In Rousseau and German Idealism James argues that Kant and the German idealists tried but failed to solve a problem identified by Rousseau. Unfortunately, it is unclear what the problem is, doubtful whether solving it was a central concern of Kant and the German idealists, and James’s argument that they failed to solve it rests on an assumption that Rousseau, Kant, Fichte and Hegel would all reject.

My discussion gives disproportionate attention to James’s treatment of Kant since, first, I have serious reservations about it and explaining and justifying these reservations require work; second, the readership of this journal is likely to be especially interested in it.

In chapter 1 James argues that Rousseau’s attempt to make freedom and interpersonal dependence consistent leaves a key problem unsolved. The main source of tension is ‘a complex form of dependence: dependence on other human beings as mediated by dependence on things’ (p. 38). This dependence on others is not as such a threat to freedom, but in a society that allows for material inequalities it can become the source of domination which, in turn, creates vice and destroys moral freedom (pp. 35, 38, 45). Of course Rousseau indicates in book 2 of the Social Contract that material inequalities are impermissible insofar as they lead to domination, but James argues that this solution fails, since only virtuous citizens would vote for the necessary legislation, while the very same legislation is a precondition for virtuous citizens (p. 48). This Catch-22 has been the topic of much scholarship on Rousseau and readers as diverse as Althusser, Durkheim and Shklar have used it to support the claim that Rousseau was pessimistic about the possibility of virtue and moral freedom. James is thus in good (though unacknowledged) company.

I do, however, think that James fails to distinguish adequately between three versions of the problem (these are helpfully distinguished by Cohen Reference Cohen2010): first, is it conceptually possible to find moral freedom in civil society? Second, if it is conceptually possible, and given what we know about human nature, are humans the sort of being that can realize this possibility? Third, what are the institutional preconditions for realizing moral freedom in society, and are these preconditions accessible for us, as we are now and in the world as we know it?

The modality of Rousseauian pessimism varies with the answers one gives to these questions: a negative answer to the first means that moral freedom is impossible for social beings; a negative answer to the second means that moral freedom is impossible for human beings; a negative answer to the third means that moral freedom is inaccessible to us, given our actual situation (though we might, for example, have followed another branch of history where moral freedom is accessible). Rousseau defends optimistic answers to all three questions, but James’s argument might lead us to the conclusion that Rousseau should have embraced a pessimistic answer to the third question. Yet it is less clear that James’s argument supports pessimistic answers to the first two. At least Rousseau’s essays on Corsica and Poland, his Letters Written from the Mountain, the sections on a just republic in Emile and his discussion of the legislator in book 2 of the Social Contract suggest that Rousseau is firmly committed to an optimistic answer to the second question.

In any case, it seems that the correctness of James’s claim that the German idealists tried but failed to solve Rousseau’s challenge varies with how we understand the challenge and that, insofar as they were occupied by the question of the consistency of freedom and society at all, the German idealists were concerned primarily with the first and second questions.

This lack of clarity bleeds into chapter 2, where James argues that Kant tried but failed to solve Rousseau’s challenge. If I understand James, his argument is that Kant tried to answer the second question, that is, to show that moral freedom is realizable for human beings, but Kant failed, since his theory of human evil precludes the solution he defends. In more detail, James’s argument is: (1) because of how ‘Kant’s theory of radical evil shapes his liberalism’ (p. 84), and (2) because Kant fails to recognize ‘the way in which dependence on other human beings … is liable to produce forms of domination and moral corruption’ (p. 55), (3) we should reject ‘the inevitability of the transition from the civil condition to a truly ethical community’ (p. 78). Therefore, (4) Kant fails to reconcile freedom and dependence and so does not find the way to virtue and moral freedom.

I have difficulties with each step in James’s argument.

(1) James’s main support for his claim that Kant’s theory of radical evil shapes his liberalism is the passage in Perpetual Peace where Kant asserts that ‘the problem of establishing a state … is soluble even for a nation of devils’.Footnote ¹ James uses this passage to support three significantly different interpretative claims (cf. pp. 54–5, 61–4, 76, 84–5, 87): first, that a Kantian republic is especially suited for evil beings, that ‘human beings get to live in the kind of society and under the kind of constitution that is best suited to their radically evil nature’ (p. 84); second, that a Kantian republic is acceptable to evil beings, that ‘Kant describes the constitution to which even a nation of devils could agree’ (p. 76); third, a claim about the maintainability of a republican society: that even devils could support and maintain a just constitution. However, even if a just constitution is suitable for, acceptable to, or maintainable by devils, this does not show that Kant’s theory of evil shapes his liberalism. To support that conclusion, James needs to show that Kant’s principles of right are determined by reference to what is suitable for, agreeable to, or maintainable by evil beings. Even worse, there is a fourth and more straightforward reading of the devils passage that James overlooks. On this reading, Kant is merely responding to the worry that a just republic is impossible, since it requires a superhuman degree of virtue of citizens so that it ‘would have to be a state of angels’ (Kant Reference Kant1996: 335; ZeF, 8: 366), by pointing out that, since justice requires only outward conformity with duty and this outward conformity can be secured by coercion, a just republican society does not presuppose a high degree of virtue of the members of society – indeed, even a nation of devils could be made to act justly, so the fact that human beings are less than fully virtuous does not imply that justice is impossible.

