Many philosophical accounts of human rights are currently presented as Kantian or strongly based on Kant’s assertion of human beings’ inherent worth. Many more are fashionably announced as anti-Kantian, or at least post-Kantian, for their firm refusal of adopting tentatively universal yet ultimately parochial standards of normativity. Despite this vast array of positive and negative references, it is surprising that no commentator has ever attempted to reconstruct what would be an authentically Kantian theory of human rights. To be sure, we have at our disposal sophisticated interpretations of Kant’s theory of justice (Recht), of rights in general, of cosmopolitan law, of humanity  and of many other concepts that are clearly relevant for any theory of human rights. Yet, with the exception of few remarks scattered here and there, and a recent attempt by Gunnar Beck, no one has ever read Kant’s moral and political thought to find in it what we would call today a philosophical theory of human rights.
There are at least two motives for this gap in the literature. To begin with, Kant seldom uses the expression “human rights”. Moreveor, when he comes closest to it, in his account of humans’ innate right to freedom, it is not clear whether such right can be safely considered as equivalent to a “human right”. More precisely, it is not clear whether Kant’s idea that humans have an innate right to freedom “by virtue of their humanity” is sufficiently similar to the central intuition behind the culture of human rights, i. e. that certain rights are bestowed on humans independently of their belonging to political groups, cultural traditions, religious loyalties, races, but merely by virtue of their membership in the human species. In fact, ‘humanity’ is a technical notion in Kant’s system and is rightly viewed as potentially divisive once the metaphysical apparatus on which it rests (our alleged membership to a noumenal world) is brought to light. Call this the terminological motive.
On a more profound level, Kant’s ethics is perceived as diametrically opposed to the effort, recommended by virtually any expert of human rights, to find some inter-cultural, non-parochial, possibly a-metaphysical, basis on which representatives of profoundly different cultures and traditions can give their assent. There are at least two features of Kant’s account of basic rights that seem to run counter this pluralism-sensitive approach. To begin with, Kant’s notion of humans’ inherent worth depends on a controversial belief on our peculiar characteristic of being transcendentally free. For Kant, humans are the only entities on earth capable of rational agency, namely capable of taking distance from sensible impulses to choose reflectively on their course of action. Practical freedom, and even more autonomy distinguish humans from animals, bestow on them inherent dignity, dictate proper respect for any human being. Leaving aside the difficulties with Kant’s proof of our freedom to which in a sense his entire moral thought is devoted, the problem here is that autonomy is not given the same importance everywhere in the world. Especially non-western cultures, even if convinced that humans are free (something that can be conceded merely for the sake of the argument) would likely remain unimpressed by this feature. Filial piety, honour, obedience to the established authority, loyalty to a religious belief are commonly viewed as elements of human worth as important as autonomy, to say the least. Moreover, Kant’s implicit assumption according to which the individual is the sole legitimate subject of ethics is questioned by alternative, non-western approaches that are told to give priority to the group, as exemplified by the so called East Asian challenge to human rights.
As a result, there are very few thinkers who venture to use Kant’s practical philosophy as an explicit intellectual basis for establishing a philosophical foundation of human rights. In this regard, Alan Gewirth is probably the last, controversial exception. His quasi-transcendental argument turning on the notion of purposive agent elicited robust and convincing criticisms that probably discouraged any other attempt along similar lines [9, p.1—24]. Kant is thus considered as useful only for those who already believe in human rights because of their liberal commitment. At best, Kant’s ethics seems to spell out specific reasons why liberals take human rights seriously. Given a pre-decided commitment in favour of the individual and of certain liberties, Kant is perceived as capable of reinforcing this commitment by cashing out what is there in the individual that commands respect and dignity (two key notions in the human rights discourse). Beyond this contribution, however, Kant seems to have little to offer to thinkers who work toward a broadening of the world consensus on human rights. Actually, in this regard, some feel that Kant’s emphasis on the individual and his liberties is not just unhelpful, but turns out to be an authentic obstacle. If this impression is right, the above mentioned gap in the literature comes as no surprise. Call this the philosophical motive.
Notwithstanding the terminological and philosophical motive, and their strong plausibility, the absence of a truly Kantian theory of human rights in the literature remains problematic, especially because we are deprived of a powerful basis from which popular approaches to human rights can be assessed and criticised. The fast and ready disqualification of Kant as a source of inspiration for a widely shared conception of human rights by virtue of his alleged partiality toward a western understanding rests on a superficial and selective reading of Kant’s texts that needs to be corrected. On close consideration, Kant interestingly manages to combine rigidity on certain universal standards (in particular freedom and equality) with a surprising flexibility concerning the implementation of these standards vis à vis the moral pluralities of our world. Kant would condemn the tendency to water down the universals of justice, and consider the worldwide violations of the right to freedom and formal equality as inexcusable for cultural or religious reasons (a position that virtually no theoretician today would dare embracing). At the same time, however, he would also condemn the contemporary tendency to impose violently these minima moralia — a prudence that many think incompatible with an commitment to universal standards. Kant does not deflect from considering maximum liberty for each individual and perfect formal equality as rights inherently human, but allows great flexibility to each society in their progress toward the implementation of these standards. This combination of rigidity about the principles and flexibility about the times and manners of implementation deserves more attention than is usually conceded. Kantian motives could be of great help for those who attempt to bring philosophical clarity in the field of human rights.
