A prominent feature of the landscape in moral philosophy and its history during the past forty years has been the simultaneous flowering of scholarship on Kant, alongside Kantian approaches to contemporary ethical theory. Kant's legal and political philosophies have fared less well, however. With some notable exceptions, they have attracted less sustained scholarly interest and inspired nothing like the contributions to current debates of Kantian moral philosophers such as Herman, Hill, and Korsgaard.Footnote ¹ Arthur Ripstein's Force and Freedom goes a long way to redressing this imbalance. It provides both a beautifully clear and insightful interpretation of the relevant Kantian texts as well as a sympathetic and forceful presentation of their central claims and arguments as Ripstein interprets them. It is a remarkable achievement.

Ripstein's focus is on Kant's Doctrine of Right, a notoriously difficult work both on its own and in relation to Kant's ethical writings. Some of the Rechtslehre's arguments are couched in the vocabulary of the transcendental idealism of the first Critique, and others in the terms of Roman private law. Moreover, Kant provides no clear road map of how his theory of right relates to his familiar ethical ideas, such as the Categorical Imperative.

Ripstein has a nice explanation for the former. Kant says that “the concept of right” is concerned only with “external” practical relations between people (6:230).Footnote ² In Ripstein's words, “principles of right gover[n] persons represented as occupying space” (12). It is to be expected, therefore, that Kant's theories regarding the concepts and categories that regulate our apprehension of physical objects in space will also show up in his view of how we should think about legal relations.

Ripstein is also sensitive to the relations between Kant's legal and political philosophy and his ethics. He rejects the idea that the former can be understood simply as an application of the Categorical Imperative to the political case. Rather, the doctrine of right is to be understood as a development of the concept of right, conceived as governing external practical relations between free persons (in time and space). Given Kant's views of the nature of persons, there are of course deep connections. The Groundwork's Formula of Humanity (FH) requires that persons always be treated as ends and never simply as means, and that entails, Kant says, that anyone who “transgresses the rights of human beings, [wrongly] intends to make use of the person of others merely as a means, without taking into consideration that, as rational beings, they are always to be valued at the same time as ends” (4:430). In parallel, Kant says in The Doctrine of Right that there is “only one original right,” which belongs “to every man by virtue of his humanity” (6:237).

This is not, however, a right to be respected or treated as an end in oneself. That is an ethical or moral constraint. What we are owed by right is independence, being free to be our “own master” (6:238). Violations of this right amount to coercion. And since, for Kant, a right is an “authorization to use coercion” (6:232), it follows that “coercion that is opposed to this (as a hindering of a hindrance to freedom)” is within our rights. The fundamental right is a right not to be coerced that justifies coercion in defense of this right.

This is supposed to follow, again, not by applying Kant's fundamental moral principle, the Categorical Imperative, but by developing the concept of right itself. Ripstein is particularly good on the way Kant's reasoning continues lines of thought from the Critique of Pure Reason. As Kant puts it:

Ripstein contrasts Kant's theory with what he calls an “applied ethics approach to political philosophy” that “supposes that law and the state are instruments for approximating factors that really matter” (7). The three main non-Kantian approaches he considers are consequentialist, Rawlsian, and Lockean theories such as Nozick's. Ripstein's Kant can be distinguished from each of these in turn. Unlike consequentialist theories of rights, but like Rawlsian and Lockean theories, Kant treats rights as having a different normativity from the good and as justifiable on other grounds. Unlike Rawlsian and like Lockean theories, Kant regards rights not as deriving from a more fundamental theory of justice but as fundamental and underived in their own right. And unlike Lockean theories, Kant sees rights as unspecifiable independently of what he calls a “rightful or “civil” condition. Whereas Lockean rights to life, liberty, and property are fully determinate in a state of nature, and a justifiable state is one that best enforces and realizes these, for Kant, only the innate right of humanity (IRH) can be specified and enforced in the state of nature. Rights to property are neither determinate nor specifiable outside of the state or a civil condition. Kant argues, partly on these grounds, that individuals in a state of nature have a moral duty to leave it and form a state in which their rights can be both determined and enforced.

