Problems in Kant’s Vindication of Pure Reason

Ted Kinnaman. Journal of the History of Philosophy. Volume 39, Issue 4. October 2001.

One of the most important questions in interpreting the Critique of Pure Reason concerns the proper way of characterizing Kant’s view of the faculty of reason. Clearly, one of Kant’s intentions is to show that reason is incapable of cognition of objects such as God or the soul, because such cognition would require going beyond possible experience. At the same time, he also wants to defend reason, to “institute a court of justice, by which reason may secure its rightful claims while dismissing all its groundless pretensions” (Axi-xii). Denied the possibility of attaining theoretical knowledge, reason instead finds its fulfillment in its practical employment, striving to bring the world into conformity with the moral law. This, however, can make it appear that the task of guiding action is one that reason must merely settle for, having unfortunately been denied its true goal, knowledge of the supersensible. It is thus difficult to see how Kant can claim to have vindicated reason.

Recently, a number of scholars have suggested that a proper understanding of Kant’s account of reason can correct this purely negative view of Kant’s theory. In particular, Richard Velkley and Susan Neiman have argued separately that Kant’s emphasis on the goal-directed nature of reason’s task helps to justify Kant’s belief that the first Critique defends as well as limits reason.’ Because this reading truly does justice to Kant’s insistence on the unity of theoretical and practical reason, it goes a long way toward bringing out Kant’s overall project in the Critique of Pure Reason. But I will argue that Velkley’s and Neiman’s defense of Kant’s conception of reason, while persuasive as a reading of Kant’s intentions in the first Critique and a clear improvement over a simply negative view of reason, is ultimately unsuccessful philosophically. I will show that Kant’s defense of reason in the Critique of Pure Reason is vulnerable to two strong objections, of which Kant himself was made aware of in the years after the publication of the Critique. We have no grounds, on the basis of the argument of the first Critique, to think that the goals reason sets for us can be achieved; nor does Kant’s theory provide a basis for defending his conception of reason against other plausible and coherent sources of warranted assertion, such as religion or culture. Indeed, Kant’s appreciation of the seriousness of the deficiency of his defense of reason in the Critique of Pure Reason helps to explain the importance of the concept of the purposiveness of nature in the later Critique of Judgment. First I explain in detail why it appears that Kant’s analysis of reason produces only a negative result-that reason cannot attain knowledge of its objects. Second, I will look at how the readings of Velkley and Neiman attempt to defend Kant’s account of reason’s positive use. I will then argue that without some argument for the suitability of appearances to systematization according to ideas of pure reason, we can be assured neither of the possibility of reaching the goals reason sets for us nor of the superiority of Kant’s views over competing views on morality and religion. In particular, I will look briefly at the writings of Johann Georg Hamann and Johann Gottfried Herder, with the intention of showing that they offer at least prima facie plausible alternatives to Kant’s view of reason.

1.

The central problem concerning Kant’s conception of reason in the Critique of Pure Reason results from the conjunction of two of his most important philosophical claims. First, it clearly is crucial to Kant’s philosophical outlook that “we cannot cognize any object that is thought except through intuitions that correspond to those concepts” (B165). But, second, Kant says that unlike the understanding reason does not relate directly to intuition, and thus is not concerned with determining intuitions according to concepts. Instead, reason aims at the systematization of the cognitions produced by the understanding, the goal being a hierarchy organized around an idea of pure reason, which is defined as a “necessary concept of reason … to which no congruent object can be given in the senses” (A327/B383). And because they do not represent objects of possible experience, ideas cannot be given an “objective deduction” of the sort offered for the categories (A336/B393); Kant describes their use as being merely “regulative” rather than constitutive. It is its tendency to consideration of supersensible objects that leads reason ineluctably into antinomy. If it is true that we have cognition only insofar as we have intuitions corresponding to our concepts, and reason necessarily is concerned with concepts for which no corresponding intuition can be given, then it seems that the main lesson to be drawn from the Critique of Pure Reason is the negative one that reason is incapable of producing cognition, and to one sympathetic to the Enlightenment Kant will indeed appear, as he did to Moses Mendelssohn, as an “all-destroyer.”

If the belief that reason applies directly to objects leads to antinomial conflict, then to the extent that Kant is to give a positive account of the regulative use of reason, this account must center on reason’s task of systematizing cognition. This task in fact has three aspects, speculative, practical, and what for want of a better term might be called philosophical. In its speculative use, reason seeks the systematic unification of empirical laws discovered by the understanding, with the goal being a hierarchy of laws organized around one idea of pure reason from which the various more specific laws can be deduced. The pursuit of such a system constitutes the heart of scientific inquiry as Kant understands it; more strongly, he suggests at one point that without such unification “we would have no reason, and without that, no coherent use of the understanding, and, lacking that, no sufficient mark of empirical truth” (A651 /13679). In addition to its speculative use, reason also has a practical use. Here too we find reason using ideas of supersensible objects to create systematic order, but in this case the order that is sought is a moral order, involving a coincidence of reality and moral commands-of the world as it is with the world as it ought to be. Here there are actually two ideas involved: the idea of a “moral world,” which is simply the representation of a world in which everyone acts in perfect accordance with duty (A8o8/836); and the “ideal of the highest good,” which is a representation of a world in which moral virtue is rewarded perfectly with happiness (AS10/B838). Kant thinks that we can only conceive of the ideal of the highest good as the result of a “highest reason, which commands in accordance with moral laws,” i.e., God (A810/B838). Finally, reason’s interest in systematic unity appears again in Kant’s account of the task of the philosopher in the “Architectonic of Pure Reason.” The system in question here is a philosophical system containing both the speculative and practical systems, with the former subordinated to the latter. Similar to the speculative and practical systems, this broader system represents “the unity of the manifold cognitions under an idea”; this idea is

the rational concept of the form of a whole, insofar as through this the domain of the manifold as well as the position of the parts with respect to each other is determined a priori. The scientific rational concept thus contains the end and the form of the whole that is congruent with it. (A832 /B860)

The philosopher, too, understood as a “legislator of human reason” who assigns tasks to scientists, logicians, and mathematicians in order to “advance the essential ends of human reason,” is “still found nowhere” and thus is an ideal (A839/8867).

