Chapter 1

Kant, Idealism, and Cognitive Constructivism

Kant, who is clearly one of the several most important Western philosophers, discusses an astonishing range of topics. I will be primarily focusing on his solution to the cognitive problem. Though Kant is not well read in the history of philosophy, he is often understood, as he explicitly suggests, as responding to Hume, but also to others—for instance, Wolff and Leibniz. Kant is also often believed to synthesize empiricism and rationalism in the critical philosophy. I do not want to deny such ways of considering Kant against the historical background, but rather to add to them. I think Kant should also be seen as responding if not to Plato, at least to Platonism. The Platonic element in the critical philosophy is sometimes discussed with respect to Kant’s moral theory.1 I believe that Platonism is central as well to Kant’s theory of cognition. It is then not by accident that the critical philosophy provides a Platonic formulation of the cognitive problem. In short, as Kant reads Plato—or, if there is a difference, Platonism—the latter provides a problem to which the Copernican turn proposes a solution.

After extensive debate, there is still no agreement about even the main lines of the critical philosophy. My interpretation of the critical philosophy is based on a reconstruction of the relationship between Parmenides, Plato, and Kant. Plato takes over the Parmenidean view that there is a way the mind-independent world (or reality) is, and that a claim to know requires an identity of identity and difference—or in other words, the Parmenidean claim for an identity between thought and being. Since this is not a treatise on Plato, it is not necessary to discuss his writings in detail. Suffice it to say that Plato’s cognitive solution consists in three points: first, he invokes an ontological difference between appearance and reality; second, he denies the reverse inference from appearance to reality, or, in more modern language, from effect to cause, in disqualifying any form of the familiar representational approach to cognition; and, third, he rejects the Platonic view that philosophers can intuit reality.

Kant’s approach to the problem of knowledge closely follows the Platonic view in invoking an ontological dualism, in his case between appearances and noumena (or things in themselves), two synonymous terms that refer to mind-independent reality. He assumes that, as in Plato’s day, there are still only two main approaches to knowledge: cognitive intuition and cognitive representation. He further follows Plato in rejecting cognitive representation; unlike Plato, he also refuses cognitive intuition, which he denies to human beings, hence denying we can know reality.

Kant’s proposed solution is the so-called Copernican turn, or the view that we can claim to know only what we in some sense construct. This solution follows the Platonic way of framing the cognitive problem in suggesting that, though we cannot know the mind-independent world, we do know appearances. Copernicanism, which takes its name from the view of the Polish astronomer, is a heliocentric astronomical approach that goes back to ancient Greek cosmological speculation. Heliocentrism was anticipated, for instance, in ancient Greece by Aristarchos of Samos. In the B preface to the Critique of Pure Reason, Kant calls attention to Copernicanism in further implying that the critical philosophy is a form of cognitive constructivism. Kantian constructivism is often mentioned but only rarely studied in any detail. If the critical philosophy is cognitively constructivist, then Kant—who has been studied as much if not more than anyone else in the modern tradition—is paradoxically still not well known.

On Interpreting the Critical Philosophy

I intend to interpret Kant’s critical philosophy as constructivist for the obvious reason that Kant himself calls attention to the relation between his position and Copernicanism. If Copernicanism is constructivist and if the critical philosophy is Copernican, then it is also constructivist. Kant initially favors representationalism before turning to constructivism. Though early and late he consistently employs representationalist vocabulary, constructivism and representationalism are mutually exclusive epistemic strategies.

In different ways, constructivism runs throughout the critical philosophy. Kant takes a constructivist approach to morality,2 aesthetics, and cognition. His moral and his aesthetic theories both require the subject to ascertain a universal principle. According to Kant, a necessary condition of morality is for the moral subject to identify a universalizable maxim to guide moral action. Similarly, on the basis of taste, the aesthetic subject must infer a universal rule from the singular aesthetic object in a judgment of beauty at least in principle acceptable to all observers. If Kantian moral theory is constructivist, then so is his aesthetic theory.

I will be concentrating on Kant’s cognitive constructivism, which precedes his moral and aesthetic constructivism. It is not easy to interpret Kant’s concept of cognition, which is increasingly obscured by an enormous and rapidly growing debate. Kant provides a theory as well as a meta-theory about how to interpret the critical philosophy. He also provides a series of conflicting indications in his writings. Kant, who is a hermeneutical holist, rejects interpretation based on passages torn out of context in favor of interpretation through the idea of the whole (CPR, B xliv, p. 123). Yet there has never been any agreement about how to identify the idea of the whole of the critical philosophy or even its central theme, idea, or insight. Conflicting indications in the texts perhaps reflect Kant’s own indecision about the nature of his project. Readings of the critical philosophy tend to follow Kant’s hints that it is representational, often without any clear indication of what that entails. Does Kant think we can successfully represent, hence know the mind-external world? Or does he merely believe we can know that it exists? The difference is important and Kant’s view of the matter is unclear.

In rejecting a representational approach in favor of a constructivist reading of the critical philosophy, I will be arguing against specific indications in Kant’s own writings. His understanding of his central problem in the so-called critical period is simply but misleadingly stated in his famous letter to Herz. Here he formulates the cognitive problem in representational language by pointing to the relation of the representation (Vorstellung) to what is represented in asking the question: “What is the ground of the relation of that in us which we call ‘representation’ to the object [Gegenstand]?”3 which motivates what in the first Critique became the initial and most important statement of his critical philosophy. This representational formulation suggests his position should be understood as still another form of the causal theory of perception that dominated early modern philosophy at the time Kant was active and which remains dominant. Yet the Herz letter is misleading if Kant later changed his mind in opting for constructivism instead of representationalism.

The Representational Approach to Knowledge

It is difficult to determine Kant’s precise view of cognition. There are at least four possibilities. As concerns cognition, he could be either a representationalist, a constructivist, inconsistently favor one or the other approach, or again reject all these possibilities. He sometimes seems to favor representation; he also insists on a constructivist approach in passages about the Copernican revolution and elsewhere. It is possible that Kant, who often seems to hesitate between alternative solutions, is simultaneously attracted to different possibilities. Though the texts are unclear, probably the best way of reading the different things he says about cognition is through a later turn, sometime after he composed the Herz letter, away from representationalism to constructivism.

Kant intervenes in the late-eighteenth-century debate, which is dominated by a causal theory of perception, hence by cognitive representationalism. Kantian representationalism carries this epistemic approach to a high point—never later surpassed—while revealing its limits. If Kant had done nothing else, then since (as he points out) representationalism fails, he would probably be known as another in a long line of causal theorists of perception widely scattered through modern philosophy.

“Representation,” which means many different things in different contexts, takes on many artistic, political, psychological, mathematical, and other subforms. The term suggests an approach to cognition in which something—the representation—stands in for, takes the place of, and also points beyond itself to something else, which is said in this way to be represented. As a cognitive strategy, “representation” is understood in different ways. Thomas-Fogiel, who follows Marin, distinguishes four views of representation.4 These views include (1) to re-present or to reflect; (2) presence and absence; (3) the substitution of one thing for another; (4) to outline or trace the contours of something in according it visual form. To represent by re-presenting or reflecting something is the basis of the familiar reflection theory of knowledge—a staple of Marxism since Engels and officially adopted by Lenin, and which probably originates in book 10 of the Republic at 596D, where Socrates talks about carrying a mirror around with him.