It is worth noting the absence of devils and evil from the Doctrine of Right. Alas, James’s treatment of Kant’s political philosophy is oddly silent on Kant’s Doctrine of Right. Indeed, as far as I could tell, James only engages with the Doctrine of Right once, namely, when he draws on Kant’s distinction between active and passive citizens to argue that Kant allows for (even affirms) relations of dependence based on material inequalities. It is an obvious weakness that James presents a theory about Kant’s liberalism without even mentioning the universal principle of right, the innate right to freedom, the discussion of the consistency of freedom and coercion, Kant’s theory of property, the distinction between private and public right, the principles of private and public right, or Kant’s discussion of republican citizenship in terms of freedom, independence and equality.

Moreover, it seems that in the Doctrine of Right Kant explicitly rejects the relation between evil and right that James attributes to him. When explaining the transition from private to public right, Kant writes:

While this passage can be interpreted in different ways, it at least indicates that the civil condition is required whether or not human beings are evil. Indeed, Kant’s definition of public right leaves out altogether any reference to dispositions: public right is ‘a system of laws for a people … which, because they affect one another, need a rightful condition under a will uniting them’ (Kant Reference Kant1996: 456; MdS, 6: 311). James at least should have explained how his interpretation can make sense of these and like passages.

(2) These shortcomings of the first step of James’s argument are, however, of little consequence, since James’s critique of Kant has little need for the claim that Kant’s liberalism rests on his theory of evil. James argues that a Kantian society creates material inequalities that engender relations of dependence and domination, which in turn create vicious citizens. This argument works (or does not work) equally well whether or not Kant’s principles of right are shaped by his theory of evil. All James needs to show is that relations of dependence and domination result from the principles of public right that Kant defends, and that such relations create vice.

James’s argument that a Kantian republic engenders dependence and domination is the following: Kantian political society is designed for self-interested individuals, and for such individuals the best design of the economic system is some form of free market capitalism (p. 85). In a free market system, the distribution of economic advantages is the outcome of unplanned, spontaneous processes – think of Adam Smith’s invisible hand or Hayek’s spontaneous order. This spontaneously created distribution of economic advantages will contain inequalities of wealth and income, and these inequalities will lead to interpersonal dependence and domination (which in turn will create the vices of culture).

The key step in this argument is the move from the assumption of self-interest to a commitment to Hayekian classical liberalism and a free market capitalist system for the distribution of economic advantages that such a classical liberalism supports. So the question is whether Kant is committed to this sort of classical liberalism.

Exactly which principles of public right Kantian contractualism supports is of course a hard and contested question, and it is therefore hard to say which principles ought to structure the distribution of economic advantages. Some have argued that Kant’s contractualism supports a Rawlsian approach to economic justice; others that Kant’s contractualism supports a classical liberal approach. So James could have offered an interpretation of the doctrine of right that supports his claim that Kant is committed to classical liberalism. Alas, since James bypasses Kant’s Doctrine of Right, he does not offer such an argument, nor does he engage with the relevant secondary literature.

I am inclined to read Kant as more of a republican than a liberal (classical or Rawlsian), but even if we grant that Kant affirms some form of classical liberalism, we should not conclude that Kant must therefore embrace vice-generating relations of dependence and domination. This step in James’s argument mischaracterizes classical liberalism and so is not a fair critique of Kant, even if Kant is a classical liberal. A classical liberal cannot make economic equality an intrinsic goal of political regulation. But if some distributions of economic advantages have pernicious effects, and if some measure of economic equality is necessary for important social ends such as freedom, prosperity and social stability, then classical liberals can endorse political measures that aim to ensure the desirable outcome.

(3) James’s third claim is that, because a Kantian republic creates vice rather than virtue, Kant cannot explain how civil society inevitably leads to ethical community. Again I doubt that Kant affirms the position James criticizes. The most straightforward reading of James’s claim is that Kant maintains that a just society contains the sufficient conditions for ethical community. Yet there is no textual basis for attributing such a position to Kant, and there are reasons to reject it, for example, Kant’s insistence in the Religion that we have a duty to create an ethical community (Kant Reference Kant2001: 132; RGV, 6: 96).

Alternatively, James might be advancing the weaker claim that, according to Kant, a just society (right) is a necessary condition for ethical community (virtue). This interpretative claim can be supported by passages in the Religion, Kant’s historical essays and Anthropology. But note that, if this is Kant’s position, it is too strong to say that Kant maintains that civil society inevitably leads to ethical community.

(4) Of course James’s claim might instead be that Kant’s philosophy is tragic, because the achievement of justice (civil society) precludes the achievement of virtue (ethical community). But, for reasons indicated above, I do not think this argument works. First, it is clear that there is no conceptual inconsistency between the demands of right and virtue. Moreover, right and virtue could (and do) develop simultaneously rather than sequentially. In addition, Kant’s principles of right are flexible enough to allow for regulation of economic inequalities (if inequalities in fact have pernicious effects). So I see no conceptual or nomological inconsistency between right and virtue in Kant’s philosophy.

