater moral capacities have greater dignity, and therefore higher status?’ A conclusion that would obviously be in stark contrast with what Buchanan (Reference Buchanan2010) called the egalitarianism of human rights. Moreover, Sangiovanni warns against a possible easy way out here. If I say that the difference individuals display in moral behaviour is insufficient ground to deny that we all have the same capacity, but simply that we exercise it more or less often, or more or less well, then ‘we seem to have departed very far from a common-sense view of what a capacity is, and to have won the argument at the price of a highly controversial metaphysical view’ (p. 254). Let us see.
To begin with, it is important to realize that on my reading autonomy is ‘simply’ the capacity to overcome all sensuous impulses if that is necessary to act morally. I do not want habits, ‘second natures’, developed in an Aristotelian way through exercise, to be part of autonomy. It is simply not what I mean. For being entitled to the status that confers human rights, on my theory, you simply need to pass a minimal threshold. This threshold is precisely that you can overcome all your inclinations. Now, is it ‘self-evidently false as a matter of fact’ that all human beings have that ‘minimal’ capacity? I do not think so, and refer back to what I said above about the immediate consciousness that each of us has of possessing this capacity. I do not claim that this immediate consciousness is a rock-solid basis for proving that the opposite view is ‘self-evidently false’. After all, short of the possibility of appealing to evidence, and given that we are not talking about an a priori truth, could there be such an argument? I am happy with the result that, construed in the minimal way I just described, autonomy is not at all far away from personal experience and common sense. We find it any time we reflect on what we could do if forced into the shoes of the poor man threatened to give false testimony.
The point can perhaps be further illustrated by a thought experiment. Let us imagine that we have to judge a woman who has stolen property but claims to be affected by kleptomania. It seems that we only have two options: either we believe the story, consider her insane, hence not responsible for the crime, or we do not and, independently of all the possible mitigating circumstances, we find her guilty. We do not say that her autonomy (and therefore her accountability) was depleted. We do not punish her because she did not develop past moral fortitude. We either say that she was free/autonomous, hence she is responsible, or we decide that she was sick, hence not responsible because not an agent at all. And, if we decide that she is guilty, we do not make the gravity of punishment dependent on how autonomous she was but on how pressing were the sensible motives that suggested that kind of behaviour (she was hungry, needed the stolen property for the well-being of her kids and the like). It seems to me that this is the way we usually reason about people and their responsibility, and I propose that a minimal, non-scalar conception of autonomy is an integral part of our judgements. If that is the case, perhaps my account (and Kant’s) is not so removed from common sense or helplessly metaphysical.
Finally, is this minimal conception of human autonomy compatible with my making room for animal autonomy, that is for different degrees of autonomy among species? I think it is. In the book I attributed a certain degree of moral agency to non-human primates. However, I never suggested that they reached the ability to silence all their natural impulses, that is that they reach the minimal threshold of autonomy. Sure, I talked about ‘animal autonomy’ but that was a rhetorical way of expressing the point that animals approximate, without reaching, the capacity of ‘pure’ agency. In fact, I attributed to them the capacity of self-sacrifice (how could one deny that to dogs, for example?), but this is different. Animals seem to sacrifice themselves because the empathic impulse to save their master or their offspring can be stronger than the instinct of survival. In contrast, humans are supposed to make (or to be able to make) a conscious and detached deliberation about their course of action. In animal morality it is just this conscious, free, reflexive endorsement of such an act that seems to be missing.
The fourth and last criticism of Sangiovanni has to do with the very possibility of having a truly Kantian theory of human rights. In particular, Sangiovanni cites and endorses the six features of human rights I list at the beginning of my book, but invites me to see that five of them are clearly incompatible with Kant’s injunction against unilateral imposition, that is, his (in)famous idea that once a state is formed, one may not question its legitimacy either from inside or outside. It would follow that my idea that there could be a Kantian theory of human rights (something like the very book I wrote) looks very dubious, because human rights are precisely supposed to enable internal and external criticism.
