In my book For the Love of Metaphysics I offer a new perspective on the history of German Idealism that focuses on the role of the principle of sufficient reason and on the Kantian idea of the primacy of practical reason. In his Critique of Pure Reason, Immanuel Kant characterizes reason as a faculty that seeks the conditions for whatever is given to it as conditioned, and also the totality of such conditions, which must be unconditioned. He writes that,
the proper principle of reason in general is...to find the unconditioned for conditioned cognitions of the understanding, with which its unity will be completed.
Reason therefore seeks an explanation of everything that admits of one, and it also pursues a complete explanation in terms of something that does not itself require or admit of an explanation. I argue that Kant's view of reason manifests a commitment to a restricted version of the principle of sufficient reason. Yet in the Critique of Pure Reason Kant identifies a clear conflict within speculative or theoretical reason: while one cannot have cognition of unconditioned objects (the objects of primary interest in special metaphysics, such as God, freedom, or the soul), reason through its own internal dynamic demands that one accept the existence of such objects. His Critique of Practical Reason opens the path for a successful resolution of this conflict within the sphere of the practical. While theoretical reason is unable to provide cognition of anything unconditioned, practical reason justifies rational belief (Vernunftglaube) in precisely those unconditioned objects that are at issue in traditional metaphysics.
In Part One of my book, I show how Friedrich Heinrich Jacobi and Salomon Maimon radicalized Kant's view that only practical reason can provide objective reality and cognitive access to the unconditioned, and I explain why this led to a different diagnosis of the conflict of reason. Employing arguments that have been used in different contexts by the British Idealist Francis Herbert Bradley (and more recently by Michael Della Rocca, Peter van Inwagen, and Karl Schaffer), Jacobi claimed that the only way to meet reason's demand for the unconditioned—for explanations that terminate in something that does not itself require or admit of an explanation—would be to accept a monistic metaphysics; yet Jacobi also argued that monism would lead to nihilism and to fatalism. For post-Kantian German Idealists, solving the conflict of reason and meeting reason's demand for the unconditioned thus turned into the task of developing a form of monism that would not result in nihilism or fatalism. In the second and third part of the book I explore the different ways in which Johann Gottlieb Fichte, Friedrich Wilhelm Josef Schelling, and Franz Rosenzweig attempt to respond to, and overcome, the threat posed by nihilism that Jacobi first identified, while also accepting Jacobi's view that only a monistic metaphysics could satisfy reason's demand for the unconditioned.
Alexandra Newton's comments focus on the Fichtean interpretation of Kant's "fact of reason" that I provide in chapter three of the book. As Newton rightly notes, this interpretation of the fact of reason plays a central role in the book. This is because one of its main aims is to explore the role of practical reason within post-Kantian German Idealist attempts to rescue metaphysics from the threat of nihilism. Newton's main concern is that certain aspects of my Fichtean interpretation of the "fact of reason" amount to a radical departure from Kant's own views. On her view, this poses a challenge for my contention that post-Kantian philosophers can be seen as radicalizing Kant's insight concerning the primacy of the practical. Yet I do not see why this should be the case, for my claim that post-Kantian German Idealists radicalize Kant's view concerning the primacy of the practical implies that they adopt a more extreme version of Kant's original view. Nonetheless, in what follows I will attempt to spell out the degree of continuity between Kantian and post-Kantian views concerning the primacy of practical reason, focusing on the first two questions that Newton raises in her critique.
