The Rebirth of Metaphysics
Before true philosophy can come to life, the old one must destroy itself. (10:57)
When it came to reforming the sciences, Kant’s plans were specific. For knowledge to move forward, Kant argued, certain limits had to be set, a circumscription that began with two questions. First, one must ask what kind of knowledge would be required in order to solve a given problem, and second, one must decide whether that knowledge was in fact possible (10:56). When one searched for vital principles or inquired into the character of the spirit world, the objects of investigation were simply unknowable. Kant was, moreover, ready to diagnose and name the exact source of so much error in these sciences. “Surreptitious concepts”—or “subreptive axioms” as they would later be called—described a specific transgression: the crossing of fields of knowledge meant to be separate.187 Subreption created confusion in the sciences when investigators took concepts gleaned from experience, the experience of magnetic forces, for example, and used them to describe processes that were incapable of experience, processes like embryological formation or the means by which spirits might communicate. As Kant introduced the problem in Dreams of a Spirit-Seer, “There are many concepts which are the product of covert and obscure inferences made in the course of experience; these concepts then proceed to propagate themselves by attaching themselves to other concepts, without there being any awareness of the experience itself on which they were originally based or of the inference which formed the concept on the basis of that experience. Such concepts may be called surreptitious concepts [erschlichene Begriffe]” (2:321). Kant actually charged himself with having made this mistake as well, given that both his New Elucidation (1755) and the Physical Monadology (1756) had ascribed the forces of attraction and repulsion to spirits and monads (1:415, 484). As an act of contrition, he was now ready to declare, “It is impossible for reason ever to understand how something can be a cause, or have a force; such relations can only be derived from experience.” Indeed, “All judgments, such as those concerning the way in which my soul moves my body, or the way in which it is now or may in the future be related to other beings like itself, can never be anything more than fictions” (2:370, 371).188
These were all of course topics that had been the special province of metaphysics, so one immediate problem was to consider what might be left for the metaphysician to do without them. Kant was prepared for this. “Metaphysics,” he began, “with which, as fate would have it, I have fallen in love but from which I can boast of only a few favours, offers two kinds of advantage.” The first advantage was the ability of metaphysics to aid reason in spying out the hidden properties of things; indeed, when left unfettered, reason was unrivaled in its capacity for such inferences. It was in fact this talent for inferences that had led to the very problem facing metaphysics now. But this was balanced by what Kant took to be the second advantage afforded by metaphysics. As Kant described it
The second advantage of metaphysics is more consonant with the nature of the human understanding. It consists both in knowing whether the task has been determined by reference to what one can know, and in knowing what relation the question has to the empirical concepts, upon which all our judgements must at all times be based. To that extent metaphysics is a science of the limits of human reason. A small country always has a long frontier, it is hence, in general, more important for it to be thoroughly acquainted with its possessions, and to secure its power over them, than blindly to launch on campaigns of conquest. (2:367–368)
The key here was to see that by redefining metaphysics as a science of the limits of human reason, Kant had radically redirected investigators away from the topics of life and soul toward an examination of reason itself. Once all investigations were to be prefaced by a separate inquiry regarding the abilities of reason, moreover, the task of determining the nature and limits of cognition became both necessary and ultimately identical to the task of determining the nature and limits of scientific investigation itself.
As Kant mapped the contours of this investigation, he took its outcome, at the very least, to be the elimination of surreptitious concepts. “All of my endeavors,” Kant explained, “are directed mainly at the proper method of metaphysics (and thereby also the proper method for the whole of philosophy)” (10:56). The method would determine the scope of reason’s possibilities—and thereby also the scope of any rational investigation—and eradicate surreptitious concepts as a result. If it was certain that a genuine insight into organic processes was impossible, then the surreptitious appeal to an “irritable force,” as made by Haller for example, should be avoided when explaining muscle contraction. This did not mean an end to further investigations in the life sciences, but it did mean, as noted earlier, that only “the appearances of life in nature, and the laws governing them, [would] constitute the whole of that which it is granted us to know” (2:351). Naturalists could focus on the regularity of nature’s appearances and continue their search for mechanical causation, but the search for vital principles and the attempt to understand the mysteries of generation and reproduction were invariably riddled with surreptitious concepts and thus doomed from the start.
But while Kant was confident in his diagnosis of the need for reform in the sciences, he was still unsure of the task left before him. “My problem is this,” he wrote. “I noticed in my work that, though I had plenty of examples of erroneous judgments to illustrate my theses concerning mistaken procedures, I did not have examples to show in concreto what the proper procedure should be” (10:56). So far as the “proper procedure” amounted to precisely delineating the limits of reason, what Kant was missing, in other words, was a positive description of reason itself.
Up until now Kant had written extensively upon questions connected to cosmological origin, and, his criticisms notwithstanding, he was thoroughly versed in the leading theories of biological origins as well. As he now took on the job of re-creating metaphysics as a science of the limits of human reason, the first task concerned questions regarding the origins of knowledge. Was it the case, as rationalists had it, that true ideas were like seeds implanted in the soul by God—a strategy in some sense parallel to that adopted by the preexistence theorists—or were empiricists correct instead when identifying the senses as the true origin of ideas? Kant was long familiar with the rationalists’ reliance on the intellectual intuition of innate ideas, and as for the other option, the mid-1760s were perhaps the heyday of Kant’s engagement with British empiricism.