In this context, however, we will make only the first step toward this ambitious goal, by identifying the material from which a Kantian foundation of human rights could be developed. Once this step has been made, it will be easier to cash out a theory capable of competing with other approaches in a foundational menu that nowadays sees an anachronistic primacy of the utilitaristic school. The paper will thus analyze, in the first part, Kant’s idea that we have an innate right to freedom by virtue of our humanity. A specific concern will be with explaining Kant’s analytical inference from the right to external freedom to the right to (formal) equality. In the second, the focus will be on our “capacity to set ourselves ends” that Kant considers definitional of our “humanity”. In particular, we will deal with the question whether this “capacity” is best understood as a form of practical freedom or rather of autonomy, two kinds of freedom quite different in Kant’s ethical thought. Thirdly, we will argue that Kant identifies our ability to be moral agents, i. e. our autonomy, as the ground on which the innate rights to external freedom and formal equality rest. In the attempt to spell out precisely how autonomy grounds these rights, we will conclude by proposing an argument that mediates between two opposed hermeneutical orientations.
I. ‘There is only one innate right’
Probably the text where Kant comes closest to expressing a theory of human rights, as we would understand it today, is to be found in the Metaphysics of Morals, in particular in the section devoted to the General Division of Rights. There Kant introduces two ways in which rights can be divided. To begin with, he makes reference to the doctrinal subdivision between natural rights, based on a priori principles, and positive rights, that are created by a legislator. Secondly, with reference to rights as “(moral) capacities [Vermögen]”, understood as entitlements to bind to certain obligations, he distinguishes between innate and acquired rights. The former originate from — obviously inborn — moral capacities “independently of any act that would establish a right”, i. e. prior to the commonwealth, while the latter presuppose such an act.
Quite abruptly, Kant then claims that “There is Only One Innate Right” [AA, VI, S. 237], [11, p. 30], i. e. freedom, understood as “independence from being constrained by another’s choice” [AA, VI, S. 237], [11, p. 30]. Keeping in mind the above distinctions, we can infer that our innate right to negative (or external) freedom has three main features: (a) it is a natural right that, as such, rests on a priori principles; (b) we are entitled to it before the establishment of a commonwealth (even if a state may be required to enforce it) and (c) it stems from a moral capacity. Kant does not make it explicit that the right to freedom is a natural one, but the fact that acquired rights presuppose the will of the legislator, while the right to freedom, as innate, precedes the commonwealth, entails by elimination that this is what he means. Features (b) and (c) seem to be logically linked in such a way that (c) grounds (b). It is precisely because this right “stems from a moral capacity” that humans possess it even before the establishment of the commonwealth. But, ultimately, even (a) rests on (c). Otherwise, one could hardly understand why humans possess this right before the civil state. It follows that in order to understand our right to external freedom it is crucial to explain the moral capacity on which it rests. This is confirmed by Kant’s quick reference to the reason why we have a right to external freedom. He claims that each human has it “by virtue of his humanity”. In the Metaphysics of Morals and elsewhere, Kant construes humanity as the capacity “by which he [the human being] alone is capable of setting himself ends” [AA, VI, S. 387]. More explicitly, Kant says that “the capacity to set oneself an end — any end whatsoever — is what characterizes humanity (as distinguished from animality)” [AA, VI, S. 392]. Such a capacity is thus the ground on which our right to freedom rests.
Before we move to the analysis of this right, however, we must notice that Kant interestingly infers from it, in a straightforwardly analytical way, our right to (formal) equality. He claims: “This principle of innate freedom already involves the following authorizations, which are not really distinct from it (as if they were members of the division of some higher concept of right): innate equality, that is, independence from being bound by others to more than one can in turn bind them” [AA, VI, S. 237].
Kant’s reasoning here is quite simple. Since each individual has a right to a sphere of freedom whose extension is limited only by the condition that such freedom be compatible with that of all others, it follows that these spheres must be all equal. Otherwise we would be adding some extra condition besides that of compatibility. Before the autonomous decision of the members of the society to limit their freedom through, say, obligations, promises (legal or moral), contracts and so forth, the sphere of freedom I enjoy should be of a size identical with that of all others. And again, this seems to be a right that precedes the establishment of the commonwealth. A state is merely supposed to enforce and secure a right to equality whose ground precedes any societal compact. Our “sole” innate right has actually given birth to another crucial entitlement: perfect equality before the law, or, formal equality.
Needless to say, any cogency that our right to equality may have depends on the foundation of our right to freedom. As we saw, this turns on the idea that humans possess a capacity to set ends for themselves. We have an innate right to freedom (and to equality) because we are capable of setting ourselves ends. What does that mean? How is it that the alleged possession of a mere ability to set ends entitles me to anything? If I am capable of killing in cold blood this certainly doesn’t entitle me to any right, moral claim or the like. Why, then, does Kant believe that the capacity to set ends for themselves entitle humans to a right to external freedom? The answer to this question presupposes clarity on humans’ capacity to set ends for themselves. Since the expression, as it stands, evokes at least two kinds of freedom, practical freedom and autonomy, both present in Kant’s system, it is incumbent on us to introduce them in their bare essentials in order to decide then which of the two Kant means in this context.