What is representation? Plato and then later Locke and other modern thinkers take opposing views. We cannot know Plato’s view, if he had one. In the well-known theory of forms, which is often ascribed to him, Plato seems to identify forms (or ideas) and the real (or reality). According to the theory of forms, things are appearances that imitate, or participate in, concepts or mind-independent reality, which can only be known through cognitive intuition, hence directly. Plato’s rejection of the very idea of appearances as knowledge motivates his attack on artistic imitation as cognition, leading to his refusal of art and art objects of all kinds. The Platonic attack on representation is strongly contested in modern philosophy. The post-Platonic debate on representation as well as the later development of aesthetics can be reconstructed as a series of efforts to rehabilitate representationalism against Plato’s rejection of representation as well as his attack on art and art objects of the most varied kinds.

An anti-Platonic approach to knowledge through representation is extremely widespread in the seventeenth century—that is, prior to the emergence of the critical philosophy. Cognitive representationalism links together such different cognitive strategies as rationalism (which considers the conditions of knowledge) and empiricism (which focuses on human knowledge). In modern times, a representational approach to cognition often relies on the canonical distinction between primary and secondary qualities drawn by Galileo, Descartes, and especially Locke. Representation is routinely identified with such sources as Locke, British empiricism, and the Port Royal School.5 According to the traditional interpretation, Locke is an externalist, not an internalist.6 For Locke as well as many modern thinkers, ideas in the mind are caused by, and hence refer to, the external world. Locke’s statement “Whatever the mind perceives in itself, or as the immediate object of perception, thought or understanding, that I call idea”7 underlies his conviction that we know the world through ideas that represent it. His anti-Platonic approach to cognitive representation is hugely influential in the modern tradition. The so-called Port Royal Logic proposes a theory of signs that link the represented thing and its representation.

Kant famously calls attention to Hume as awakening him from his dogmatic slumber. The view that Hume’s influence is pervasive in the critical philosophy is supported by recent research.8 Hume belongs to British empiricism, which centers on a representational approach to cognition influentially formulated by Bacon and Locke and contested by Reid, Berkeley, Hume, and others. Representationalism, or the causal theory of perception, stands or falls on the ability to justify the crucial anti-Platonic inference from effect to cause—in one prominent version, from an idea in the mind to the world. This difficulty, which has never been resolved, undermines the later anti-Platonic effort to answer Plato’s ancient attack on representation and artistic imitation. This same difficulty returns in Kant’s more complex approach to cognitive representation.

Kant’s evolving view of representation is convoluted, unclear, and perhaps inconsistent. In the precritical Only Possible Argument in Support of a Demonstration of the Existence of God (1763), he suggests that “the word ‘representation’ is understood with sufficient precision and employed with confidence, even though its meaning can never be analyzed by means of definition.”9 But in the Dohna-Wundlacken Logic—presumably based on lectures given in the 1790s, hence in the critical period—he states that representation “cannot be explained at all.”10 In the seminal passage in the first Critique describing his relation to Plato, he indicates that sensation (a change in the state of the subject), perception and cognition (or an objective perception) all fall under the general heading of representation. Cognition, Kant insists, is in general either intuitive or conceptual. Yet, since he favors a categorial approach to experience, he rules out intuitive cognition and hence rules out a Platonic approach to cognition, which leaves only representation.

In the critical philosophy, representationalism concerns the link between noumena and phenomena, the latter a term Kant frequently seems to use as a synonym for appearance.11 By “appearance,” Kant—perhaps distantly following Plato—understands the representation of a mind-independent object, which affects the subject12 through sensation, or sensory intuition, in turn giving rise to an appearance (see CPR, B 33, p. 172).

Kantian theory of knowledge resembles classical modern causal theory of perception, hence representationalism, in two ways: with respect to vocabulary and through the causal relation between world and mind. Kant frequently uses the term “representation” in place of the modern representationalist term “idea,” a word that has another usage in the critical philosophy. Further, like classical modern representationalists, he describes the causal input, which derives from the impact of the mind-independent world on the mind. Yet the Copernican turn that arguably lies at the epicenter of the critical philosophy—a cognitive approach often supposed to be representational—is not representational but in fact is based on the failure of representationalism.

Kant’s attitude toward representationalism is inconsistent. Three intractable difficulties arise in any effort to classify Kant as a representationalist. To begin with, since Kant insists on the subjective contribution to cognition, and even though he uses representationalist terminology, the specific anti-Platonic backward inference necessary for representationalism is not possible but rather impossible in the critical philosophy. Second, representationalism and constructivism are inconsistent, and, through the Copernican turn, he is clearly committed to constructivism. Finally, it would be clearly inconsistent to claim to represent the real while denying knowledge of noumena, or things in themselves. Hence Kant is not and cannot be a representationalist, or at least he cannot be a representationalist as “representation” is usually understood.

Representationalism and the Double Aspect Thesis

It is sometimes argued that Kant is a representationalist in a nonstandard sense of the term concerning the so-called double aspect view. The double aspect view is not a theory. It is rather a cognitive thesis embedded within a theory—in this case, in the critical philosophy within which, depending on the interpretation, it is sometimes thought to play a central role.

Kant, who often has difficulty in choosing between alternatives, typically defends, consciously or unconsciously, more than one approach. I have been suggesting that, following Plato (though he perhaps also defends or earlier defended representation), through his commitment to constructivism, Kant later turns away from representation, a cognitive approach that runs throughout the modern debate. The double aspect thesis indicates his hesitation about whether to defend or to abandon representationalism. This thesis is presented in different ways in Kant’s writings, especially in respect to the possibility of morality (CPR, B xxvii, p. 115), where it plays a crucial role, and as concerns the cognitive problem (CPR, B xviii, p. 111).

There are different metaphysical and non-metaphysical ways of reading the critical philosophy.13 A metaphysical reading of the critical philosophy includes a so-called ontological commitment. In the double aspect thesis, the twofold metaphysical commitment includes the view that the mind-independent world affects the subject; it also includes the further view that the effect (or result) and the cause (or the thing in itself) are two aspects of the same thing. In that sense, the double aspect theory is an extreme form of the modern causal theory of perception—extreme in that the cause and the effect of the subject’s affection through the thing in itself (noumenon or mind-independent reality) are presumed to be identical. In the context of the critical philosophy, this amounts to collapsing the crucial distinction between phenomena and noumena on the assumption that at the limit, there is not any distinction between appearance and reality. This approach suggests that Kant intends in the critical philosophy to meet the Parmenidean requirement for knowledge as the grasp of the mind-independent world. Yet if what we perceive is constructed by the subject, then we cannot infer noumena or things in themselves from representations or appearances; hence we can make no positive cognitive claims about reality. In that case, thought can only be identical with being—and hence meet the Parmenidean criterion for cognition—if as its condition the subject must construct what it knows.

The double aspect thesis is a form of representation, which some Kant scholars defend, if necessary even against Kant. In summing up discussion that was already more than a century old, late in the nineteenth century, Hans Vaihinger identified three equally unsatisfactory ways of explaining affection in the critical philosophy as originating from: (1) the perspective of a thing in itself, (2) objects in space, or again (3) through a so-called double affection. The latter includes a nonempirical affection through things in themselves and an empirical affection through objects in space.14 Several decades later, Erich Adickes suggested that the double affection version of this thesis is central to Kant’s theory of knowledge.15 More recently, Henry Allison has in effect made this thesis central to his defense of the critical philosophy.

Allison, in defending the double aspect thesis, is countering Strawson, Guyer, and many others. In the Bounds of Sense, Strawson reads the critical philosophy without transcendental idealism, which he regards as a deep mistake.16 He and many others interpret the critical philosophy through the traditional Platonic dualism between appearances and reality. In the Critique of Pure Reason, Kant examines this distinction in detail in the chapter on phenomena and noumena (CPR, B 295–315, pp. 338–53). The concept of the noumenon can be interpreted differently; for instance, as referring to an epistemic limit or as an ontological distinction between what appears and what is. According to Strawson, who never mentions Plato, the distinction between phenomena and noumena is ontological. For Strawson and others who favor an ontological two-worlds reading of the critical philosophy, pioneered by Plato, it is not possible to make a cognitive inference from phenomena to noumena. This way of reading Kant is close to Maimon’s view of the critical philosophy as a form of cognitive skepticism. Allison, on the contrary, takes an epistemic approach—originally worked out after Kant above all by Fichte—in which noumena and phenomena are two ways of cognizing one and the same object.17 Versions of this view perhaps appear in the tradition before Kant in Spinoza, and after him in Husserl.