Now although the relation of ideas to sensation is merely indirect, they are nevertheless a priori concepts, and as such are, in their own way, both necessary and universal. They are universal simply by virtue of being principles of reason, and they are necessary insofar as they guide our investigation of nature rather than arising from it. And, says Kant, we cannot reliably use any a priori concept without having a transcendental deduction of it (A669/B697). Such a deduction cannot of course show that the ideas of pure reason are conditions of the possibility of experience, as the argument of the Transcendental Analytic showed for the categories of the understanding, but it must nevertheless vindicate the reasonableness of pursuing systematic unification of cognitions through the use of ideas. Kant makes two different attempts to argue for the ideas, one in connection with their speculative use and one in connection with their practical use. In discussing the speculative use of the ideas, Kant argues that, since ideas have a merely regulative rather than constitutive function, it suffices to show that their use helps us in our investigation and (as long as we guard against transcendental illusion) never leads us into contradiction:

But if one can show that … all rules of the empirical use of reason under the presupposition of such an object in the idea lead to systematic unity, always extending the cognition of experience but never going contrary to experience, then it is a necessary maxim of reason to proceed in accordance with such ideas. And this is the transcendental deduction of all ideas of speculative reason … [.] (A67 1 /B699)

This “deduction” consists in the entire argument of the Appendix to the Transcendental Dialectic, where Kant explains how the ideas aid scientific inquiry without being taken for constitutive principles. Clearly, no argument of this sort can show that these principles are a priori, for it seems to be just the sort of empirical argument (deriving as it does from claims about the practice of science) that Kant says in a number of places cannot support claims of necessity or universality.5 Of course, this may be the strongest justification that ideas are capable of receiving, compatibly with the distinction between understanding and reason and the crucially important claim that our knowledge is limited to experience.

In introducing the question of the justification of the practical use of ideas of reason in “On the Ideal of the Highest Good,” Kant clearly gives the impression that this is a sort of substitute for what reason could not have in the speculative sphere:

In its speculative use reason led us through the field of experiences, and, since it could never find complete satisfaction for itself there, it led us on from there to speculative ideas… Now yet another experiment remains open to us: namely, whether pure reason is also to be found in practical use, whether in that use it leads us to the ideas that attain the highest ends of pure reason which we have just adduced, and thus whether from the point of view of its practical interest reason may not be able to guarantee that which in regard to its speculative interest it entirely refuses us. (A804/B833)

While this last question may seem to suggest that Kant is raising the possibility that ideas in their practical use have objective validity,6 in fact it turns out that what is at stake is whether ideas have objective reality (A808/B836), that is, “the possibility of such an object as is thought through the concept” (A220/ B268). Although the problem Kant faces with regard to ideas of practical reason is basically analogous to that faced with regard to speculative ideas, his argumentative strategy is quite different. Here Kant argues that the objective reality of principles of pure practical reason is a necessary condition of the exercise of practical reason. The presupposition that there are such principles-that is “pure moral laws that determine completely a priori … the action and omission, i.e., the use of the freedom of a rational being in general”-can be justified by appeal to “the moral judgment of every human being, if he will distinctly think such a law” (A807/B835). These principles are thus in a sense principles of the possibility of experience, in so far as they represent actions that “could be encountered … in the history of the human being.” Kant supports this surprising assertion with the observation that “since they [the principles of pure practical reason] command that these actions ought to happen, they must also be able to happen” (A807/8835). In short, apparently, ought implies can.7 The same principle plays a crucial role as well in the argument for the ideal of the highest good. The idea of the moral world yields the moral imperative “Do that through which you will become worthy to be happy” (MM/8836). But in order for it to be rational to follow this imperative, Kant thinks, it is also necessary to believe that acting according to the moral law will in fact lead to happiness. Kant concludes that

just as the moral principles are necessary in accordance with reason in its practical use, it is equally necessary to assume in accordance with reason in its theoretical use that everyone has cause to hope for happiness in the same measure as he has made himself worthy of it in his conduct, and that the system of morality is therefore inseparably combined with the system of happiness, though only in the idea of pure reason. (A80g/ B838)

Thus both the idea of the moral world and the ideal of the highest good have objective reality, albeit in a very different sense from that in which the categories of the understanding have objective reality, insofar as it is limited to reason in its practical use. Nevertheless, as Kant makes clear, these principles of reason’s practical use have consequences for reason’s theoretical use, providing support for the ideas of God (as the cause of the harmony between the “system of morality” and the “system of happiness”) and the immortality of the soul (as necessary for representing the rewards of good action) that could not be found from a theoretical point of view.