Kant’s closest friend during this period was the British merchant Joseph Green, a merchant known for his literacy in the writings of Hume and other members of the Scottish enlightenment.189 This was surely a topic of shared interest, for in Kant’s description of his ethics lectures planned for 1765–1766, for example, he wrote, “The attempts of Shaftsbury, Hutcheson, and Hume, although incomplete and defective, have nonetheless penetrated furthest in the fundamental principles of all morality” (2:311). We know, moreover, that by the mid-1760s Kant was well acquainted with a 1755 translation of Hume’s Enquiry Concerning Human Understanding as well. “Since the Essays of Locke and Leibniz, or rather since the origin of metaphysics so far as we know its history,” Kant later declared, “nothing has ever happened which could have been more decisive to its fate than the attack made upon it by David Hume. He threw no light on this kind of knowledge; but he certainly struck a spark from which light might have been obtained, had it caught some inflammable substance and had its smouldering fire been carefully nursed and developed” (4:257). Indeed, taken as whole, 1766’s Dreams of a Spirit-Seer reverberated with the potency of Hume’s skepticism regarding the “dreams of metaphysics,” and there was thus real justice in Kant’s eventually citing Hume for having woken him from a dogmatic slumber (4:260). But while Kant’s adoption of a skeptical methodology put him in position to recognize a need for reform in the sciences, the work facing him now required him to take a positive stance regarding the workings of the mind, and for this Kant turned first to the work of Leibniz and Locke.
Like Hume’s, Locke’s works were both available and well known in Germany at this time. Georg Kypke had already been a longtime friend of Kant’s when he translated Locke’s posthumously published addendum to the Essay Concerning Human Understanding in 1755.190 And by then too Kant would have had access to the new Latin translation of the Essay itself, a book he described as “the ground of all true logica” (24:37).191 Indeed, as Kant later explained in his course on logic, “Some books are of great importance and require considerable inquiry; these one must read often, e.g., Hume, Rousseau, Locke, who can be regarded as the grammar of the understanding, and Montesquieu, concerning the spirit of the laws” (24:300). By the end of his career, references to Locke would be peppered throughout Kant’s notes, lectures, and published writings.192
In Kant’s notes, in particular, Locke would frequently be paired with Leibniz. The two were typically cited for their investigations into the origin of ideas, and in the years after Dreams of a Spirit-Seer this made them significant interlocutors for Kant in his attempted reform of the sciences. Although it is unknown when exactly Kant read through the posthumous publication of Leibniz’s New Essays (1765), it seems likely that by 1770 he was at least familiar with the first page of it, since Leibniz’s opening formulation of the divisions of philosophy would be subsequently repeated by Kant on numerous occasions.193 As Leibniz positioned himself there against Locke
Although the author of the Essay says hundreds of fine things which I applaud, our systems are very different. His is closer to Aristotle and mine to Plato. . . . Our disagreements concern points of some importance. There is the question whether the soul in itself is completely blank like a writing tablet on which nothing has as yet been written—a tabula rasa—as Aristotle and the author of the Essay maintain, and whether everything which is inscribed there comes solely from the senses and experience; or whether the soul inherently contains the sources of various notions and doctrines which external objects merely rouse up on suitable occasions, as I believe and as [does] Plato.194
Kant accepted Leibniz’s divisions, often visually grouping them—Aristotle in a column with the empiricists, Plato with the intuitionists—when writing out his lectures. But where did that leave Kant? In 1769 Kant appeared to be torn as to how to proceed with his investigation into the origin of knowledge, accepting, on the one hand, Locke’s dictum regarding sense as the necessary occasion for all thought, and Leibniz’s admonishment, on the other hand, regarding the impossibility that a concept of God could ever arise from the senses. As Kant outlined his own view of cognition, the picture thus presented an amalgamation of Leibniz and Locke:
Some concepts are abstracted from sensations, others merely from the law of the understanding for comparing, combining, or separating abstracted concepts. The origin of the latter is in the understanding; of the former, in the senses. All concepts of the latter sort are called pure concepts of the understanding, conceptus intellectus puri. We can of course set these activities of the understanding in motion only when occasioned to do so by sensible impressions and can become aware of certain concepts of the general relations of abstracted ideas in accordance with the laws of the understanding; and thus Locke’s rule that no idea becomes clear in us without sensible impression is valid here as well; the notiones rationales, however, arise no doubt by means of sensations and can also only be thought in application to the ideas abstracted from them, but they do not lie in them and are not abstracted from them (sind nicht von ihnen abstrahirt). (17:352)
Kant thus remained in keeping with Locke insofar as even the “pure concepts of the understanding” were concepts empirically gleaned through reflection on the contents of sense. These concepts were to be distinguished—as “abstracted” from sense—from those rational notions whose origin would have to be different.
When Kant was offered a chair in logic and metaphysics the following year, he was required to present an “inaugural dissertation,” a piece that would offer him the opportunity to fulfill his promised reform of metaphysics. When it was completed, Kant’s solution to the problems of origin and subreption stood at the forefront of the project. It was a solution, at least with respect to subreption, that could be accomplished in one stroke by the radical division Kant now proposed between the faculties of sense and intellect. Insisting that the method of metaphysics concern itself wholly with the prevention of “subreptive axioms,” Kant argued that only a radical separation between sense and intellect could avoid the possibility of such cross-contamination; a separation effectively announced in the book’s title: On the Form and Principles of the Sensible and the Intelligible World.
