(Dis)avowal of Science
“Back to the things themselves!” (Zurück zu den Sachen selbst): Such is Edmund Husserl’s famous battle cry for the phenomenological movement. That return to things means climbing down from the unassailable position of pure thought, which Ludwig Wittgenstein, the other great figure of the twentieth century’s philosophical revival, called “slippery ice,” where “conditions are ideal, but … we are unable to walk.”1 Husserl’s injunction, powerful in its impetus but vague in its direction, becomes more concrete when it is explained, to borrow Wittgenstein’s words once again, as a “return to the rough ground” (Zurück auf den rauhen Boden!).2 From the start, the rough ground Merleau-Ponty sought was that of the perceptual situation. Although the question is at work across the entire arc of his philosophy, it is possible to distinguish a first phase—extending from approximately 1933 until 1945—when the problem of the nature of perception constituted the guiding thread of his thinking. Merleau-Ponty, taking an interest almost from the first in the new results of experimental psychology, became particularly aware of the question of the perceptual through writings in Gestalt psychology, which he studied systematically. In 1933 or 1934, he began work on a thesis that would concern the nature of behavior. His requests for grants from the Caisse Nationale des Sciences reveal the driving force behind a line of research that would persist until Phenomenology of Perception.3 The originality that can already be glimpsed in these grant proposals is manifest in a bibliography filled with works on psychology, neurology, and psychiatry, at the expense of philosophy (no classic author appears on it). But though Merleau-Ponty thumbs his nose at academia, which was dominated at the time by idealism and French neo-Kantianism, he does not bid philosophy adieu, since he will be quick to declare his disagreement not with the object of science but with its methods. The thesis proposal asserts the irreducibility of the perceptual world to scientific epistemology: “The universe of perception could not be assimilated to the universe of science” (PrP, 13/75). Phenomenology of Perception is even more explicit, beginning with an apodictic and puzzling statement: “Phenomenology … is first and foremost the disavowal of science” (PP, ii/lxxi).
What are we to make of this assertion? Does not Merleau-Ponty fall into the trap of the same intellectualist philosophies he denounces in these lines? But we must first agree on what that “disavowal” means. Merleau-Ponty explains himself on that count with reference to Husserl’s own approach: It is not that phenomenology should be “indifferent” to the empirical sciences and psychology (PrP, 21/77) but that it must avoid modeling itself on their method. “It is a matter of renewing psychology on its own terrain, of bringing to life the methods proper to it by analyses which fix the persistently uncertain meaning of fundamental essences, such as ‘representation,’ ‘memory,’ etc.” (PrP, 22–23/78). When Husserl—and Merleau-Ponty as well—criticizes psychology, he hardly does so to call into question the legitimacy of an inductive approach; rather, he seeks to shore it up with an eidetic approach. There is unquestionably a continuity between The Structure of Behavior, the first book that emerged from Merleau-Ponty’s reflections in contact with the sciences, and Phenomenology of Perception (1945). In the earlier work, published in 1942 but already completed in 1938, he attempts to conceive of perception with and against science; in the second, he radicalizes that movement while making an effort to identify what science assumes and what, qua assumption, remains its unthought—the lived experience of the sensible world. In the preface, he maintains that “if we wish to think science rigorously, to appreciate precisely its sense and its scope, we must first awaken that experience of the world of which science is the second-order expression” (PP, iii/lxxii).