Kant’s arguments, then, for the objective reality of both the idea of the moral world and for the ideal of the highest good rely on the ‘ought implies can’ principle: In order for practical reason to pursue its ends, the world must be such as to allow these ends to be achieved.8 This principle as it is used here can be understood in two ways. On the one hand, we might take Kant to be asserting that the independent determination that one has a duty to perform an action entails one’s ability to perform this action. This reading is I think the one most obviously suggested by the text of “On the Ideal of the Highest Good,” and it also allows him to accomplish what he needs, which is to show that the ends reason prescribes can be reached. But it has the disadvantages that it contradicts the normal construal of “ought implies can,” according to which “can” is a necessary condition for “ought,” and that it saddles Kant with an extremely implausible claim. If ought implies can, after all, and we can determine our duties without first determining whether we are able to perform them, then from the awareness that I ought to save a drowning man I could learn that I can swim! More plausibly, we can read Kant as making a weaker assertion about the relation between “ought” and “can,” roughly that the recognition that one has a duty to perform an action creates an “interest of reason” in becoming able to perform that action, or perhaps in believing that one is capable of performing it.9 But if Kant is not making an assertion about what is actually the case in the world, then the fundamental problem remains: We cannot know that we can do what reason commands us to do, and thus (on the more common interpretation of “ought implies can”) is it possible that what we think are moral duties will turn out not to be such. In short, in either way that we read Kant’s argument in “On the Ideal of the Highest Good,” he lacks a convincing argument that reason’s goals can be achieved.

It is easy, I think, to see how someone reading the first Critique might view the practical employment of reason as merely a poor substitute for the knowledge denied it in the theoretical realm. The central principle of Kant’s theory of knowledge is that “without intuition all of our cognition would lack objects, and therefore remain completely empty” (A62/1187). But ideas, whether in their speculative or practical use, represent objects that cannot possibly be given intuition, and thus these ideas cannot constitute cognition. In providing the “transcendental deduction” of the ideas of pure theoretical reason, Kant does show a certain relation between these ideas and intuition, insofar as they assist in scientific inquiry. Ideas of pure practical reason are absolutely necessary, but only from the practical point of view. This means that it may be necessary for the exercise of moral reason to believe that, for example, good conduct is rewarded with happiness, but strictly speaking it does nothing to show that it is true that good conduct is rewarded with happiness. If one takes reason to be concerned with the production of knowledge, or more generally of justified assertability, then Kant’s theory must be seen as offering an overwhelmingly negative view of reason. It is in response to this expectation that Velkley and Neiman offer their interpretations of Kant’s account of reason.

2.

In his book Freedom and the End of Reason, Velkley argues that the claim that reason is not in “immediate relation” to any object is central not only to the argument of the Appendix to the Transcendental Dialectic in the Critique of Pure Reason but to Kant’s entire critical enterprise. Kant was impressed, he says, with Rousseau’s argument that the development of reason had led not to greater human happiness and moral virtue but instead to misery and vice. Reason, as slave of the passions, had served only to sharpen the divisions between individual rational human beings, each bent on pursuing their own personal happiness. This is what Velkley calls the “teleological problem of the end of reason,” the question whether reason can demonstrate its own goodness, its own usefulness for promoting human happiness. Specifically, Velkley sees Rousseau as raising three questions for Kant: First, whether it is possible to give a self-consistent account of reason; second, whether reason contributes to human happiness and virtue; and third, whether philosophy serves to promote the “emancipation and enlightenment” of humanity.

As Kant (on Velkley’s reading) saw it, the central mistake of both Rousseau and the Enlightenment thinkers he criticized was their reliance on an instrumental conception of reason, that is, that they assumed that reason is a mere means to achieving human happiness. As an alternative, Kant tried to develop an account of reason according to which the ends of reason-the goals of both our moral and theoretical striving-are determined not by the passions and hence by nature, but rather by reason itself. He wanted, in other words, an account and defense of the autonomy of reason, both speculative and practical, and this is just what Velkley reads Kant as offering in the Critique of Pure Reason.

Kant’s doctrine of the autonomy of reason is of course familiar from his works in moral philosophy. Kant’s view is that the only valid moral laws are ones that reason gives itself, for otherwise its commandments will tell us not that we have an unconditional duty to do x but only that it would be prudent to do x if we want to achieve a certain end given not by moral reason but by nature. In Kant’s own terms, such imperatives would be merely hypothetical rather than categorical. Velkley’s contribution is that he moves this emphasis on the autonomy of reason to the very center of Kant’s critical project: Rather than seeing theoretical and practical philosophy as distinct but somehow rather awkwardly linked parts of the critical system, he argues that even the foundational arguments of the first Critique are part of an attempt to free reason from mere subservience to ends given it by nature. By denying that reason has any direct application to objects of experience, Kant is, on Velkley’s reading, “removing … obstacles that stand in the way of reason’s grasp of its true essence,” namely that it is “a self-determining power that alone is responsible for the projection and achievement of humanity’s final goal.” The goal in question here is the ideal of the highest good, which represents the unification of speculative and practical reason. Thus Kant’s response to Rousseau’s criticism of Enlightenment reason (to return to the three specific aspects of this criticism) is, first, that reason’s autonomy provides us with a self-consistent view of reason’s task; second, that by understanding reason’s relation to the ideal of the highest good as teleological rather than epistemic (pointing us toward this ideal rather than producing knowledge of it) Kant grounds an account of the autonomy of reason that avoids the problem of reason’s ability to produce happiness; and finally, because it is the task of the philosopher to outline this system, Kant’s account assures the importance of philosophy for culture.