When Kant first broached the problem of subreption in 1764–1765, he had sent his thoughts—including those already mentioned regarding the problem of having nothing more than negative examples to illustrate his point—to J. H. Lambert, a philosopher and mathematician living in Berlin. Lambert had agreed with Kant’s call for reform, observing in reply that “whenever a science needs methodical reconstruction and cleansing, it is always metaphysics” (10:62). Turning to the subject of Kant’s proposal, Lambert had made a number of remarks:
The first concerns the question whether or to what extent knowing the form of our knowledge leads to knowing its matter. This question is important for several reasons. First, our knowledge of the form, as in logic, is as incontestable and right as is geometry. Second, only that part of metaphysics that deals with form has remained undisputed, whereas strife and hypotheses have arisen when material knowledge is at issue. Third, the basis of material knowledge has not, in fact, been adequately shown. . . . Fourth, even if formal knowledge does not absolutely determine material knowledge, it nevertheless determines the ordering of the latter, and to that extent we ought to be able to infer from formal knowledge what would and what would not serve as a possible starting point. (10:64)
Lambert’s emphasis on the connection between form and matter proved highly influential for Kant, as was clear from the start of Kant’s finished text. The Dissertation opened, for example, with a discussion of the concept of a world, in particular of “its two-fold origin in the mind” (2:387). While Locke had been correct with respect to the matter of sensations, Kant now argued, form was a result of mental determination and thus lay in the mind. This was not to end up on the side of Leibniz against Locke, however; on the contrary, Kant was suggesting something new. The imposition of form and the supplying of content described a division of labor that would now be applicable to both sense and intellect alike.
Kant started with the case of sensitive knowledge, explaining that in “representations of sense there is in the first place something that we may call matter, i.e., sensation [sensatio], and something else that we may call form, i.e., the sight [species] of sensible things, which obtains when various things which affect the senses are co-ordinated by a certain natural law of the mind” (2:392–393). The process by which form was imposed upon sensation followed Locke’s model of reflective comparison, a process described by Kant as the result of the “logical use” of the understanding, whereby sensations could be classified or subordinated under common class concepts. Unsorted sensations remained at the level of appearance (apparentia), according to Kant, whereas “the reflective cognition which arises from the intellectual comparison of a number of appearances is called experience” (2:394). This somewhat borrowed account of sensible cognition sat alongside Kant’s genuine innovation, the identification of space and time as the “schemata and conditions of all human knowledge that is sensitive” (2:398).
Two years earlier Kant had argued that space was absolute; now space and time were jointly identified as the formal, yet subjective, principles of the phenomenal world. These were the principles underlying mechanics and geometry, fields that each yielded undisputable truths. With this move Kant repositioned the status of sensitive knowledge. For although he insisted that we remember the sensible origin of even the most abstract laws of sensible phenomena, his insistence concerned the specter of subreption, not the quality of sensible knowledge. “There is thus a science of sensible things,” as he put it, one that “yields us quite genuine knowledge, and at the same time furnishes a model of the highest certainty for knowledge in other fields” (2:398). Against a critique like Wolff’s regarding the confused perceptions of the senses, therefore, Kant presented a theory of sensible cognition that explained the success of geometry and mechanics—at the same time that it could be validated by that success—and grounded the possibility of certainty regarding sensible phenomena.
Kant’s discussion of sensible cognition proceeded in stages. The first concerned Kant’s shifting the focus away from objects of perception toward our mental representations of them, since representations alone were susceptible to the mind’s imposition of form. A representation, Kant explained, “indicates a certain aspect or relation of the sensa and yet is not properly an outline or schema of the object, but only a certain law inborn in the mind coordinating with one another the sensa arising from the presence of the object” (2:393). The task for Kant was to balance the quality of the real, one granted by the material content of a sensation, with the opportunity for control of that content through the mind’s inborn laws. Once laws for the mental construction of phenomena became too thorough, Kant realized, the account would risk charges of idealism (2:397). Leaving these difficulties aside, Kant concentrated instead on his argument for sense certainty.
A genuine knowledge of sensible phenomena was possible, according to Kant, because judgments about sensible objects fell under the purview of the logical use of the understanding, and the logical use of the understanding was concerned only with determining the internal agreement between subjects and predicates in judgments. By focusing, therefore, on the internal relationship between subject and predicate over the supposed, but unknowable, external connection between subject and object, certainty regarding phenomena could be guaranteed by the proper functioning of the mind’s laws for construction. “Consider judgments about things sensitively known,” Kant began. “The truth of a judgment consists in the agreement of its predicate with the given subject. But the concept of the subject, so far as it is a phenomenon, can be given only by its relation to the sensitive faculty of knowledge, and it is also by the same faculty that the sensitively observable predicates are given. Hence it is clear that the representations of subject and predicate arise according to common laws, and so allow of a perfectly true knowledge” (2:397). Phenomena, for Kant, were thus the synthetic result of sensible matter and the mind’s imposition of form, a synthesis accomplished through laws grounding the certainty of sensible experience.
But what, precisely, was Kant’s understanding of these laws for the logical coordination of sensible data, laws that were said to be inborn (innatas) in the mind? Were they meant to balance Kant’s deference to Locke regarding the independent reality of material sensations? For now, at least, Kant left the status of these laws unexamined. Not so the concepts of space and time. These would fall into a third category, somewhere between the sensible acquisition of empirical concepts and the mental recovery of innate ideas; space and time, according to Kant, were “originally acquired” so far as they were generated by the mind itself. Asking rhetorically whether space and time were connate (connatus) or acquired (acquisitus), Kant immediately rejected the possibility of their empirical acquisition so far as that would render geometry contingent, something it clearly was not. The alternative was “not to be rashly admitted,” either, however, “since in appealing to a first cause it opens the path to that lazy philosophy which declares all further research to be in vain” (2:406). Instead, Kant argued that the origin of space and time lay between these alternatives:
Both concepts are without doubt acquired, being abstracted not from the sensing of objects (for sensation gives the matter, not the form, of human cognition) but from the action of the mind in coordinating its sensa according to unchanging laws—each being, as it were, an immutable type to be known intuitively. Though sensations excite this act of the mind, they do not influence the intuition [non influunt intuitum]. Nothing is here connate save the law of the mind, according to which it combines in a fixed manner the sensa produced in it by the presence of the object. (2:406)
It was clear that much rested, therefore, on the laws of the mind: the regularity of their logical operations generated space and time as the pure forms of sensible intuition, and through their subsequent coordination of sense data, they grounded the possibility of sense certainty.