Newton's first question concerns my two-step reading of Kant's deduction of freedom through the fact of reason in the Critique of Practical Reason. She writes,
Nisenbaum's position in the scholarly debate about the relation between the Groundwork and the second Critique is not entirely clear. [NFR 45]
More specifically, it is not clear on Newton's view whether I agree with many scholars that Kant's appeal to the fact of reason amounts to a great reversal in his views concerning the justification of morality. While I see the source of the confusion when I write
that on a Fichtean interpretation of...the fact of reason, a form of self-relation that Fichte calls self-positing is shown to be the ground of moral obligation (LM 147),
it is worth noting that the title of the relevant chapter is "Kant's Deduction of Freedom," which indicates that the chapter explains Kant's deduction of human beings' faculty of freedom (rather than a deduction of the moral law). As is well known to Kant scholars, in the Critique of Practical Reason Kant claims that one should abandon the "vainly sought deduction of the moral principle," and he asserts that the moral law "has no need of justifying grounds." In my book, I acknowledge this shift in perspective when I state:
In works leading up to the second Critique, Kant first sought to establish that we are free or spontaneous...and then use the fact of our spontaneity to ground morality. Yet in the second Critique...the moral principle...serves as the principle of the deduction of...the faculty of freedom. [LM 132]
While I recognize the significant reversal in Kant's argumentative strategy, my aim in that chapter is to show how Kant's appeal to the moral law as an undeniable fact of reason can be understood, without giving rise to the charge of dogmatism. Notably, Karl Ameriks alludes to this charge when he writes that Kant "can be said to have encouraged the return, at least in Germany, to a kind of dogmatic metaphysics." As I explain in the first part of the chapter, the form of skepticism that I believe Kant's deduction of freedom is meant to address is a form of normative justificatory skepticism. This kind of skeptic would accuse one of dogmatism by arguing that "nothing entitles us to hold onto our beliefs" (LM 113). This is one way in which my Fichtean reading of Kant's deduction of freedom shows consistency with the spirit of Kant's critical philosophy. Kant's deductions address the question: quid juris? This is a question of justification or entitlement. An interpretation of Kant's deduction of freedom that is faithful to the spirit of his critical philosophy should therefore avoid any form of dogmatism.
Newton also worries that my reconstruction of the performative and first-personal aspect of Kant's deduction of freedom (step two in my two-step interpretation) is,
hard to square with Kant's suggestion that the moral law is the "self-consciousness of a pure practical reason, this being identical with the positive concept of freedom" (KpV 5:29). [NFR 45]
According to Newton's view "there is no conception of the moral law that is not a self-conception" (NFR 45). She argues that my Fichtean interpretation of the fact of reason suggests that,
I could know what it would be to be subject to the moral law, and thereby know what being free would entail, without knowing that I myself am subject to the moral law, or that I am free. [NFR 45]
Eric Watkins and Marcus Willaschek have addressed in detail the Kantian distinction between cognition (Erkenntnis) and knowledge (Wissen), and Peter Kain provides an insightful discussion of practical cognition. It is important to note that in the first step of the argument I do not describe one's representation of the moral law as being an act of knowledge. Instead, at this point in the argument I note that Kant is operating at the level of conceptual analysis. Kant asks:
supposing that a will is free: to find the law that alone is competent to determine it necessarily. [KpV 162, 5:29]
Then, he states the reciprocity thesis: "Thus freedom and unconditional practical law reciprocally imply each other" (KpV 162, 5:29) and explains that one immediately becomes aware of the moral law as soon as one draws up maxims of the will for oneself. The moral law is thus what first reveals itself to humans and leads to the concept of transcendental freedom. Crucially, the argument begins with the above stated hypothetical claim. If the aim of the argument is to address the challenge of a naturalist such as David Hume, who denies that humans possess this form of freedom, it would be a clear failure. Read in this way, the argument amounts to nothing more than a dogmatic assertion that humans possess transcendental freedom. For this reason, Kant added the second step in the argument, where he invites his readers to actualize or realize their own transcendental freedom, by generating respect for the moral law. Given that Kant has argued that if I am a transcendentally free rational agent, then I am bound by the moral law, and if I am bound by the moral law, then I am a transcendentally free rational agent (the reciprocity thesis), his deduction of freedom also provides a "kind of credential of the moral law" (KpV 178, 5:48), even if this does not amount to a deduction of the principle of morality.
Newton's second challenge is that my reconstruction of the second step in the deduction of freedom, which establishes the actuality of freedom, seems to
conflict with Kant's claim that the existence of persons is not the realization of an end or concept, but that their existence is in-itself an end. [NFR 46]
If I understand her concern, the worry is that I seem to suggest that humanity is an end in the sense that it is an effect of an action—it is something to be realized. Yet this is clearly not Kant's view. I agree with Newton that when Kant speaks of humanity as an end, he is not using the standard conception of ends as effects of actions. Kant makes this clear, for example, when he says that humanity or rational nature must not be thought of
as an end to be effected but as an independently existing end, and hence thought only negatively, that is, as that which must never be acted against and which must therefore in every volition be estimated never merely as a means but always at the same time as an end.