Velkley’s principal interest in Freedom and the End of Reason is in defending a reading of Kant’s philosophy rather than in arguing that this philosophy achieves its goal. As such, I think it is quite convincing. Seeing the accounts of theoretical and practical reason as linked in a project of vindicating human autonomy helps to explain the relation between the first and second Critiques, as well as that between the Doctrine of Elements and the Doctrine of Method in the first Critique, in a way that makes clear how they fit together in one coherent philosophical project. Furthermore, emphasizing the way in which Kant’s philosophy is intended as a response to Rousseau allows us to see Kant not only as limiting reason but also, as he clearly intended, as defending it. What I am interested in is not the interpretative question of Kant’s intention in formulating the Critique, but rather the philosophical question of the extent to which Kant’s defense of reason is a convincing one. As I will argue later, the chief difficulty is that Kant can offer no reason for thinking that the goals reason sets for us can be achieved, or that reason as Kant construes it can be defended in comparison with other possible views on the basis of morality and religion.

Like Velkley, Neiman in her book The Unity of Reason argues that Kant must be seen as offering a teleological conception of reason. As Neiman reads Kant, the account of the regulative use of the ideas of pure reason in the Appendix to the Transcendental Dialectic is part of Kant’s broader project of replacing the traditional view of reason, according to which reason is concerned primarily with knowledge, with his own theory of the unity of theoretical and practical reason-what is known as the doctrine of the “primacy of the practical.” This involves reconceiving both the theoretical and the practical use of reason: in the former case, because reason’s task is to guide us toward the goal of systematic unification of cognition, a task that it fulfills not through speculation but through practice; in the latter case, because “the practical is as thoroughly rational as the rational is practical.”

The doctrine of the unity of theoretical and practical reason is intended, Neiman argues, to circumvent the questions of the limits of reason’s ability to know the world and of its ability to guide human action. In both connections, the crucial point is that reason’s task is to “propose ends,” whereas the understanding is concerned with “the simple recording of what is. “ It makes sense to ask whether the understanding is capable of knowledge of the world, a question to which Kant gives a positive answer in the Transcendental Analytic. Kant’s conception of reason, on the other hand, avoids this question. Neiman explains: An end is necessarily something beyond that which is already given. In proposing ends, therefore, reason declares its right to make demands on experience in a manner forbidden to the understanding. The concepts of the understanding give order to experience; the principles of reason are the standard by which it is judged. In a conflict between the two, it is not the principles which require revision, but experience which is inadequate to the principles of reason.

Nevertheless, Kant requires that all philosophically legitimate use of concepts involve some relation to experience. It is for this reason that Kant introduces the notion of a regulative principle, which “ensure [s] that reason’s activities are grounded in the world without being determined by it.”

Neiman makes it clear that her interpretation of Kant’s account of reason is also a defense of that account. Specifically, she thinks that if Kant’s view is suitably understood, then questions about reason’s authority over human life will appear as irrelevant or beside the point.  In relation to the role of ideas in speculative reason, for example, Neiman argues that we ought not to see ideas as “failed constitutive principles” because “the questions they are intended to settle are not questions about the world but about the behavior of human beings in the world.” What is needed to justify the use of these regulative principles is thus only an argument that science is impossible without them, which is what she takes Kant to be offering in the Appendix to the Transcendental Dialectic. With regard to practical reason, the case is more complicated. The justification of the moral law as such is made simpler by the fact that (in the doctrine of the “fact of reason”) Kant does not undertake to defend the moral law against moral skeptics, nor against someone who thinks that moral laws are not universal. Given this limitation, Neiman takes Kant’s justification of the categorical imperative to be coherentist, involving a demonstration that the various components of Kant’s moral theory (e.g., transcendental freedom, the different formulations of the categorical imperative, etc.) mutually entail each other. This having been demonstrated, there remains only the problem of whether the goals moral reason sets for us can be reached. This problem is easily solved, Neiman says, because it deals only with a fact about ourselves. As she explains in a slightly different context:

Practical reason’s objects… are breathtakingly simple to realize. For human nature is our nature; the barriers to realizing the ideas of practical reason are internal ones. The fact that the good will of all rational beings is all that would be needed to establish peace on earth tomorrow is the fact to which Kant points in proclaiming the primacy of practical reason. The distinction between theoretical and practical reason is meant to underline how much lies in our hands alone.

The relative ease with which moral goals may be realized is emphasized in Kant’s discussions of moral education, both in the Critique of Practical Reason and in the Metaphysics of Morals, which are intended to show that proper understanding of examples of noble human behavior helps each of us to recognize our own capacity for such behavior.

But beyond the issue of the validity of the moral law-the question “What ought I to do? “-there is the further question of the extent to which it is rational to hope that the ideal of the highest good can be realized in this world. As Neiman points out, the ideal of the highest good has been viewed by many commentators as being incompatible with the heart of Kant’s ethical doctrine, which flatly rejects all teleological ethical conceptions. It is possible to avoid this conclusion, she says, if we look on the categorical imperative as a ground of action, and take the ideal of the highest as an end of action “constructed by reason itself. As I will argue in the detail in the next section, however, if the ideal of the highest good is constructed by reason, and reason is concerned with how the world ought to be rather than with how it is, then we are still left with a deep problem of what grounds one might have for believing in a systematic relation between virtue and happiness. Kant makes it clear that the ideal of the highest good plays a role in both the practical and theoretical uses of reason. Neiman’s response, on the other hand, seems to involve denying this connection:

Nothing else in Kant’s work expresses so clearly the gap that he believes separates reason from nature. Morality, the product of pure practical reason, is free of all natural conditions, happiness is wholly dependent on the natural world. Our desire to become the authors of our own happiness is a desire to overcome that separation. But despite suggestions to the contrary, Kant’s notion of the Highest Good is riot a means by which to do so.