The notion that concepts might be generated or “originally acquired” through the workings of cognition marked Kant’s major advance from the position he had outlined in 1769. At that point, even the “pure concepts of the understanding” fell under the Lockean model of concepts gleaned by abstraction from sense (17:352). On Locke’s view, it was in fact this “gleaning,” so to speak, that constituted the main work of the understanding. Things had clearly changed for Kant by the time he composed the Dissertation. The difference between intellectual concepts—“possibility, existence, necessity, substance, cause, etc. with their opposites or correlates”—destined for the “real use” of the intellect (2:395) and the sensitive concepts of space and time controlled by its “logical use” (2:398) were enormous, according to Kant, but their birthplace was the same. Like the concepts of space and time, Kant considered intellectual concepts to have been given “in the very nature of the pure intellect, not as concepts connate to it, but as concepts abstracted (by attention to its actions on the occasion of experience) from laws inborn in the mind, and to this extent, as acquired concepts” (2:395).195 The difference between sensible and intellectual concepts lay, therefore, in their objects, not their origin, even if “each kind of knowledge preserves the mark of its descent, so that the former kind, however distinct, is on account of its origin called sensitive, while the latter, however confused, remains intellectual” (2:395).
While the sensitive concepts of space and time grounded an experience of phenomena that was capable of staving off skepticism—“The laws of sensibility will be laws of nature, insofar as nature falls within the scope of the senses” (2:404)—the case was different for intellectual concepts. Sensitive concepts were applied to sensible intuition, the “apparentia” waiting to be organized into a coherent experience; intellectual concepts, by contrast, had no intellectual intuition with which to work. This explained the clarity of geometry when compared to the obscurity surrounding the traditional content of metaphysics. Kant emphasized this restriction, moreover, for it was precisely such overreaching that had opened the door to the surreptitious application of sensitive concepts in the first place. “No intuition of things intellectual but only a symbolic knowledge of them is given to man,” he declared, for “thinking is only possible for us by means of universal concepts in the abstract, not by means of a singular concept in the concrete” (2:396). While sensation provided direct contact with its contents, the intellect had to work discursively, either through intellectual concepts or through its generation of moral exemplars to guide actions, the exemplars of God and moral perfection, for example. And neither of these possibilities contained the kind of content boasted of by sense. “All the matter of our knowledge is given by the senses alone,” Kant concluded, “whereas a noumenon, as such, is not to be conceived through representations derived from sensations. Consequently, a concept of the intelligible as such is devoid of all that is given by human intuition” (2:396).196
For someone newly interested in reorienting metaphysics toward an account of the extent and limits of human reason, Kant had thus made a good start. Cognition was now described as a twofold exercise, one that was both sensitive and intellectual. The intellect was described as having both a logical and a “real” use, with the former devoted to the task of logical subordination according to laws inborn in the mind. This subordination was responsible for both the discrimination of sense data in the generation of empirical judgments and the logical exercises associated with reflection on the concepts and exemplars generated by the intellect’s real use. “The logical use, but not the real use,” Kant explained, “is common to all the sciences” (2:393). By the real use of the intellect, “the very concepts of objects or relations” were acquired through the nature of the intellect itself (2:393). The intellectual concepts generated in this manner provided concepts of objects in terms of their existence, substance, possibility, necessity, cause, and number; in their so-called dogmatic use, they issued moral exemplars.197 By denying the intellect any content for its intellectual concepts, Kant could argue that he had staved off the path leading investigators to the use of surreptitious concepts. And the attention Kant paid to the difference between an abstractive process yielding empirical concepts and one that could, by its own workings, actually generate or “originally acquire” concepts, identified Kant’s new solution to the problem of understanding the origins of knowledge.
From Original Acquisition to the Epigenesis of Knowledge
Given the focus of the Dissertation, Kant must have been tempted when that year’s topic for the Preisschrift was announced: an essay that could reconcile Descartes and Locke on the origin of ideas.198 But whether Kant’s research agenda for 1770 left him inclined to take up the topic or not, the real question is how he had arrived at his solution to the problem. What models were there for his description of an original acquisition of sensible and intellectual concepts? Kant had worked closely with Buffon’s text when preparing his account of space, but 1768 was also the year that Kant had reported “a deep indifference towards my own opinions as well as those of others” (10:74). And what resources existed for Kant’s discussion of the laws whose workings generated concepts in the first place? Laws like these, or certainly processes with similar functions, were assumed by Locke and Leibniz both; indeed the Leibnizian quip that “for Locke nothing is in the understanding—except the understanding itself” turned on that fact.
As for Leibniz’s account, Kant seemed determined to avoid innate ideas, reproaching, on the one hand, such “rash” recourse to innatism as a type of lazy philosophizing and eliminating, on the other hand, the kind of intellectual intuition that would be required for them. Leibniz’s innatism had turned in part on connections Leibniz saw between Plato’s doctrine of recollection and Leibniz’s own sense that “the seeds of the things we learn are within us—the ideas and the eternal truths which arise from them.”199 Like the divinely implanted seeds, the mind too came from God, an origin explaining its capacity to realize eternal truths in the first place. “So it is not a bare faculty,” as Leibniz characterized the mind in his New Essays, “consisting in a mere possibility of understanding those truths: it is rather a disposition, an aptitude, a preformation, which determines our soul and brings it about that they are derivable from it.”200 As was seen earlier, the preexistence theory of encasement operated for Leibniz as a biological analog to his own theory regarding the formation of ideas. In each case there was a “virgin birth,” so far as both individuals and ideas regarding eternal truths were generated from seeds implanted by God at the creation of the world. But if by 1770 Kant wanted something more than Locke’s account of empirical concepts abstracted from the senses, it is also clear that he wanted something less than the harvesting of truths grown up from seeds that God had sown in the mind.
