I The Unresolved Problem of the Grundlegung
The announced aim of Kant's moral philosophy, as stated most clearly and unambiguously in the Foundations of the Metaphysics of Morals, is to discover unconditionally valid principles of practical reason. Kant conceives his task as falling into three parts, exactly corresponding to the three Sections of the Foundations. First, he must identify and state the Moral Law, the highest principle of practical reason, by which all rational agents, merely in virtue of being rational agents, are bound. Secondly, he must demonstrate that the Highest Moral Principle can be derived, entirely a priori, from a conceptual analysis of Practical Reason. Finally, in light of the teaching of the First Critique, Kant must connect up his conclusions concerning the principles guiding the choices of rational agents with our experience of ourselves as causally determined beings in the realm of appearance.
In sum, then, the Foundations attempts to identify and state the principle governing the actions of rational agents; to derive that principle from an a priori analysis of rational agency, thereby proving that all rational agents, simply in virtue of their rational agency, necessarily act out of respect for that principle; and to demonstrate that we, as conditioned beings in the realm of appearance, must act as though we know ourselves to be rational agents, even though such knowledge is, strictly speaking, unavailable to us.
It is, of course, a matter of considerable controversy whether Kant succeeds in achieving all three, or indeed any, of these goals. The first is the least controversial, for it merely involves showing that Kant and his audience of North Prussian Pietists share a belief in a particularly rigoristic formalism. The achievement of the third goal, involving as it does the entire complex metaphysical and epistemological theoretical structure of the First Critique, is best left to one side in a discussion such as this. But the second of Kant's undertakings warrants our closest attention, for it is here that we find the greatest controversy arising among students of Kant's moral philosophy.
The controversy surrounding Kant's derivation of the Categorical Imperative concerns not the validity of the derivation but its significance. Many students of Kant's philosophy, and more broadly many students of moral and political philosophy, will readily grant that some principle of formal consistency of willing can be derived a priori from an analysis of agency and rationality as such. What makes Kant's moral theory so controversial, especially in regard to his second undertaking, is the extraordinary claim that the purely a priori formal principle of consistency in willing is sufficient to identify, from among the host of possible moral rules, just those which reason commands, and which therefore constitute the substance of morality.
Kant actually tries three times in the Foundations to establish the universal validity of unconditional moral principles. The first attempt is the famous four examples of the Categorical Imperative. Kant's arguments against suicide and self-indulgent laziness are hopelessly inadequate. Both make explicit and utterly unjustified appeal to the supposed objective natural purpose of some human capacity or psychological tendency. The argument concerning ungenerosity is rather interesting, but it is ultimately inadequate. For a variety of reasons, most scholarly and philosophical attention has focused on the second argument, concerning false promising.
The standard reconstruction of the false promising example, with which I entirely agree, is to construe promising as a social practice defined by a system of implicit, but clearly understood, rules, among which rather prominently featured is a rule against false promising. On this view, when I utter the words "I promise" in the appropriate social context, I am implicitly endorsing, or committing myself to, or willing, in Kant's language, the entire system of rules that constitute the practice of promising. But it takes only a little reflection to recognize that the injunction against false promising is not actually a categorical substantive moral command. Rather, it is hypothetical in form. What Kant's argument, suitably reconstructed, demonstrates is that false promising is incompatible with the practice of promise-making, from which it follows that we must, in all consistency, choose either not to endorse, participate in, and commit ourselves to the practice of promising or else choose not to make false promises. But there is nothing in Kant's argument to dissuade us from forswearing the institution of promising altogether. The crucial point is that Kant has no plausible a priori argument in support of the claim that a rational agent will necessarily enter into the practice of promising, or more generally into whatever overarching practice is implicit in the truthful use of language, from which the obligation in particular to participate in the practice of promise-making might be derived. In effect, Kant is at this point in his exposition unwittingly assuming that the agents of whom he is speaking have some minimal commitment to honorable social interaction with which, as he quite rightly argues, false promising is logically incompatible. To the thorough reprobate who simply rejects such a commitment, however, Kant has no valid objection.