II. Two Kinds of Freedom
Notoriously, for Kant human actions are not fully determined by the sensuous inclinations that normally motivate us. Rather, given any inclinations, no matter how strong, it is always up to the individual to “endorse” them or to resist them. With the rather obvious exception of non-voluntary responses to stimuli (such as the familiar knee coming up on a hammer strike), Kant thinks that human behaviour as a whole is subject to free rational deliberation. This means that all voluntary actions stem from the individual’s free evaluation of a certain maxim, understood as a subjective rule of action. Thus humans scrutinize through their reason the opportunity (moral or prudential) of a certain maxim and are free to adopt it or reject it. Kant at times expresses this crucial point with reference to the notion of an arbitrium liberum, distinguished from the arbitrium brutum typical of animals [A533—534/B561—562]. Humans, on this theory, set for themselves the ends of their life without being driven or fully determined by desires and needs. At most, desires and needs suggest a certain path of action. A free decision on the part of the actor to endorse or reject such a path, thereby making it one’s own end, is, however, always necessary. This explains the force of the reflexive clause (“capable of setting himself ends”) that appears in the formula above. Humans set for themselves ends. These ends are in no way imposed on them by external forces or internal passions. Humans enjoy what Kant at times calls “freedom in the practical sense” [A 534/B 562].
The Kantian notion of practical freedom comes very close to what one would call rational agency. This can be captured by two essential characteristics: a) independence from pathological necessitation and b) capacity to act on the basis of imperatives (rules of action) in the pursuit of a give goal. Thus humans are rational agents for Kant because they select the rule that guide their behaviour (they do not act randomly) and they are free to select such a rule. To give an obvious example, if I am thirsty, and I see no reasons why I should resist or delay the satisfaction arising from the extinguishing of my thirst, I can freely select the maxim “any time an agent X is thirsty, she should drink.” As emphasized by Henry Allison, although apparently plain, this account of rational agency is already highly controversial in that it expresses an incompatibilist account of human freedom. In fact, for Kant the decision to drink is not fully determined by my inclination, nor is it determined by a more sophisticated pleasure calculus or the like. The decision to endorse the maxim in question is irreducible to one of the many natural causes we experience in the sensible world. The decision is a radically free act on the part of the subject. As Kant sometimes puts it, it is only on this condition that I can say that I perform any action, as opposed to “something in me led me to act” or “that particular objective state of affairs evolved in a certain direction”.
As an historical point, this is in a nutshell the theory of freedom Kant presents in the first critique and that circa 1781 he thought to be sufficient for the sake of morality. In the Groundwork (1785), however, we assist to the grand entrée of the concept of autonomy that makes Kant’s theory of freedom even more problematic for the contemporary philosopher’s sensitivity. Autonomy, as defined in the Groundwork and in the Critique of Practical Reason, entails more than independence of pathological necessitation, a feature underlying all kinds of rational behaviour and equivalent to the notion of arbitrium liberum. It entails more than the ability to take distance from our contingent inclinations in view of some more distant, yet still empirically motivated end (I resist my desire to smoke in view of the higher goal of avoiding cancer). For the mature Kant, an agent whose freedom is limited to this ability is free but irremediably heteronomous. The agent is free because her inclinations (no matter how strong) do not exhaust the causal story behind her actions (it always takes a free rational act of endorsement), yet the agent is heteronomous because inclinations are a necessary element of the motivational story behind the action. Their influence on the agent’s action can take different shapes. It ranges from a simple “giving in” to a given inclination, to a delay of its satisfaction, resistance for some greater good, dismissal for a sense of shame in having that inclination and so on. All these forms of responses share the feature of being reflective responses ultimately motivated by some empirical interest. And this renders the actor a free rational, yet heteronomous agent.
To the contrary, being autonomous for Kant entails the ability to act in complete independence from inclinations. Positively expressed, this means to be able to find a sufficiently strong motivation in a very special kind of non-empirical interest, that is, obviously the respect for the moral law, or ability to be determined by the authoritative force of morality. As Allison aptly puts it, a “will with the property of autonomy is one for which there are (or can be) reasons to act that are logically independent of the agent’s needs as a sensuous being” [1, p. 97]. An autonomous agent doesn’t merely give herself the rule of her action (this is spontaneity or practical freedom). She does so independently of any inclination. This is what Kant expresses with a slightly different language in one of the official definitions of autonomy in the Groundwork where this form of freedom is presented as “the property the will has of being a law to itself (independently of every property belonging to the objects of volition)” [AA, IV, S. 440]. While the human will is always a law to itself, because even heteronomous behaviour presupposes that one makes a certain sensuously motivated maxim one’s maxim, i. e. that one gives a law to oneself, it is only in autonomous agency that this law is self-imposed without sensuous influences. This is the force of the parenthetical clause in the definition of autonomy.