It might seem, therefore, that for Kant the legitimate role of the state would be more expansive than it is for Lockean theories. And there is a sense in which that is correct. According to Kant, a just state will have powers extending significantly beyond those necessary to adjudicate, legislatively specify, and enforce natural rights—including powers to relieve poverty and to afford formal equality of opportunity for official positions. But for Ripstein's Kant, the just society has no purpose beyond establishing rightful relations between individual persons. It might be thought of as a form of what Nozick calls an “entitlement theory,” concerned with how persons relate to one another individually rather than with the overall “pattern” or structure of their relations.Footnote ³ In particular, Ripstein is at pains to distinguish Kant's account from a Rawlsian one that seeks to specify individual rights as part of a larger conception of a basic structure assuring “justice as fairness.” According to Ripstein's Kant, “the only basis for the state to make, enforce, or apply law at all,” is “the systematic realization of” rights rooted in the “innate right of humanity” not to be dependent upon or subject to “the private purposes of another” (218).

As Ripstein emphasizes, however, Kant argues that a civil condition is necessary to specify what counts as one individual subjecting another to his private purposes, since that will depend, as Ripstein also emphasizes, on their respective property relations, and that cannot be specified in the state of nature. Property cannot be acquired with a merely “unilateral will,” but only if it is authorized by the “omnilateral will” or “united will” that is necessary for civil condition. Moreover, Ripstein maintains that Kant holds that it is the need for a united will that requires the state, among other things, to alleviate poverty, since poverty can create a kind of dependence on others that is inimical to the forming of a united will (273–284). Pushing on these aspects of Kant's view may lessen the differences with Rawls's views, however, as I argue below.

We might best begin, though, by considering the elements of Kant's legal and political philosophy that seem least like Rawls's. The IRH, again, is a right to independence, a right to be free of having one's choices wrongfully determined by others. It is a right to be one's own master and not a slave or a servant of anyone else. But what does this involve? Uncontroversially, for Kant, persons are end setters. So the IRH entails that each person gets to decide which ends or purposes he or she will pursue, so long as that does not violate others’ rights. As Ripstein sometimes puts it, if you are your own master, then “nobody gets to tell you what purposes to pursue with your means” (34).

But usurpation of choice is not the only, or perhaps even the most usual, form that violation of the IRH takes. It is important to Ripstein's account that if I take something to which you have a property right, which you might use as a means to accomplishing some purpose, then I am in effect telling you what purposes to pursue. Your choices and will have been made, by a deed of mine, wrongfully dependent on my will and choice. The point is a delicate one, but it is utterly central both to Ripstein's interpretation of Kant and to his own Kantian position.

There are two important ideas underlying the Kantian point as Ripstein interprets it. One is a distinction between mere desires or wishes and choice or will. The second is Kant's distinction between the form and matter of choice.

Setting an end, in the sense of choosing or intending it, is not the same thing as simply desiring that it be realized (14). It is committing oneself to it in a way that is subject to constraints of means/ends rationality, as, for example, Michael Bratman emphasizes in his work on intention. There is nothing irrational in failing to desire the necessary means to merely desired ends, though there is, arguably, in failing to intend necessary means to ends one has set in the sense of chosen or intended. The reason that what is sometimes called “the Hypothetical Imperative” is supposed to be analytic is that in the setting of an end, as Kant puts it in the Groundwork, “my causality as acting cause, that is, the use of means, is already thought” (4:417). So if one knowingly lacks or cannot acquire the means necessary to pursue some end, then one cannot intelligibly set that end for oneself. Therefore, taking someone's means from her affects her as an end setter.