Thus, Kant's first two attempts to extract substantive moral principles from his purely formal analysis of rational willing fail because of two fundamental problems: First, his failure to demonstrate that rational agents have a standing, unconditional obligation to enter into, or participate in, rule-governed social practices in the context of which the notion of contradictory willing can be fleshed out and given substance; and Second, his failure to identify obligatory ends, adoption of which follows necessarily from the mere fact of being a rational agent, and the existence of which would, in a different but equivalent way, provide substantial content for the purely formal principle of consistency in willing. Kant's third attempt in the Foundations is directed precisely at compensating for both of these failures - namely, his invocation of the social contract tradition of political theory under the guise of what he calls a "Realm of Ends."
By "realm" I understand the systematic union of different rational beings through common laws. Because laws determine ends with regard to their universal validity, if we abstract from the personal difference of rational beings and thus from all content of their private ends, we can think of a whole of all ends in systematic connection, a whole of rational beings as ends in themselves as well as of the particular ends that each may set for himself. This is a realm of ends, which is possible on the aforesaid principles. For all rational beings stand under the law that each of them should treat himself and all others never merely as means but in every case also as an end in himself. Thus there arises a systematic union of rational beings through common objective laws. This is a realm that may be called a realm of ends (certainly only an ideal), because what these laws have in view is just the relation of these beings to each other as ends and means.
A rational being belongs to the realm of ends as a member when he gives universal laws in it while also himself subject to these laws. He belongs to it as sovereign when he, as legislating, is subject to the will of no other. The rational being must regard himself always as legislative in a realm of ends possible through the freedom of the will, whether he belongs to it as member or as sovereign. He cannot maintain the latter position merely through the maxims of his will but only when he is a completely independent being without need and with power adequate to his will.
Morality, therefore, consists in the relation of every action to that legislation through which alone a realm of ends is possible. It must be able to arise from his will, whose principle then is to take no action according to any maxim which would be inconsistent with its being a universal law and thus to act only so that the will through its maxims could regard itself at the same time as universally lawgiving.
This language is, of course, a direct echo of Rousseau's characterization of the republic brought into existence by the social contract. It specifies the procedure by which a collection of self-interested individuals can transform themselves into a republic by entering into the mutual agreement referred to as a social contract.
At the close of the Foundations, despite Kant's introduction of the evocative notion of humanity as an end in itself, and his invocation of a Rousseauean conception of a republic regulated by a social contract, we are left with the two problems outlined earlier: First, how to demonstrate that rational agents as such must, in all consistency, enter into collective agreements that establish structures of social practices in the context of which substantial meaning can be given to the notion of contradictory willing; and Second, how to demonstrate that rational agents who have thus constituted themselves a realm of ends or republic will, qua rational, arrive at a single universal, necessary, and therefore objective set of substantive laws as the content of their collective rational willing.
Kant's most successful attempt to solve these two problems appears not in the Foundations but in the Metaphysics o f Morals for which the earlier work is intended as a foundation or groundwork. Before turning to that text, however, it might be worth pausing for a moment to observe the relationship between the unresolved problems just stated and the work of the most prominent contemporary political theorist, John Rawls. In his widely read work, A Theory of Justice, Rawls undertakes to demonstrate that rationally self-interested individuals, placed in a situation designed to mimic that of noumenal agents, will necessarily choose to commit themselves to a set of general principles regulating the basic structure of any society of which they may be members. By virtue of the conditions of deliberation - characterized fancifully by Rawls as consisting in a "veil of ignorance" - the choices made by these individuals will be universal and necessary, hence objective, and will at the same time be sufficiently specific to yield substantive social imperatives.