Thus for Kant humans display two kinds of freedom. Through practical freedom, they set ends for themselves and do it without being fully determined by inclinations, even if the motivational package that leads to action is sensibly influenced. Through autonomy, they set ends for themselves independently of any empirical motive. Since for Kant any form of agency, including moral agency, presupposes that the agent is moved by some interest, so that the question regarding the principium executionis be answered, this means that moral/autonomous agency will be motivated by the only form of pure (i. e. non-empirical) interest open to humans, i.e., respect of the moral law. On this reconstruction, autonomous behaviour is equivalent to moral behaviour, although, obviously, moral behaviour is only a subset of free behaviour (that includes heteronomous actions).
It is important to realize that both forms of freedom are for Kant peculiar of human beings. They can be predicated neither of animals who possess an arbitrium brutum nor of purely rational beings (putative entities like angels) who, lacking a sensibility, are not required to deal with any sensible inclination, let alone silence the demands of self-love if incompatible with those of morality. In fact, not only moral agency under the categorical imperative, but also practically spontaneous decisions taken under the guidance of the hypothetical imperative trespass for Kant the limits of what can be explained through natural laws. In fact, the sensible motives at the basis of an heteronomous action do not exhaust all there is to explain about the individual’s decision to pursue that action. The endorsement of those sensible motives would still be an irreducibly free act.
Clearly, this is a mere exposition of the forms of freedom Kant assigns to human beings and nothing has been said about the controversial arguments proposed to justify the ascription of these properties to our will. Kant’s arguments vary from the Groundwork to the second critique, and are sophisticated and intriguing, but there is no way we can deal with them here. Thus we are forced to bracket the question of whether these arguments are sound and postpone to another occasion the confrontation with the sceptic. For the moment, in order to complete Kant’s argument, we need to clarify which of the two forms of freedom defined above is alluded to in the definition of “humanity” as “the capacity to set oneself an end — any end whatsoever”.
III. Autonomy as the basis of the right to external freedom
Both practical freedom and autonomy appear as respectable candidates for the role of basis of our innate right to freedom because both are peculiar of human beings and satisfy Kant’s definition of humanity. The reference to “any end whatsoever” suggest that Kant means practical freedom, because even immoral ends seem to be contemplated. Other considerations, however, of systematic and textual nature, lead us much more convincingly in the opposite direction. Starting from the systematic reasons, the fact that a certain capacity is peculiar of a species hardly grounds any right of that species. As we said above, the human species is arguably the sole capable of killing in cold blood or even for amusement, but this hardly grounds a human right to kill for these purposes. Thus, the nerve of the argument cannot be the exclusive possession of an ability, but its intrinsic worth. Kant’s argument must be that freedom, as a property of our will, displays something intrinsically good about humans, something that grounds human dignity and a fortiori lays the foundation for our right to external freedom. This something, as any student of Kant knows, is precisely autonomy, understood, as a property necessary for a will to become a good will, i. e. the only thing in the world that is “good without qualification” and that, like a jewel, shines “by its own light as something which has its full value in itself” [AA, IV, S. 394]. Practical freedom at most displays our nature of rational beings, a feature that in a sense already positions us above the natural world. Practical freedom, however, does not establish our greater worth compared to any entity of the sensible world. For this, the ability to follow the moral law even at the detriment of any empirical interest (including our survival) is required. While practical freedom makes us kings and queens of the sensible world, autonomy makes us demi-gods, inhabitants of this world, but at the same time qualified members, or, perhaps more modestly, qualified applicants to another kingdom.
Moving to the textual reasons, we find confirmation of the insufficiency of practical freedom when Kant affirms that the sheer capacity to set ends for themselves makes humans extrinsically more valuable than animals or things, but not superior to them as for their intrinsic, unconditioned value. He writes:
“In the system of nature, a human being (homo phanomenon, animal rationale) is a being of slight importance and shares with the rest of animals, as offspring of the earth, a ordinary value (pretium vulgare). Although a human being has, in his understanding, something more than they and can set himself ends, even this gives him only extrinsic value for his usefulness (pretium usus); that is to say, it gives one man a higher value than another, that is a price as of a commodity in exchange with these animals as things, though he still has lower value than the universal medium of exchange, money, the value of which can therefore be called preeminent (pretium eminens).
But a human being regarded as a person, that is, as the subject of morally practical reason, is exalted above any price; for as a person (homo noumenon) he is not to be valued merely as a means to the ends of others or even to his own ends, but as an end in himself, that is, he possesses a dignity (an absolute inner worth) by which he exacts respect for himself from all other rational beings in the world. He can measure himself with every other being of this kind and value himself on a footing of equality with them” [AA, VI, S. 434—35].
Notice how Kant denies explicitly that practical freedom confers an unconditioned value to humans. At most, understanding and practical freedom make humans particularly efficient and functional entities, goods more valuable than any other in the world, yet inescapably mere goods. To the contrary, considered as subjects of a morally pratical reason, i. e. as autonomous, humans elevate themselves above any price, become ends in themselves, and legitimately demand respect from similar creatures. Kant re-emphasizes the point when he claims: „Humanity in his person is the object of the respect which he [main] can demand from every other human being” [AA, VI, S. 434—35]. Humanity as the basis for the only innate right to external freedom is ultimately our being subject to the moral law, our being autonomous.
At this point, it remains to be explained how precisely our autonomy is supposed to ground external freedom. The inference is less direct and clear than it appears at first sight. In fact, interpreters have divided themselves in two main groups in the attempt to fill the logical gap left by Kant.