Before we arrive at the question of primordial expression—the subject of my next chapter—let me insist once again on the unitary aim of that entire first phase but also on the differences at work. Whereas The Structure of Behavior takes a negative path, insisting on why a phenomenological approach cannot be reduced to behavioral psychology, Phenomenology of Perception seeks to elaborate a positive account of the life of the embodied subject. The first book seeks, according to the explanation that Merleau-Ponty himself gives, to determine the sense (or nonsense) of an approach that considers the human being “from the point of view of the disinterested onlooker” (P2, 12), whereas the second, “placing itself within the subject” (P2, 13), brings to light the unthought, that is, experience, on the basis of experience itself. When we consider these two works side by side, we discover a chiasm: Where Phenomenology of Perception belongs fundamentally to a mode of thinking from inside experience and places positive knowledge in a position of exteriority, The Structure of Behavior does not yet situate itself within an experiential perspective but rather argues from within scientific discourse, producing a critique of the sciences that is paradoxically informed by these same disciplines.4
That interpretation may grant too much weight to Merleau-Ponty’s later efforts in 1951 to reconcile the arguments in his previous books in support of his pending candidacy for a position at the Collège de France (one cannot fail to observe the conceptual evolution that occurred between The Structure of Behavior and Phenomenology of Perception). In its intentions, however, the earlier book already brings about a radical displacement of the philosophical milieu within which Merleau-Ponty sought to speak. As a result, it reactivates the problem of the relation between body and soul—since, of course, it is this old problem that is, once again, at stake. And it does so without trudging through the aporias of the tradition. Rather, it installs itself from the start within scientific debate, which carefully avoids these metaphysical dichotomies. It raises once again the question of the relation between activity and passivity, not reducing them to the subject-object duality but analyzing what precedes that distinction, namely, behavior. In that sense, behavior is less the book’s theme—which is already and still perception—than its strategic apparatus. In opting for the scientific domain, then, Merleau-Ponty already distinguishes his thought from an intellectualist philosophy that dissolves perception into “thinking about perception.” In taking behavior as a shortcut across the field of the empirical sciences, he already chooses a path that will allow him to arrive at his object, experience proper.5 For the moment, however, let us remain within the layout of The Structure of Behavior. In it Merleau-Ponty studies very heterogeneous concepts one by one, still struggling to identify his own point of view, which has led a number of commentators to set aside that first book in favor of the second. I believe, however, that by 1933–38, when he was working on The Structure of Behavior, the critical perspective that would persist throughout all the different inquiries to which Merleau-Ponty subsequently dedicated himself was already taking shape: namely, a philosophy that denounces any philosophy of transparency. Although the term “transparency” does not yet have the value of an operative concept that it will later assume,6 the motif is already undeniably present. In addition, the opacification of perception and its reinsertion into the life of the embodied subject, which will lead to Phenomenology of Perception, is no doubt the unifying moment of the disparate analyses of the earlier book.
Between the Mechanical and Gestalt
Like Husserl, who recommended beginning from a naïve attitude and from the “representations” we form of things, Merleau-Ponty, in this treatise on behavior from the vantage point of science, feels obliged to identify the doxa in the “scientific representation” of behavior (SC, 199/184). He starts by identifying the naïve attitude of classic science, which later returns under the designation “philosophy of causality.” Although never developed in detail, that expression is used to stigmatize any position that trusts in the possibility of a direct, immediate, and linear action on an object. Such a philosophy understands the cause as a “constant and unconditioned antecedent” event (SC, 7/9), the necessary and sufficient condition for the effect to occur and thus always verifiable, provided it is isolated correctly. As a result, what presents itself as a coherent realism amounts merely to an atomism that dissects processes into so many independent links; the task is to eliminate the external elements to reach a pure relation whereby the cause unconditionally gives rise to the effect. And that causalism is in no way restricted to the physical world. Animal reflexology, to which Merleau-Ponty devotes a substantial part of the book, purports to include the physiological, but without doing away with the causalist postulate, which it in fact continues to embrace. In studying the instincts of dogs, Ivan Pavlov, the Russian physiologist, even broadens its domain, since he posits a direct action of the physical on the physiological. Pavlov’s theory of reflexes would generate absurd extensions: From the moment this theory presupposes an original state of direct correspondence between stimulus and response, it is constrained to invent “inhibiting powers” that “interfere with” the transparency of the immediate reflex. The problem is insurmountable so long as behavior is conceived as the correlate of a reflex produced in an empty channel. Even when an interaction between different channels is posited, the existence of a constituted world with preestablished connections (SC, 35/35) is never questioned. Through that decomposition of the world into all its fixed connections, causalism becomes engulfed in interminable Zeno’s paradoxes.