Rawls' treatment differs fundamentally from that of Kant, of course, inasmuch as Rawls posits rationally self-interested, which is to say in Kant's language heteronymous, individuals. Nevertheless, as Rawls has refined and revised his theory, he has moved more and more in the direction of a Kantian reinterpretation of his central ideas. Rawls' theory is a good deal more technically sophisticated than Kant's, involving as it does notions drawn from modern neo-classical economic theory and the branch of mathematics known as Game Theory. But, not surprisingly, Kant's theory is a great deal more profound than Rawls', for whereas Rawls posits a society of rationally self-interested agents, thereby giving up entirely any attempt to identify unconditional principles of morality, Kant holds firm to the idea of rational agents as such, abstracting even from their self-interest, and appealing only to what can be derived from their character as agents, which is to say from the fact that they possess practical reason.
II. The Resolution of the Problem in the Rechtslehre
It is in Part I of the Metaphysics of Morals, the "Theory of Right" or "Theory of Justice" [Rechtslehre] that Kant finally mounts a full-scale frontal assault on the problems left unresolved at the end of the Foundations. This fact - assuming for a moment that my reading of the situation is correct - has a very interesting significance. Contrary to Kant's own conception of the relationship between his moral and political theory, it would appear that they are not separate and co-equal branches of the Metaphysics of Morals. Rather, they are a single integrated theory, in which the central thesis of the political theory is required to complete the argument of the moral theory. In this regard, it is suggestive to compare Kant both with Rousseau, who influenced him, and with Rawls, whom he in turn influenced.
Kant begins the Rechtslehre by introducing the concept of justice. In a section entitled "What is Justice?" he writes:
The concept of justice, insofar as it relates to an obligation corresponding to it (that is, the moral concept of justice), applies [only under the following conditions]. First, it applies only to the external and - what is more - practical relationship of one person to another in which their actions can exert an influence on each other (directly or indirectly). Second, the concept applies only to the relationship of a will to another person's will, not to his wishes or desires (or even just his needs), which are the concern of acts of benevolence and charity. Third, the concept of justice does not take into consideration the matter of the will, that is, the end that a person intends to accomplish by means of the object that he wills; for example, we do not ask whether someone who buys wares from me for his own business will profit from the transaction. Instead, in applying the concept of justice we take into consideration only the form of the relationship between the wills insofar as they are regarded as free, and whether the action of one of them can be conjoined with the freedom of the other in accordance with a universal law.
Justice is therefore the aggregate of those conditions under which the will of one person can be conjoined with the will of another in accordance with a universal law of freedom.
Thus, as Kant states two paragraphs later, "the universal law of justice is: act externally in such a way that the free use of your will is compatible with the freedom of everyone according to a universal law."
Kant glosses this, almost immediately, as follows:
[T]he concept of justice can be held to consist immediately of the possibility of the conjunction of universal reciprocal coercion with the freedom of everyone. Just as justice in general has as its object only what is external in actions, so strict justice, inasmuch as it contains no ethical elements, requires no determining grounds of the will besides those that are purely external, for only then is it pure and not confused with any prescriptions of virtue.
There are a number of problems in Kant's doctrine here, arising principally from his insistence on speaking as though the distinction between the noumenal and the phenomenal [or the internal and the external, as he puts it here] can actually be drawn within experience. All such claims, implicit or otherwise, are, of course, strictly incompatible with the teaching of the First Critique. The real difficulty for my present purposes is the fact that this conception of justice as justified universal reciprocal coercion does not provide the unconditional a priori substantive content for moral principles for which we are searching.
The problem, very simply, is that despite the appearance of Kant's formulation, which is cast in categorical language, the injunction is still hypothetical. IF you choose to coerce others, THEN you yourself must submit to a like coercion. Note, by the way, that it is as yet unclear how narrowly this injunction constrains us, even should we choose to coerce. It would appear that there is a very wide range of reciprocal coercions compatible with the injunction, including some that Kant would presumably not find attractive. For example, would his principle be compatible with a system of laws that authorizes blood feuds and duels?