While it has been fair game to speculate on the source of the “great light” that the year 1769 brought to Kant (18:69), attending to the problem of origin at least points one past the usual suspects. Leibniz was hardly uncommon in his liberal use of vocabularies drawn from both religious and scientific discourses, and his appreciation for Plato aside, Leibniz’s strategy was in fact deeply suggestive of Kant’s own solution to the problem of origin. But whereas Leibniz had appealed to preexistence theory as a biological analog, it seems likely that Kant had some form of epigenesis in mind when describing the mind’s generation or “original acquisition” of concepts. When Kant proposed in 1763 that we forgo supernatural accounts of generation, and mechanical views as well, he had argued that what science needed instead was an explanation that “granted to the initial divine organization of plants and animals a capacity, not merely to develop their kind thereafter in accordance with a natural law, but truly to generate their kind” (2:115). By 1770, Kant was convinced that such an explanation could come only at the cost of subreption. He seems to have felt, however, that the two-step model of divine formation and organic generation could be safely mapped onto a theory of cognition aimed at explaining the generation of concepts from innate laws. The details were still fuzzy. It was not yet clear to Kant, for example, how these concepts were specifically connected to the implanted laws for logical subordination from which they arose, but the strategy epigenesis offered for discovering an origin that was neither supernatural nor empirical was clearly promising.201
In 1769 Kant introduced an explicit discussion of epigenesis in his course on metaphysics. Kant always used A. G. Baumgarten’s Metaphysica as the basis for his course, and the topics concerning the soul ranged from discussions of human understanding to mind-body interaction and the afterlife.202 In a section devoted to the origin of the soul, Baumgarten had rehearsed the reigning theories of organic generation: preexistence, spontaneous generation—Baumgarten’s example here was infusoria—creation ex nihilo, and finally, “concreationism,” according to which the soul was produced through some sort of transfer accomplished by the parents, a position derived from Aristotle’s treatment of the matter. When preparing his own notes for this section, Kant wrote out the questions that would be addressed in his lecture: Was the soul a pure spirit before birth? Had it lived on the earth before? Did it live in two worlds—the pneumatic and the mechanical—at once? The questions were accompanied by a quick list of the various theories of generation, with Kant noting that the central division was between supernatural approaches to the question of origin and a naturalistic account, an account Kant described as an “epigenesis psychologica” (17: 416). The majority of Kant’s commentary, however, was devoted to the comparative advantages of the preexistence theory of generation, in either its spermist or ovist variation, over the system proposed by epigenesis. In contrast to the preexistence theory, for example, the naturalistic system of epigenesis assumed material contributions from each of the parents, and this, Kant observed, required that prospective couples consider each other with greater care when planning to marry and reproduce.203
In later years, Kant would use this section of Baumgarten’s text to discuss the properties of the soul and would invariably dismiss the possibility of its epigenesis.204 In 1769, however, Kant’s commentary focused on the physical aspect of generation, identifying epigenesis with a theory of blending that was in line with what he knew of Maupertuis’s and Buffon’s use of heredity as a basis for their arguments against preexistence theory. The next time Kant came to add notes to this section, epigenesis was again considered in terms of its biological claims, with Kant now explicitly linking the theory to the desired account of species generation he had first sketched in 1763. In his words, “The question is whether nature is formed organically (epigenesis), or only mechanically and chemically. It seems that nature does have spirit, given that in the generation of each individual there is a unity and connection of parts. And is there not also such a spirit, an animating essence, in animals and plants. In this vein one would have to assume an animating Spirit, operating within an original chaos, in order to explain differences between animals which can now only reproduce themselves” (17:591). This two-step model was the same proposed in Kant’s Only Possible Argument, so far as an initially divine organization—out of an “original chaos”—was followed by the organic capacity for reproduction within the divinely delineated species lines. These two sets of comments, dated by Erich Adickes as having been written in 1769 and 1772–1776, respectively, demonstrate that during a period of crucial formation with respect to the development of Kant’s system of transcendental idealism, Kant was actively aware of the epigenesis alternative to preexistence theories of generation.