This passage makes it evident that humanity or rational nature is not, on Kant's view, something that one is to realize or bring into existence. As Christine Korsgaard has argued, Kant thinks that the end of humanity "functions in our deliberations negatively—as something that is not to be acted against." On Kant's view, morality is not a means to any end, not even the end of humanity.
By contrast, Fichte believes that human beings' moral goal is the greatest development of human nature. I have argued elsewhere that Fichte's normative ethics should be seen as a form of moral perfectionism, that is, a moral theory according to which one's good consists in the perfection or full realization of one's essential nature and capacities. On Fichte's version of perfectionism, the perfection of every individual also entails a commitment to help others perfect their own nature. Our moral goal is thus the greatest development of human nature, both in ourselves and in all other individuals. For Fichte, humanity is the end of each human being, in the sense that the greatest development and perfection of human nature is the aim of one's moral actions; it is something one ought to strive to attain or produce. As Michelle Kosch has also argued,
Fichte's ethics...is teleological, insofar as duties are understood in terms of progress toward a positive, substantive moral end that is taken to be partially constitutive of rational agency.
This clearly amounts to a significant departure from Kant's view concerning the sense in which humanity is an end in itself, but it can be seen as consistent with Kant's view that humanity is that for the sake of which humans act. On Kant's view, an end in general is that for the sake of which one acts. In this sense, I agree with Julia Markovits, who argues that Kant conceives ends as
worth-bestowers—as things that bring value to our actions, or...to our other ends.
Kant's conception of humanity as an end is thus understood as that
to which the worth of all actions and conditional ends may be traced back, but whose own worth is conditional on nothing. [MR 93]
Like Newton, Kraus' comments center on the degree of continuity between Kantian and post-Kantian views. Her questions focus on the second chapter of my book, where I argue that,
by emphasizing the regulative role of the ideas of pure reason in Kant's account of empirical cognition, Maimon enables a rereading of the argumentative structure of the first Critique that reveals the relationship between sensibility, understanding, and reason. This rereading brings Kant closer to Maimon and to the post-Kantian German Idealists, showing that Kant has the resources to address Maimon's key challenges. But this reading also puts pressure on Kant's discursive account of human cognition. [LM 56-7]
As I explain in the book,
while it's obviously true that Kant is committed to regulative ideas of reason, it's not obvious that the ideas play any role in his account of empirical cognition. Many of Kant's interpreters believe it's possible to offer a reading of the Transcendental Analytic (where Kant develops his account of empirical cognition) that makes little to no mention of the Transcendental Dialectic (where Kant develops the regulative role of the ideas of pure reason). [LM 56n4]
Kraus largely agrees with my Maimonian rereading of the argumentative structure of the Critique of Pure Reason, which highlights the role of the ideas of reason in Kant's account of empirical cognition and presupposes an identity-in-difference between the human and divine intellects. Yet she argues that, if one reassesses the regulative use of ideas of reason and distinguishes
between a semantic function of ideas in the formation of empirical concepts for human cognition and an epistemic function that projects the ultimate goal of human cognition as it would be grasped by a divine intellect (NKM 49)...both Kant's and Maimon's account of human cognition appear to be even closer in spirit than Nisenbaum is willing to admit, even if not in letter. [NKM 53]
The distinction Kraus makes between semantic and epistemic uses of the ideas of reason in Kant's Critique of Pure Reason seems plausible to me. Yet I do not believe it reduces the distance between Kant and Maimon in the manner that she suggests. More specifically, I do not see how the distinction between semantic and epistemic uses of the ideas of reason reduces the distance between Kant's transcendental idealism and empirical realism and Maimon's rational dogmatism and empirical skepticism. As I argue in the book, Maimon's rational dogmatism leads him to reject Kant's answer to the quid juris question, the question concerning why humans have a right to apply certain a priori concepts (the categories of the understanding) to sensible intuition to arrive at experience, or empirical cognition, of sensible objects. Let me briefly reiterate why.