Kant, who is ambivalent, formulates the widely known double aspect thesis in the B preface to the Critique of Pure Reason at a time when, as noted, in virtue of the so-called Copernican turn, he has apparently moved away (or is in the process of moving away) from representationalism. Thus in an important footnote, in reference to the validation of the propositions of reason, he suggests the cognitive object can be considered from two perspectives as both an appearance and a thing in itself.18 It is difficult to know what this claim signifies since, as already mentioned, representationalism is inconsistent with constructivism. It is possible that at the time he was formulating the constructivist approach Kant was still ambivalent about representationalism. His suggestion that the cognitive object can be understood from two perspectives is arguably a fall-back effort, either on his part or on the part of those committed to representation who intend to save Kant’s version of the modern causal theory of perception through an undemonstrated and in fact indemonstrable claim. Since “officially” there cannot be any cognition of things in themselves, neither Kant nor anyone else can demonstrate the same objects are the objects of sensation and of thought. Hence, the double aspect thesis does not provide an acceptable analysis of the relation of representations to objects called for in the famous Herz letter (1772).

In considering this thesis, it is useful to distinguish between Kant and those influenced by him, who are sometimes more orthodox about the critical philosophy than he was. Kant, who hesitates to make up his mind, on occasion defends incompatible alternatives, and also sometimes changes his mind. His remarks in the B preface about Copernicus suggest either that he later gives up representationalism—his initial cognitive strategy, which he adopts in following numerous modern thinkers—for a “replacement” constructivist strategy for knowledge, or again, as mentioned above, that he is simultaneously and inconsistently committed to both approaches. The crucial passage on the “Refutation of Idealism,” inserted in the second edition of the Critique of Pure Reason, suggests we are affected by and in fact know a mind-independent external world. Yet this view is incompatible with the constructivist view that we experience and know only what we in some sense construct.

Allison’s defense of Kantian representationalism through a reading of the double aspect thesis19 is influential.20 Thus Beiser distinguishes between objectivist and subjectivist interpretations of the double aspect thesis. He attributes a quasi-subjectivist interpretation to Allison referring to the operations of the human mind,21 but objects that this reading is inconsistent with Kant’s identification of representations and appearances at A 492. Here Kant, in a passage on transcendental idealism, writes in part that “all appearances, are not things, but rather nothing but representations, and they cannot exist at all outside our mind” (CPR, A 492, p. 511).

Certainly Kant, in some of his moods, identifies appearances with representations. This claim is consistent with Allison’s reading of Kant. Yet even if this were a correct interpretation of the critical philosophy, it would fail to demonstrate the double aspect thesis. The difference between appearances and phenomena is clear. All appearances are phenomena, but only some phenomena are appearances, or, if the cited passage is in fact Kant’s considered view, representations. Appearances would be representations if and only if noumena appear, which is the point at issue. But it cannot be shown that appearances are representations without showing that the double aspect thesis is correct.

In the familiar double aspect thesis, to which Kant alludes in passing, he draws attention to the distinction between objects of sensible intuition, or appearances, and things in themselves before indicating that the human subject is both phenomenally determined and noumenally free. In this context, he writes: “But if the critique has not erred in teaching that the object should be taken in a twofold meaning, namely as appearance or as thing in itself” (CPR, B xxvii, p. 116). In this complex passage, Kant is concerned, since he believes in strict phenomenal causality, with staking out room for morality, which he defends through noumenal freedom. There are other similar passages. For instance, in reference to the distinction between rain and a rainbow, Kant writes: “Thus, we would certainly call a rainbow a mere appearance in a sun-shower, but would call this rain the thing in itself, and this is correct, as long as we understand the latter concept in a merely physical sense, as that which in universal experience and all different positions relative to the senses is always determined thus and not otherwise in intuition” (CPR, B 63, p. 170).

Beyond the defense of the possibility of morality, this claim is important with respect to cognition. To demonstrate that one and the same object is both an appearance and a thing in itself would go a long way toward making out a representational form of the modern causal theory of perception. If Kant’s version of the causal view of perception could be defended, it would vindicate, through the claim that “appearance” and “representation” are synonyms, an entire line of argument that otherwise, if Kantian representationalism fails, no longer seems promising.

The difficulty lies in maintaining a form of the double aspect thesis that allows Kant to make out the relation of appearances to objects without violating the central claim “that we have no concepts of the understanding and hence no elements for the cognition of things except insofar as an intuition can be given corresponding to these concepts” (CPR, B xxv–xxvi, p. 115). Yet if there is no cognition of things in themselves—which we can only think of, but neither know nor even experience—then it is not possible but rather impossible to demonstrate the double aspect thesis.

Allison’s non-metaphysical reading of the critical philosophy, which looks away from any ontological commitment, is not at all obvious. A more obvious way of reading the critical philosophy might be as a theory about the affection or causal impact of the mind-independent but uncognizable external world on the cognitive subject. “Things in themselves,” from this perspective, would have ontological weight since this term would literally refer to things. There are numerous passages that might be read in this way. For instance in a remark on speculative theology, where Kant seems once again to equate appearances and representations, in reference to “things in themselves,” he writes: “Such an intelligible cause, however, will not be determined in its causality by appearances, even though its affects appear and so can be determined through other appearances” (CPR, B 565, p. 535). Kant seems here once again to be saying that the mind-independent world affects the subject in producing through the latter’s activity representations and appearances.

Allison rejects an ontological approach in favor of an epistemic reading of the concept of the thing in itself. According to Allison, who perhaps has in mind the familiar contrast between ordinary (or everyday), unsophisticated views and philosophical views, the term “thing in itself” simply specifies the epistemic conditions of cognizing spatiotemporal objects.22 He intends to make it plausible to understand representations as both appearances as well as things in themselves. Hence, if the double aspect approach could be made out, representationalism would be plausible because noumena would in fact appear.

When taken in isolation, the double aspect thesis, as Allison’s critics (e.g., Ameriks, Guyer, and Langston) point out, undermines the familiar ontological reading of the Kantian cognitive object.23 Numerous passages support an ontological reading of the thing in itself, which, if sustained, would run against a purely epistemic reading of this concept. Kant consistently describes the thing in itself as an entirely unknown and unknowable cause of representations, which is not a thing but a concept.24 It follows that it is not in either space or time, that it cannot be an object of sensible intuition (see CPR, B 522, pp. 512–13), and that it ought not be accorded the status of a self-subsistent reality (see CPR, B 594, p. 550).

Certainly Kant cannot claim that reality is beyond cognitive reach and defend the double aspect thesis. Both the ontological and the epistemic readings of the distinction between noumena and phenomena have textual support, but the epistemic reading is cognitively indefensible for two reasons. First, though representationalism is often asserted, there is no known argument that demonstrates the transition from the representation to what it represents, hence apparently no way to justify the claim to represent reality. This point supports the Platonic rejection long ago of a backward inference from effect to cause. The best way to justify Kant’s Copernican turn is as an alternative approach to cognition (a second-best theory) in place of representationalism, which fails. Second, the double aspect thesis is a form of representationalism. Since representationalism and constructivism are incompatible, the turn to representationalism mandates a turn away from constructivism. Now Kant is perhaps inconsistent, since he appears on occasion to defend both representationalism and constructivism. Yet since representationalism is both inconsistent with constructivism and, as Kant himself correctly indicates, fails as an approach to cognition, he cannot consistently defend and also cannot consistently be defended as a representationalist thinker.