But what if property is taken from someone that is not necessary for any end she currently has? How could that violate her right to humanity as an end setter? How could that make her choices wrongfully dependent on or subservient to the choices of another person? The key here is a second Kantian distinction, between the matter and form of choice. Though stolen property may not be a necessary means to any end the owner has already chosen or set, it will nonetheless be necessary to some end she might set. So taking the property from her affects her as a setter of her ends no less than it would if she had already chosen or set the end. It makes ends that would otherwise have been eligible, ineligible. “Limits on independence,” Ripstein writes, “abstract from what Kant calls the matter of choice” and focus on their “form,” on a person's “capacity to set purposes without having them set by others” (16; see also 104). Taking means that are necessary to some purpose someone might have undermines her ability to set that end and so makes her choices wrongfully dependent on one's choice.

Of course, something can be a necessary means to an end that someone has set or might set even if she lacks a property right to those means. So it would seem that whether a person is affected as a setter of ends, whether her choices are independent of or dependent upon the will or choices of others, need not depend on what she and others respectively own. Here we confront a point that is absolutely critical to Ripstein and to Kant as Ripstein understands him. Kant's innate right of humanity to freedom is, Ripstein says, the right “to use the means that you have to set and pursue whatever purposes you see fit, restricted only by the entitlement of others to do the same with their means” (63, emphasis added). In other words, the IRH and its requirements presuppose already existing property relations and cannot be specified independently of them.

Because the point is so important, it is worth quoting Ripstein at some length:

This means, again, that what counts as violating a person's right of humanity depends on what powers or means are that person's, either means that she owns, in the case of property that could be “mine or thine,” or means that partly constitute the person she herself is (in the case of her body and its powers) (e.g., 12, 91). If you need a drug to survive, and I own it but have no use for it, then I do you no wrong if I refuse to give it to you. Even if there is a clear sense in which your choices now depend upon mine, they will nonetheless be independent of mine in the sense of the IRH on Ripstein's interpretation. I would not be making you my servant or subjecting you to me. Were you to take the drug from me, however, you would be subjecting me to your will and so violating my right of humanity.

To see how Kant can be interpreted in this way, it is worth taking some pains with the opening paragraphs of the Introduction to The Doctrine of Right. “The concept of right, insofar as it is related to an obligation corresponding to it (i.e., the moral concept of right) has to do,” Kant says, with three things:

1. 1. “The external and indeed practical relation of one person to another, insofar as their actions, as deeds, can have (direct or indirect) influence on each other.”

2. 2. “It does not signify the relation of one's choice to the mere wish (and hence also to the mere need) of the other, . . . but only a relation to the other's choice.”

3. 3. “In this reciprocal relation of choice no account at all is taken of the matter of choice, that is, of the end each has in mind with the object he wants. . . . All that is in question is the form in the relation of choice on the part of both, insofar as choice is regarded merely as free, and whether the choice of one can be united with the choice of another in accordance with universal law.” (6:230)

Kant thus takes it to be a conceptual truth that rights of the kind that entail correlative obligations concern relations between persons, more specifically, that they concern persons’ “reciprocal relation of choice,” that is, how their choices “influence” the other as a chooser or setter of ends, irrespectively of the specific ends she might set or choices she might make. If a person's choices are in a rightful reciprocal relation with another person, then his choices can be “united with the [the other's] choice” and so “coexist” with his “freedom.” So if a person's choices are in a rightful reciprocal relation with every other person, then they can “coexist with everyone's freedom in accordance with a universal law.”

Kant then draws “the universal principle of right” (UPR) as a consequence: “‘Any action is right if it can coexist with everyone's freedom in accordance with a universal law” (6:230). “Right” in this formula means rightful, that is, the action wrongs no one. We wrong others when our actions cannot coexist with theirs in a reciprocal relation of freedom.

From the UPR, Kant infers “the universal law of right” (ULR): “so act externally that the free use of your choice can coexist with the freedom of everyone in accordance with a universal law” (6:231). Violating the ULR is, perforce, violating someone's freedom; it is “a hindrance to freedom in accordance with universal laws” (6:231). And wronging someone in this way is also wrong (period, as we might put it); it violates the ULR, which “law lays an obligation on me” (6:231).