The irony here is that I agree with Sangiovanni. The five features of human rights in question are incompatible with Kant’s prohibition to resist the supreme authority. What I would like to emphasize, in a way that I must not have succeeded in doing in the book, is that I subscribe to Buchanan’s distinction between human rights (HR) and international human rights (IHR). HR are those rights that we may have all sort of philosophical or non-philosophical reasons to consider as human rights, independently of whether they are already part of international law. IHR are those rights already present in official legal documents. The two sets do not coincide and this is good because, as Buchanan emphasizes, this allows a dialectic between the theory of HR and the theory and practice of IHR. If we come up with a list of HR too distant from the practice of IHR, we may want to rethink the philosophical reasons that led us to that list. Conversely, if IHR are in sharp contrast with what we have strong reasons to believe should count as HR, we may want to change international law. Now, in the passage that Sangionvanni cites I was speaking about human rights in general (both HR and IHR). But I thought it was clear that most of what I say in the book concerns HR, not IHR. I am – in a very traditional manner – offering a philosophical thesis that tells us why certain things should count as rights we have merely by virtue of our humanity, leaving aside the question of the fit between this theory and IHR. In this sense I will concede to Sangiovanni that there cannot be a truly Kantian theory of IHR because too many features of them conflict with Kant’s prohibition against resistance. But I insist that there could (and should) be a theory of HR, like the one I tried to offer, from which, among other things, we may get guidance on how to change IHR, if the misfit between the two domains is too large.
In addition, it is important to remember that Kant qualifies his unconditional duty to obey. In the relevant passage in the Metaphysics of Morals he writes: ‘Obey the authority who has power over you (in whatever does not conflict with inner morality)’ (MS, 6: 371). One reasonable reading of the parenthetical clause is that that if the despot demands of me that I do something that conflicts with morality – say give false testimony against an innocent man – then the non-conditionality of obedience starts shaking. Whether this is meant as a full-fledged authorization to rebel or resist is open to debate. What interests me is to see that even the rigid Kant of ‘obey no matter what’ leaves some room for having an external, moral perspective from which legitimate authority may and should be evaluated, criticized and possibly resisted. This suggests that even the just conceded incompatibility between Kant’s theory and IHR may need qualifications.
2. Williams
Howard Williams esteems my work probably beyond its merits, and I am not going to fight hard to change his mind. Since he does not have much to quarrel with in my interpretation, and I have little to quarrel with in the general framework in which he nicely inserts it, I hope he will excuse me if my reply is rather short. Williams, however, does have three points of disagreement with my account. The first relates to my reading of the status of Kant’s thesis that there is a guarantee of perpetual peace. The second has to do with my ‘isolated view’ of the moral politician. The third concerns my failure to insert in fine enough detail Towards Perpetual Peace within Kant’s general theory of law.
Regarding the first point, Williams thinks that I am making the ambitions of the guarantee thesis too high. He agrees with me that Kant wants to get his hands dirty with some empirical considerations, but thinks that anything Kant says to convince us that perpetual peace is possible, or worse, necessary, is to be taken as nothing more than a counterargument against the sceptic. As such it is not meant to have any more empirical solidity than the argument presented by the opponent. In a similar vein, Williams argues that the (sole? main?) point of the guarantee thesis is to ‘clear space for full deployment of practical reason in politics’ (p. 267). Williams explains this by equating the role of the guarantee thesis with that of the highest good: ‘It [the guarantee thesis] is a future related way of conceiving our political actions that gives them moral coherence. Just as “the highest good is the necessary highest end of a morally determined will and is a true object of that will” ([KpV] 5: 115), so perpetual peace is a highest end for a morally motivated political leader’ (p. 267).
I have very little to argue against the ‘practical’ reading which Williams defends and I think that it is perfectly compatible with my more theoretical one. I simply think that Kant wants to do something more than point to a necessary object we need to conceive of in order to make our practical (moral or political) agency coherent. He provides some essentially theoretical, broadly empirical reasons to believe that progress in human affairs is more likely than any other outcome. And he does so by referring to natural mechanisms that, combined with background assumptions that Kant does not make explicit, but evidently assumes, are supposed to back up this view. If the guarantee thesis were merely to play the role Kant reserves for the highest good in our moral life, then it would be hard to explain why he talks in Towards Perpetual Peace of a visible (hence rather empirical) ‘mechanical course’ of nature that exhibits ‘purposiveness … letting concord arise by means of the discord between human beings even against their will’ (ZeF, 8: 360). And why he makes an effort to detail how concord is to spring from discord, by mentioning unsocial sociability, the unintentional benefits of linguistic and religious differences, the equally unintended peaceful significance of greed (trade as an unconscious force for peace) and so on. Williams says that the guarantee thesis is ‘is neither wholly an empirical prediction nor merely a source of reassurance for the right-minded’ (p. 267). I agree. I would simply add that if it is to be more than a reassurance for the righteous, then some quasi-empirical claims have to be given (and in fact are). One of the main goals I had in interpreting this controversial part of Kant’s political thought was precisely to spell out as clearly as possible these quasi-empirical claims.