IV. Two viewpoints on the foundation of our right
to external freedom
Starting from the first group, Mary Gregor claims that external freedom is presupposed in the very concept of autonomy. For humans to be autonomous, Gregor thinks, it is necessary they be given the possibility to choose among alternative courses of action. Thus, humans must be free externally, free from constraints imposed by other humans or institutions. As she puts it: “It is morally necessary that each of us, as a free agent, be able to express this freedom in external actions” [10, p. 27]. The body of laws in a commonwealth has as its main task that of guaranteeing and defending this external freedom which is nothing but the necessary means for the exercise of internal freedom, or autonomy.
Recently this interpretation has been endorsed by a number of commentators who insist on the necessary link between autonomy and external freedom, the latter construed as a condition of the possibility of the former. For example, Leslie Mulholland claims that the principal duty of the juridical system is that of creating the conditions of possibility of the exercise of autonomy[15, p. 402n]. Similarly Ernest Weinrib affirms that “The principle of right is therefore the external aspect of practical reason, or as practical reason as it pertains to interaction among free wills” [19, p. 27]. Charles Taylor does not construe external freedom as a condition of the possibility of autonomy, but argues that the former is a legitimate possession of an entity capable of “originating ends” and as such enjoying the status of moral agent. States that fail to dispense external freedom and equality to humans violate this status. Roger Sullivan continues on this line by affirming that autonomy confers on humans moral authority and binds states to treat individuals according to their status as moral agents. States may interfere with the lives of individuals as long as they respect their external freedom. Finally, one can say that Rawls, especially in A Theory of Justice, lets his two principles of justice, as well as the priority of the first principle over the second, follow from a Kantian conception of individuals as autonomous. In fact, in the original position Rawls assumes that individuals are “free and equal,” which interestingly parallels the two inborn natural rights we saw Kant grants to humans. Moreover, it is the assumption that individuals care about their autonomy (what Gerard Doppelt calls the assumption of a Kantian ideal [8, p. 259—307]) that leads Rawls to the lexical order between the two principles of justice. We rank the first principle of justice higher than the second because (see the Kantian interpretation of justice as fairness) we are supposed to prefer a political system wherein our liberties are fully guaranteed over one in which, say, we have higher economic benefits, but an unequal share of freedom. Thus Rawls, like all other interpreters, albeit through a different justificatory device, assumes that autonomy is the crucial element for a foundation of basic negative liberties. After all, the first principle of justice is nothing but a reformulation of the Rechtslehre universal principle of right.
Moving to the second group, Thomas Pogge, probably the most influential interpreter within this hermeneutical orientation, criticizes the Rawlsian reconstruction of Kant’s liberalism as “comprehensive,” i. e. as based on moral/metaphysical tenets, by arguing in favour of the possibility and opportunity to separate ethics from politics in Kant, and a fortiori, to keep the right to external freedom (as expressed in the universal principle of right) as logically independent of autonomy (the property of the will that ultimately binds us to the categorical imperative). External freedom (this is the gist of Pogge’s interpretation) can be grounded independently of any appeal to our autonomy. But if Kant can justify our liberties independently of the metaphysical apparatus on which his ethics rests, it follows that his liberalism can be “salvaged” from the uncomfortable qualification of being “comprehensive” [16, p. 133—158].
As textual evidence in favour of this heterodox reading, Pogge cites a passage form the Introduction to the Doctrine of Right in which Kant affirms: “Thus the universal law of right, so act externally that the free use of your choice can coexist with the freedom of everyone in accordance with a universal law, is indeed a law that lays an obligation on me, but it does not at all expect, far less demand, that I myself should limit my freedom to those conditions just for the sake of this obligation; instead reason says only that freedom is limited in its idea to those conditions and that it may be actively limited by others; and it says it as a postulate that is incapable of further proof”  [ AA, VI, S. 231].
On Pogge’s reading, Kant here severs the bonds between the categorical imperative and the universal principle of right by stating that the “bindingness” of the latter does not depend on that of the former. To the contrary, the obligation to conform our actions to the universal principle of right follows from the very idea of freedom (“in its idea”), because this idea already contains the restriction that the freedom of one be compatible with that of all others. The main task of the Rechtslehre would thus be to come up with a set of rules enabling the “co-existence game”, under the specified restriction. Such restriction, however, and this is the crucial point, is already contained in the definition of the game and is not imported from external sources. In this sense, the normativity of the universal principle of right would be contained in the very idea of freedom without owing anything to Kant’s (or others’) morality.
More recently, Gunnar Beck has reached a similar conclusion on the separability between Kant’s ethics and politics. Beck moves two fundamental objections to the traditional interpretation. On the one hand, he claims that this reading has scant textual evidence. As Beck puts it. “Nowhere […] in his political writings does Kant justify man’s right to external liberty by express reference to the requirements of autonomous agency” [2, p. 381]. On the other hand, Beck continues, it is not evident that one needs external freedom to be autonomous. An individual can be autonomous even if she is deprived of her external freedom. As he puts it: “As long as the agent possesses a morally good will, no external force can obstruct his internal state of autonomy, even though restrictions on his external liberty may prevent the agent from exercising or acting upon his autonomy, i.e., to act in accordance with what he nevertheless accepts as his duty ad wills to act on” [2, p. 383].