More significant than Kant’s commentary on Baumgarten for our purposes, however, is the set of notes Kant composed shortly after finishing his Dissertation. For in these notes, Kant explicitly connected theories of generation to systems of reason and to claims regarding the origin of ideas in particular. Distinguishing empiricists from rationalists, Kant identified his own position with the most radical possibility of all. As he sketched it, “Crusius explains the real principle of reason on the basis of the systemate praeformationis (from subjective principiis); Locke on the basis of influx physico like Aristotele; Plato and Malebranche, from intuit intellectuali; we, on the basis of epigenesis from the use of the natural laws of reason” (17:492). It was epigenesis, therefore, that Kant identified with the theory of “original acquisition” for explaining the generation of sensitive and intellectual concepts from the mind’s own laws in the Dissertation. While it cannot be said for certain that Kant took epigenesis as his model when first drawing up his account of the origin of knowledge in 1770—though the evidence from 1769 certainly suggests this—it is certain that in the months following the Dissertation’s completion the connection had been made, that by then Kant had, to paraphrase Darwin, “at last got a theory by which to work.”205
Concepts and Objects: Kant’s Letter to Herz, 1772
Kant had presented his Inaugural Dissertation—with his former student Marcus Herz playing the role of disputant—on August 21, 1770. Twelve days later, Kant sent copies of the Dissertation off for feedback and, as with their earlier exchange, it was J. H. Lambert’s response that would prove again to be of the greatest significance for Kant’s developing project.206 Kant had, in fact, never replied to Lambert’s letter regarding the distinction between form and matter. “The reason,” Kant now explained, “was none other than the striking importance of what I gleaned from that letter, and this occasioned the long postponement of a suitable answer” (10:96). Having dismissed the first and fourth sections of the Dissertation as discussions to “be scanned without careful consideration,” Kant wanted Lambert’s thoughts on the remainder of the work, for in Kant’s own estimation of the remaining sections, “there seems to me to be material deserving more careful and extensive exposition” (10:98). Sections 1 and 4 of the Dissertation had covered topics that were traditional for metaphysics: the problem of intuiting the world as a whole versus as an aggregate and the difficulties in accounting for interaction between substances. The truly innovative work of the Dissertation appeared in the remaining parts of the text. Section 3 presented Kant’s account of space and time as the originally acquired forms of sensible intuition. Section 2 laid out the strategy for certainty regarding empirical knowledge and introduced Kant’s distinction between the laws at work in the “logical use” of the intellect and the “real use” by which pure concepts could be generated by attention to the working of these laws. The last section of the Dissertation, section 5, outlined Kant’s method for metaphysics in light of the mind’s susceptibility to surreptitious concepts or, as he had renamed these in the Dissertation, “subreptive axioms.” These were the three sections meant for Lambert’s inspection, and Kant summarized his general results for Lambert in a few lines:
Space and time, and the axioms for considering all things under these conditions, are, with respect to empirical knowledge and all objects of sense, very real; they are actually the conditions of all appearances and all empirical judgments. But extremely mistaken conclusions emerge if we apply the basic concepts of sensibility to something that is not at all an object of sense. . . . It seems to me . . . that such a propaedeutic discipline, which would preserve metaphysics proper from any admixture of the sensible, could be made usefully explicit and evident without great strain. (10:98)
Kant had described his project as just such a “propaedeutic” (2:395) in his Dissertation, locating its success regarding the prevention of subreptive axioms in the radical break between sense and intellect.
Lambert replied within a matter of weeks, generously discussing the sections inquired after by Kant. From Lambert’s perspective, the main challenges for Kant’s theory lay in the ideality of space and time described in section 3 (this was the focus of Moses Mendelssohn’s response to the Dissertation as well). But while Kant was willing to incorporate some of this in his later discussions of space and time (e.g., A36–37/B53–54), his general position regarding their transcendental ideality would not change. The importance of Lambert’s letter for Kant lay rather in the remarks concerning section 2 of the Dissertation. Here Lambert was direct regarding what he saw as a problem facing the heterogeneity of sense and intellect as independent sources of knowledge: “My thoughts on this proposition have to do mainly with the question of universality, namely, to what extent these two ways of knowing are so separated that they never come together. If this is to be shown a priori, it must be deduced from the nature of the senses and of the understanding. But since we first have to become acquainted with these a posteriori, it will depend on the classification and enumeration of their objects” (10:105). Lambert’s remark raised two concerns: first, the seeming impossibility of an a priori demonstration of sense and intellect’s universal separation and, second, the need, as a consequence of that impossibility, to turn to their respective objects for evidence of their separation—a turn that would limit Kant to an a posteriori proof. Resorting to experience like this had been essential in Lambert’s own ontological investigations, as he made clear further on in his reply: “It is also useful in ontology to take up concepts borrowed from appearance [Schein], since the theory must finally be applied to phenomena again. For that is also how the astronomer begins, with the phenomenon; deriving his theory of the construction of the world from phenomena, he applies it again to phenomena and their predictions in his Ephemerides [star calendar]” (10:108).207 “In metaphysics, where the problem of appearance is so essential,” Lambert advised, “the method of the astronomer will surely be the safest.” For the metaphysician could also “take everything to be appearance, separate the empty from the real appearance, and draw true conclusions from the latter. If he is successful,” Lambert concluded, then “he shall have few contradictions arising from the principles and win much favor” (10:108).
In the Inaugural Dissertation Kant had in fact emphasized the importance of distinguishing between empirical concepts garnered along the lines now suggested by Lambert and the “original acquisition” of pure concepts, concepts that, as he had put it, would “never enter into any sensual representations as parts of it, and could not, therefore, in any way be abstracted from it” (2:395). But Lambert had raised an important point nonetheless. How could one understand the fact that an intellectual concept like “cause,” for example, should seem so readily applicable to experience and yet belong, by definition, to an entirely separate realm of knowledge? Sensitive knowledge, so neatly accounted for in the Dissertation via the forms of intuition and the processes of logical subordination, suddenly seemed deficient when explaining the experience of causal relations. Subreption had served as the catalyst for Kant’s attempt to redefine metaphysics as a science of the limits and extent of human reason, but the radical separation of sense and intellect—the key to Kant’s solution to the problem of logical subreption—might have to be rethought after all.