Regarding the second point, Williams agrees with me that the life of the moral politician is more difficult than that, already not easy, of the moral person, mainly because the former has a responsibility to foresee the consequences of his or her decision, while the moral person is relatively free from that responsibility. This Weberian contrast between an ethics of responsibility and an ethics of principles was mainly meant to underscore that in order to predict the consequences of moral political actions a good deal of (good) empirical science and, most importantly, of phronesis is necessary. This led me to conclude in a rather pessimistic tone (at times even my optimism wavers) that it is very rare to encounter moral politicians. But Williams reassures me that there is no need to construe the moral politician as an isolated hero expected to be near-omniscient. A moral politician, quite simply, may ask for help from counsellors and experts. This is part of the reason why Kant cared about the freedom of the pen. Experts have to be able to enrich the public debates, even if the final decision remains in the hands of politicians. I think Williams is right. Even when I try to be a pessimist, I do not succeed.
Regarding the last point, I am also inclined to concede the point, if this means that I may have not sufficiently emphasized the parts of the Metaphysics of Morals relevant for the peace project, or the continuity between Towards Perpetual Peace and the later work. But I reject the criticism if this is meant to charge me with having paid scant attention to the role that the law is supposed to play in Kant’s project. While I resist its full reduction to an institutional recipe, because I think that the morality of common people and leaders remains an essential part of the story, nonetheless I do see Kant as arguing for a priority of the law over morality in the progress towards our cherished end. As he puts it in a passage in Towards Perpetual Peace, the importance of which is difficult to exaggerate: ‘it is not the case that a good state constitution is to be expected from inner morality; on the contrary, the good moral education of a people is to be expected from a good state constitution’ (ZeF, 8: 366). My interpretation not only of the 1795 essay, but also of Idea and of On the Common Saying, remains faithful to this principle (Caranti Reference Caranti2017: 121–2, 126, 138).
3. Guyer
Paul Guyer honours me with a careful reading of my work, aptly combined with clear statements on the points in which our views differ. I will try to do justice to each of them.
The first minor disagreement has to do with the permissions Kant infers from the innate right to external freedom. Guyer thinks that these permissions are three, not four as I suggest. The authorizations to be one’s own master, to be beyond reproach and to communicate one’s thought freely would be nothing but articulations of the right to formal equality. So neither do they add anything to formal equality nor does formal equality add anything to them. Now, from a textual point of view, it is true that Kant introduces the three authorizations through a mithin (hence), which may be taken to suggest that they ‘follow’ what has just come before, that is, the right to equality. And I am ready to concede that equality may stand to these authorizations as a kind stands to its species. What I am not convinced of is that innate equality can be ‘exhausted’ by the three authorizations. The notion of (formal) innate equality seems to entail more. It entails, among other things, that no discrimination of sex, race, religion is allowed, that we are equal before the law, that no noble rank is recognized and so on. This seems to go way beyond the three specific authorizations that Kant cites. Indirect evidence of this comes from the 1784 Naturrecht Feyerabend where, as Guyer knows well (Guyer Reference Guyer2016), the authorizations Kant infers are neither three nor four, but five (Kant adds now the right to acquire property). This suggests that the list of rights or authorizations one may derive from the innate right to freedom (and equality) is open-ended, something hardly compatible with the notion of equality being exhausted with three (or four, or perhaps any) of them.
More important is our disagreement about the role that the notion of ‘humanity’ plays as basis for the innate right to (external) freedom. Guyer alerts me to the fact that Kant’s official definition of humanity (‘the capacity to set oneself an end – any end whatsoever’) does not seem to square with the role that I cut out for it in the argument. Having the capacity to set any end is what enables us to set moral as well as immoral ends. Hence in humanity, no intrinsic worth is to be found.
