The Death of Philosophy
Page 28
2. The second example with which I can establish the synonymy between knowing and depicting in Kant is precisely the theory of the subject. If we read section 16 of the Critique of Pure Reason literally, the transcendental subject = x is unknowable. This is because it is not representable, that is to say, it cannot be schematized in space, like a triangle, or in time, like rational numbers. Consequently, the fundamental presupposition from which the subject’s unknowability is pronounced is indeed that all knowledge is representational. The subject’s unknowability is declared in the name of the figure. In this sense, Kant is not the heir of what we traditionally call the philosophy of the subject, because his problem is above all the problem of objectivity.12 He was trying to answer the question, “How can we know an object?”—and this is why the Kantian theory of representation is presented as a theory of different possible judgments about the object.
Having clarified these points, I can say that the Kantian theory is a theory of representation in three respects: (1) in its general definition, since each aspect of cognitive activity is defined as representation; (2) in its definition of knowledge, since knowing consists of representing in a certain way; and (3) in this definition’s presuppositions, namely, that all knowledge is of an object (even if the object is defined as a phenomenon) and that knowing an object boils down to delineating it in time or space. From this clarification of the precise meaning of “representation” in Kant, I can situate his project with respect to the problem of reference. The Kantian theory is undeniably an idea of reference—in that what is to be known is the object, and only objects can be known. In this sense, critique and the beginnings of analytic philosophy can be indexed, without historical distortion or philological eccentricity, to a single encompassing category, “ideas of reference.” The difference between these approaches hinges on their definitions of the object, not the key thesis that only objects can be known. The common genus, ideas of reference, is thus divided into different species: objects as empirical entities or as figures constructed a priori. This designation of critique within the general problematic of reference is confirmed by Kant’s definition of the concept of reflection. Critique—far from being an idea of reflection—turns out rather to be the moment of its impeachment. Let’s explain this seeming paradox.
Reflection
In comparison to Kant’s interest in representation, his interest in reflection is in fact much less. To become convinced of this we need only note Kant’s difficulty in giving this concept a precise and univocal definition. Indeed, if representation is very clearly defined (genus, then species, and within this species, valid knowledge), reflection on the other hand is the object of multiple definitions that are not at all compatible with one another. In fact, Kant gives at least four meanings to reflection:
First of all, he gives it a very broad meaning, writing in the Lectures on Logic that “the origin of concepts as to mere form rests on reflection and on abstraction from the difference among things that are signified by a certain representation.”13 In this framework, any conception is reflection.
He uses it in “The Amphiboly of Concepts of Reflection”—here discussing transcendental reflection, the faculty that enables a philosopher to understand representation in terms of the concepts of form/material, internal/external, etc. “The act by which I confront the comparison of representation with the cognitive faculty to which it belongs, and by means of which I distinguish whether it is as belonging to the pure understanding or to sensible intuition that they are to be compared with each other, I call transcendental reflection.”14
He quite obviously uses it in the Critique of Judgment. Reflection, in this text, is defined as investigation of a universal from a given particular. Reflection results in the regulative judgment that commands an “everything happens as if.”
Finally, Anthropology makes reflection a purely empirical faculty.
From this enumeration we can note (1) the scattershot nature of these definitions, (2) the incompatibility of these definitions, and (3) the way in which Kant distances himself from the classic definition of reflection. We must pay attention to this third dimension in that it shows how Kantian thought moves closer to an idea of reference in the current sense of the term. Indeed—and in my view, this has not been sufficiently appreciated—by advancing the term “representation” and obscuring the term “reflection,” Kant breaks with the tradition born in Descartes, namely, the study of one’s relation to self understood as going back over the operations of a thinking subject. In Descartes, this review was possible only through a suspension of one’s belief in the world (the first Meditation)—for the possibility of reflection to emerge, the object and one’s relation to the object must be momentarily abolished. Beyond Descartes, Locke, too, made a basic distinction between impressions of perception by which I know an object and the impressions of “reflection” by which I go back over the operations I have effected: “When we see, hear, smell, taste, feel, meditate, or will anything, we know that we do so. Thus it is always as to our present sensations and perceptions.”15 This definition of “reflection” as knowledge of our mental operations was also put forward by Leibniz: “Thus it is good to distinguish between perception, which is the internal state of the monad representing external things, and apperception, which is consciousness, or the reflective knowledge of this internal state, something not given to all souls, nor at all times to a given soul.”16 And yet—a revealing fact—in Kant, apperception is defined as the condition of possibility of objective unity, no longer as a relation to oneself.