The problem was as follows: Subreption, as Kant initially conceived it, was unidirectional. It focused on the prevention of sensible concepts, concepts like “causality” and “force,” being surreptitiously applied to objects of what would have to be a nonsensible intuition, objects like angels and souls. When Kant sat down to write the Inaugural Dissertation, however, his theory of cognition had outstripped the earlier conception of the problem. As Kant had explained in his first letter to Lambert, all he really had in 1765 was a negative account—a kind of “what not to do” for anyone interested in reconstructing metaphysics as a science of limits set by the boundaries of the human mind. By 1766’s Dreams of a Spirit-Seer, this committed Kant to denying, for example, the possibility of direct knowledge of either human souls or the principles of life within nature. This prohibition automatically eliminated, therefore, the explicit objects to which sensible concepts were supposed to have been surreptitiously applied. The prohibition was carried over to the Dissertation, where, despite Kant’s characterization of the break between sense and intellect as similar to the ancient distinction between the worlds of phenomena and noumena (2:393), there were in fact no noumenal objects to be found, and the intellectual intuition of such objects was flatly rejected (2:396). With noumenal objects thereby out of reach, subreption—still described as the central problem facing metaphysics—was reconceived as the result of misunderstanding the subjective nature of space and time as forms of human intuition.
According to the Dissertation, then, subreption appeared in three guises: it occurred when asserting that space and time could be applied to nonmaterial objects, as in attempting to spatially locate the soul within the body; it occurred when asserting that because we only experience objects in space and time, all objects are necessarily spatiotemporal, a fallacy in line with demanding that the universe have a beginning in time; and, finally, it occurred when asserting that intellectual concepts could only be applied to experience via space and time.208 This last example was a surprise, for here, in the closing moments of the Dissertation, Kant was discussing a case of intellectual concepts being applied to experience after having expressly forbidden it in section 2 (2:395). Was this a slip?
For the third type of subreption, Kant had taken his example from Crusius, who, according to Kant, illicitly filtered the intellectual concept of “existence” through the lens of temporality when declaring that “whatever exists contingently has at some time not existed” (2:417). Subreption occurred in this case by supposing that intellectual concepts required sensible intuition for their application. Crusius’s “spurious principle,” Kant explained, “arises from the poverty of the intellect, which for the most part discerns the nominal marks of contingency or necessity, seldom the real ones. Since, therefore, we can scarcely hope to determine, through marks derived a priori, whether the opposite of some substance is possible, we shall be able to do so only insofar as we have evidence that at one time the substance did not exist” (2:417). In the absence of any a priori discovery, in other words, the intellect turned to experience and, borrowing the concept of temporal change, illicitly declared it to be necessarily connected to the concept of contingency. But contingent existence, as Kant had already understood it in The Only Possible Proof for the Existence of God (1763), was also a way for seeing effects to be the result of God’s free choice. Since God’s activity was not susceptible to temporal laws, this marked the subreptive fallacy in connecting contingency and time. The argument seems to have distracted Kant from the fact that up until now in the Dissertation he had denied any possible connection between intellectual concepts and sensible experience, a denial motivated with respect more to maintaining the pure status of the intellectual concepts and moral exemplars than to the problem of subreption itself. In 1770, Kant still understood subreption to be a unidirectional problem, and the possibility that he might need to apply intellectual concepts to experience had simply not occurred to him.209
When Lambert questioned the universal separation of sense and intellect as independent modes of cognition, therefore, he might have pointed to Kant’s discussion of Crusius, but his focus on section 2 pointed Kant back to his account of the intellectual concepts themselves. For here Lambert must surely have noted that “causality” was no longer considered a sensitive concept at all. Kant had indeed moved causality to his list of intellectual concepts, a decision undoubtedly reflecting the influence of Hume’s skepticism regarding necessary connection. But while the a priori status of causality “protected” it from Hume’s skepticism, without a connection to sensible phenomena, Lambert seemed to suggest, metaphysics would remain not only sterile but ultimately useless in the face of the empiricist challenge.
In the wake of Lambert’s response, it was clear to Kant that he would need to reconsider whether the separation of sense and intellect could be maintained at all. As he described this realization, “I noticed that I still lacked something essential, something that in my long metaphysical studies I, as well as others, had failed to pay attention to and that, in fact, constitutes the key to the whole secret of hitherto still obscure metaphysics” (10:130).210 When Kant went on to describe this “key to the whole secret” of metaphysics to Marcus Herz, it turned on the problem of connecting intellect and sense. It was a problem of maintaining that “pure concepts of the understanding must not be abstracted from sense perceptions,” that they “have their origin in the nature of the soul,” and that “they are neither caused by the object nor bring the object itself into being,” while also explaining how such pure concepts could be connected to objects at all (10:130). Kant could not have been clearer regarding the status of the concepts under consideration: their origin lay in the nature of the soul; they were neither abstracted from nor caused by the object; they were, in keeping with the Dissertation, original to the mind itself. This much had not changed. “And if such intellectual representations depend on our inner activity,” Kant continued, “whence comes the agreement they are supposed to have with their objects—objects that are nevertheless not possibly produced thereby? And the axioms of pure reason concerning these objects—how do they agree with these objects since the agreement has not been reached with the aid of experience?” (10:131). This was the change. The trajectory of Kant’s thinking since writing the Dissertation, so far as he now recounted it for Herz, turned on the problem of connection.211 The problem of origin, by contrast, was no longer an issue.