Having the right not to be wronged in this way, however, is not the same thing as others simply being obligated (period) not to do this, that they do wrong if they do. It might be wrong to act in some way in respect of a person without its being the case that the action thereby wrongs that person and so is something the person has a right that one not do. What brings in the additional element of right, according to Kant, is how one is entitled to act if one's right is violated and one is wronged or if someone is attempting to wrong one.

We are wronged when our freedom is hindered. But just as “resistance that counteracts the hindering of an effect promotes this effect and is consistent with it,” so also must “hindering of a hindrance to freedom” be “consistent with freedom in accordance with universal law” (6:231). So Kant concludes that it follows from the UPR that hindering violations of the ULR is right, in the sense of not violating anyone's rights. Since “coercion is a hindrance or resistance to freedom,” it follows further that coercively resisting or reversing coercion is within our rights. Kant concludes: “Hence there is connected with right by the principle of contradiction an authorization to coerce someone who infringes upon it” (6:231). In other words, authorization or justification for coercing those who violate “the only original right belong to every man by virtue of his humanity,” the right to “independence from being constrained by another's choice,” follows from the very concept of right. If rights are postulated, so also is the authorization to coerce, on pain of contradiction. For Kant, this is the “additional element” in which rights consist. “Right and authorization to use coercion therefore mean one and the same thing” (6:232).

It is worth noting that Kant's analogy of hindering hindrances to freedom to hindering the promotion of an effect is tightest only if being “consistent with freedom” is the same thing as promoting freedom or at least not promoting its opposite. This conceives of freedom very differently from being something to be honored or respected, as something to which people have a right. Though the causal effects of two opposing forces cancel one another out, it does not follow, in the order of right, to put the point tendentiously, that two wrongs make a right. Even if we remove the tendentious rhetoric, it is hard to see how it can possibly follow simply from the concept of right that if one has a right to something, one is justified in using coercion that would otherwise violate someone else's right to protect one's having it. There must be some additional element, some moral or normative power, involved in having a right to something that extends beyond the simple fact that another would do wrong if he took that thing from one. One must have some power or authority one would not have otherwise have had, but it hard to see how the authority to coerce simply follows.

Worse, Kant says that “right and authorization to use coercion mean one and the same thing” (6:230). So note also that if this is true, then what it means to have a right to something is simply that one can, consistently with others’ rights, use coercion to protect one's having that thing. Despite Kant's saying that the moral concept of right “is related to an obligation corresponding to it,” no such obligation follows from the concept of right as an authorization to coerce. When one has such an authorization, hence right, it will no doubt be independently true that others are obligated to provide it by some moral principle, but this obligation will not be entailed by the right. All that is entailed by the right is that one can, consistently with others’ rights, coerce them to do what they are independently obligated to do.

If the concept of right is the same as that of an authorization to coerce, then no Hohfeldian correlative obligation follows. Though others may be obligated to comply with our rights, and though we know that he thinks this is so from both the ULR and Kant's Groundwork claim that transgressing others’ rights treats them as “mere means,” no directed or bipolar obligation to the right holder follows from rights as Kant defines them.Footnote ⁴ To put the point another way, the ULR cannot follow from the concept of right, on Kant's definition, as the line of thought sketched above might suggest. It would seem that the ULR must be an independent normative principle. Perhaps this is why Kant says it is “a postulate incapable of further proof” (6:231).

But now, how does Kant get from the line of thought we have been sketching to the IRH as Ripstein interprets it, namely, in terms of already presupposed property relations? Recall Kant's idea that like mathematical concepts, the concept of right is constructable “with a priori intuition” (6:233). Just as the idea of a “right [straight] line” is given through an a priori construction “between two given points,” so “analogously to this, the doctrine of right wants to be sure that what belongs to each has been determined (with mathematical exactitude)” (6:233). On Ripstein's interpretation, the IRH is each person's right to use his means for whatever ends he pleases, independently of others’ will and pleasure. Kant seems to be saying that a construction of “law of a reciprocal coercion necessarily in accord with the freedom of everyone” must be in terms of “what belongs to each.”