Kantian Subjectivity and Cognition

Kant’s conception of the subject is central to his constructivist approach to cognition. The subject is one of the most difficult but also most important themes in modern philosophy, as well for Kant. Early in the medieval period, a view of the subject was formulated to provide for an account for human responsibility in the context of the Christian view of original sin. This basic view of the subject is carried over into the modern period—for instance, through Descartes, who transforms a Christian conception of the subject morally responsible for its acts25 into a cognitive subject capable of apodictic knowledge able to defeat the most rigorous forms of skepticism.

In the complex modern debate on subjectivity, the two main variants consist in an anthropological approach to knowledge, based on conceptions of the nature and limits of finite human being; and a speculative approach, linked to the requirements of knowledge in general. These two approaches to the cognitive subject respectively lead to views of human knowledge, or again, knowledge in general.

The two-dimensional Cartesian subject, which combines both approaches, is both a finite human being as well as a rational construct, capable of apodictic knowledge through reason in restraining the will. The British empiricists stress human knowledge through a concept of the subject as finite human being. Kant, who anticipates the Husserlian critique of psychologism, stresses the purely rational aspect of the subject throughout his position. Kant was interested in anthropology and wrote a book on the topic. He was also one of the first to teach anthropology in Germany. He often but perhaps inconsistently refers to the capacities of finite human beings in working out his transcendental deduction. For instance, he notes more than once that finite human beings do not have intellectual intuition, hence must rely on sensory intuition (see, e. g., CPR, B 135, p. 248; B 139, p. 250; B 145, p. 253; B 146, p. 254; B153, p. 257). Yet the Kantian cognitive subject is simply “deduced” as the final step in the transcendental deduction—and, since this is the conceptual heart of the first Critique, as the copingstone of the critical philosophy, as it were.

A rapid reference to the role of the subject in cognition does not exhaust the theme in the critical philosophy. One can interpret the overall position through the three Critiques, each of which proposes an account of a basic type of experience in terms of a specific kind of activity. The difficulty lies in bringing together an analysis of aspects of the subject’s activity in an overall concept of the subject. If he could accomplish this task, then Kant could answer the deep question, which, in the Jäsche Logic, he indicates is more important than any other, that is: what is man? In retrospect, Kant seeks but fails to combine abstract approaches to cognition and morality—or theoretical and practical reason—in a single concept of the subject. Despite his best efforts, this synthesis remains no more than a promissory note, which is redeemed only in the later discussion by Fichte, and then in a related but different way by Marx.

Kant is critical of other modern theories of the subject. He stresses, for instance, Hume’s influence and criticizes Descartes. Yet he rejects Hume’s bundle theory of the subject while following the Cartesian emphasis on the subject, or “I think,” an obvious translation of the French thinker’s “cogito” (or “je pense”), as central in cognition. Classical modern representationalism follows the Cartesian insight that the road to objectivity leads through subjectivity while comparatively deemphasizing the subject when emphasizing the object as the unimpeachable source of objective knowledge claims. At most, as Descartes emphasizes, the subject chooses among different ideas an acceptable candidate for knowledge.

The traditional minimalist modern approach to subjectivity sacrifices the active role of the subject in cognition, which Kant hastens to restore. For Descartes, error arises through a failure to bring the will under control, which Kant stresses in his rational conception of the moral will. And for Hume, the mind leads us astray in thinking we perceive causal connections where there are none.

By virtue of his constructivist approach to cognition, the Kantian subject cannot be passive or only passive, but must also be active in constructing what it knows. According to Kant, cognition requires both the causal impact of the world through sensation, which, since it lacks form, is uncognizable, and which acquires form, hence becomes cognizable, only through the activity of the subject in constructing, producing, or making a cognitive object.

In simplifying, we can say that Kant, like Descartes, understands the subject on two levels as both finite and infinite: as a human being, a theme to which he refers in the Jäsche Logic and discusses in the Anthropology; and as the philosophical subject he describes in “The Transcendental Deduction of the Categories” as the highest concept in his cognitive theory. Kant’s conception of the subject changes over time. In both the first and second Critiques as well as in the Groundwork, the philosophical subject is not an individual but a mere conceptual placeholder whose contours are not based on observation, but are rather deduced in terms of the requirements of the critical philosophy, above all in “The Transcendental Deduction of the Categories.”

For purely cognitive reasons, Kant ascribes epistemic activity to the subject, which is not just passive, but passive as well as active. The view of the subject as also active alters the conception of the subject as well as cognition. Before Kant, early modern conceptions of knowledge feature passive reception of what is as a faithful representation of the world. After Kant, knowledge depends on a subject, which, in being affected, constructs cognizable cognitive objects, or so-called appearances of a mind-independent world it can neither experience nor know.

Some Views of Kant’s Copernican Revolution

In turning now to the Copernican revolution, it is useful to distinguish between the astronomical theory formulated by the Polish astronomer, Kant’s interpretation of it, and the constructivist approach to cognition it inspired in the critical philosophy. “Copernican revolution” is now used indiscriminately to refer to the views of both Copernicus and Kant. By this term I understand an approach to experience and knowledge of objects as dependent on the subject, illustrated by Copernicus, Kant, the post-Kantian German idealists, and many others.

Kant indicates that what quickly became known as his Copernican revolution is central to critical philosophy. But, in a Kantian debate where even seemingly minor themes are analyzed in detail, there is surprisingly comparatively little attention devoted to the Copernican revolution.26 Discussion of Kant’s Copernican revolution, is often very brief and confined to the epistemic consequences of a change in orientation. According to Paton, Kant—like Copernicus, who explains apparent motions of the planets through the motions of the observer—describes reality as it appears through the subject.27 Ewing contends that in introducing a distinction between appearance and reality, Kant, like Copernicus, attributes reality to the subject.28 In Höffe’s view, for Kant the objects of knowledge appear by virtue of the subject.29 In his recent study, Friedman, who distinguishes between appearance and experience, mainly stresses the role of the Copernican astronomical revolution in Kant’s concept of nature in looking away from the constructionism featured in the title of his book.30

There seem to be two main approaches to Kant’s Copernican revolution in the debate. Cassirer and other German neo-Kantians, who provide epistemic readings of the first Critique, regard the Copernican revolution as key to Kant’s approach to cognition.31 On the contrary, phenomenologists and those influenced by a phenomenological perspective tend to downplay the interest of Kant’s remarks on Copernicus, which they regard as furnishing false or misleading suggestions about how to interpret his theory. Heidegger, an opponent of German neo-Kantianism, substitutes a metaphysical for an epistemic interpretation of the first Critique, which he regards as basically mistaken.32

The core claim common to both perspectives is that in Kant’s Copernican revolution lies a change of perspective based on a distinction between appearance and reality. As a result, what earlier appeared to be reality is now known to be mere appearance. This interpretation of Kant’s so-called Copernican revolution captures it in part. Yet Kant does a good deal more than that: his interpretation of Copernicus’s change of perspective crucially affects his reading of Newton and of the rise of the new science as well as his own distinctive theory of the conditions of knowledge.

Copernicus, Kant, and the New Science

There is a difference between a specific astronomical use of the term “revolution,” and the way this term functions in Kant’s critical philosophy. In De revolutionibus orbium coelestium, Copernicus uses “revolution” to refer to the circular orbits of the planets. As concerns the critical philosophy, the term does not refer to Copernicus’s understanding of revolution, but rather to the influence of the Copernican astronomical revolution on Kant’s cognitive theory.