Focusing on the problem of connection, then, Kant listed the kinds of relations that were easy to grasp. One could easily see, for example, how sensible content was connected to sensible representations, and it was also clear how an “archetypal” intellect could serve as the ground for its own representations. In mathematics, it was possible to understand how to connect axiom and intuition without experience because in this case “the objects before us are quantities only because it is possible for us to produce their representations,” a production guaranteeing their connection (10:131). But the question of understanding how a concept like causality, for example, could both be generated a priori and yet conform to sensible experience “remained in a state of obscurity” (10:131).212
Moving on from the “obscurity” surrounding the problem of connecting sense and intellect, Kant proceeded to review theories regarding the origin of concepts. Since he had already listed the ease in understanding the relationship between sense data and sensible concepts, he now limited himself to theorists describing a priori concepts, since the locus of the problem of connecting them to sensible phenomena lay precisely in their purity. As Kant rehearsed the list, “Plato assumed a previous intuition of divinity as the primary source of the pure concepts of the understanding and of the first principles. Malebranche believed in a still-continuing perennial intuition of this primary being. . . . Crusius believed in certain implanted rules for the purpose of forming judgments and ready-made concepts that God implanted in the human soul just as they had to be in order to harmonize with things. Of these systems, one might call the former the Hyperphysical Influx Theory and the latter the Pre-established Intellectual Harmony Theory” (10:131). Kant dismissed such theories immediately, acidly noting that “the deus ex machina is the greatest absurdity one could hit upon in the determination of the origin and validity of our cognitions,” a recourse encouraging “all sorts of wild notions and every pious and speculative brainstorm” (10:131).
The list Kant rehearsed was, of course, the same breakdown he had previously outlined for himself regarding theories of origin (minus the cases presented by Aristotle and Locke regarding empirical concepts): “Crusius explains the real principle of reason on the basis of the systemate praeformationis (from subjective principiis); Locke on the basis of influxu physico like Aristotle’s; Plato and Malebranche, from intuit intellectuali; we, on the basis of epigenesis from the use of the natural laws of reason” (17:492). When employing biological vocabulary in his own notes, Crusius’s belief in “implanted rules,” for example, was identified with preformationism. In the letter to Herz, however, Kant was entirely focused on the question of connection, and the examples of theorists arguing for a nonempirical origin were therefore schematized in terms of their means for connecting a priori concepts and objects. The “Hyperphysical Influx Theory” defined systems where concepts and objects maintained connection because of their effective identity in God’s mind. The “preformationist” theory maintained by Crusius relied on “Pre-established Intellectual Harmony” given God’s work to establish all future potential connections between concepts and things at the moment of creation. For his own part, Kant was still clear regarding the epigenetic origin of concepts, concepts whose source lay “in the nature of the soul,” but he had yet to discover a basis for connecting these to sensible objects.213
Kant was, however, ready to announce the progress he had made with respect to his understanding of the concepts themselves. Whereas the earlier list of concepts had been both short and somewhat vague—“possibility, necessity, substance, cause, etc. with their opposites or correlates” (2:395)—Kant now appeared to have in mind not only a specific number of concepts but, more importantly, a basis for their organization. And Kant took this advance to be important enough that he was ready to tell Herz that “so far as my essential purpose is concerned, I have succeeded, and I am now in a position to bring out a critique of pure reason that will deal with the nature of theoretical knowledge” (10:132). What precisely was this advance? Kant explained, “As I was searching in such ways for the sources of intellectual knowledge . . . I sought to reduce transcendental philosophy (that is to say, all the concepts belonging to completely pure reason) to a certain number of categories, but not like Aristotle, who, in his ten predicaments, placed them side by side as he found them in purely chance juxtaposition. On the contrary, I arranged them according to the way they classify themselves by their own nature, following a few fundamental laws of the understanding” (10:132). The origin of the intellectual concepts would no longer be generally based on the workings of the mind; they would from now on be indexed to particular mental laws as a means for their specific classification. If this was not yet to directly identify the intellectual concepts and the mind’s laws for logical subordination, Kant was certainly very close to making this connection. This was significant, for it demonstrated that while Kant might still have been uncertain regarding the means for connecting a priori concepts and sensible objects, he was apparently close to adopting the successful model provided by sensitive knowledge when approaching intellectual cognition.
In the Dissertation, sensitive knowledge could be called “genuine knowledge” so long as the truth of a judgment of experience was determined by the inner coherence of the mental laws connecting subject and predicates in the judgment itself. As Kant put the point in 1771, “All truth consists in the correspondence of all thoughts with the laws of thinking and thus among one another” (17:524). Were Kant to embrace this model for cognition in general, to identify the laws of logical subordination with the intellectual concepts themselves, then the purity of the intellectual concepts would not be compromised, and the connection to sensible experience could be explained by their application to the “apparentia” delivered up by space and time. Embracing this model, however, would also mean accepting that the objects of cognition—objects dependent upon the mind in order for us to experience them—would henceforth be redefined as objects of knowledge. Nonetheless, it was precisely to this model, with these consequences, that Kant turned. And he did so in short order. By 1773 Kant summarized his position as follows:
If certain concepts in us do not contain anything other than that by means of which all experiences are possible on our part, then they can be asserted a priori prior to experience and yet with complete validity for everything that may ever come before us. In that case, to be sure, they are not valid of things in general, but yet of everything that can ever be given to us through experience, because they contain conditions by means of which these experiences are possible. Such propositions would therefore contain the condition of the possibility not of things but of experience. (17:618)
This passage was Kant’s response to the question he had just posed for himself: if “there are judgments whose validity seems to be established a priori, but which are nevertheless synthetic, e.g., everything that is alterable has a cause, whence does one arrive at these judgments?” (17:617). How do we achieve certainty, in other words, with respect to judgments that contain a synthesis of pure concept and sensible intuition? As Kant’s response made clear, we achieve certainty by understanding such synthetic a priori judgments to be the means by which experience—so far as it is possible to know it, at least—becomes possible at all, a conclusion based, significantly, on the newly asserted identity between laws for logical subordination and concepts for conceptual determination. As Kant put it, “The concepts of the understanding express all the actus of the powers of the mind, insofar as representations are possible in accordance with their universal laws, and indeed their possibility a priori” (17:622). The “key to the secret of metaphysics,” at least insofar as Kant had outlined the problem for Marcus Herz, appeared to have been found.