Like Béatrice Longuenesse, I argue that Kant's answer to the quid juris question relies on the view that,
the unity of space and time is the product of a figurative synthesis of the imagination through which the understanding determines sensibility, it is clearly in accord with the intellectual unity of the understanding. [LM 73]
Kraus helpfully summarizes this point when she writes,
the figurative synthesis of the imagination explains how space and time are unified in accordance with the categories such that whatever is given in space and time will be determinable by the categories, too. [NKM 51]
In this way, Kant's commitment to the transcendental ideality of space and time plays a crucial role in his Transcendental Deduction. Since Kant is not a rational dogmatist, he is also not committed to the view that all aspects of experience are in principle intelligible, and therefore he can accept the idea that space and time are the contingent forms of human sensibility. Yet since Maimon is a rational dogmatist, he is also committed to the view that all aspects of experience are in principle intelligible, and therefore he cannot accept Kant's view that space and time are the contingent forms of human sensibility. This leads Maimon to develop the view that the specific features of space and time can be derived from an absolute or self-evident epistemological principle, the principle of determinability, and to the idea that sensibility and understanding "both flow from one and the same cognitive source." If the forms of sensibility and the categories of the understanding both flow from the same cognitive source, then the question regarding one's right to apply the categories of the understanding to sensible intuition in order to arrive at experience or at empirical cognition becomes irrelevant.
In other words, within the parameters of Maimon's rationalist epistemological framework, the Kantian version of the question quid juris does not even arise. I explain this point in the book as follows:
Maimon attempts to justify the objective validity of the categories by showing that we can derive them from the principle of determinability. Because the principle of determinability is the most general law for the permitted combination of concepts, and because, on Maimon's view, a real object is nothing more than the chain of conceptual determinations that compose it, there is no longer a question about our entitlement to apply the categories to sensuously given objects, since now even matter has been fully conceptualized. [LM 90]
Nonetheless, Maimon does need to explain how, when making a judgment, one knows that two concepts have been brought into a real synthesis. Only if this is the case, is one in possession of actual experience or cognition.
Maimon's answer to this question relies on the view that there is an identity-in-difference between the human finite intellect and the divine intellect. He writes:
We assume an infinite understanding (at least as idea), for which the forms are at the same time objects of thought, or that produces out of itself all possible kinds of connections and relations of things (the ideas). Our understanding is just the same, only in a limited way. [TP 38]
The version of the question quid facti that affects Maimon's view concerns whether the infinite understanding exists, and whether there is in fact an identity-in-difference between the human and divine understanding. As I explain in the book, one must be careful here when speaking of God's existence. On Maimon's view, the claim that God exists means that the idea of an infinite intellect has actuality—that one adopts as an end of one's own the task of the complete explanation of experience, the kind of explanation one would have if the human intellect were infinite (LM 104). This aspect of Maimon's skepticism concerns one's warrant to affirm the condition that would guarantee the correctness of human judgments—namely, the existence of the infinite understanding and the identity-in-difference between the human and divine understanding.
Kraus appears to understand Maimon's version of the question quid facti differently from the above when she asks:
How can one ever be certain that one's actual empirical cognition is in fact true of the perceived object? [NKM 51]
Maimon's answer to this question is that one can be certain if one assumes the identity-in-difference between the human and divine intellects. Yet Maimon's version of the question quid facti would be: but is this assumption warranted? Kraus's distinction between a semantic function of ideas in the formation of empirical concepts for human cognition and an epistemic function that projects the ultimate objective of human cognition as it would be grasped by a divine intellect might help to answer Kant's version of the questions quid juris and quid facti. Yet, I do not see how this distinction would help to answer Maimon's version of these questions.