Kant’s interpretation of Copernicus follows from a conception of science that was far from standard even when he was writing. This is not the place to discuss Kant’s conception of science in detail. For present purposes, it is sufficient to describe that approach as including three main factors: To begin with, there is the discontinuity, or scientific revolution, that Kant thinks separates Copernicus’s theory and earlier science; then there is the way that this discontinuity was brought about by Copernicus, who made possible the rise of the new science which, as Kant understands it, is literally built on a Copernican foundation. And finally, there is the way that Copernican astronomy functions as a basis for the epistemic revolution Kant intends to carry out in philosophy. Each of these points is problematic.

The idea of a Copernican revolution in astronomy—which for many years was regarded as obvious,33 even as the central event of modern times34—now appears questionable in virtue of recent discussion of the very idea of a scientific revolution. It has been seriously suggested that there was no scientific revolution.35 And if there was no scientific revolution, then it would be meaningless to make claims about Copernicus’s supposed revolution in astronomy.

The case for Copernicus’s astronomical revolution has to be made with respect to the difference between Copernican and other types of astronomy. Those who believe in a Copernican revolution see his contribution as breaking with preceding astronomical views. Those who argue against the idea of a scientific revolution regard his theory as another in a long line of incremental changes in which there was no decisive discontinuity, and hence no break with preceding theories.

It seems difficult to deny that there was a scientific revolution. At best, one could deny that any single event brought about the scientific revolution about. Thus it would be implausible, say, to equate Aristotelian science, which is qualitative, with the new science, which is quantitative. Yet this does not suffice to explain Copernicus’s scientific advance.

Copernicus theorized in the context of the reigning geocentric theory of astronomy, which was formulated by Ptolemy and was based on Aristotelian cosmology. Late scholastic thinkers such as Grosseteste, Bradwardine, Buridan, Oresme, and Nicholas of Cusa were attracted to a heliocentric hypothesis but impeded by established dogma from accepting it. Prior to Copernicus, there was no single astronomical theory covering all the observed phenomena. At the very least, his astronomical contributions include: a heliocentric theory to replace the geocentric view then in vogue; a single global theory to replace all the ad hoc hypotheses (in practice, the epicycles required in the Ptolemaic system to account for observed motion); a systematically simpler theory than the preceding theory; and the interpretation of observed motion in terms of our own motion.

Kant certainly thinks Copernicus’s contribution constitutes an astronomical revolution. In displacing the geocentric hypothesis with a heliocentric hypothesis, everyone knows Copernicus displaced not only the earth, but also the human being, from the center to the periphery of the universe. Yet Kant was not attracted to Copernicus because the latter displaced either the earth or human being from the center to the periphery, or even through his creation of a new absolute center occupied by the sun. Kant was rather attracted to Copernicus by the counterintuitive way in which the latter related observed motion—what is now known as the apparent retrograde motion of the planets—to the spectator, as the result of our motion (see CPR, B xxii, p. 113).

In re-situating the earth from the center of the solar system, Copernicus notes that it rotates on its axis and revolves around the sun. This point affects the interpretation of Kant’s view of Copernicus, which remains difficult. It is not clear whether Kant’s understanding of the Copernican astronomical revolution refers to the rotation or revolution of the earth; or, on the contrary, to the change in point of view that allows Copernicus to explain perceived motion; or finally, to both of the above.

Kant’s conviction that Copernicus made possible the rise of modern science rests on his reading of the link between the Copernican explanation of the kinematics and the Newtonian explanation of the dynamics of the solar system. According to Kant, Newton’s theory, which is the high point of natural science, depends on the change of perspective in Copernican astronomy. For Kant, Newton’s discovery of the law of gravitation was possible only because of Copernicus’s change in perspective.

In the preface—added to the second edition of Newton’s Principia in 1713—Roger Cotes suggests36 that Newton’s astronomical contribution lies in proving from appearances that gravity belongs to all bodies.37 Kant apparently generalizes Cotes’s suggestion to relate Newton to Copernicus. According to Kant, Copernicus put forward as a hypothesis a theory that offers a physical explanation for the perceived phenomena, but which he could not prove and which was only finally proven through Newton’s law of gravitation. This suggests if Newtonian mechanics depends on Copernican astronomy, and if the latter only finally emerges through a basic change in perspective, that in this way Copernicus makes possible the emergence of modern science. Thus Newton, in formulating the law of gravitation, finally solves Copernicus’s problem in his theory of mechanics, which represents the high point of modern science, and even the end of natural science (if, as Kant is persuaded, the theory is valid).

Kant’s view of the transition from Copernicus to Newton is innovative in the role assigned to Copernicus and in the idea that scientific theories can be proven. This view derives support from two factors: first, the way that Copernicus, though still a medieval figure, displaced theology from nature; and second, the introduction of a different mechanical model. Modern science is typified by the introduction of a simplified mechanical model for more complex phenomena, which, through this strategy, receive a mathematical interpretation.

One should not overlook the originality of Kant’s approach. In privileging the role of Copernicus in the rise of the new science, Kant differs from the more standard view; for instance, in the version suggested later by Husserl, according to which the main impetus in the development of modern science lies in Galileo’s application of mathematical techniques to the understanding of nature.38

It is clear that in supplanting the geocentric astronomical view with a heliocentric view, Copernicus helps displace theology from nature. His mechanical model was in turn reformulated by Kepler, whom Hegel defends against Newton (who, by Hegel’s account, routinely receives credit for something he did not do), then by Newton. Kepler, who built on Copernicus, is usually regarded as the first modern astronomer. In practice, his three laws of planetary motion led directly to Newtonian mechanics. Kant, who believes that the laws of nature can be proven (see CPR, B 198, p. 283), suggests that, on the basis of the Copernican hypothesis, Kepler formulated the laws of planetary motion, which were then proven through what he calls Newton’s “central laws of the motions of the heavenly bodies” (see CPR, B xxii, p. 113)

Kant’s claim remains unclear. The difficulty with this way of reading the relation of Newton to Copernicus does not lie in the relation of a dynamic to a kinematic analysis, but rather in the very idea that the laws of nature, hence natural science, can be proven. Kant claims that Newton proved gravitation to be more than merely phenomenal, which Copernicus only assumed as a hypothesis. Yet there are at least two reasons to doubt this is the case. First, it seems odd and difficult to understand that something can be proven at one time but later refuted. This suggests that the law of gravitation as formulated in Newtonian mechanics could not have been proven, even were such proofs possible, since it was later abandoned in general relativity. The point is not to compare two theories that are different (and perhaps for that reason incomparable), but to note that from the current fallibilist conception of modern science, theories are not and can never be proven, and hence are never beyond the possibility of refutation.39 It follows that Newton, who could not have proven Copernicus’s hypothesis, could at most have demonstrated reasons for adopting it, such as its heuristic usefulness in calculating the planetary orbits.