The general impression is that there are pieces of truth scattered in the two hermeneutical schools, although the first group of interpreters seems closer to Kant’s spirit and, pace Beck, Kant’s text. To begin with, although Kant does not justify humans’ right to external freedom through an explicit reference to autonomy, it is sufficient to investigate in his notion of “humanity” to find autonomy around the corner. As we saw, humanity is the explicit ground Kant offers for our right to external freedom. Man has that right “by virtue of his humanity,” says Kant. If our reconstruction of Kant’s argument is correct, however, the essence and worth of our humanity consists of our capacity to conform to the moral law. That is: it consists of our autonomy. Analogously, Pogge’s sterilized reading of the Rechtslehre and of Kant’s liberalism shows at most that if we all decided to play the coexistence game (say, for prudential reasons), then this game would obviously have its rules. In other words, if it is assumed from the beginning that free rational agents must coexist in such a way that their freedoms be compatible, then of course the rule that will best organize this coexistence is the universal principle of right. As pointed out by Bernd Ludwig, however, Pogge cannot give an authentically Kantian reason why these free individuals should play the Rechtslehre game, as opposed to, say, bellum omnium contra omnes game. Notoriously, for Kant the obligation to leave the state of nature and to enter a commonwealth is not merely “a good idea” all things considered. It is rather a moral duty. Exeundum est e statu naturali is for Kant a moral command. It is in fact the only way to respect, in us and in ourselves, of the human being’s intrinsic dignity that is already there before the establishment of the commonwealth. If Kant’s political philosophy is severed from his moral basis, however, we are deprived of this reference to the intrinsic value of individuals that in turns provides the non-prudential answer to the question as to why we should play the Rechtslehre game.
The bit of truth to be found in the second orientation, and in particular in Beck’s reading, is that external freedom does not really seem to be a presupposition, let alone a condition of the possibility, of autonomy. Let us imagine the most extreme case of denial of external freedom: a slave. Now the question is: Would a slave be an autonomous individual given Kant’s definition of autonomy? If by autonomy we mean a capacity to submit to the moral law the answer is clearly yes. Obviously, as a slave there are many ends that I cannot pursue, but there is still a number of decisions to make that would decide my moral status. I can, for example, decide to follow the duties towards myself as much as possible in those conditions. I can decide to rebel against the master or find less violent means of protest. I can treat my fellow slaves as ends in themselves or as mere means. In short, there is still a plenty of space in which I can exercise my capacity to live up to moral standards. The mere fact that this behavioural space is significantly smaller than that of a free man clearly does not entail that it is insufficient for exercising my autonomy.
Here is precisely the problem with the first group of interpreters. Although they rightly view external freedom as somehow dependent on our autonomy, they exaggerate when they construe this logical dependence as if the latter were a condition of possibility of the former. As in the case of the slave, one can remains an autonomous individual even if deprived of external freedom. Autonomy, as a property of our will, can always be exercised, even when our external freedom is most compressed. (It is something, so to speak, “nobody — individual or state — can take away from you”). Incidentally, this is one of the features of autonomy that makes it attractive as a candidate for the foundation of human rights. These in fact are universally considered as something that pertains to humans independently of the circumstances in which they happen to live. Autonomy grounds a dignity for the individuals that even they could not alienate, remove, compromise through immoral behaviour. This is another advantage of having the dignity defended by human rights rest on autonomy. Human rights should in fact protect even the most heinous criminals, dictators, abusers, in a word, those who have done their best to remove any trace of decency from their lives. Few human rights advocates, in fact, would deny that Saddam Hussein’s execution has been a violation of his inalienable entitlements.
This does not mean, however, that autonomy does not play any role in the foundation of our right to external freedom. It means merely that we have to abandon the hope to find a transcendental argument that moves from autonomy to external freedom as its condition of possibility. Hence our task is to fare between the Scylla of the complete separability between autonomy and external freedom and the Charybdis of the attempt to construe the latter as a presupposition of the former.
Perhaps one way to sail safely through this strait is to investigate into the presuppositions, not directly of autonomy, but of the peculiar respect we have by virtue of being autonomous. Notoriously for Kant a clear measure of respect is whether individuals treat themselves or others as ends in themselves and not as mere means. Since any arbitrary limitation of the freedom of one amounts to treating that individual as a mere means, it follows that any social organization in which the individual is not given the largest amount of freedom compatible with that, of equal size, of all others fails to respect at least one individual. The argument is roughly this:
1)Humans are autonomous (in the Kantian sense)
2)This property entitles them to respect, understood as an guarantee not to be treated as mere means
3)Any arbitrary limitation of the freedom of some members of the society amount to treating them as mere means, i. e. as a failure to respect them in the required manner
4)The only a priori limitation of external freedom permitted is that necessary to make the external freedom of one compatible with that of all others.
5)Any individual has a (pre-political, inborn) right to the largest amount of freedom compatible with the same freedom of all others
6)All humans have a pre-political, inborn right to external freedom and formal equality.
