One of this book's many important contributions is Rosen's argument that the German Idealist conception of freedom arose from a theological attempt to make the workings of divine justice intelligible in the world. I suggest a possible amendment to this profound suggestion. In chapter 4, Rosen takes the fundamental conception of freedom at the heart of the German Idealist enterprise to be the absence of arbitrariness. But might freedom from arbitrariness be, for Kant and Hegel, a species of a more fundamental genus—freedom from external determination? Since Rosen develops his account of German Idealist freedom mostly in relation to Kant, I focus on him but conclude with a note about Hegel.

In Rosen's account, to justify the goodness of God in a world of evil, Kant designed a theodicy to show that desert is proportional to merit, that people get what they deserve, even if they get their just deserts only in the afterlife. The notion of the afterlife is a crucial element of Rosen's larger project of showing that post-Kantian thinkers converted personal immortality into historical immortality—our identification with the human species. Rosen notes that the very possibility of just deserts meted out by a divine judge requires us to be the kind of creatures who can justly be punished for our evil deeds. So Kant developed a moral metaphysics that justifies divine punishment by focusing on our freedom to choose good or evil. Only if we are free can our punishment be deserved, and only then is God's justice assured.

However, as Rosen shows, Kant also argued that if our choices are subject to arbitrary determinants, then we are not wholly responsible for them. If we are to be free in such a way as to deserve divine punishment, our choices must stem from our wills, and from nothing else. This is the doctrine of freedom as self-determination. At the same time, though, it does not seem that the elimination of arbitrariness, for Kant, is sufficient to make a choice fully ours. For example, our choices may be determined by nonarbitrary factors, such as, for moral realists, moral facts, or by the divine laws of a rationalist God—the type of nonvoluntarist God Kant envisioned, who is not arbitrary. If freedom from arbitrariness were the nub of freedom, Kant would not have seen these nonarbitrary determinants as interferences with our freedom. It would seem that the problem in such cases for Kant is that regardless of whether the determinants are nonarbitrary, they are external to the self, and thus detract from self-determination.

If so, it would be plausible to conclude that Kant conceived of freedom as fundamentally the elimination of all sources of external determinants on us, rather than as just the elimination of all sources of arbitrariness. It would seem that Kant assumed that if we are free from all external constraint or determination, then the only possible determinant of our choices is our own will. For Kant, then, we are responsible for our choices if we alone will them—which is not to say that our wills are the source of the value of the choice: only the good will, which wills the moral law, is unconditionally good.

It would appear that Kant believed that freedom as self-determination rests on a more primitive, negative type of freedom, freedom from external determination. In considering the possibility that someone may lack freedom yet think he has it, Kant writes in the second Critique that such a person would be “a thinking automaton, but the consciousness of his own spontaneity, if taken for freedom, would be mere delusion inasmuch as it deserves to be called freedom only comparatively, because the proximate determining causes of its motion and a long series of their determining causes are indeed internal”—and thus might seem to be a case of self-determination—“but the last and highest is found entirely in an alien hand [in einer fremden Hand]” (quoted at Rosen, 71), and is thus not a case of true self-determination. In the Groundwork Kant writes that “will is a kind of causality of living beings insofar as they are rational, and freedom would be that property of such causality that it can be efficient independently of alien causes determining it, just as natural necessity is the property of the causality of all nonrational beings to be determined to activity by the influence of alien causes” (quoted at 202, emphasis original).

It would seem, then, that Kant describes freedom as not only the absence of arbitrariness but the absence of external determination. But this raises the question of what he took the relationship between them to be. As Rosen sees it, “what makes a cause ‘alien’ is the arbitrariness that it shares with an uncaused, random choice” (202). In this view, the external is a species of the arbitrary. But I wonder whether Kant might have thought that the arbitrary is a species of the external, especially if we allow that rationalist divine laws or moral-realist facts are nonarbitrary but external determinants that would count, in Kant's eyes, as impingements on self-determination. In this view, Kant allowed that arbitrary determinants do interfere with our freedom, but they make us unfree, not necessarily because they are merely arbitrary, but because they are intrusions stemming from an alien cause acting on us without our consent. Their arbitrariness, then, is symptomatic of their externality to our wills. In this account, freedom as nonarbitrariness is not merely semantically but conceptually distinct from freedom from external determination.

Two implications of treating nonexternality as the core of German Idealist freedom arise. First, it may enable us to better situate Marx within the Kantian tradition. Rosen notes that Kant's language of “alien causes” “anticipates the famous formulation by Marx in ‘On the Jewish Question’ in which he describes human beings in civil society as ‘the playthings of alien powers’ (Spielball fremder Mächte)” (202). In the Grundrisse, too, Marx says that “the social character of activity . . . appears as something alien and objective [Fremdes, Sachliches]” and that individuals’ “collisions with one another produce an alien social power [fremde gesellschaftliche Macht] standing above them.”Footnote ¹ Similarly, in the first volume of Capital he writes that the laws of capitalist production “confront the individual capitalist as an external coercive law [äußerliches Zwangsgesetz]” and that the connection between workers is, when brought about by capitalists, “outside [außer] them” and is the “power of an alien will [Macht eines fremden Willens].”Footnote ² Such quotations might be multiplied, for Marx's texts are replete with the assumption that freedom requires the absence of externality. Yet as far as I can tell, he was less concerned with arbitrary determinants, except as symptoms of an external will.

Perhaps Marx simply departed from Kant in this respect, as he departed from Hegel in many respects. But the notion of freedom from external determination also illuminates Hegel's work, particularly his view that freedom is being with oneself in one's other (bei sich selbst in seinem Anderen sein). This is not an obvious conception of freedom, but I think it can be shown that Hegel derives it, ultimately, from a conception of freedom as the absence of external control. In his early essay on natural law, he maintained that freedom's “essence and its formal definition is just that nothing is absolutely external [nichts absolut Äußeres ist].”Footnote ³ Freedom is not simply the absence of external determination but the absence of externality, requiring that a free being be one that lacks an absolutely external other, presumably because this makes external determination impossible. From this view of freedom, Hegel derived the notion of absolute totality,Footnote ⁴ which he identified with God. The logic is that if freedom requires the absence of external determination, then only a being that incorporates its other inside rather than outside of itself can be free and thus wholly self-determining. Such a self-determining being, in turn, is at home in its other precisely because its other is internal rather than external to it.

I recognize that it is a bit deflationary to locate the core of Kantian, Hegelian, and Marxian freedom in freedom from external determination rather than freedom from arbitrariness. For this appears to assimilate them to a Hobbesian account of freedom as the absence of external constraint, a republican account of freedom as the state of not being subject to domination by an external will, or some mixture of the two. I see this, though, as one of the strengths of this alternative view, for it situates Kant, Hegel, and Marx within a larger historical frame, while suggesting that the importance of a conception of freedom may lie less in the conception itself than in how one grounds it and in the implications one derives from it. Kant, for example, grounded this view of freedom in the notion of moral responsibility that is required if we are to be justly punished by God, while Hegel derived from it the notion of totality and the unexpected conclusion that freedom requires an internal other.