It is now usual to hold that science justifies but does not prove itself. On this view, Newtonian mechanics justifies, say, Newton’s law of gravitation, which cannot further be justified. Kant’s contention that philosophy must ground, hence justify, science has its beginnings in the Platonic idea that philosophy demonstrates its own and all other claims to know. Yet from the philosophical angle of vision, the ancient Platonic view of science as requiring a form of justification it cannot itself provide was already outmoded in Kant’s time. It was clearly rejected by Hegel, who accepted the view, typical for modern science, that modern science need not appeal to any further form of cognition to secure its knowledge claims. Even Husserl, who otherwise strongly defends Kant, rejects the Kantian idea that natural laws can be known other than through empirical induction.40

Kant, Copernicus, and Newton

Kant’s interpretation of Newton as building on Copernicus suggests natural science presents a project that develops through successive theories. This interpretation provides insight into his own effort to develop Newtonian mechanics beyond Newton through a cosmological theory. Kant’s early work on Universal Natural History began an effort to extend Newtonian mechanics, which Newton limited to the solar system, to the entire universe. Newton did not distinguish between the solar system and the universe, as the full original title of his treatise makes clear. Kant made the necessary distinction in the course of working out his theory of the genesis of heavenly bodies and their motions through mechanical laws.41

In the Principia, Newton famously contends at the end of the “General Scholium” that God widely separated the stars to prevent gravity from causing them to collapse.42 Kant, who relies solely on Newtonian principles, takes a wholly secular approach to science. He removes God as an explanatory factor by contending that nature is self-sufficient,43 hence in no need of divine interference. In arguing that the fixed stars move, he posits as a second basic force the tendency for bodies in motion to continue in a straight line, which counteracts gravitational attraction (a concept apparently similar to our present concept of conservation of momentum). The two forces, when taken together, form a universal system of orbital motions.44

The same intention to generalize Newtonian mechanics to the universe in general is visible many years later in the first Critique, for instance in a passage on the regulative employment of ideas. Here, in a reference to Kepler’s revised understanding of elliptical planetary orbits as grounded in the Newtonian theory of gravitation, Kant suggests its application to comets as well as the entire universe (des unbegrenzten Weltsystems) (CPR, B 691, p. 601).

Kant’s Copernican Revolution

Kant’s interpretation of Newton is later contested in Hegel’s critique of Newtonian mechanics. As Kant reads modern science, Copernicus introduced a hypothesis that Newton demonstrated on empirical grounds with certainty (hence apodictically), but which still must be grounded philosophically through a general, or a priori, demonstration of the laws of science. Kant claims to do this through the famous Copernican turn, which, he believes, alone shows the possibility of objects of experience in general and knowledge of them in particular.

Kant’s Copernican turn is his alternative to representationalism, which depends on the possibility of representing reality. Constructivism, on the contrary, denies that mind-independent reality can be successfully represented by contending we can only cognize what we in some sense “construct.”

When Kant was active, representationalism was familiar, but constructivism (which is still not well known) was unusual. It is possible, as Kant himself suggests, that as concerns constructivism he is working with an original idea he knows how to use but does not know how to describe. It is also possible that he is not entirely clear about the differences between representationalism and constructivism, or even that he sometimes inconsistently espouses both approaches.

Though Kant insists on a priori cognition, he bases the argument for a new approach squarely on experience. Despite his “official” reliance on theory formulated prior to and apart from experience, Kant inconsistently believes we can judge a science by its results. He clearly bases his Copernican experiment on the lack of results following from the venerable assumption, consistent with the ancient Parmenidean view, that cognition must conform to its object. It is because there has never been progress on this assumption that he proposes that, as an experiment, we assume that the object must conform to our cognition.

Kant’s suggestion is similar to the indirect mathematical proof, which relies on excluding one possibility to demonstrate the other. Since he has formulated the theory based on the hypothetical assumption that the object must conform to cognition, Kant—perhaps with the indirect mathematical proof in mind—thinks this approach is no longer hypothetical but rather true.

Kant does have in mind historical precedent—in the contributions of Galileo and Copernicus—in creating modern science. He remarks that Galileo and others understood in the seventeenth century in creating modern natural science that “reason has insight only into what it itself produces according to its own design” (CPR, B xii-xiii, p. 109). If reason can know only what it constructs, then it obviously cannot rely on a representational strategy to cognize the external world. Yet Kant’s proposed solution via the construction of cognitive objects is unclear. In the B preface, he surveys four forms of cognition. Kant (who here abandons his preference for the a priori) suggests that, in their own way, logic, mathematics, and natural science are already on the secure road of science. This is an a posteriori criterion Kant here apparently adduces, by contradicting his claim to base knowledge on a priori grounds, in order to ensure the acceptability of any and all claims to know. Kant’s approach to placing the future science of metaphysics on this road, which appears to be multiply-determined, includes: his concept of knowledge as apodictic, hence unrevisable; his view of mathematics, especially ancient geometry; and his reading of the rise of modern natural science, as well as other factors.

According to Galileo, the universe “is written in the language of mathematics, and its characters are triangles, circles, and other geometric figures.”45 Many observers believe that mathematics is central to the rise of modern science. Kant’s conception of mathematics is controversial. In the account of “The Doctrine of Method” late in the first Critique, he affirms the traditional view of mathematics as an exceptional example of pure reason—in short, as what is sometimes still called the so-called queen of the sciences. Kant further draws attention to the distinction between philosophy, which analyzes concepts, and mathematics, which constructs concepts (CPR, B 741, p. 630).

Though he stresses the difference between philosophy and mathematics, Kant apparently takes mathematics, especially geometry, as his philosophical model. According to Kant, mathematics and philosophy are mirror images of each other. Mathematics considers the universal in the particular, and philosophy considers the particular in the universal (CPR, B 742, pp. 630–31). Construction takes place through intuition; that is the only way an object is given (CPR, B 747, p. 633). Geometrical construction yields a particular, which is an instance of the general concept.46 This point is further supported by Kant’s claims in Metaphysical Foundations of Natural Science (1786) that because, as he writes, “the possibility of determinate natural things cannot be cognized from their mere concepts . . . it is still required that the intuition corresponding to the concept be given a priori, that is, that the concept be constructed, which is a task that requires mathematics.”47 This passage, which is Kant’s justification for the important claim that “in any special doctrine of nature there can be only as much proper science as there is mathematics therein,”48 is clearly basic to his general view of knowledge.

The discovery of non-Euclidean geometry required changes in claims about the relation of geometry to the world. Before the discovery of non-Euclidean geometry, it seemed obvious to observers that there was only a single kind of geometry, which was also the geometry of the world. Later mathematical developments cast doubt on aspects of Kant’s view of mathematics, especially geometry. Kant not surprisingly shares the view of mathematicians of his time that Euclidean geometry is a priori and synthetic. He relies on the inference, which was correct at the time but later shown to be mistaken, that there can be only a single geometry to infer that it describes real space. The rise of non-Euclidean geometry demonstrates that alternatives are possible which can describe the properties of physical space as accurately as Euclidean geometry.49 In Einstein’s general theory of relativity, for instance, the curvature of light rays in a gravitational field is often believed to show that space is non- Euclidean. Yet it is unclear that Kant would need to revise the view I am attributing to him that geometrical construction is the exemplar of a constructivist approach to cognition.

A reading of geometry as the cognitive exemplar agrees with Kant’s reading of mathematics, natural science, and the future science of metaphysics. In the B preface to the first Critique, he contends that mathematics and physics determine their objects a priori, the former entirely a priori and the latter partly so. As concerns mathematics, he gives as an example the a priori construction of a figure. In natural science “grounded on empirical principles,” he thinks we must rely on principles according to so-called “constant laws,” or laws of nature, when depending on “the agreement among appearances.” At stake is the constructivist view that “reason has insight only into what it itself produces according to its own design” (CPR, B xiii, p. 109). If we recall the importance of geometrical construction, then it is plausible Kant may be thinking of the successful application of mathematics to nature in the seventeenth century.

The difficulty evidently lies in ascertaining what kind of cognitive inference, if any, can be drawn through such an approach. There are many views of intuition. Kant’s constructivist approach relies on intuitive construction in mathematics and natural science. Plato relies on intellectual intuition, which Kant later rejects, for knowledge of the forms. In Rules for the Direction of the Mind, Descartes distinguishes between deduction and intuition as the two main approaches to knowledge. For instance, in the third rule he calls for the examination of what we can intuit or deduce with certainty. Modern intuitionism is an approach to mathematical thinking based on mental construction, invented by Kronecker, Poincaré, and especially Brouwer. Euclidean geometry is constructive, as already noted, in a mathematical sense. Kant apparently believes Euclidean geometry is true a priori for two reasons: it can be formulated, and there is no possible alternative. Since he fails to consider alternative scientific theories about the same phenomena, he similarly seems to believe that agreement among appearances can count as laws. By implication, from the Kantian perspective, a cognitive alternative—for instance, an alternative to Newtonian physics that later emerged in relativity theory—is impossible.