This criticism is similar to one of Sangiovanni’s we have already discussed, and I take the opportunity to expand my reply. To begin with, I recognize that Kant’s term for our capacity to set ends in complete independence from sensuous inclinations, that is, for autonomy, is personality, not humanity. Even in the passage from the doctrine of virtue I cite, it is the human being considered as a person that has dignity and is set above any price. It is also true, however, as I say in a footnote, that Kant is not always consistent in the use of these two terms and sometimes does use ‘humanity’ when he clearly means ‘personality’. In the Groundwork, for example, Kant is explicit that what confers worth on humans is our personality understood as ‘freedom and independence from the mechanism of the whole nature’ (G, 4: 87). But only a few lines later it is ‘humanity’ that we must consider as holy in us: ‘A human being is indeed unholy enough but the humanity in his person must be holy to him’, and of course, yet another few lines down, it is humanity (not personality) that is ‘an end in itself’. This terminological inaccuracy is understandable because not only human beings, but all rational beings are persons: ‘rational beings are called persons because their nature already marks them out as an end in itself’ (4: 428). Hence it is reasonable to assume that sometimes Kant uses the term ‘humanity’ as a short name for that subset of persons that are human. My suggestion is that it is precisely in this sense that he uses ‘humanity’ when he refers to the basis for the innate right to external freedom. Moreover, as I said in response to Sangiovanni, it is true that humanity in a sense encompasses personality because, as Guyer correctly notices, it is the capacity to set any ends, which of course includes moral ones. And yet, at least from 1784 on, autonomy is sharply distinguished from the freedom we have to set ends that are sensuously determined, and there is no doubt that Kant sees the former capacity, not the latter, as the seat of our worth.
To back up my textual reasons, in the book I offer few systematic reasons. If the basis in question were humanity, understood as what makes us superior to animals and yet having ‘ordinary value’ or merely ‘extrinsic value’ (MS, 6: 434), I do not see how Kant could claim that humans acquire any right whatsoever because of (kraft) it. Guyer thinks that the right in question arises not from humanity itself, but from the injunction to treat humanity never merely as a means, that is, from the categorical imperative. However, this only seems to push the question one level up. Why should we use humans never merely as means? I imagine Guyer thinks that Kant is allowed to borrow here the results of his foundation of morality in Groundwork and the second Critique. Well, possibly so. After all, I am a ‘dependentist’ like him in the reading of the relation between morality and law in Kant. On my interpretation, however, the mature Kant grounds morality itself on the fact that we are autonomous. Autonomy is the necessary and sufficient condition of morality and the ultimate basis on which the whole edifice of Kant’s juridical thought rests. So, even if Guyer is correct, this does not change the basic fact that it is autonomy that does the normative work.
This leads me to Guyer’s philosophical worry. The problem he sees is that the fundamental principle of morality cannot be respect for the capacity to be moral because this is stained with circularity. If I am supposed to respect those entities that are capable of being moral, I need an independent criterion that tells me what is moral, and then apply it to the appropriate individuals. My short answer here is that, as Guyer knows well, for Kant the content of the moral law, what is moral and what is not, is clear to any common human understanding (G, 4: 454). We have an immediate access to that criterion. Equally evident for us is the fact that we are persons, in the sense of being able to follow what we intuitively know is obligatory (KpV, 5: 87). Now, one can disagree with Kant that this content and our capacity to act on it are so evident. But I think this suffices to show that Kant’s argument, on my reconstruction, can be charged with intuitionism, not circularity. And I confess that, given Kant’s emphasis on how we discover the reality of freedom through the immediate consciousness of duty (the fact of reason, as I read it), I feel quite confident in attributing to Kant this bit of intuitionism.
What about Guyer’s interesting alternative to my foundation? He suggests that Kant derives the fundamental principle of morality from rationality itself, in particular from the principle of non-contradiction. As evidence, he cites some very early annotations on the theme of morality, handwritten on the margins of Kant’s own copy of On the Feeling of the Beautiful and Sublime (1764), according to which there is something not just ‘ugly’, but also ‘contradictory’ and ‘absurd’ in treating someone who has ‘a soul’ as if he or she did not have one. Kant seems to be suggesting that all we need to ground morality is to appeal to the basic principle of rationality, that is, to the principle of non-contradiction.
In proposing his alternative foundation, Guyer recognizes that it presupposes that it is evident to everyone that each human being has a ‘soul’ or a ‘free will’. But this is exactly the point that I am defending. My foundation says that morality rests on the recognition that we are all autonomous, that this entails awe and respect, and – here comes Guyer’s point – that we cannot, on pain of contradiction, treat people as if they did not have that capacity. Now Guyer may protest: No, the fundamental assumption is that we have a free will – the freedom to set any end – not that we are autonomous (the freedom to set moral ends). But now we are back to a point already made. I would reply that it is probably true that by ‘having a soul’ Kant meant in 1764 a ‘freedom of the will’ in the sense of what I call ‘practical freedom’. But the great progress Kant makes from the pre-Critical to his mature view of morality is precisely that he stops thinking, as he still did in 1781, that practical freedom, free will in the generic sense, is sufficient for the sake of morality. As I have already mentioned, from the Groundwork on (1784, so twenty years after that handwritten remark) Kant introduces the crucial idea that it is autonomy, not mere ‘practical freedom’, that is both a necessary and a sufficient condition for morality.