Since the right concerns practical external relations of persons occupying space, the “basic case” of the innate right to freedom is a person's right to determine the use of her own body (12). “In the first instance, your right to your person is your right to your body,” since “your body is the sum of your capacities to set and pursue your purposes” (40). The idea is not that there is a fundamental property right in the body; rather, “your body is your person” (91). Property is a right in something that can be “mine or thine.” My body can only be mine.

What, then, grounds property rights? If I am holding something you do not own, and you try to take it away from me, then you will be using force on my body and so on me. But the wrongness of forcefully taking something I have in my physical possession does not entail that I own it. As a conceptual matter, property requires not physical possession but what Kant calls “intelligible” or “noumenal possession” (95). But how does one acquire a property right in something “external” and possess it noumenally? As an act of freedom that purports to give one a right to it, it must be through some “sign” that publicly expresses one's will “to exclude everyone else from it” (6:258). This much is guaranteed conceptually. Moreover, Kant takes it to follow from the same line of thought that leads to the UPR that if an “external” thing is unowned, then it must be possible for individuals to acquire a property right in it. This is what he calls the “postulate of practical reason with regard to rights” (6:250). If it were “not within my rightful power to make use” of what “I have the physical power to use,” then “freedom would be depriving itself of the use of its choice with regard to an object of choice,” which cannot, Kant thinks, be consistent with the idea of reciprocal freedom under universal law (6:250).

We can, for these reasons, interpret the UPR, the ULR, and the IRH as Ripstein proposes. The fundamental right that individual persons have against each other is a right to use their means as they choose, independently of the choices of others. We can even suppose that any right that individual persons have against one another must be grounded somehow in this right and can assume that acquired rights are restricted to those of property and contract. It will follow that individuals can wrong each other only by “physically constraining” them in the use of their means, “by depriving [them] of the ability to use their means,” or by failing to follow through on any contracts they have made with one another. Nothing else, including need, however great, can ground the claim that someone is owed something as a matter of right.

Reading Kant in this way gives him, as I say above, a kind of entitlement theory in Nozick's sense. What we owe another person as a matter of right depends only on our choices in relation to that person, given his right to use his means as he will for ends he determines. As I also mention above, however, there are elements in Kant's theory, most especially his idea that property requires an “omnilateral will” and so a public political order that could result from a “united will,” that push his thought much more in the direction of ideas we associate less with Nozick than with Rawls.

It is worth recalling, in this connection, that Rawls's theory of justice is consistent with the idea that individuals do one another no wrong, they violate no individual right, so long as they respect one another's person and personal property. Justice, for Rawls, is the “first virtue of social institutions,” and what individuals owe one another as a matter of right must be reckoned relative to a just basic social and political structure. At the level of individual transactions, a Rawlsian theory may not differ from an entitlement theory such as Nozick's. Similarly, Kant's theory might have an entitlement structure at the level of individual transactions, but since, unlike a Lockean theory such as Nozick's, no determinate property relations can be specified outside of a political or “civil” condition subject to a “united,” “omnilateral,” or “general will,” it might end up with significant Rawlsian elements. To this we now turn.

First, note why, according to Kant, property relations require a civil condition that is grounded in an “omnilateral will.” The deepest reason is because property relations “put everyone under obligation,” and it is impossible for a “unilateral” will to do this (6:263). The point is clearest in the case of acquisition. To acquire something as one's own is to change the normative relations that others have to that thing. Others can no longer make use of it without one's consent without violating one's property right and their obligation to respect that. It follows, as shown above, that property can be acquired only by a public expression of will. But though a publicly expressed will is necessary to acquire a property right, it is not sufficient, since simply expressing one's will to others that they be bound does not necessarily bind them.

Intuitively, property acquisition involves a “normative power,” the exercise of which can alter normative relations.Footnote ⁵ But it is a perfectly general feature of normative powers that they can be exercised only in “transactions” in which participants reciprocally recognize their powers to enter into them.Footnote ⁶ No normative power can be exercised in isolation or unilaterally. Just as it takes two to tango, it takes at least two to exercise a normative power. It is thus conceptually impossible that property could be acquired by a “unilateral will.”