Kant, who thinks the constructivist approach has brought mathematics and physics to the secure path of science, relies on this approach as the basis of the future science of metaphysics. He undertakes a revolution to put metaphysics on the secure path of a science. His Cartesian strategy consists in going beyond constant disputes through the appropriate method, whose correct application guarantees objective cognition. His constructivist insight, which is apparently the same idea underlying his understanding of Euclidean geometry, is that construction in intuition functions as a proof of the existence of the object, in this case as a proof of a priori knowledge in metaphysics.

We have noted above that Kant’s approach is analogous to indirect mathematical proof, which supposes that there are two and only two possibilities and that one is false. Kant similarly claims that in metaphysics that there are two and only two cognitive strategies. He further claims that our experience so far has failed to reveal, and is unlikely to reveal, a way to demonstrate the claims of metaphysics on the assumption that cognition conforms to the cognitive object. In a seminal passage, he writes: “Up to now it has been assumed that all our cognition must conform to the objects; but all attempts to find out something about them a priori through concepts that would extend our cognition have, on this assumption, come to nothing. Hence let us once try whether we do not get farther with the problems of metaphysics by assuming that the objects must conform to our cognition which would better agree with the requested possibility of an a priori cognition of them” (CPR, B xvi, p. 110).

This passage is significant in a number of ways. It points to a deep tension in Kant’s position, since he clearly but perhaps inconsistently intends to ground a priori knowledge on the a posteriori plane—more precisely, on the results of experience. Second, it initially presents Kant’s novel approach to metaphysics as nothing more than an experiment. Then, it introduces an anthropological element in positing that cognitive objects must correspond to the structure of the human mind. Further, it follows Kant’s understanding of Euclidean geometry as the geometry of the world. In constructing geometrical figures, it is possible to anticipate a priori what according to Kant must hold true of experience, and hence must be confirmed on the a posteriori plane. Finally, it crucially suggests the utility of reversing the usual approach in which knowledge depends not on the subject but rather on the object.

This reversal has a famous precedent, to which Kant immediately refers in likening his procedure to Copernican astronomy. Kant compares his metaphysical approach to “the first thoughts of Copernicus, who, when he did not make good progress in the explanation of the celestial motions if he assumed that the entire celestial host revolves around the observer, tried to see if he might not have greater success if he made the observer revolve and left the stars at rest” (CPR, B xvi, p. 110). Though elsewhere Kant espouses an a priori approach, he seems here to be suggesting that cognition, which is a priori, is based on a trial-and-error process in selecting the appropriate method. He calls attention to the similarity between his metaphysical view and the Copernican astronomical approach in famously writing that “If intuition has to conform to the constitution of our faculty of cognition, then I do not see how we can know anything of them a priori; but if the object (as an object of the senses) conforms to the constitution of our faculty of intuition, then I can very well represent this possibility to myself” (CPR, B xvii, p. 110).

Kant’s argument obviously rests on a complex analogy linking his approach to metaphysics with Euclidean geometry and Copernican astronomy. The reactions to this analogy are very disparate. Suffice it to say that according to Kant, Euclidean geometry and Copernican astronomy both employ constructivist cognitive approaches. From the former he takes the view that the a priori construction of instances is necessarily confirmed in experience. From the latter he takes the idea that experience, hence knowledge, is not independent of anthropological considerations since it depends on the constitution of the human mind. Kant sums up the argument in writing that “we can cognize of things a priori only what we ourselves have put into them” (CPR, B xviii, p. 111).

Kant’s constructivist approach, which is not often analyzed in detail, is often misunderstood. His interesting suggestion that someone who discovers an original idea often does not fully grasp it perhaps applies to Kant himself. It is arguable that Kant does not understand or fully understand how to expound his Copernican insight (see CPR, B 862, p. 692). It is plausible that his original idea must, as he says, be explained according to reason, but not according to the way he formulates it.

Cognitive constructivism is based on a type of identity—more precisely, a view of identity in difference. Identity in difference, which differs from identity tout court, goes all the way back at least to early Greek philosophy. The difference lies in the stress on activity—more precisely, the subject’s activity as the unity, which is neither subjective nor objective, and which subtends the difference between them. This concept appears as early as Aristotle. Aristotelian activity (energeia) is apparently Aristotle’s answer to Plato’s supposed inability to grasp the relation between forms and appearances. The Aristotelian view of unity as subtending difference anticipates a theory of practice, which later reappears in many related guises. Aristotle, for instance, understands the many episodes of an individual life as forms of activity, as well as human life in general as an activity. This concept of unity in difference later becomes the basis of cognitive constructivism.

In one form or another, identity theory runs like a red thread through the views of the main post-Kantian German idealists. This claim, which is easily misunderstood, requires clarification. There are different kinds of identity theory and different kinds of identity. Plotinus is said to deny any difference between the representing and the represented.50 Through the principle of the identity of indiscernables, Leibniz contends that no two distinct things exactly resemble each other.51 Heidegger insists on an ontological difference. Identity theory is now frequently linked to reductionism; for instance, the reduction of the mind to the brain.52 Thus the correspondence theory of truth is sometimes criticized on the grounds that statements about reality fall short of it. Frege, for instance, distinguishes between things and ideas in denying that things can correspond to ideas.53

It is useful to call attention to the link between constructivism, or the claim that the subject constructs its cognitive object as a minimal condition of knowledge, and so-called identity theory.54 This claim refers to a form of identity between the subject that knows and a cognitive object that is known—or in another formulation, between thought and being. Kant’s suggestion that a cognitive subject cannot know a mind-independent object points toward a theory based on the opposite claim: in knowing, one knows oneself, or more precisely, oneself in the form of externality. In other words, Kantian constructivism supposes a form of identity.

In different ways, this concept echoes through post-Kantian German idealism. The post-Kantian German idealists each restate a form of the Copernican idea that we know only what we construct. Kant’s transcendental deduction culminates in a conception of the cognitive subject, or synthetic unity of apperception, as the highest point of the understanding, logic, and transcendental philosophy (see CPR, B 134, p. 247). If a minimal condition of knowledge is that the subject must construct the object, then in a sense, in order to be specified, subject and object are both different as well as identical. In other words, there is a metaphysical identity in the difference between the subject that knows and the object that it knows.

We can illustrate this concept with respect to the cognitive object, which derives from two sources: sensation through which the subject is affected, and which provides the matter of the cognizable object; and the form of the object that arises from the imposition of the categories situated in the understanding. Hence, by knowing, the subject in part knows what is external to it (sensation), which it cognizes as the object; and in part knows itself—or is self-conscious, so to speak, of what it does by constructing the cognitive object. In other words, according to the constructivist approach, the cognitive subject is both conscious and self-conscious.

German idealist constructivism, which is routinely described as an identity theory (Identitätstheorie), in fact rejects it. Though it appears to require an identity between subject and object, knower and known, thought and being, it refuses a pure unstructured identity in favor of an identity between identity and difference. In different ways, this approach runs throughout the entire Western tradition. It is, for instance, central to Platonism, which appears to build on or at least to presuppose the Parmenidean view that to know requires a grasp of mind-independent, unchanging reality lying beyond mere appearance, presumably through the identity between one or more ideas in the mind and mind-independent external reality. Parmenides is often depicted as the main progenitor of the Platonic view that we cannot know appearance. This Platonic view is clearly denied in the Kantian claim that knowledge is restricted to appearance only. “Appearance” in this context designates what is given in experience in phenomenal form as distinguished from a representation of mind-independent reality.