I appreciate the detailed attention Richard Eldridge gives to the role of existential commitment in the main argument of my book. In what follows, I will focus on two of his three questions on specific points of scholarship. Eldridge's first question concerns my interpretation of Kant's Transcendental Deduction, which is the focus of the chapter on Maimon's legacy and his reading of Kant. The aim of the Deduction is to justify the objective validity of the categories or pure concepts of the understanding. To provide such a justification, Kant would have to show that everything that may become an object of experience for humans must be subsumed under the categories. By "experience," I do not mean the facts of science, but empirical cognition, understood as the activity of determining conceptually what is given to humans in sensibility. In agreement with Ameriks, I present the Deduction as a regressive argument that moves from the assumption that there is empirical knowledge to a proof of the preconditions of that knowledge. Understood in this way, the Deduction has the form of a transcendental argument. Eldridge's objection to my interpretation of the Deduction seems to concern the kind of justification that transcendental arguments provide and the sort of skepticism that they are meant to address. He phrases his worry regarding the conclusion of the argument as follows:
It sounds like an answer to the quid facti: humans do in fact thus represent objects to themselves and describe matters of fact as states of affairs involving substances and their alterations under causal laws. We are compelled to do this. But it seems not to answer the quid juris—namely, are we entitled to do this? Does empirical nature really consist of a quantum of substance, of particulars which always undergo alterations in accordance with a causal law? [ME 41]
Eldridge seems to think that, in order to show that humans are justified or entitled to use the categories or pure concepts of the understanding, Kant must show that,
empirical nature really consists of a quantum of substance, of particulars which always undergo alterations in accordance with a causal law. [ME 41]
My objection to this is that it is unclear what Eldridge means by "nature." Does he mean empirical objects, considered as things in themselves or as mind-independent entities, or does he mean empirical objects, considered as appearances or as entities that are fully mind-dependent? Moreover, if humans are compelled to describe matters of fact as states of affairs involving substances and their alterations under causal laws, why does this fact not amount to a justification of one's use of these concepts? I find it plausible that it does amount to one.
In the third chapter of the book, I explain my views concerning the kind of justification that transcendental arguments provide and the sort of skepticism that they are meant to address. Since Eldridge's dissatisfaction with my interpretation of the Deduction concerns this question of justification, I will briefly summarize my perspective on these two issues. I argue that transcendental arguments address what Robert Stern calls "normativist justificatory skepticism." Normativist justificatory skepticism needs to be distinguished from epistemic skepticism. The epistemic skeptic accuses one of possible ignorance by arguing that humans cannot know with certainty that their beliefs are true. By contrast, the normativist justificatory skeptic accuses one of dogmatism by arguing that nothing entitles humans to hold on to their beliefs. As Robert Stern notes, the normativist justificatory skeptic is in a "dialectically stronger position" (TA 18) than the epistemic skeptic. This is so because one can grant the epistemic skeptic that humans are fallible, that one cannot know with certainty the truth of one's beliefs, yet one can still justify or safeguard those beliefs by showing why one is entitled to them. By contrast, the normativist justificatory skeptic offends one's cognitive self-image by arguing that human beliefs are grounded on nothing but one's dogmatic assertion of them. Stern explains the greater dialectical power of normativist justificatory skepticism as follows:
We may be prepared to admit that we are cognitively limited and hence open to error; it is less easy to grant that we are epistemically irresponsible, governed by caprice, wishful thinking, or habit, rather than reason and rational principle. The justificatory skeptic therefore has a position of much greater dialectical power. [TA 18]
In other words, on Stern's view, and on the view I wish to defend, the aim of a transcendental argument such as Kant's Transcendental Deduction is to show why humans are entitled to hold on to certain beliefs, such as the belief that the natural world is governed by the causal principle. The question quid juris? is a question concerning this sort of justification. Yet this sort of justification still leaves open the possibility that we are cognitively limited—in Kantian terms, it still leaves open the possibility that objects considered as things in themselves might not be governed by the categories of the understanding.
In his influential essay, "The Goal of Transcendental Arguments," Barry Stroud observed that Kant's aim was to infer certain necessary features of the world from the conditions of our thinking about or experiencing the world. Stroud argued that without Kant's transcendental idealism in place, or without invoking a verification principle, transcendental arguments could at most establish that one must believe that the world has the features that make thought, or experience, or meaningful discourse possible. Contemporary analytic philosophers tend to accept Stroud's claim that if one employs transcendental arguments, one must give up on Kant's ambitious aim to reform metaphysics. One must instead limit oneself to the modest aim of revealing relationships of implication amongst one's beliefs by showing that certain beliefs have a special status or position in one's thought that renders them invulnerable to skeptical doubts. Does Eldridge believe that my reading of the Transcendental Deduction results in something like Stroud's modest version of a transcendental argument? If so, he seems to disregard the fact that my reading of the Deduction leaves Kant's transcendental idealism in place.