This reflects the general approach of both Neiman and Velkley to the problem of reason’s relation to reality. I think it is very persuasive as a reading of Kant’s understanding of his own view as expressed in the Critique of Pure Reason. But it may solve the problem only by postponing it to another level.

3.

Aside from differences in emphasis, the interpretations of Velkley and Neiman share the thesis that Kant’s account of reason can be made convincing by understanding reason as concerned not with knowledge of what is but rather with pursuit of what ought to be. Seeing reason as a faculty for setting goals has the advantage of rendering insignificant worries about its inability to produce knowledge, but it opens the door to a new worry: If the philosophical legitimacy26 of the ideas of pure reason depends on their role in the systematization of experience, and this systematization is viewed as a goal to be pursued, then the question of their philosophical legitimacy becomes the question of whether there can be a justification for pursuing the goal of systematic unity among cognitions. The question here is not whether it is rational to pursue systematization, which would be, so to speak, an internal question concerning Kant’s theory, but rather whether reason’s goals are good ones to adopt. This can be broken down into two questions: First, is it possible (in some sense of ‘possible’) to attain the goals reason sets? Second, even if the goals of Kantian reason are attainable, what grounds can Kant offer for allowing reason rather than say, tradition or religious authority to set goals for human life?

The first question is one the difficulty of which Kant himself underestimated, at least in the Critique of Pure Reason. As we have already seen, in discussing the possibility of a “transcendental deduction” for the ideas of pure speculative reason, Kant suggests that it is sufficient for establishing our right to use these ideas as a priori regulative principles to show that the ideas “lead to systematic unity, always extending the cognition of experience but never going contrary to experience” (A671/13699). I think that Kant’s confidence that use of the ideas will never “go contrary to experience,” and thus that the goals they represent can be attained, stems from his holding the view, in line with the reading offered by Velkley and Neiman, that because the ideas are regulative rather than constitutive principles their use cannot come into conflict with the understanding. But even once we recognize that reason is more concerned with setting goals than with knowing facts, Kant’s defense of the ideas in both the speculative and the practical context still runs afoul of the fact that the possibility of attaining reason’s goal of systematic unification entails a substantive and possibly false claim about experience, or more specifically about appearances, namely that they are susceptible of such unification. The property of appearances which renders them capable of such systematization Kant calls their “purposiveness” [Zweckmassigkeit] (A686/11714); the concept of the purposiveness of appearances thus assumes a role in relation to the deduction of the ideas that the “affinity of appearances” has in relation to the Deduction of the Categories, as that property of appearances that renders them susceptible of subsumption under a concept (Al 22). Just as it is conceivable that appearances might be so disorderly-cinnabar might appear sometimes red, sometimes black, and so on that it is impossible to apply concepts to them, so it is also possible that appearances might be so disparate as to render impossible reason’s task of ordering them into a coherent system (A653-4/B681-2).

The possibility that appearances cannot be systematized threatens the employment of reason in all its manifestations. In relation to speculative reason’s goal of unifying scientific cognition, this possibility corresponds to the possibility that the discourses of various scientific disciplines might be incommensurable, that is, incapable of being expressed in neutral terms common to each. In relation to practical reason’s goal of creating a world in which everyone does what she ought, there are three respects in which the possibility of systematization depends on the purposiveness of appearances. First, human beings might be constitutionally unable to live in perfect accordance with the moral law.

Second, even if human beings are capable of following the commands of moral reason, it is possible that the nonhuman world is such that it thwarts our efforts to create a moral world.30 Third, even if it is possible to create a world in which everyone acts according to duty, this virtue may not be rewarded with happiness as the ideal of the highest good postulates. And finally, if there are doubts about the possibility of achieving the goals set respectively by speculative and practical reason, then the overarching philosophical system, elaborated in the “Architectonic” and elsewhere, is undermined as well.

It is therefore crucial to Kant’s account of reason that we have some reason to believe that appearances are purposive, that is, that reason’s goal of systematization can be achieved. But any attempt to argue for the claim that appearances are purposive faces a serious dilemma. This claim must be either analytic, synthetic and a posteriori, or synthetic and a priori. It is not analytic, for certainly there is no contradiction entailed by the claim that appearances cannot be systematized. It must therefore be synthetic, and either a priori or a posteriori. If the purposiveness of appearances were synthetic a priori, then their justification would have to take the form of a transcendental deduction of the sort given for the categories of the understanding, arguing that their objective validity is a condition of the possibility of experience. But this is out of the question, for this would make constitutive principles of the ideas of pure reasons, whereas it is the point of Kant’s argument in the Appendix to the Transcendental Dialectic to show that they are (merely) regulative. There remains therefore only the possibility that the purposiveness of appearances is a synthetic a posteriori proposition, and requires empirical support from observation of nature. But clearly it cannot be a posteriori, because the ideas of reason are a priori, and thus cannot admit of empirical support.