But neither is any number of individual or “particular” wills sufficient. For these purposes, Kant says, even a “bilateral, but still particular will is also unilateral” (6:263). The reason is that property relations “put everyone under obligation,” where the “everyone” is not just every existing individual but anyone who might happen to exist. It is a familiar theme in Kant's thought that nothing genuinely universal can arise contingently. Property acquisition is possible, therefore, only through “a will that is omnilateral, that is, united not contingently [for example, as in any actual contract] but a priori and therefore necessarily” (6:263). Therefore, “a right against every possessor” requires an “authorization” that “can be thought as contained in a synthetic general will” (6:269).

The authorization necessary for property relations, without which, as shown above, the IRH cannot even be interpreted, can therefore be provided only in a civil condition in accordance with a public rule of law. In this way, Kant's view is unlike Nozick's and more like Rawls's. As noted, however, Ripstein stresses that Kant's theory of right differs also from Rawls's because although Rawls understands relations of right and justice as what would result from an ideal hypothetical agreement or choice, Kant sees these as whatever emerges from the actual workings of public political and legal institutions, so long as these meet certain constraints that are necessary so that the people can give themselves the relevant law (215–219). As Ripstein puts it, “if a validly enacted statute could be agreed to, citizens are required to regard it as having received” the a priori “omnilateral authorization” necessary for it to create the requisite rights and obligations (214).

But what are the real structural differences between Rawls's theory and Kant's on this interpretation? It is impossible even to discuss, much less to settle this question in any detail here, but I would like to give some reasons for thinking that Ripstein may be overplaying the differences. To begin with, Rawls holds that questions of justice can arise at different levels. In addition to his best-known idea that a just political order has a basic structure satisfying two principles of justice as fairness that would be agreed to in the original position, the perspective of free and equal citizens, Rawls holds that questions of justice arise also at the constitutional and legislative stages: What would be a just constitution for our society (informed by particular social facts)? What would be a just law for our society (informed by particular constitutional facts)?Footnote ⁷ Although the right to own personal property is assured by a first principle of justice that is applicable to any society, this right must be specified by actual constitutions and legislations to create specific property rights and obligations.

In this respect, Kant's and Rawls's theories have a similar structure. Granted, it is possible in Rawls's scheme for “duly enacted law” to be unjust. But that does not commit Rawls to holding that citizens cannot be obligated by them. And Kant, as Ripstein points out, requires that actual laws meet the test that they be an a priori possible legislation, something the people can agree to and thus that the omnilateral will authorizes. (Alternatively, we might say that this test must be met for them actually to be law. This would be a difference with Rawls, but a semantic one.)

The crucial point, one that Ripstein himself emphasizes, is that a people cannot be understood to agree to just any public laws, since some are inconsistent with their “duty of rightful honor” that they not be subservient or “mere means for others” (6:236). As Ripstein notes, Kant holds that any public realm capable of being authorized by an omnilateral will must include provisions for poverty relief and equal opportunity, since, Ripstein stresses, leaving the meeting of human needs to private charity makes the poor depend upon the rich in ways that are inconsistent with their “rightful honor” (272–284). And Ripstein makes similar arguments for why any state authorizable by an omnilateral will must have public roads, spaces, health care, and education. Without significantly more analysis and argument, it is not obvious how the idea that obligating political institutions and laws, including property arrangements, must be those people can agree to consistently with their duty of rightful honor differs from the Rawlsian claim that they must be consistent with constitutions that are themselves consistent with a basic social structure that would be rationally agreed upon by free and equal citizens.

But whatever the relation to Rawls, there is no doubt that Ripstein's Force and Freedom enables the reader to get inside Kant's theory of right in its own terms in a way that no previous treatment has been able to. The lines of thought it opens up, both scholarly and philosophical, will give us much to think about for long time to come.