The return to Kant is in fact the choice of a single path that brings an end to the tension in the critical project. It is a matter of “returning” to the question of representation as an explication of the relation between a subject and an object. Let’s first of all recall that, from 1810 to 1850, Hegel and his disciples were the main figures on the philosophical scene. Henri Dussort points this out, “From 1800 to about 1840, speculative thought, its famous developers and their disciples occupied the center stage.”1 Friedrich Engels himself noted that this enthusiasm for the Hegelian doctrine was at its peak between 1830 and 1840, “It was precisely from 1830 to 1840 that ‘Hegelianism’ reigned most exclusively, and to a greater or lesser extent infected even its opponents.”2 But beginning in 1850, this widespread Hegelianism rapidly broke down, and as Ernst Cassirer notes, “In place of the metaphysical orgy inspired by post-Kantian philosophy, a complete sobriety appears.”3
Why did it break down? In fact, it is traditionally attributed to the development of and the orientation adopted by the positive sciences, which seemed to undermine the Hegelian analysis of the sciences and the ostensible absence of mathematics in Hegel despite the fact that the age witnessed a clearer and clearer mathematicization of all the positive sciences. This is an established historical point, in the sense that it is indeed what his contemporaries said—it was necessary to distance philosophy from Hegelianism and tie it to the so-called exact sciences. However, we could maintain that it was not only about Hegelianism’s ostensible “weakness” or nonconformity with the science of the time but also about a choice for a kind of rationality (representation, to the detriment of reflection). The opposition between Wissenschaftslehre and Erkenntnistheorie is less the opposition between one conception that will turn out to be false (German idealism) and another one, true (positivism), but rather the choice between two possible orientations for philosophical questioning.
What launched this return to Kant, apart from Eduard Zeller’s famous discourse, was Hermann von Helmholtz’s simultaneously scientific and philosophical elaboration as a whole. From a general point of view, Helmholtz presented a positivist reading of Kant that tended to “naturalize” the critical project. First of all, this resolute naturalism reflected Helmholtz’s scientific training, since, apart from logic (which would not be reinvigorated until after his time), he distinguished himself in all the distinct sciences of the nineteenth century—the biological, physical, and chemical sciences, and geometry.4 Helmholtz’s scientific viewpoint can be expressed relatively simply: his fundamental approach consists in applying strictly mechanistic and quantitative models to all phenomena, including biological phenomena. In doing so, he disagreed with his era’s physiologists’ vitalism. In his eyes, a Newtonian mechanism could account for the totality of natural phenomena. What is of interest to me is the way that he thought Kant could justify this entirely Newtonian conception of science. His decisive reinterpretation of Kantian concepts is always carried out along the same axis: a concealment of the reflexive dimension and a naturalization of Kantianism. On this point, we should note the shift in meaning that occurs in Helmholtz’s use of the terms “transcendental” and “a priori.” In all his analyses, Helmholtz considers these concepts to be strictly equivalent. In fact, he reduces knowledge to two dimensions: knowledge is either empirical or a priori, that is, it is either dependent upon or independent of experience. With this reduction, he purely and simply identifies a priori knowledge with transcendental knowledge. But in Kant we could enumerate not two but indeed three ways of knowing: the transcendental, which is knowledge about knowledge, not a mode of knowledge of objects; then a priori knowledge, defined as a structure independent of experience; and finally, a posteriori knowledge. Thus, here we should note the impact of Helmholtz’s shift, for we can read the first eclipse of reflection in it. Indeed, to reduce the transcendental to the a priori amounts purely and simply to concealing the possibility of a metacognitive questioning (knowledge of knowledge). This first shift of meaning that paves the way for the naturalization of critique is echoed by another, just as important, namely:
Here we must grasp the movement of Helmholtz’s interpretation—it first effects an initial reduction, of the transcendental to the a priori, and then understands the a priori in psychophysiological terms. This is indeed what I have characterized as a naturalization of critique. This second shift can be seen above all in Helmholtz’s theory of perception. This theory of perception is based on the fundamental law of the physiology of perception, articulated by Johannes Müller, called the “law of specific nerve energies.” This law explains that our sensible impressions do not depend upon the type of stimulation but exclusively upon the stimulated nerve. Each kind of nerve provokes in us a unique, specific, and incommensurable form of sensation, regardless of the kind of external stimulation. Inversely, the same stimulation, in contact with different nerves, causes different sensations. For example, the same sensation can be caused by a given source of light (electromagnetic waves) but also by pressure on the eyeball, or even, Helmholtz explains, by a displacement of the optic nerve caused by a brusque movement of the eye, etc. Thus, various sources can have the same effect. On the other hand, the same electromagnetic waves will prompt a sensation of heat if they are in contact with a nerve of the skin and a sensation of light if they are in contact with the nerves of the eye. It follows, from a general point of view, that the type of our sensible impressions does not depend on the type or the origin of the stimulation but on the stimulated nerve—such is what the “law of specific nerve energies” says. But Helmholtz interprets this physiological law as a proof of the Kantian thesis about the a priori forms of sensation. Sensations do not depend upon the object (since, in the last example, it is the same object—namely, electromagnetic waves) but upon the subject (the affected nerve). There is thus a psychophysiological predisposition of the subject that explains that the sensation will be of one sort or another. This is why, Helmholtz explains of the discovery of this law of specific nerve energies, that “in a certain sense, it is the empirical fulfillment of Kant’s theoretical concept of the nature of human reason.”5
These two displacements (the shift between transcendental and a priori and the shift between a priori and psychophysiological) illustrate Helmholtz’s relation to critique, which he summarizes in a passage in “The Facts of Perception”:6
That is the answer we must give to the question: what is true in our ideas? In giving this answer we find ourselves at the foundation of Kant’s system and in agreement with what has always seemed to me the most fundamental advance in his philosophy.
I have frequently noted in my previous works the agreement between the more recent physiology of the senses and Kant’s teachings.7
Thus, Helmholtz claims in fact to have arrived at the same conclusions about the a priori as Kant but through a scientific demonstration, drawing upon experimental verification. Not only does he judge that the transcendental derivation on the one hand (Kant) and the empirical derivation on the other (here, the law of specific nerve energies) give the same status to the propositions that they yield (which is inaccurate in Kant’s eyes, since the transcendental deduction is absolutely necessary while the empirical is always hypothetical), but he also considers the results of these deductions to be identical. In doing so, he draws an interpretation of the forms of sensibility in terms of the origin, or even the cause, of our affections—this is characteristic, in my view, of the naturalization of the transcendental, because precisely the question of the source, of the origin, should be avoided in Kant through recourse to the concept of transcendental.
Thus we have here a dual movement of naturalization: first, Helmholtz encompasses the a priori and the transcendental within one term, thereby denying the transcendental’s specificity; next, he assimilates the empirical results of physiology to Kant’s philosophical statements. As Moritz Schlick saw it, Helmholtz believed his theory of knowledge to be in better agreement with Kant’s than it really was.8
To be sure, it could be retorted that this naturalization is already present within critique. Remember, on this point, that Husserl reproached Kant for having naturalized the subject—but Helmholtz incontestably accentuated what was only one tendency among others in critique. He embodies the first great naturalist reading of critical statements. This naturalist reading has the advantage of erasing the contradiction induced by the Kantian use of the term “intellectual,” but it has the disadvantage of cementing the distinction between two kinds of questions: “How do we know?”—which leads to investigation into the nature of our knowledge (is it dependent upon experience or not, a priori or a posteriori?)—and “How do we know that we know?”—which leads to an investigation into the very structure of knowledge, as knowledge of knowledge. Helmholtz’s system marginalizes the second question so much that it eliminates it.
This naturalization becomes even more apparent if we consider a third example: Helmholtz’s transformation of the Kantian distinction par excellence, the distinction between things in themselves and phenomena. Indeed, his naturalism culminates in this reinterpretation—which, moreover, allows us to understand how this reading of Kant had such a decisive influence on the Vienna Circle and logical positivism at the beginning of the century, thus determining what would become the “race to reference.”