In The Structure of Behavior, Merleau-Ponty also discusses the philosophical positions that reject such a materialist atomism. He shows that, ultimately, they too still operate within the very system they reject. On the one hand, vitalism claims to reintroduce vital force where there is only mechanical force. Although that vital force evades the classic physicalist explanation, its effects must be reintegrated into the mechanical and turn out to be only a supplementary factor (SC, 149/138). As a result, vitalism paradoxically takes the forms of a “physics in the living being,” without really managing to conceptualize the “physics of the living being” (SC, 164/151). On the other hand, intellectualism moves beyond the juxtapositions of atomism, but only to displace the sensible to the sphere of understanding. Given the heterogeneity of essence between the sensible and the perceived, there would no longer be a relation of contiguity but rather a relation of conformity. Neo-Kantian thought—strongly rooted in interwar France and Merleau-Ponty’s target whenever he uses the term criticisme—would simply reduce perception to one mode of judgment among others (SC, 217/201). That prevents it from putting “consciousness in contact with an opaque and foreign reality” (SC, 283/224). In contrast to mechanistic philosophy, which understands behavior in terms of an initial state of transparency—and whose approach is not truly called into question by vitalism but rather confirmed by it—and in contrast to intellectualism, which reduces behavior to the instrument of an ideality, the key issue for Merleau-Ponty is to conceptualize the organization of the sensible in terms of perception. Contrary to any reductive doctrine, which, in the case of intellectualism, gives rise to a transparent model of thought and, in the case of mechanism, to a “mosaic” of sensations (here Merleau-Ponty is using Max Wertheimer’s expression), he believes one must consider “the total ‘image’ of the organism” (SC, 22/23). He finds the principle for that comprehensive approach in Gestalt psychology, to which he will ceaselessly return until the last of his writings. Whereas materialist atomism claims to provide explanations on the basis of a fragmentation of the causal chains, the Gestalt school’s credo is that the whole is always greater than the sum of its parts (SC, 49/47). Even if there truly is a distinction of the parts within the whole, these parts cannot be separated out, because they constitute an organic totality. Form (Gestalt) is neither in things nor in consciousness: It organizes their relation. Above all, it is structure.
Much later, at a conference on the word “structure” held in Paris in January 1959, Merleau-Ponty will provide a definition of that term that applies equally well to “form”: Structure is an “internal principle of an observable distribution” (S, 154).7 The emphasis is on “observable,” since, contrary to the more common usage of the term by what is usually classified as “structuralism,” this is not a structure underlying the world of the senses but a structure within the world of the senses. As such, it is constitutively sensible. Even though, by virtue of the terms “form” and “structure” (often used interchangeably in The Structure of Behavior), a number of inherited forms of reductionism are swept aside, other difficulties persist. For Merleau-Ponty, the school of Kurt Koffka and Wolfgang Köhler tends to substantialize form, to abstract it from the real interactions of experience. Although he saw the research of the Gestalt theorists as “relieving neurologists of the task of looking for decals of mental functions on localized anatomical functions” (PrP, 12/74), any theory of form detached from the permanent readjustment processes involved in perception runs the risk of establishing in turn a new apriorism. The proof, if any were needed, consists of the few recurrent and canonical examples of visual forms appearing in the works of Gestalt theory, which risk placing form before its perception and of bringing an insidious comeback of Platonism. The Gestaltist approach, moreover, even if it rejects all idealism and resolutely locates forms in the immanence of the physical world, does not see that preestablished forms do not exist in the world; rather, forms are always already the emergence of a world. The formulation of Phenomenology of Perception is explicit: Gestalt “is the very appearance of the world, not its condition of possibility. It is the birth of a norm, not realized according to a norm” (PP, 74/62).
In chapter 4 of The Structure of Behavior, Merleau-Ponty thus shows that there can be no alphabet of symbolic forms, no order to preestablished elements, but only structured structures, which are at the same time always already structuring structures. Hence perceptual consciousness is neither the receptacle of that “life of forms” nor a duplicate of the structures of the world but rather the site of its emergence (SC, 157/145). In that sense, there is neither exteriority nor indifference between the sensible world and consciousness. The relation between consciousness and one’s own body—Merleau-Ponty reminds us from the outset that this is what is at stake in The Structure of Behavior—is not an instrumental relation of “means” but an embeddedness in a “milieu.” Consciousness is a “milieu of the universe,” and phenomenology, as a science of what is given to consciousness, is an “inventory” of that milieu (SC, 215/199). Many have insisted—and rightly so—on the fecundity of these ideas, which would give rise in particular to the theory of the body as milieu and then to the theory of creative expressiveness. However, only a few have looked into the genesis of the vocabulary used here.
Milieu
In his use of the term “milieu,” Merleau-Ponty was primarily inspired by the ethologist Jakob von Uexküll8 and the psychopathologist Kurt Goldstein.9 The word belongs to a long-standing tradition on which it would be worthwhile to pause. Paradoxically, the notion of milieu, used by Merleau-Ponty in his critique of mechanistic determinism, is itself derived from mechanics. It is essentially in that sense that it appears in Diderot and d’Alembert’s Encyclopédie: “Milieu, noun, masculine (Mech.): in mechanical philosophy, it signifies a material space through which a body in motion passes, or in general, a material space in which a body is situated, whether or not it is moving.” More exactly, the French term milieu is a translation of the English “medium,” which Newton used to explain the phenomenon of action at a distance. Unlike Cartesian mechanics, which—as Merleau-Ponty will show in Eye and Mind—understands action in terms of “impact” (vision, the sensation at a distance par excellence, is explained through the analogy of the blind man’s cane feeling out the surface of the ground and of objects), Newton believes that every action is produced in “fluid mediums.” That so-called fluid medium, however, is still strictly determined by its physical properties. Therefore, the “resistance of the medium” in question in the entry of the Encyclopédie is associated with the physical “density” of the parts composing the medium. Air, water, and glass are given as examples.