But even if Kant can demonstrate that a group of individuals, by committing themselves to the fundamental principle of justice, thereby so severely constrain their subsequent legislative choices that only a single system of laws is compatible with that principle, that system will still have a merely hypothetical status, for it will command only those who have chosen to enter into the social contract. What Kant needs - what he has needed from the very start - is an argument designed to show that failure to enter a social contract can only issue from an internal contradiction in willing. In short, Kant must show that a rational agent as such necessarily seeks to enter into a social contract, and does so as soon as possible.
To return for a moment to the failed example of false promising from the Foundations, if Kant could show that the institution of promising is required by the fundamental principle of justice [not, one would imagine, too difficult a task], and if Kant could also show that a rational agent as such necessarily enters a social contract, then he could conclude that rational agents as such are not only, in all consistency, required by mere reason alone to keep such promises as they make, but that they are also required, by the dictates of a priori reason, to adopt the practice of promise-making. He then really could conclude, as he wishes to, that false promising is an example of contradictory willing all the way down.
I do not believe that Kant accomplishes these extraordinary tasks. If I did, I would, in all consistency, forthwith embrace his ethical theory. But I think I can show that he makes an extremely imaginative stab at the second of them in the Rechtslehre, where, as I shall suggest, he advances an argument designed to show that we have an unconditional obligation to enter a social contract.
The key to his argument is the concept of property.
An object is mine de jure (meum juris) if I am so bound to it that anyone else who uses it without my consent thereby injures me. The subjective condition of the possibility of the use of an object is possession.
An external thing is mine, however, only if I can assume that it is still possible for me to be injured by someone else's use of the thing even when it is not in my possession. Consequently, there would be a self-contradiction in the concept of possession if it did not have two meanings, namely sensible possession and intelligible possession- Sensible possession means the physical possession of an object, whereas intelligible possession means the purely de jure possession of the same object.
Kant goes on to discuss the distinction between sensible and intelligible possession in ways that are thoroughly problematic, involving as they seem to the legitimacy of a distinction between the phenomenal and the noumenal within experience. We can leave that difficulty aside for our purposes, for it is of course the concept of intelligible possession, or possession de jure, that is relevant. Almost immediately, Kant states what he calls the Juridical Postulate of Practical Reason, which asserts that "it must be possible to have any and every external object of my will as my property." In other words, as Kant explains, "a maxim according to which, if it were made into a law, an object of will would have to be in itself (objectively) ownerless (res nullius) conflicts with Law and justice.”
Before analyzing how Kant justifies this postulate, and uses it to accomplish his fundamental aim in his moral theory, it is worth pausing to remind ourselves just what is being claimed here, for a great deal of contemporary importance is at stake. It is not too much to say that Kant is here laying the groundwork for the refutation of all manner of environmentalist and ecological doctrines, as well as a number of nationalist doctrines based upon a conception of the objectively privileged territory, homeland, or place of a people. Kant himself, of course, is looking backward, not forward. His intention is to destroy the last vestiges of feudalism, and lay the groundwork for a thoroughly rational commodification of natural objects.
What the principle says is that anything can, in principle, be someone's possession. There is nothing unownable. It does not follow, needless to say, that everything is actually owned; only that there is nothing - no tree, no river, no plot of land, no species of animal or plant, no planet, no solar system - that by its nature resists ownership, that is such that it cannot be the rightful possession of some individual or group of individuals. And possession here implies rightful use of the possessed thing, by the owner, in pursuit of the owner's purposes, and also alienation or legal transfer of the possession of the possessed thing to another rational agent.
Needless to say, this conception flies in the face of the pre-capitalist traditions against which Kant is arguing. To suggest that the Earl of Northumberland owns Northumberland, and can sell it to the King of France, should he choose, is fundamentally to undermine the notion of hereditary family possession implied in the familiar allusion to the Earl as simply "Northumberland," as in one of Shakespeare's plays. Donald Trump, on the other hand, can perfectly well sell the Trump Shuttle to Delta, should he find himself a bit short of cash.