Thus, Kant had to break with a process that modern philosophy has advanced, even if it never managed to think this process all the way through, for we find a temptation, for example in Descartes, after having highlighted reflection (the subject’s review of its own operations), to understand it as a relation between two terms,17 a subject (“I”) looking into a subject res (“me”). But for Kant, the “I” can never turn its attention onto itself, because the object cannot be bracketed. The bracketing of the world is a procedure that is missing in Kant’s text. In section 16 of the Critique, to make the unity of the subject appear, he uses neither a reflective method nor a method of suspension (epoché or doubt) but employs a synthetic procedure whose avowed goal is to enable the constitution of a known object. How is the subject to be obtained? By showing that it is a necessary term for understanding objectivity.
And yet the chosen route—eliminating a reflective process—is clearly dependent upon the advancement of a representational paradigm. This desire to understand knowledge solely from knowledge of the object is later reiterated by the neo-Kantians and colors the Kantian legacy. On this point, the way that the neo-Kantians understood Husserl’s transcendental doctrine—whose epoché revives a Cartesian type of reflective process—is extremely revealing of this demand for referentiality. Let’s quickly consider this point, as a supplemental confirmation of the thesis that Kant’s philosophy, as much in its content as in its historic reception, is given as a philosophy of representation, as an idea of reference.
Indeed, it is extremely indicative to see how the later Husserl—the one that embraces the transcendental problematic—was criticized by the Kantians of his era (the neo-Kantians). In fact, they accused him of ontologizing the sphere of validity. This dispute, which arose between orthodox Kantians at the beginning of the twentieth century and Husserl regarding the categorical intuition, unmasks the theory of knowledge that critique leads to. Husserl, who was considered, in some of his texts, as a Kantian writer, was accused by Friedrich Kreis, Rudolf Zocher, and others of “spilling,” “falling,” etc., into ontologism, because he “extends” the concept of intuition to “essences.” (This controversy was launched by Rickert and his school: Friedrich Kreis attacked Husserl’s intuition in a 1930 text Phänomenologie und Kritizismus.18 Eugen Fink responded to his criticisms in a 1933 article “The Phenomenological Philosophy of Edmund Husserl and Contemporary Criticism.”19 Paul Ricoeur takes up thi
s debate in “Kant and Husserl.”20) The concept of a view of essences is understood by the neo-Kantians as the representation of idealities comprehended in the same way as objects of experience. The Kantian conditions of possibility, “the components of valid meaning,” are thus, in the Kantians’ eyes, objectivized by Husserl into objects posited as of an ontically superior order. This criticism, as Eugen Fink showed, rests on a total misunderstanding of the Husserlian project. Indeed, for Husserl it is not a matter of transferring empirical knowledge’s or scientific knowledge’s mode to essences that are given to the philosopher like a table or a triangle. It is rather a question, as he will later attest in The Crisis of European Sciences and Transcendental Philosophy,21 of employing a different kind of rationality. The roots of the Kantians’ misunderstanding of Husserl’s project take their origin in the Kantian conception of validity: Kant, obsessed by the theme of knowledge of objects, came to either conceal or to prohibit knowledge of the subject’s actions, though it was required for the completion of his project. As Ricoeur argues, precisely with respect to this dispute between the neo-Kantians and Husserl, there is a rough or “implicit phenomenology of Kantianism”22 simultaneously presupposed and concealed by the problem of objectivity. What is important here is that we see the extent to which the problem of objectivity dominates the center stage in forbidding the use of another path. The Kantians’ edict against any other approach is no less severe than the analytics’. Having established this point and clarified the nature of the Kantian doctrine, it remains for me to show how an impossible reconciliation of these two terms gives birth to a serious aporia within the Kantian system.
Use of the Term “Intellectual Representation” as an Expression of the Tension Between Representation and Reflection
This valorization of representation over reflection would not have posed an immediate problem in Kant if he had asserted that he privileged one approach—here, emphasizing the object or reference—while leaving it to other books or other philosophers to explore the other approach. But that is not what he did, and Kant himself makes the impossibility of his viewpoint appear, through the use of a term that causes the system to implode in that it contradictorily says one thing and its contrary. This term—the extreme point of the tension between representation and reflection—is “intellectual representation.” When Kant uses this term in a positive sense, it is to try to express a point of view that his own doctrine forbids. To put it more precisely, with this term “intellectual representation,” Kant twice puts his own doctrine at risk. Let’s analyze the use of this term within the Critiques:
In L’élucidation critique du jugement de goût selon Kant, Louis Guillermit notes that “the word ‘intellectual’ is very rarely used in Kant.”23 The philosopher’s avoidance of the term is such that not only does he use it in often negative contexts (as, for example, in the expression “intellectual intuition”) but he even modifies the canonical expression of his time, “intellectual world,” using instead the uncommon phrase “intelligible world.” The term’s rarity in the Kantian lexicon serves only to accentuate its few occurrences. In this respect, it seems legitimate to divide them into two distinct groups. The first usage of the term is in the Critique of Practical Reason and the Critique of Judgment. For example, in section 42 of the Critique of Judgment, he is giving a definition of an intellectual interest in beauty. In this context, “intellectual” is in contrast to the sensible and the empirical. Moreover, this adjective refers to a character of immediacy—Kant writes of someone who takes pleasure in contemplating nature that “such a person is taking a direct interest in the beauty of nature, and this interest is intellectual.”24 Finally and above all, the term “intellectual” characterizes the will’s capacity to be determined a priori by reason: “something intellectual, viz., the will’s property of being determinable a priori by reason.”25 In a general way, the term thus refers to spontaneity and immediacy. If this first usage does not pose any particular problems, the second group of occurrences turns out, on the other hand, to be more problematical. This no longer concerns the practical domain or the aesthetic field but only the theoretical domain. Thus, the adjective “intellectual” is used in the Critique of Pure Reason with respect to the representation of the “I,”26 and, in a letter to Johann Schultz, with respect to the properties of arithmetic.27 And yet, in both cases, Kant meets with the same difficulty because he introduces a mode of validity that transgresses the definition given in the Critique of Pure Reason. To demonstrate this claim, I shall examine each of these occurrences separately.