Now, it should be noted that later, in the Critique of Judgment, Kant seems to countenance a posteriori evidence for a priori claims of purposiveness. This is apparent, for example, in his discussion of the partial contingency of the systematic unity of empirical laws, expressed in the “principle of the purposiveness of nature” with which the work is essentially concerned. While it is necessary for reflective judgment, in its investigation of nature, to take the systematicity of nature as a “subjective principle (maxim),” still “we rejoice (actually we are relieved of a need) when, just as if it were a lucky chance favoring our aim, we do find such systematic unity among merely empirical laws” (V:184). This is because even though the systematicity of nature is a necessary principle of judgment, it is nevertheless to be regarded as “contingent objectively” (V:185). Similarly, in the “Critique of Aesthetic Judgment,” although the judgment of taste “rests on a priori bases” insofar as its determining ground is the suitability of a presentation for “cognition in general,” (V:221-2) these judgments are singular rather than universal. This is because the beauty of a particular object can only be judged by “hold[ing] the object directly up to my feeling of pleasure and displeasure,” and cannot be derived from a rule saying, for example, that all roses are beautiful (V:215). But while in these cases it appears that we discover contingent evidence for an a priori proposition regarding purposiveness in nature, they are the subject of Kant’s critique precisely because, as a priori, they require a transcendental grounding in the human cognitive powers. And in any case this line of thought is not pursued in the first Critique. The result is that, at least in the Critique of Pure Reason, Kant cannot provide a convincing argument for the claim that reason’s goals can be achieved, leaving open the possibility that the ideas merely represent, so to speak, figments of reason’s imagination. This is what Kant’s contemporary Thomas Wizenmann had in mind when he compared reason (as Kant conceives it) with a man who fell so thoroughly in love with his imagined beloved that he came to believe she must really exist. And without a good response to Wizenmann, Kant lacks the needed grounding for the ideas of pure reason, and thus for his defense of reason in general.

Since Neiman puts more effort than Velkley into arguing for the success of Kant’s defense of reason, it instructive to look to her work for a possible response to this criticism on Kant’s behalf. Although Unity of Reason is divided into considerations of the various aspects of reason’s activity-in science, morality, religion, and so on-a number of passages in different contexts bear on the question of the possibility of achieving reason’s goal of systematic unification. I have already mentioned Neiman’s claim, following Rawls, that Kant aims at giving a “self-authenticating” or coherentist account of reason, and therefore regulative principles, being “directives for behavior rather than items of knowledge … are not the sorts of thing for which objective justification is coherent or desirable.” But I am arguing that because the possibility of achieving systematic unity among cognitions depends on the truth of a contingent claim about the world, namely that appearances are purposive, an argument for this claim, and thus for reason, is not only desirable but necessary. Any number of conceptions of reason, or more generally of the basis of warranted belief, whether in regard to morality, science, or religion, might be equally coherent and yet quite incompatible with Kant’s view. The failure to argue for the purposiveness of appearances thus leaves Kant unable to defend his account of reason against these other views. A second relevant passage in Unity of Reason has to do with the postulates of pure practical reason as the basis of “rational faith.” Neiman considers Wizenmann’s objection that this seems to encourage us to believe in anything that we really want to exist. In response, she notes Kant’s claim that the postulates are based on a “need of reason,” which rules out the possibility that one could be justified, on Kant’s theory, in believing in the existence of anything not required for the exercise of practical reason. But as Neiman correctly notes, this defense “seems to rule out the possibility of individual fantasy but not that of mass delusion.” Her answer is to appeal again to the notion that “rational faith, like other regulative principles, does not concern truth conditions at all. But again, if I am right that the legitimacy of the ideas of pure reason depends on the purposiveness of appearances, this is simply not true. If ought implies can-if we can be morally obligated to perform only those actions we are capable of performing-then the possibility that reason’s goals cannot be achieved threatens Kant’s entire account of practical reason. Finally, in discussing the ideal of the highest good and the question ‘What may I hope?,’ Neiman again denies that Kant’s view in any way depends on a claim about the world.

Kant does not argue for the claim that reason has a need to see happiness apportioned to virtue, nor is it easy to imagine how he might do so; his strategy is merely to show that it is this assumption that underlies naive moral intuition as well as most of religious and philosophical tradition. His claim is not that something is (let alone must be) the case in the world we know but that reason has a need for it to become so. At issue is not a statement about the world but a demand that we make upon it.

Viewing the ideal of the highest good in this way as a construction reflecting only the need of practical reason to represent the state of its fulfillment helps to avoid attributing to Kant the claim that something is because it ought to be. But if this is all that the ideal is-if there are no rational grounds for believing not only that the ideal ought to be but that it can be realized-then Kant has failed in his stated goal of unifying practical and speculative reason.

The second question about Kant’s defense of reason concerns his ability to justify adopting as the leading force in human life reason as he conceives of it rather than some other source of authority. This question requires answering both because Kant’s view has such momentous consequences for human life, and because there were and are quite plausible alternatives to this account. To appreciate the former point, consider the truly momentous and controversial nature of the view of morality Kant is proposing. First, it is clearly Kant’s view that moral laws are universal. This is a consequence of the fact that for Kant the moral law is derived from reason, and thus describes the nature of obligation for every rational being as such. Second, Kant holds the view that in and of itself human happiness is of no moral significance. Admittedly, this is clearer in the Critique of Practical Reason and in the Groundwork of the Metaphysics of Morals than it is in the Critique of Pure Reason, for in the “Ideal of the Highest Good” Kant formulates the moral law as the commandment “Do that through which you will become worthy to be happy” (A80g/ B837). This contains a reference to human happiness, but as something that is hoped for rather than as constitutive of moral obligation. Finally, and most importantly, the very autonomy of reason so important to Velkley’s and Neiman’s readings has the quite radical consequence that the moral law derives its validity entirely from reason, and thus not from religious or secular authority, divine commandment, or tradition. Despite the important role of the idea of God in Kant’s description of the various uses of reason, religious thinkers suspicious of the Enlightenment were right to see in Kant’s theory a threat to the sovereignty of religion over human life.