In a more modern vein, any suggestion that the human race stands in a symbiotic, or fiduciary, or other moral relationship to nature is completely incompatible with Kant's Postulate. Equally incompatible, of course, is any form of religious or quasi-religious privileging of species or things other than the human species or other species of rational agents, should they exist. In the coin of Kant's Realm of Ends, the principle "Treat humanity always as an end, and never simply as a means" is inscribed on the obverse. On the reverse, however, is found the correlative principle, "Anything else may be treated as a means only."
Kant now offers an extremely strong interpretation of the concept of intelligible possession. "A thing is externally mine," he says, "if it is such that any prevention of my use of it would constitute an injury to me even if it is not in my possession (that is, I am not the holder of the object)." What Kant is speaking of here, as he indicates immediately, is intelligible possession, or de jure possession. The language might lead us to conclude that Kant is deliberately trying to construct a justification for the most thoroughgoingly unregulated period of capitalist expansion, but that would be a trifle hasty, I think. Kant's Postulate is perfectly compatible with positive legislation to constrain the ways in which property owners deal with their property - zoning laws, and so forth. What the Postulate says is that all such laws must be acts of a legislature constituted by a social contract. They cannot be deduced, independently of legitimate legislation, from the nature of the objects themselves. It is not that an owner must be allowed to do with his or her property whatever he or she wills, but that such freedom must at least be possible, in order for there to be de jure possession.
Having defined the concept of de jure possession, Kant immediately makes what is for him, by this late stage in the unfolding of his system, an entirely predictable move. He asks for a deduction of the concept of purely de jure possession of an external object [what he calls, parenthetically, possessio noumenon). That is to say, he seeks to show that the concept finds legitimate employment, indeed must find legitimate employment, within the realm of experience.
Put as simply and clearly as I am able, Kant's argument for the possibility of de jure possession is this: Since I am a phenomenal being - since I am, in other words, a rational agent that manifests its agency in the realm of appearance - my will at least potentially requires the cooperation of nature for the fulfillment of its purposes, whatever they may turn out to be. Even if I adopt the extreme stoicism of an Epictetus, seeking only virtue and not the powers or pleasures of the world, nevertheless I shall find myself compelled to employ some portions of nature as means to my ends.
But the laws of nature are such that I can use a portion of nature as a means to my ends only by appropriating it, and thereby excluding others from a like appropriation of those same portions. In short, for the accomplishment of what I will, for the enactment of my maxims, I require property.
Were I an incorporeal being, not manifesting my agency in space and time, I might have neither the need for, nor indeed the possibility of, property. Imagine, for example, that I were merely a noumenal rational agent whose acts consisted in the contemplation of pure ideas or the endless elaboration of the relationships among abstract logical constructs. In that case, my appropriation of modus ponens or the law of the excluded middle would in no way exclude others, for the contemplation of a timeless truth of logic does not require that others be denied its use or enjoyment. But because we are phenomenal beings whose agency is manifested in space and time, my appropriation at least potentially excludes you.
Kant's argument thus far can be summarized in a series of conditionals, preceded by a declarative assertion that comes as close as he thinks possible to the flat claim that I am a rational agent. The argument looks like this:
1. Insofar as I act, I must assume, though I cannot know it, that I am a rational agent, which is to say, that I am free. [This is the conclusion of the Third Section of the Foundations.]
2. If I am a rational agent, then willing an end, I necessarily will the means. [This, Kant has persuasively argued in the Second Section of the Foundations, is analytic.]
3. If I will the means to the fulfillment of my ends, then, as a phenomenal being - one whose agency is manifested in the realm of appearance - I must legitimately appropriate, which is to say take de jure possession of - portions of the spatiotemporal realm as means.
4. If I must take de jure possession of some portion of the spatio-temporal realm as means, then it must be possible for me to do so. What is more, there must be, in principle, no portion of the natural world that it is not possible for me to possess, inasmuch as there is nothing in the nature of willing as such that places limits on what might, according to the laws of nature, serve as means to my ends.