Use of the Term “Intellectual” with Respect to Numbers
In a November 25, 1788, letter to the mathematician Johann Schultz, Kant wrote with respect to the properties of arithmetic:
Time, as you correctly notice, has no influence on the properties of numbers (considered as pure determinations of magnitude), as it may have on the property of any alteration (considered as alteration of a quantum) that is itself possible only relative to a specific state of inner sense and its form (time); the science of number, notwithstanding succession, which every construction of magnitude requires, is a pure intellectual synthesis that we represent to ourselves in thoughts.28
To say that the properties of numbers are not at all dependent upon time seems like an astounding concession. With this text, Kant puts his own doctrine in jeopardy and goes “so far as to contradict the letter of the Critique”29 in that he seems here to recognize the existence of propositions that, although authentic knowledge (that is, not merely analytic, as is indicated by the use of the word “synthesis”), are not given in an intuition. We should be even more wary about this anomaly, for arithmetic is constantly defined by Kant as what is closest to the laws of pure thought. Kant’s letter to Schultz is, in its entirety, the epitome of the difficulties that Kant faces in accounting for arithmetic with his own principles.
Let’s first clarify the meaning of the restriction in the passage I quoted—“notwithstanding succession, which every construction of magnitude requires.” In the Critique of Pure Reason, a number, defined as the unity of a multiplicity, has a need for time in order to be constituted. The arithmetical judgment 7 + 5 = 12 is a synthetic a priori judgment because it presupposes the application of a category (quantity) to an intuition (succession in time). Every accumulation of a multiplicity is the generation of a unity from another unity. This is expressed, in the “Introduction” to the Critique of Pure Reason, by the still-famous example of the generation of 12 as the addition of 7 + 1 + 1 + 1, etc., a generation that can be depicted with the fingers of one hand or with marks on a page. Taken literally, these examples are undeniably disastrous. They justify Louis Couturat’s fury, who reads only an obscene empiricism in the expression; they authorize Frege’s irony, who on this precise point puts Kant and John Stuart Mill together in the unflattering category of those who do arithmetic with cakes and stones; they inevitably provoke the massive objection that William and Martha Kneale have summarized30—one could never, by simply counting on one’s fingers, obtain the proposition 135,664 + 37,863 = 173,527. To save the Critique from such a reading, we clearly ought, as did mathematicians like Johann Schultz as well as philosophers like Salomon Maimon,31 to understand the unity that Kant discusses as unspecified, that is to say that any number (for example, 1,000) could be taken as a point of reference. This makes it possible to explain the addition of large numbers while still maintaining the proposition that numeration presupposes a generation in time. Numbers become the understanding of the method thanks to which I can construct the multiplicity and are thus no longer the intuition’s generation of a real succession that would seem to be implied by the unfortunate example in the “Introduction.” In this way, the too-empiricist character of the genesis of numbers can easily be neutralized. But this reconstruction, which all his disciples will immediately implement, comes with its own cost: arithmetic becomes an exclusively symbolic form of knowledge.32 But if symbolic knowledge is not absolutely opp
osed to ostensive construction, the former is nevertheless distinguished from the latter by a greater degree of intellectualism. The genesis of numbers thus reveals the dual obstacle facing the Kantian theory of rational numbers—taken literally, this theory illustrates Locke’s principles better than Kant’s; but in a parallel way, its necessary and legitimate reformulation makes arithmetic into a science far removed from ostensive demonstration, which in its obvious sense no longer applies to anything but geometry. The second stage of the Kantian theory of arithmetic only accentuates the trouble—a truly intermediate level, it leads to an astounding final concession.