Given the controversial nature of his account of reason, it is not surprising that in the period immediately following the publication of the Critique of Pure Reason Kant was confronted with quite plausible opposing views, specifically those of J. G. Hamann and J. G. Herder. Both Hamann and Herder reject Kant’s attempt to establish a timeless and universal basis for morality and religion, and endorse instead an account of both as historically contingent. Furthermore, each view is intended as a critical response to Kant’s theory, and was understood by him as such. It is impossible for me to do more here than simply to show that each of these thinkers offers a plausible view incompatible with Kant’s, and thus that Kant was aware of such alternatives in the period immediately after the publication of the first Critique. But even a brief examination of their writings will make it clear that Kant’s inability to give a good argument for the purposiveness of nature leaves him without a convincing rejoinder to these alternative views.

Hamann, a near-contemporary of Kant and fellow Konigsberger, spent much of his intellectual career opposing the Aufklarung in general and Kant in particular. His review of the Critique of Pure Reason was the first to appear in print, and his small essay of 1784, “Metakritik fiber den Purismus der Vernunft,” though unpublished, exerted a strong influence on German thought. The essay’s thesis is that “language … is the center of reason’s misunderstanding with itself.” Specifically, whereas Kant imagines that “the universal character of a philosophical language, hitherto sought, is already found,” Hamann emphasizes the necessary dependence of philosophical language on natural language. What he means by this is that it is impossible to invent a purely rational language, one that is equally and immediately accessible to any rational creature; rather, any such “philosophical language” must derive from, and be translatable into, some natural language, which is of necessity historically and geographically bounded. But, one might reasonably ask, why can’t one invent a rational language, simply defining one’s terms carefully to avoid confusion? A close examination of the text of the Metacritique shows that Hamann’s real point is not that such a language cannot be created, but that it can be assured of its connection to reality only by the possibility of being translated into a natural language. The essay opens with a reference to Berkeley’s rejection of abstract ideas, which Hamann apparently endorses:

[A]ll general terms are nothing but particular ones, annexed to a certain term, which gives them a more extensive signification, and which makes them recall upon occasion other individuals, which are similar to them.

Berkeley’s point here is that, contrary to what Locke had claimed, abstract ideas are not needed in order to explain the use of language. More broadly, the rejection of abstract ideas is an expression of Berkeley’s idealism, insofar as it dismisses all terms that do not refer to sensible particulars. The remainder of the Metacritique is an elaboration of the idea that Kant’s critical philosophy runs afoul of Berkeley’s criticism of abstract ideas. Hamann sees Kant as attempting to “purify” philosophy by eliminating all dependence on tradition, custom, experience, and language. The problem with this is that, while Kant emphasizes the principle that all our cognition must relate to experience, the very language Kant uses to construct the Critique-the critique, that is, undertaken as part of reason’s task of systematizing cognition-cannot pass this test, meaning that Kant’s system cannot meet its own standard for cognition. Although it is merely alluded to here, Hamann does have an alternative view of his own. He sees both nature and language as parts of Gods revelation, “a speech to creation through creation,” and although this Book of Nature can certainly be understood by human beings, it can only be understood “by plowing with some other calf than reason.” Hamann is thus a skeptic about reason, but not in general, nor is he a relativist.

In contrast to Hamann, Kant’s former student Johann Gottfried Herder achieved notoriety more for his own theories than for his criticism of Kant. Herder came into conflict with Kant after the publication of his Ideen zur Geschichte der Philosophie der Menschheit in 1784 and 1785. Herder’s project in the Ideen can be understood as the attempt to conceive physical and human nature as a unified, teleologically organized whole, subject to laws which, if not the same, at least have the relation of close analogy. The key concept in this effort is that of “organic forces”: All progressive development, whether in the natural or the human realm, is a manifestation of the tendency of such forces to organize themselves into ever more complex systems. The intent is to show both the naturalness of morality and the morality of nature: Morality and religion are part of nature in so far as they arise from the natural human tendency to reason, through which instincts are suppressed (though not eliminated), and freedom, and thus morality and religion first made possible. Nature is moral, on the other hand, insofar as human moral reason is presented as the goal and completion of the hierarchy that characterizes the natural world as a whole.

Herder’s Ideen thus represents a challenge to two distinctions crucial for the critical philosophy, namely those between theoretical and practical reason and between immanent and transcendent concepts. Kant responded sharply to Herder in a review of the Ideen published shortly after its publication. He disparaged Herder’s theory of organic forces as the attempt “to explain that which one does not comprehend by that which one comprehends even less” (VIII:54). This accusation is justified, however, only if we take Herder to hold, first, that organic forces are unseen entities which are the cause of moral order in the universe, and second, with Kant, that our knowledge is limited to objects of possible experience. But Herder does not accept any sort of correspondence theory of truth at all, and Kant’s distinction is therefore of no relevance for him. Instead, Herder’s theory is pragmatist and historicist, forming a coherent philosophy self-consciously opposed to Kant’s.