Clearly, the nerve of the argument is here the notion of respect. Respect bridges the logical gap, by and large gone unnoticed by the first group of interpreters, between autonomy and external freedom. From respect originates the prohibition to limit arbitrarily, not autonomy itself, which is strictly speaking, immune from restriction, but humans’ external freedom.
Even if this reading appears closer to Kant’s spirit and text, one needs to realize that our foundation of human rights faces two main defects. On the one hand, it moves back the logical cogency of the whole argument to Kant’s proof of the reality (from the practical point of view) of our autonomy. Since both the arguments offered in the Groundwork and those of the second critique are controversial, to say the least, the price to pay is considerable. On the other hand, even granting that these or other similar arguments succeed, the problem of the significance of autonomy in non-western cultures remains untouched. If we are proven to be autonomous, but being autonomous is not seen as the sole, or even as the main source of human worth, then Kant’s arguments will be ineffective as a first step toward an intercultural consensus on human rights. Beyond the relatively restricted circle of Kantian liberals, few would be ready to attach to autonomy the overarching importance one finds in Kant’s system. I don’t mean to suggest that these difficulties are insuperable. To the contrary, I think that something can be said in favour of this foundation. Most needed in this regard would be an argument that shows how autonomy is not, as commonly perceived, one value among the many considered as important by different cultures, but as a sort of condition of possibility of all possible values. After all, even religious loyalty, or attachment to group traditions and the like have some significance if and only if they are “endorsed” by free, autonomous individuals. In short, there is no genuine attachment to a tradition, even the most illiberal, that is not the result of the individual’s autonomous decision. Hence, it may be possible to show that independently of what value a group ranks as first, a condition of its cogency is that it is autonomously endorsed by the members of that group. If this transcendental uncovering of the logical dependence of all value from autonomy is combined with the notion of respect along the lines above suggested, then we may end up with a powerful foundation of human rights capable to cut across the moral plurality of our world. Needless to say, this is simply an unstructured intuition, and much more needs to be said to produce a convincing Kantian ground for human rights. The objective of this paper, however, was to move only the first step in this direction, by reconstructing an authentically Kantian interpretation of human rights. Our polemical target, in fact, was not the representative of a non-western, illiberal culture, but simply those who claim to be proposing or opposing a Kantian orientation in the philosophical reflection on human rights without clarity on what this means. There is nothing wrong in finding in Kant a source of inspiration for laying a foundation of human rights that bears, at the end of the day, a vague resemblance with Kant’s thought. Attributing to Kant attempts that are far from his intentions, however, does nothing but increase the conceptual confusion that already infects a good deal of contemporary speculation on human rights.
1. Allison H. Kant’s Theory of Freedom. Cambridge: CUP, 1990.
2. Beck G. Kant’s Theory of Rights // Ratio Juris. 2006. Vol. 19. № 4. P. 371—401.
3. Bielefeldt H. Autonomy and Republicanism: Immanuel Kant’s Philosophy of Freedom // Political Theory. 1997. Vol. 25. P. 524—58.
4. Caranti L. Moralità senza l’imperativo categorico? Risposta a Philippa Foot // Allison H. — Caranti L., Libertà trascendentale ed autorità della morale in Kant Roma: Luiss Edizioni, 2002. P. 63—95.
5. Carr C. Kant’s Theory of Political Authority // History of Political Thought. 1989. Vol. 10. P. 719—31.
6. Danto A. Constructing an Epistemology of Human Rights: A Pseudo Problem? / E. F. Paul, F. D. Miller, Jr., J. Paul (Ed.), Human Rights, cit. P. 25—30.
7. Dershowitz A. Rights from Wrongs. A Secular Theory of the Origins of Rights. New York: Basic Books, 2005.
8. Doppelt G. Rawls’ System of Justice: A Criticism From the Left / Nous.1981. Vol. 15. № 3. P. 259—307.
9. Gewirth A. The Epistemology of Human Rights // E. F. Paul, F. D. Miller, Jr., J. Paul (Eds.), Human Rights. Oxford: Blackwell, 1984. P. 1—24.
10. Gregor M. Laws of Freedom. Oxford: Blackwell, 1963.
11. Gregor M. The Metaphysics of Morals (henceforth MS). Cambridge: CUP, 1996.
12. Ignatieff M. Human Rights as Politics and Idolatry. Princeton: Princeton University Press, 2001.
13. Kant I. Grounding for the Metaphysics of Morals. Indianapolis: Hackett, 1781.
14. Kant I. Political Writings. Cambridge: CUP, 1970.
15. Molholland L. S. Kant’s System of Rights, New York, Columbia University Press, 1990.
16. Pogge T. Is Kant’s Rechtslehre a ‘Comprehensive Liberalism’ // M. Timmons (Ed.), Kant’s Metaphysics of Morals. Interpretative Essays. Oxford-New York: Oxford University Press, 2002. P. 133—158.
17. Reiss H. (Ed.), Kant. Political Writings. Cambridge: CUP, 1970.
18. Sciacca F. Il concetto di persona in Kant. Normatività e politica. Milano: Giuffré, 2000.
19. Weinrib E. J. Law as Idea of Reason, in Essays on Kant’s Political Philosophy, Cardiff, University of Wales Press, 1992.