In “The Facts of Perception,” Helmholtz reformulates the distinction between things in themselves and phenomena in physiological terms—phenomena are what the perception tells us about the nature of the real. He extends the thesis of the thing in itself as “unknowable,” because perception does not give an image of the real but only consistent signs. Indeed, Helmholtz, taking the law of specific nerve energies as the basis for his thesis about access to the real, shows how this law in itself implies the rejection of a theory of perception as a simple reflection of the real. The specific action of nerves (that is to say, our psychophysiological constitution) is interposed between the objective cause of a sensation (for example, electromagnetic waves) and its transformation into a sensation. It follows that this sensation does not give us a reflection, nor an image, nor a copy of the real but is the regulated result of stimulation of the nerves. This leads Helmholtz to reject any naive realism or natural and immediate empiricism, for which sensations give us a trace of the real and in which perception would be considered as a copy of exterior things. The only correspondence that we can establish between perceptions and the causes of stimulation is, Helmholtz tells us, a relation encrypted by signs. Nevertheless, once this naive realism has been rejected, the specter of radical skepticism emerges, as it does in all Kantians. But Helmholtz averts the possibility of skepticism (according to which, because our sensations are not images of things, they are dreams) by advocating a new form of “scientific realism.” Indeed, insofar as signs are consistent, they tell us that the real processes are also. We have to acknowledge here that the structure of such reasoning actually appears very Kantian: it is neither naive realism (perception as a reflection or copy of the very thing) nor for all that radical skepticism (there is no connection between our perception and things). The connection is in fact a relation through the law; the signs’ regularity informs us about things’ regularity and about the lawfulness of real events.
That said, before I can systematically reconstruct the “scientific real,” I must discuss an additional stage of encryption between sensations and thought. This intermediate step completes the theory of perception and shows us how Helmholtz interprets, in the final analysis, the difficult Kantian distinction between a “thing in itself” and a “phenomenon.”
The necessary intermediary is what Helmholtz calls “unconscious thought,” which is exerted, despite ourselves, on the information delivered by the nerves.9 Between what the nerves give us and what we perceive, an unconscious processing takes place. This is particularly obvious for binocular vision: each of the optic nerves gives a different image of the thing we see, but we see only one image (except in the case of excessive—reprehensible—drunkenness). For Helmholtz, the explanation of this well-known physiological phenomenon is found in “unconscious inferences,” a doctrine that furnishes the basis for his theory of perception. With this doctrine, Helmholtz physiologically reinterprets the famous chapter of the Critique on “Anticipations of Perception.” Despite appearances, he remains Kantian here, for in upholding this theory of “unconscious inference,” he disagrees with a much more deterministic vision of humanity, known in his era as “nativism.” Nativism, in physiology, claimed that we are born with a given psychophysiological structure; this constitution was the absolute framework in which the real is given. Thus nativism maintained that the structure of our impressions (for example, spatial) is directly ordered by our physiological organization. In this sense, because he is unilaterally determinist, the nativist is the diametrical opposite of critique. In contrast, Helmholtz’s theory of perception could seem to him to be a kind of Kantianism. To give a precise example, for the nativists, our visual image’s uniqueness results from the connection of each of the two nerves corresponding to the same location on the retina. The exact physiological conjunction of these nerves results in the construction of a unique image. The definitive structure of spatial intuition would be innate, that is, organic. But Helmholtz argued at length against this theory (which was supported by Ewald Hering10)and stigmatized it as “explain[ing] nothing,” “rash and questionable,” and “unnecessary.”11 He observes that animals, and in particular humans, need a long apprenticeship to be able to use their faculties or to adapt them to new conditions. Far from being innate, the spatial structure of perception (for example, the correspondence between different senses) is acquired through experience, throughout one’s existence, by means of unconscious inferences. These inferences are not inferences in the strict sense (in the sense of reasoning of the sort “if … then”) but are unconscious mental processes with which we structure the information transmitted by the nerves. These processes, from the viewpoint of their results, coincide with the processes of conscious thought. If the existence of these inferences makes it possible to oppose the nativist and “innate” conceptions, it also provides a very precise reformulation and reinterpretation of the “Anticipations of Perception,” unconscious anticipations by which, in Kant, our relation to the real is forged prior to conscious thought. Here again, this is an extremely significant expression of the naturalization of critique, because the chapter in which Kant goes furthest into the question of the origins of sensation is formulated here in psychophysiological terms, in terms of unconscious inferences and processes of encryption of various sensations.