Eldridge's second question concerns my claim that Kant "fails to account for the possibility of evil as a positive capacity" (LM 186). In answer to this let me first briefly explain what role the discussion of evil plays in the trajectory of the book. As I have mentioned above, I understand the project of post-Kantian German Idealism as the attempt to provide a comprehensive rational explanation of all aspects of human experience, without falling prey to the nihilistic and fatalistic consequences of a commitment to the principle of sufficient reason. In chapter four of my book, I explain why, to meet these two requirements, Fichte develops a monistic metaphysics according to which there is a single fundamental entity, the "absolute I," that is constituted by two forms of activity, namely real and ideal activity. Yet, at least in the Jena period, Fichte does not seem to provide an adequate explanation for the relationship between these two forms of activity, and that seems to be a limitation of his philosophical system. In chapter five of the book, I argue that Schelling's Philosophical Investigations into the Essence of Human Freedom and his The Ages of the World fragments are motivated by his attempt to provide an explanation for both the distinction and relation of dependence between real and ideal activity. Given that Fichte and Schelling hold that the relation between real and ideal activity is another name for the relation between subject and object that characterizes all states of human consciousness, these two works offer an answer to the question that Schelling considered to be the central question of philosophy: "Why is there a realm of experience at all?" To answer this question, both Schelling and Rosenzweig develop the view that human experience is grounded in three irreducible elements—God, the natural world, and human beings, which relate to one another in three temporal dimensions: creation, revelation, and redemption.
The discussion of evil is part of this account, since Schelling holds that human beings determine or individuate themselves in relation to or in opposition to the divine person, so evil can be understood as a form of defiance and goodness as a form of love; and the existence of the finite world is the condition of possibility for human individuation.
I turn now to Eldridge's question concerning my claim that Kant "fails to account for the possibility of evil as a positive capacity" (LM 186). Eldridge notes that in Religion within the Boundaries of Mere Reason, Kant explains how he conceives the volitional structure of moral evil, and he seems to provide an account of its possibility as a positive choice: moral evil has its source in the manner in which a person incorporates the moral law and integrates the principle of self-love into one's maxims. While a morally good person subjects to the moral law the incentives of their sensuous nature; the evil person does the reverse and regards these incentives
as of themselves sufficient for the determination of [a human being's] power of choice, without minding the moral law.
Some of Kant's early and contemporary readers believe that the Wille/Willkür distinction that Kant introduces in Religion within the Boundaries of Mere Reason (and that he makes more explicit in the Metaphysics of Morals) explains how evil can be conceived as a positive choice. I disagree with this view. Wille is practical reason, it legislates the moral law. Willkür is the power of free choice which forms the maxims that guide one's actions. It might seem like this distinction makes it easier to see how an action can be freely chosen whilst not being in conformity with the moral law (Willkür is capable of forming maxims that prioritize the incentives of the agent's sensible nature over the moral law's demands). Yet if one goes down this road, one could not explain how the moral law is categorical or unconditionally binding, which is a core Kantian commitment. Let me explain why.
I want to say that moral evil can be conceived as a free choice. If one asks in virtue of what evil or immoral action is free, the person who relies on the Wille/Willkür distinction will say that it is in virtue of the freedom of Willkür (it cannot be in virtue of Wille, for Wille legislates the moral law). But this stance implies that the law of Wille is not Willkür's own law, that the moral law is not in fact self-legislated by Willkür. Yet I need to hold on to the claim that the moral law is self-legislated by Willkür to explain why the moral law is categorical or unconditionally binding (which is what Kant wants to say). Another way to put this is that if one relies on the Wille/Willkür distinction to justify how free immoral action is possible, one effectively ends up introducing something like the heteronomous structure of divine command theory within the agent. Wille legislates the moral law, but whether one's Willkür obeys the law depends on its interest in upholding the authority of Wille. This means that the moral law is not in fact self-legislated by the agent.
I find Eldridge's proposal appealing that moral evil involves a kind of mistake concerning what the power of choice is for, and I would love to find an opportunity to discuss his view at greater length. In one of my current projects, I argue that Kant conceives moral action as the actualization of human beings' rational capacity to will, and in that way practically cognize, the good.