More importantly, leaving the evolution of Kant’s thought aside, if the ‘soul’ we are all supposed to see clearly were the sheer capacity to set ends in an intelligent, rational way, then there would be no contradiction in recognizing this ‘soul’ and treating people who have it merely as means. These people would be, as the mature Kant would say (MS, 6: 434), commodities, extremely precious ones, and yet still commodities. Hence, the most I feel confident to concede to Guyer here is that the principle of contradiction enters the picture once what really does the foundational work – autonomy – has already been secured.
Let us now move – I am afraid briefly – to Guyer’s two other criticisms of Part 1 of my book. The first is that I would need to do more than cite a few lines from sacred texts to bring home the point that adherents to any of the major religious traditions would – properly guided through Socratic questioning – recognize autonomy as a condition on the very possibility of having a significant religious life. Yes, as I already did with Sangiovanni, I plead guilty here. Let me just clarify that, contrary to what Guyer thinks, I was not expecting believers to show ‘habituation to establish a degree of detachment from [their] deepest comprehensive convictions in favour of a negotiation with other individuals who adhere to reasonable comprehensive doctrines radically different from the one[s they] uphold’ (Caranti Reference Caranti2017: 92). This is the expectation that proponents of global reason as the ultimate normative compass for human rights have to form. I argued that this expectation is unrealistic and expressed my preference for reflective equilibrium.
But Guyer’s main worry is that I cannot get a specific list of human rights from my refined Kantian account. He argues convincingly that I drew too little from Kantian resources in the move from the higher and fundamental innate rights to freedom and equality (possibly with the few ‘authorizations’ Kant infers from it) and left too much to each society’s freedom to come up with a specific list of basic rights. Guyer shows me how one could infer most, if not all, the rights that Kant lists in the Doctrine of Right from the innate right to external freedom. I have nothing to argue against Guyer’s expansion from this basic tenet. In fact, not only do I take it as fully compatible with my argument, but I also think of it as an important and welcome strengthening. I simply wish to clarify the following point. In the book I was mainly trying to show that Kant provides a powerful and non-parochial way of bringing home the fundamental principle that lies behind the human rights culture, that is, the idea that humans have rights by virtue of their humanity. This is why I did not focus enough on how to infer extra rights from the right to freedom and equality. But if there is a Kantian way to get from these two overarching and very generic rights of external freedom and equality closer to the specific and numerous human rights we are accustomed to, this is great news for my approach.
Guyer then moves to discuss the third part of my book, and he focuses on my defence of the guarantee thesis. The first concern, it seems, is that I do not sufficiently define the thesis I want to defend, because I oscillate between the idea that there is a guarantee of peace and the weaker thesis that peace is merely likely. Well, I do oscillate because I think, in a way that I should have explained better, that the two theses, properly qualified, are compatible, and I want to assert both. My position is that by pointing to some natural mechanisms Kant shows that the world has a propensity towards peace. A propensity, in the Popperian sense, is an objective tendency which ‘rigs’ a natural system towards a certain outcome. Now saying that a system has a propensity is fully compatible with saying that in certain specific instances the system will evolve (temporarily) in a direction opposite to that indicated by that propensity. So, given any time T1 in world history, we would have no certainty that the world at T2 will not progress, rather than regress. Nonetheless, we still have the certainty (guarantee) that the system will reach the condition of stable peace at Tx, if x is a point in the future remote enough to give the system the time necessary to adjust any possible involutions running against its natural propensity. Obviously, Kant thinks that we cannot know when this will happen. Now, when I say that peace is merely likely I am saying it is likely that the world will evolve in the near future towards a higher level of security. When I say that peace is guaranteed I am talking about a remote point in the future (unknown to us) when the process (whose mechanisms we do know) will be completed. In short, peace is merely likely if we are talking about any specific time before the system stabilizes itself. Peace is guaranteed if we enlarge our view toward an indefinite future point.