These three aspects of the reinterpretation of Kant show the extent to which the question of knowledge as a relation to the real is taken up and accentuated. This orientation is adopted to the detriment of self-reflection and the type of argumentation connected with it, namely, transcendental arguments. Helmholtz abandons the question, “How do we know that we know?”—the question of one part of German idealism. In doing so, he naturalizes Kantianism and, paradoxically, makes empiricism and Kantianism into two doctrines that are no longer antithetical but relatively close, because Kant’s discoveries or conclusions are experimentally verifiable (through psychophysiology), and, reciprocally, the experimental results are corroborated by the Kantian deduction.
In conclusion, it is clear that Helmholtz quite well embodies a positivist reading of critique, which he specifies as a naturalization of critique (psychophysiology). Nevertheless (and this is the interesting point for my inquiry into the “race to reference” in the twentieth century), this authentically positivist conception is broad enough to be claimed by both the classical neo-Kantians, like Ernst Cassirer, and by the Vienna Circle. In this respect, we can read in “The Vienna Circle Manifesto” that “Epistemological analysis of the leading concepts of natural science has freed them more and more from metaphysical admixtures which had clung to them from ancient time. In particular, Helmholtz, Mach, Einstein, and others have cleansed the concepts of space, time, substance, causality, and probability.”12 In the same way, Cassirer salutes “Helmholtz, who quite fiercely asserts a theory of knowledge proper to physics … and who, in order to accomplish this task must return to Kant.”
To be sure, Hermann Cohen severely criticized what he called Helmholtz’s psychologism, rejected his concept of “representation” as vacuous, and meant to give back to the term “transcendental” its methodological, not physiological, significance. However, Cohen’s reading was nonetheless marked by this positivism. Two essential traits of positivism can indeed be noted in his doctrine. On the one hand, as Cassirer himself noted, “Cohen thought that the transcendental method’s essential point is that it begins with a fact in order to investigate what makes the fact possible … But he limited this general definition by only offering the mathematical natural sciences as really worthy of this investigation.” And in fact Cohen, disputing the idea that philosophy is the analysis of representations (a term that is so vague, in his eyes, as to mean nothing) or even the analysis of consciousness,13 defined philosophy as the clarification of knowledge, whose only model is furnished by the mathematical and physical sciences. On the other hand, stigmatizing the conception of a Cogito as a mental event, Cohen makes the “I think” into a simple methodical principle. This second trait is undeniably true to Kant but at the same time entails the risk of a simply logical, even superfluous, subject. This second interpretation of the subject in Kant has been, by far, the most historically common, for, beyond the Marburg school, its traces can be found in the philosophy of the subject. Frédéric de Buzon, in an important article, “L’individu et le sujet,” shows how positivism, in almost all its forms, follows this Kantian line, “Philosophies of the concept are not philosophies of the absence of the subject, … these philosophies only assert a subjective lack of differentiation.”14 This influence can be detected even in the early Wittgenstein. Indeed, in the Tractatus Logico-Philosophicus, he still maintains a “sort” of transcendental subject, an empty form, the other name for the structure—in other words, a subject that is the world’s horizon, about which nothing can be said. Be that as it may, with this dual choice, Cohen follows Helmholtz’s positivist reading, even if for Cohen positivism takes on a different coloration, which we will examine in the next chapter.
In summary, Helmholtz’s Kantianism has chosen a particular orientation, investigation of the relation of our knowledge to things (the problem of reference). This orientation can either take the form of naturalization or Cohen’s more methodological form. But be that as it may, this orientation interprets critique as a form of positivism. And yet the beginning of the twentieth century will witness another reading of Kant, and from that, an apparent “bifurcation” in its interpretation through an overcoming of the positivist temptation. Nevertheless, this bifurcation takes place, in my view, within a single orientation, namely orientation toward the object. If Heidegger and neo-Kantianism certainly embody two different ways of reading Kant, these two ways nevertheless have a point in common: they reconstruct the entirety of critique exclusively from the problematic of the Critique of Pure Reason, and consequently do not plow the second path, of metacognitive justification. I will try to explain this proposition by reconstructing the Heideggerian reading of Kant and comparing it with Cohen’s.