This article was firstly published in collected articles «Klassische Vernunft und Herausforderungen der modernen Zivilisation»:
Caranti, Luigi. Kant’s Theory of Human Rights// Klassische Vernunft und Herausforderungen der modernen Zivilisation. Hrsg. Von Prof. Dr. Wladimir Bryuschinkin. Bd.1. Kaliningrad, 2010. S. 57– 78.
 The only occurrence I could find is in To Ward Perpetual Peace. In the context of the discussion on the relation between motality and politics, Kant claims, “The rights of man must be held sacred. However great a sacrifice the ruling power may have to make” [17, p.125]. Notice, however, that the original has it in the singular “Das Recht der Menschen …” [AA, VIII, S. 381].
 We will explain later the distinction between these two forms of freedom.
 Also see the interesting criticism by A. Danto , and the less than convincing reply by Gewirth [6, p. 25—30].
 One of the most popular foundations of human rights is offered by Michael Ignatieff who argues that the only plausibile grounding is prudential. In his view, we should stop asking why we have human rights, and concentrate on what they do for us. Similar is the approach by Alan Dershowitz in Rights from Wrongs, who argues that the only valid legitimation of human rights rests on the experience of the atrocities that infallibly take place in political regimes where human rights do not inform the constitutional law. See respectively [7; 12].
 Since this “right” is affirmed before the existence of the political structure needed to declare any right, it is to be understood as a moral entitlement, a legitimate claim each human can make to a certain form of treatment.
 Although many commentators are inclined to identify practical freedom with autonomy, and the hermeneutical issue would deserve a long discussion, there is a variety of good reasons (historical, systematic, textual) to resist this tendency. The historical reason is that Kant introduced the notion of autonomy relatively late in his career, i. e. after 1781. Until the first critique, in fact, Kant believes that all morality requires is practical freedom. And since practical freedom presupposes that some empirical end determine (without necessitating) my path of action, by Kant’s own later standards, this moral theory would count as heteronomous. The systematic reason is that the distinction makes room for free (hence imputable) immoral behaviour, thus salvaging Kant from the usual charge of embracing the grotesque view that on his account only moral behaviour count as free. The textual reason are, quite simply, the definitions of autonomy from the Groundwork on, that insist on the human will’s capacity to be a law to itself independently of any sensuous stimuli (i. e. without those empirical motives that determine practically free behaviour) as the hallmark of autonomous agency.
 One could say, albeit paradoxically, that heteronomojus agency contains in it a quota ofautonomy.
 On this see my reply to Philippa Foot’s thesis that moral agency remains mysterious in Kant because duty should be obeyed in absence of any motivation [4, p. 63—95].
 By reading Kant’s theory of freedom along these lines, at least we avoid attributing to Kant, as it has been done so often, the grotesque view that only moral agency is free and non-moral agency is not accountable precisely because not free. Needless to say sometimes Kant lends himself to such an interpretation when, for example, he claims: “what else, then, can freedom of the will be but autonomy, i.e., the property that the will has of being a law to itself” [AA, IV, S. 446—447] or “a free will [one with the property of autonomy] and a will under the moral law are one and the same [AA, IV, S. 447].
 And even if some behavioural regularities could be noticed — given similar stimuli and similar psychological settings of the agent, similar and therefore predictable responses follow — this would not change much. For Kant, in order to be able to say that “the agent X did the action Y”, as opposed to “the objective state of affairs Z, which includes object X, evolved toward the state of affairs W which includes Y”, namely, in order to save the presence of a subject, the agent’s making of a certain rule of action (e.g. “if thirsty, find the nearest fountain”) his maxim is still required. In the same way in which we would not consider the inference from two premises in a syllogism to the valid conclusion as a mechanical act, so we should not consider the individual’s selection of the (let’s assume, right) rule of action given certain circumstances as mechanically determined. The sheer fact that, given a certain knowledge of the subject, of the circumstances and perhaps of past token of responses to relevantly similar stimuli, we can predict how the individual will behave is clearly insufficient to downgrade that act to a mechanical response. Ultimately for Kant freedom, even in the “weaker” sense of practical spontaneity — a sense, to be sure, already way too strong for contemporary naturalistic standards — is not merely a presupposition for accountability. Being a necessary condition of subjectivity, it rules out any talk of a subject doing something mechanically as based on a categorical mistake.
 As explained, both practical freedom and autonomy share this feature. The difference is that autonomy adds to this ability the mark of complete independence from sensuous motivation.
 I am in debt with Gunnar Beck for the list of the authors who fall in the first group. See . In particular [2, p. 374—79].
 A similar assessment is to be found in [5, p. 719—31] and in [3, p. 524—58]. They both believe that Kant aims to infer external freedom from autonomy. The argument would be this: to treat someone as an end in herself means to treat her as an agent capable of self-direction. Any arbitrary infringement of her external freedom would diminish her autonomy (here understood as identical with the capacity of self-direction).
 See in particular [16, p.141—42].
 I here modify slightly Sullivan’s translation.
 On this see the already quoted criticism of Gewirth by Danto .
 At most it is the exercise of autonomy that can be limited, but merely in the sense that, if our external freedom is severely restricted, the occasions in which our choice to live up to the standards of morality (i.e., as autonomous beings) is to be exercised are less numerous.