Perhaps the thesis can be further clarified (again in a Popperian key) by focusing on what would falsify it. If tomorrow a foolish democratic leader waged war against another democracy after 200 years of peace, this does not falsify my thesis. The thesis would be refuted if the world today were not less bellicose than it was 200 years ago, given reasonable standards of measurement. Now since the opinions of social scientists differ on this point, I confess I cannot offer non-controversial evidence. Maybe wars in absolute terms, that is in the world at large, are neither fewer nor less cruel than what they used to be. It is undeniable though (and evidence is abundant and clear) that within the part of the world where Kant’s three definitive articles have been implemented most, war is less likely than it was in the past, in fact, than it has ever been. I do not know if this is enough to ground the guarantee thesis. I hope it is enough to discourage the thought that it is just a simple-minded rosy prediction.
What I just said should also help in responding to Guyer’s specific point about radical evil. I do not need to deny that humans, being capable of subverting the path of morality even if it serves their best interests, could in specific circumstances act in a direction opposed to the one predicted by the thesis. I do not need to deny the possibility cited above of foolish leaders. For my thesis to be confirmed, these bits of agency that run both against morality and against the long-term best interests of agents have to become more and more rare. Now, I think Guyer will concede that people living in liberal democracies, at least those which are part of large international organizations such as the EU, and which enjoy a high degree of interdependence and permeability with neighbouring countries, are less exposed to wars waged ‘as a kind of amusement’ (ZeF, 8: 350) or as ‘whims’ than the people living in the nation states in Kant’s time. Sure, there are still today dictators of despotic states (Putin, for example) and the probability of having foolish democratic leaders is not zero (Trump is in power). But after the longest period of peace in European history, I would resist the idea that humanity is not less exposed to wars today than at the time when Kant wrote his ‘philosophical sketch’, at least that part of humanity that has, so to speak, ‘learned their lesson’.
I do not think that this comes even close to settling the question about whether we are in fact approximating perpetual peace. My goal in this part of the book was far more modest. I wanted to take Kant’s arguments seriously, try not to neutralize their theoretical potential, and show that they are not metaphysical and simple-mindedly optimistic, because they are meant to spot a tendency in an objective system, admittedly a rather large system: the world as it is and as it will be.
Guyer finds the second part of my book the most convincing. I can therefore be brief in my response. First, I would like to clarify that I do not attribute to democratic peace scholars the view that majority rule is all that matters for the sake of peace – no serious theorist in that tradition thinks so, and certainly Doyle and Russett do not. In fact, what is now known as ‘democratic’ peace was originally called ‘liberal’ peace because crucial for Doyle, besides wide suffrage, were liberal institutions and values (Caranti Reference Caranti2017: 166–8). Moreover, Doyle’s and Russett’s ways of replying to generations of critics of the liberal peace thesis makes clear that the idea was (and has always been) that only when the three Kantian factors are combined is there ‘certainty’ of peace (191–4). My deepest criticisms of the democratic peace scholars are rather (a) that even if enriched by liberal elements, democracies should not be equated with republics – the crucial missing element being a republican ethos – (b) that they have misunderstood the nature and role of the international Federation, and (c) that they have reduced the significance of cosmopolitan right to the classical liberal idea that international trade promotes peace.
This leads me to Guyer’s final criticism. He thinks that all there is in the third definitive article is ‘an anti-colonialist restriction of any presumptive right of persons of one nation to claim and occupy other already occupied regions of the earth for their own benefit’ (p. 287). Sure, there is that element, as Kant’s harsh and merciless criticism of colonialism in the very discussion of the third definitive article proves beyond doubt. But is there not also a positive side? If all the third article is supposed to accomplish is to counter colonialism, why not say that frontiers should be impermeable? Kant is explicit that the way in which the right to visit promotes peace is first by allowing the creation of economic interdependence (‘nature … unites nations … by means of their mutual self-interest. For the spirit of commerce sooner or later takes hold of every people’, ZeF, 8: 368), and secondly because through economic and cultural interdependence people enter into a ‘universal community’ (8: 360), which Kant thought was so developed already in his times that ‘a violation of rights [we would say today human rights] in one part of the world is felt everywhere’ (8: 360). In short, cosmopolitan right counters colonialism and exploitation, but also enables commerce and the creation of a global moral conscience, which are the peace-promoting factors that the third definitive article adds to those of the first two. This positive side is not only compatible with anti-colonialism, but also complements it by showing how humanity is bound to progress towards a universal consensus (I call it a ‘global moral conscience’) on fundamental normative standards that we would call today human rights. And of course human rights are primarily meant to rule out things such as colonialism and aggression. Perhaps human rights, peace and progress in Kant’s political thought have to do with one another even more than I suspected when I wrote the book.