From all of which it follows that de jure possession must be possible.
The question remains, however: How is de jure possession possible? What is required for such legitimate possession to be actual? Locke, of course, had begged this question in the Second Treatise of Government by simply asserting the right of property as a truth revealed by the natural light of reason. But Kant, correctly in my view, recognizes that this is a radically unsatisfactory grounding for the right of property. Instead, like both Hobbes and Rousseau before him, Kant grounds the right to property in a prior mutual agreement, or social contract, among all those who, in the pursuit of their ends, may come into conflict with one another in the appropriation of portions of nature.
Legitimate ownership involves the exclusion of others from the use and enjoyment of a portion of nature, an exclusion that may be instituted by force if necessary. Such a use of force, Kant argues, if in all consistency universalized, entails mutual constraints among all the members of a society - where society here can be understood quite simply as the totality of persons who, in the pursuit of their ends, are likely to interfere with one another. So Kant concludes that the possibility of property entails the existence of a social contract.
And now Kant concludes his argument in a strikingly bold and imaginative fashion. The central text is §8 of the First Chapter of the First Part of the Rechtslehre. Here it is in its entirety:
When I declare (by word or deed), "I will that an external thing shall be mine," I thereby declare it obligatory for everyone else to refrain from the object of my will. This is an obligation that no one would have apart from this juridical act of mine. Included in this claim, however, is an acknowledgment of being reciprocally bound to everyone else to a similar and equal restraint with respect to what is theirs. The obligation involved here comes from a universal rule of the external juridical relationship [that is, says the translator, the civil society]. Consequently, I am not bound to leave what is another's untouched if everyone else does not in turn guarantee to
me with regard to what is mine that he will act in accordance with exactly the same principle. This guarantee does not require a special juridical act, but is already contained in the concept of being externally bound to a duty on account of the universality, and hence also the reciprocity, of an obligation coming from a universal rule.
Now, with respect to an external and contingent possession, a unilateral Will cannot serve as a coercive law for everyone, since that would be a violation of freedom in accordance with universal laws. Therefore, only a Will binding everyone else - that is, a collective, universal (common), and powerful Will - is the kind of Will that can provide the guarantee required. The condition of being subject to general external (that is, public) legislation that is backed by power is the civil society. Accordingly, a thing can be externally yours or mine only in a civil society.
[And now, Kant concludes with a dramatic flourish:]
Conclusion: If it must be de jure possible to have an external object as one's own, then the subject must also be allowed to compel everyone else with whom he comes into conflict over the question of whether such an object is his to enter, together with him, a society under a civil constitution.
The conclusion of the argument prior to this point, you will recall, was that de jure possession must be possible, or, as Kant puts it in the text before us, "it must be de jure possible to have an external object as one's own." Combining this with the conclusion of the argument just quoted, we can now conclude that it must be allowed to compel others to enter with one into a society under a civil constitution. Since this necessity is universal, it of course follows that others have an equal right to compel me to enter civil society.
We can now see that in the text of the Rechtslehre, Kant finally provides what was missing from the argument of the Foundations, namely a demonstration that such obligations as faithful promising are not hypothetically, but are rather categorically, imperative.
Kant's argument is open to criticism at every stage, of course, though it is nonetheless, in my estimation, extremely interesting. Before sketching some of those criticisms, let me call attention once again to an implication of the argument that goes directly counter to the most common impression concerning Kant's ethical theory. Kant is generally viewed as a rigorist, an objectivist, a universalist, a theorist who claims to be able to demonstrate a priori the universal validity of very powerful moral principles that are binding on all agents regardless of their nature or circumstances. And this impression is quite correct. But readers of Kant almost always draw the natural conclusion that he believes these moral principles to be binding on us even in the absence of a legitimately established state. In the familiar language of liberal political theory, Kant is seen as siding with Locke on the question whether the moral law is binding on agents in a state of nature. But it should now be clear that Kant's argument actually entails the opposite conclusion. The procedural obligation to enter into a social contract is certainly binding upon agents in a state of nature, but the moral principles enacted into law by the Legislature of a state thus established are binding only after, and on the condition of, the establishment of the legitimate state.