Kant’s Transcendental Deduction, Identity in Difference, and German Idealism

“The Transcendental Deduction of the Categories” illustrates Kant’s constructivist approach to cognition. It is then no accident that §14, in which Kant describes “The Transition to the Transcendental Deduction of the Categories,” and §27, in which he presents an account of “The Result of the Deduction of the Concepts of the Understanding,” both begin by restating Kant’s Copernican insight: it is not the mind-independent object that makes cognition possible; rather, it is the construction of the object in the mind—the result of bringing the contents of sensory intuition under the categories—that makes the object possible.

To conclude the discussion of Kant, it will be useful to say a few words—but no more than that—about how his constructivist approach supposedly works in practice with particular reference to the transcendental deduction. I have suggested that, as the name suggests, Kantian constructivism turns on the construction of a cognitive identity in difference. Different conceptions of identity in difference run throughout Kantian and post-Kantian constructivist approaches to cognition. Kant works out his view of this identity throughout his position. His metaphysical and transcendental deductions both serve to justify a priori cognitive claims. The two deductions together form the Kantian answer to the transcendental question about the conditions of the possibility of knowledge. Speaking generally, the metaphysical deduction constitutes Kant’s effort to identify the categories or pure concepts of the understanding. According to Kant, there are as many logical functions of the understanding as there are what he calls logical functions of all possible judgments. Kant’s claim that there are in fact twelve and only twelve logical functions of judgment as well as twelve corresponding categories suggests closure.55 The transcendental deduction, which follows the metaphysical deduction, is intended to exhibit the possibility of “a priori cognitions of objects of an intuition in general” (CPR, B 159, p. 261); in other words, the construction of the so-called appearance—or more precisely, the phenomenal object—on the basis of sensation in principle derived from the mind-independent object (which, however, does not and cannot appear).

The metaphysical deduction supposedly identifies the categories of the mind and the transcendental deduction informs us about how it works. The transcendental deduction, which presupposes the metaphysical deduction, is central to the critical philosophy as well as one of the most complex and frequently analyzed parts of the text. The transcendental deduction, which is often regarded as the centerpiece of the Critique of Pure Reason, has been extensively studied in what has by now become a cottage industry. It would go beyond the limits of the present account to propose still another reading of this complicated passage. It will suffice at present merely to bring out that the Kantian account of the construction of the cognitive object in fact turns on the construction of an identity in difference and hence, at the risk of repetition, justifies the theoretical claim embodied in the Copernican turn: we know only what we construct.

In simplest terms, the transcendental deduction provides an account of the cognitive object as the product of an interaction between the mind-independent real (variously designated as the thing in itself or noumenon), and the cognitive subject (or again, the transcendental subject). This interaction results in an identity in difference through a conceptually unified and knowable cognitive object including subject and object, form and content, or again external input through sensation and internal output through the activity of the understanding, through which, in bringing sensations under the categories, or rules of synthesis, the cognitive object is constructed.

Kant’s argument in “The Transcendental Deduction” resembles his view of Euclidean geometry, which depends on the possibility of constructing a single instance of a class. If it turned out that Kant could deduce the categories but could not justify their function in the construction of the cognitive object, then he would be unable to justify the Copernican turn.

Kant’s approach is synthetic rather than analytic, hence based not on the analysis of a preexisting mind-independent object, but rather on its construction as a condition of its cognition. There is a superficial analogy between the Kantian approach and certain forms of twentieth-century phenomenology, which restrict themselves to descriptions of what is present to mind only. Kant’s deduction is not and in fact could not be based on description; for instance, on the careful description of how the psychological knowing process unfolds. The transcendental machinery Kant discusses within the framework of the transcendental deduction must operate in order for the cognitive subject to be conscious of the cognitive object or objects. In other words, the transcendental deduction is intended to identify what in Kant’s opinion must occur on the preconscious, rather than on the conscious, level as its necessary precondition. It follows that the transcendental deduction is not descriptive but rather based on Kant’s speculative reconstruction of what according to him must be the case in order for cognition to be possible.

I will end this account of Kant’s cognitive theory in noting his own estimation of his achievement. There is a modern idea of philosophy as a historical discipline. According to this view, philosophy arises in an ongoing debate in which observers react to others in seeking to advance the discussion. This conception of philosophy as a historical discipline arose only in the debate after Kant. Kant holds an ahistorical—even antihistorical—view of his accomplishment and of philosophy in general. He thinks his preferred cognitive approach is not one among other possible approaches but rather the only possible solution. In this sense, Kant is basically opposed to, say, Hegel—who is often credited with the view that in his position, philosophy comes to a high point and to an end, but who also rather thinks that his position (like all others before and presumably after it) belongs to the history of the tradition.

Philosophers are not notable for their conceptual humility. Perhaps under Kant’s influence, Husserl obviously pretends finally to make a true beginning. Kant clearly casts himself in the extraordinary role of the thinker who both finally begins and ends the philosophical tradition. He supposedly begins the philosophical tradition since all prior philosophical theories are merely dogmatic but undemonstrated assertions. The critical philosophy differs from its predecessors in that it provides an antecedent critique of its own capacity, or the capacity of pure reason (CPR, B xxxv, p. 119). If all prior philosophy is dogmatic, then by definition, philosophy worthy of the name would begin with Kant. His suggestion that there can never be more than a single true theory indicates that his position is the only one worth taking into account. In short, Kant can be understood as claiming that in the critical philosophy he both initiates and brings to an end philosophy worthy of the name through the absolutely definitive solution of the cognitive problem, a solution that since it is correct cannot be revised, a solution intended to stand forever. Since the critical philosophy is the first example of a philosophy that is critical, it also marks the true beginning of philosophy. His critical philosophy further brings the tradition to an end since it in fact presents the unique solution to the cognitive problem. According to Kant, his position is final and unalterable, so that the cognitive debate, which he claims to end, will not and in fact simply cannot be resumed at some future time. He concedes that his presentation can be improved (for instance, stylistically) while calling attention to the link between his system and human reason in all its many forms. According to Kant, to change anything—anything at all—in his position would introduce contradictions into his system as well as so-called universal human reason (CPR, B xxxviii, p. 120).

There is an obvious difference between a theory that advances the ongoing debate and one that effectively brings the discussion to an end. Kant’s view that he has ended the philosophical tradition in the critical philosophy appears to be multiply determined. This reflects his normative views of mathematics, science, and philosophical method. His interest in a supposedly unique solution to philosophical problems may also reflect his natural scientific background, according to which there is supposedly a unique solution for every problem. Though Newton had his critics, it is conceivable that when he brought out the Principia, at least some observers thought that physics—hence natural science—had reached a high point and an end. Kant seems to hold a view that a form of cognition that is really on the secure road to science would never later need to be revised. This conviction almost certainly also reflects his view of the a priori approach to cognition as a source of apodictic, or universal and necessary, cognition.

Like the statue of Ozymandias, Kant’s ahistorical view that he has constructed a position that will stand forever was later eroded. There is a clear difference between Kant’s belief that he proposes the permanent, unalterable solution to the cognitive problem and the shared post-Kantian idealist view that he advances the debate in important ways. Kant obviously thinks he has decided the cognitive question once and for all. However, the post-Kantian German idealists believe he put on the table an important suggestion that constitutes a huge step forward, but does not end the debate. His successors believe his proposed solution must neither be ignored nor accepted as is, but rather needs to be criticized, reformulated, and hence carried forward and completed. The post-Kantian German idealists were, in different ways, all occupied with this general task.