Herder’s theory is pragmatist in that for him the primary virtue of propositions is not their correspondence to reality but rather their “Bequemheit,” which is roughly their usefulness or convenience. Herder extends this pragmatism to a general view of human reason: Much as a child connects sensations by means of thoughts, so philosophers connect propositions and theories into the greatest possible coherence and order. As befits a consistent pragmatism, neither basic concepts nor the broader connected whole we make of them is seen as closer to reality than other propositions or theories. This is crucial for understanding why Kant misinterprets Herder when he accuses him of trying to “explain that which he doesn’t comprehend through that which he comprehends even less”: The appeal to “organic forces” is not supposed to explain the hierarchical organization of nature; rather, the connection of the two concepts is an attempt at a more “bequem” relation among ideas, and neither these concepts nor the connection between them need be thought of as having a privileged link with reality. Herder’s historicism is expressed in the claim that knowledge, morality and religion are “children of practice, tradition and habit.” Herder’s view is that progress, or, as he puts it, “Erziehung,” takes place through the handing down of both how to reason and the results of previous reasoning:

Man is thus an artificial machine, albeit gifted with genetic dispositions and a quantity of life; but the machine does not play itself, and even the most capable person must learn how to play it. Reason is an aggregate of remarks and practices of our soul, the result of the education (Erziehung) of our race, which the one being educated, copying given alien models, finally completes in himself.

Furthermore, the fact that every human culture reasons in a way that reflects its climate and surroundings explains the diversity in human scientific, moral and religious beliefs.

Of course, Hamann and Herder are hardly the only thinkers in the Western tradition to have been skeptical about the power of reason to guide human moral conduct. I refer to their positions at some length because each to a large extent formulated his position in response to Kant’s account of reason, and because Kant was familiar with each. The crucial thing about the view that tradition or authority is a better source of moral guidance is simply that it is a plausible view incompatible with Kant’s. The lack of an argument in the Critique of Pure Reason for the purposiveness of appearances thus leaves Kant with little room for demonstrating the superiority of his view to this one-at least to someone who does not already share Kant’s philosophical outlook. Neiman and Velkley are surely right that the conception of reason as a faculty for setting goals rather than for producing knowledge is crucial to Kant’s attempt to vindicate reason, and this insight helps us to accord Kant’s view the significance and the credence it deserves. In order, however, to defend reason as he conceives it not only against radical skepticism and relativism but also against a view like Herder’s, Kant faces a delicate task. He must show either i) that the goals reason sets for us can be achieved; or 2) that adoption of these goals is necessary even for those who claim, as one quite reasonably can, to reject his view of reason; or 3) that even if it is not necessary to adopt the conception of reason outlined in the Critique of Pure Reason, doing so is somehow preferable to other viable alternatives. In either case he must argue in some way for the purposiveness of appearances, but he must do so without negating the distinction between constitutive and regulative principles. It was to this end, I think, that Kant returned to the concept of purposiveness in the Critique of Practical Reason and above all in the Critique of Judgment. But in the Critique of Pure Reason, this task goes unfulfilled.

Appreciation of Kant’s urgent need for an argument for the purposiveness of nature helps to explain the focus on the concept of purposiveness in the 1790 Critique of Judgment. There, in the published Introduction, Kant raises two problems. One concerns the question of the relation between theoretical and practical reason, or, as Kant refers to them here, between understanding and reason. Although each has its own proper “domain,” nevertheless, because practical principles are supposed to influence the real world with which the understanding is concerned, we need to know whether it is possible to unite theory and practice (V: 176). This problem is then related, in somewhat mysterious fashion, to a second problem: While the Critique of Pure Reason has shown that nature is subject to certain very broad laws specified in the pure principles of the understanding, there remain indefinitely many natural laws that are not entailed by the possibility of experience, but which are nevertheless laws, and as such require some a priori element. Kant accordingly proposes an a priori principle of the faculty of judgment, that of the principle of the purposiveness of nature. That principle states that

empirical laws must, as regards what the universal laws have left undetermined in them, be viewed in terms of such a unity as they would have if they too had been given by an understanding (even though not ours) so as to assist our cognitive powers by making possible a system of experience in terms of particular natural laws. (V: 180) Kant says that this is a “transcendental principle,” and thus “requires a transcendental deduction” (V: 181- 2).

A number of features of the argument of the Introduction to the Critique of Judgment indicate that the problem Kant is taking up there is the very problem we have been examining, namely the problem of establishing that there are grounds for thinking that reason’s goals can be attained. The most obvious and important such features are simply the terms in which Kant presents this problem: it concerns the purposiveness of nature, insofar as this is a necessary condition for the achievement of a philosophical system constructed around ideas of pure reason. The concern with system is clearest in the discussion of the system of empirical laws, which is strictly a concern of speculative reason or, as Kant is now calling it, of the understanding. But as we have seen, Kant is also committed to the view that reason aims at a philosophical system in which speculative and practical reason are united. Furthermore, after explaining the problem of the unification of theory and practice, Kant makes it clear that he sees this problem as being essentially connected to the use of ideas, because the need to see the moral law applied to reality is a need of reason which nevertheless goes beyond possible experience. The principle of the purposiveness of nature might therefore be seen as concerning equally the formation of a system of empirical laws, and uniting theory and practice, because in each case we are confronted with the problem of pursuing a goal of reason, the possibility of which surpasses our ability to know. A less obvious reason for reading the Introduction to the third Critique in this way is that the epistemic situation Kant faces with regard to the principle of the purposiveness of nature is precisely that which he faced with regard to the ideas of pure reason in the first Critique. The problem of the ideas of pure reason is that these are “necessary concept[s] of reason … to which no congruent object can be given in the senses” (A327/ 8383; emphasis added). They are therefore incapable of the sort of “objective deduction” (A336/B393) offered for the categories. Similarly, the problem of the purposiveness of nature is that empirical laws “carry necessity” with them, but nevertheless “we cannot cognize them a priori” (V: 183). Thus Kant’s attempt, in the Critique of Judgment, at a transcendental deduction of the principle of the purposiveness of nature might plausibly be understood as an attempt to provide the grounding for the ideas of pure reason lacking in the first Critique.