Kant clearly does not intend this consequence of his moral theory. Quite to the contrary. But his failure to provide a satisfactory theory of obligatory ends forces anyone wishing to embrace his moral theory to rely upon the legislation of the legitimate state as the only source of morally binding substantive principles of practical reason.
And this leads us ineluctably to the question sketched above, whether a legitimately established state, based upon a unanimous social contract and committed to embodying the fundamental principle of justice in its laws, is thereby constrained to a single structure of justice that thus constitutes the objective, universal, unconditionally binding system of principles of practical reason? This, I take it, is essentially the question originally posed by John Rawls in the earlier versions of his theory, before he drained it of its force and interest by an endless series of ad hoc adjustments, concessions, baroque elaborations, and qualifications.
If the answer to the question is yes - if, let us say, a legitimately established republic of rational agents in search of principles compatible with the fundamental postulate of justice must necessarily come upon and agree to Rawls' Two Principles of Justice - then by combining Kant's argument and Rawls', we would have a very powerful defense indeed of a universal system of moral principles. The argument, in a nutshell, would go like this:
1. Rationality as such entails consistent willing.
2. Consistent willing, for a phenomenally appearing agent, entails the possibility of property.
3. The possibility of property entails the necessity of establishing a state through a social contract.
4. A legitimate state composed of rational agents will necessarily enact one and only one set of fundamental principles of justice, namely the Two Principles of Justice.
5. Therefore, we are all, as phenomenally appearing rational agents, obligated universally and unconditionally to form legitimately grounded political communities with those with whom we come into contact, and in those communities to enact the Two Principles of Justice as the fundamental laws governing our interactions.
There are three stages in this argument: the derivation of the requirement of consistent willing from an analysis of what it is to be an agent; the deduction of the necessity of entering a social contract from an analysis of the preconditions of property; and the demonstration that reason dictates one and only one system of principles as the content of a legitimate state.
My own view is that the first stage in the argument, which as I have indicated we find in the pivotal portion of the Second Section of the Foundations, is essentially sound. To be an agent is to be moved by reasons, and the logic of reasons requires consistency of willing. What counts as a good reason for me necessarily counts as a good reason for any agent in relevantly similar circumstances.
The third stage of the argument, I am convinced, is mistaken. Neither Rawls nor anyone else has, to my knowledge, made a convincing case for the claim that free and equal rational agents must, insofar as they are rational, coordinate on a single system of substantive principles regulating their interactions with one another.
It is the second stage that I find especially interesting, in part, I confess, because I have managed to make it clear to myself only recently. Can the material circumstances of human existence - our spatio-temporality, our dependence for the pursuit of our ends on inanimate nature - ground an argument that we have a standing procedural obligation to attempt rational community with those agents with whom we interact? Can we, as Kant
claims, compel such rational community?
I think we can conclude immediately that we cannot compel others to enter with us in a social contract, for the essence of such an agreement is that it represents mutual willing, and that implies that it is voluntary. But am I required, simply by the constraints of rational consistency, to seek such community with others, or is it consistent with my status as an agent to adopt nothing more than an instrumental stance vis-a-vis those with whom I interact? Is this, after all, the real content of the injunction to treat humanity as an end and not simply as a means?
I honestly don't know. My pre-philosophical inclination is to believe that the answer is yes, that there is something about construing myself as a rational agent that requires me, in all consistency, to attempt to achieve rational community with others whom I consider to be agents as well - though I do not consider myself bound, should they reject that attempt, to treat them as I would have agreed to treat and be treated, had we actually achieved rational community. But at this point, I do not see an argument that can make that conclusion plausible.