The main message of chapters 3–5 is that, like Kant, Fichte and Hegel tried but failed to solve Rousseau’s problem. In chapter 3, James argues that Fichte offers a theory of property that Rousseau could endorse and that Fichte’s theory of property solves the problem of dependence and domination by introducing collectivized ownership of property and a state-guaranteed distribution of equal economic rights and advantages (pp. 110–18). However, this solution replaces the problems created by economic inequality with problems created by coercive state power, which equally threatens the moral freedom of citizens (pp. 130–41). In chapter 4 James argues that Hegel’s theory of right presents ‘an unstable synthesis of the position represented by Kant’s philosophy of history, on the one hand, and the position represented by Fichte’s theory of right, on the other’ (p. 142). Like Kant, Hegel underestimates how the inequalities created in civil society (the market) can undermine the possibility of the sort of political society (state) that will lead to ethical community (ethical life) (cf. pp. 155, 182, 186, 192, 221). In chapter 5 James discusses the place of idleness and leisure in the philosophies of Rousseau and Fichte.

The problem that the German idealists failed to solve can be presented in the form of a dilemma, and the general argument of Rousseau and German Idealism is that the failure of German idealism to solve this dilemma helps explain why we face the same dilemma today. The dilemma is that the corrupting effects of dependence and domination force upon us a choice between two unacceptable options: either we allow that economic advantages are distributed by a spontaneous process (i.e. the free market) or we take political control of the distribution of economic advantages. If we follow the first route (as Kant allegedly did), we create a system of material inequality that leads to dependence, domination and vice, and moral freedom becomes inaccessible. If we follow the second route (as Fichte allegedly did), we create a state that secures the unwilling cooperation of citizens by force and hence undermines moral freedom. Since both options are destructive of moral freedom, it appears that we cannot have both moral freedom and civil society.

James’s dilemma assumes that social dependence forces us to choose between an unregulated spontaneous process and a coercive system of regulation. But we should reject this assumption: a spontaneous process can be regulated, and regulation need not be coercive in a way that conflicts with freedom.

First, it is easy to see why one might assume that a spontaneous order must be unregulated – it is, after all, spontaneous. Yet a spontaneous order develops through individuals acting within a system of roles and rules, and we can regulate a spontaneous process by controlling these roles and rules. To illustrate, if you want to make rock crystal, you insert a string into a solution of water and sugar; the crystal is created by individual atoms crystallizing on the string. The process is spontaneous, and you do not control the behaviour of the individual atoms. But you can control the process in a number of ways by controlling the conditions – if for example you add food colouring you can control the colour of the crystals, if you keep the string and glass clean you get clearer crystals, and so on. Similarly, we can control what spontaneous order a free market generates by controlling the roles and rules that define the interactions of individuals within that market. Controlling these conditions does not mean that the individuals are not freely choosing or that the result is not a spontaneous order. Incidentally, this insight was used equally by Smith, Hayek and Rawls. Indeed, the system generated by the principles of Rawls’s justice as fairness provides a counterexample to James’s dilemma. In such a system, the distribution of material advantages results from a spontaneous order process (individuals interacting in the free market), yet by controlling the institutional background conditions of this process we can secure a fair distribution of benefits and burdens, without having to decide who gets what.

Second, the political regulation of external freedom need not be unduly coercive and so need not be inconsistent with moral freedom. This insight is of course captured in Rousseau’s idea of the general will: when laws are just and are generated by a democratic process, they can reliably claim to express the general will of the people and thus of every citizen. In actual bodies politic, citizens can reasonably doubt both the democratic genesis and the justice of legislation, and citizens will be tempted to break even laws that they believe are valid. So to offer assurance and secure compliance laws must enjoy what Habermas calls dual validity: laws must express the general will (and so be just), while at the same time carrying sufficient sanctions to make it rational for citizens to obey them independently of their justice. The point remains: a system of just laws does not force the wills of those subjected to it in a manner that is inconsistent with their moral freedom, for such laws express only what citizens will, when considering the matter from the standpoint of the general will.

So there is a path between the horns of James’s dilemma, and it is, I believe, a path followed by contractualists like Rousseau, Kant and Rawls: to construct a society in which the market is regulated so that the spontaneously generated material inequalities are acceptable to all. If it turns out that some inequalities create corruptive relations of dependence and domination, then that is sufficient reason for disallowing these inequalities, and we would then structure the system of rules to prevent them. In such a system external freedom is maintained, since the laws express the will of those subject to them, and virtue is possible, since the corrupting material inequalities are avoided.

In sum, James argues that the German idealists tried but failed to solve Rousseau’s challenge, but it is unclear exactly what the challenge is and doubtful that Kant, at least, was terribly worried by it. (I have similar concerns for Fichte and Hegel, but have not discussed these.) I have also voiced concerns about James’s interpretation of Kant. Moreover, even if James is right about the motivating concerns and positions of the German idealists, his argument that they failed to solve the problem assumes a choice between individual freedom and political coercion – an assumption that Rousseau, Kant, Fichte and Hegel would all reject.Footnote ²