That entry from the late eighteenth century, however, already contains the germ of what will become, in the nineteenth century, the behaviorist theory of the milieu. “Water,” writes the author of the entry, “is a milieu in which fish live & move.” It is this aspect that will attract the attention of Jean-Baptiste de Lamarck, the French biologist, who was deeply interested in milieux (the expression is always in the plural) such as water, air, and light. He draws from them his conception of “influencing circumstances,” the set of factors that determine the organism from the outside. Léon Brunschvicg (whose courses Merleau-Ponty attended while studying at the École Normale Supérieure between 1926 and 1930) had revealed that the explanation for the living thing proposed by Lamarck is nothing other than a transformation into biological terms of the physico-mathematical system of Newton’s “external interactions.”10 In that theory of the living milieu, the organism is thus exposed to action at a distance, in the sense that, through the intervention of need, the medium or milieu conditions the organism. Lamarck insists, however, that the organism, while always tending toward the finality of conformation, is the source of all motion. As a result, Lamarck’s theory of evolution, far from being reducible to Darwin’s attacks, is an attempt to conceptualize a dynamic and temporal relation between the organism and the milieu. Whereas, in the nineteenth century, Auguste Comte’s and Hippolyte Taine’s philosophical positivism would insist on the determining aspect of the milieu, making human beings mere “products of the environment”11 (twentieth-century behavioral biology would inherit that adaptionist and creativist tradition),12 Merleau-Ponty would return in his own way to Lamarck’s dynamic intuition. The sociological perspective, which postulates that to be a human being means to be among others, and to be shaped by their social milieu, then appears to be related in some way to a more primordial dimension: Being human is being not only in a milieu, amid a milieu, but rather being through a milieu, via a milieu. (Note that parmi, the French word for “among” or “amid,” is derived from the classical Latin per medio, “through the medium.”)13
In following that intuition, Merleau-Ponty will take a particular interest in the psychobiological studies of Jakob von Uexküll and Kurt Goldstein, which establish that biology cannot be modeled on physics: That is, it cannot consider the relation of the organism to its outside in terms of an sheer physical causality. Whereas the relation between the milieu and physical objects is quantitative, that between the milieu and its organism is qualitative. It is this quality that is missing from behaviorism, which, to borrow Bergson’s definition of the comic, “plasters the mechanical onto the living [plaque de la mécanique sur du vivant].” Merleau-Ponty will articulate his objections more succinctly in his Sorbonne lectures collected in Child Psychology and Pedagogy, which are nothing other than a reformulation of the results of his first two books. Behaviorists believe they can situate themselves within the geographical setting and deduce scientific data from it (PPE, 432ff./343ff.). Here Merleau-Ponty is using the distinction, proposed by the Gestalt psychologist Kurt Koffka in his Principles of Gestalt Psychology, between “geographical environment” and “behavioral environment.”14 The notion of behavioral environment supposedly proposes a qualitative theory of behavior to counter scientistic behaviorism, but it overlooks what it is that behaves, in this case, an organism. It seems, therefore, that Merleau-Ponty’s major objection to Koffka’s Gestalt psychology would be that it remains trapped within a notion of the environment that—in not taking the dynamism of the organic into account—does not really manage to come up with a theory of the milieu. In other words, Gestalt psychology neglects the dialectic between the living thing and its milieu because it lacks a way of thinking about life.
Two conclusions can be drawn from this analysis of those theories of perception that deem themselves anti-intellectualist, and aren’t at all, according to Merleau-Ponty. First, a mechanistic interpretation presupposes an immediate action between the perceived and the perceiving. Its paradigm is the theory of reflexes, which assumes there is a direct transmission from the starting point to the endpoint. Laughter could then be reduced to an electrical stimulation of the facial nerve, as in the experiment of Georges Dumas (PPE, 554/446–47). And yet, explains Merleau-Ponty, basing himself on Kurt Goldstein in particular, if there truly is a relation between a stimulus and the response, that relation is not direct: It “necessarily occurs within a milieu: a field of forces” (PPE, 433/346). Second, Gestalt psychology attempts to conceptualize the organization of the field and its lines of force by means of the concept of Gestalt, or form. Although Gestalt avoids causalism, according to Merleau-Ponty it quickly loses its roots in perception and tends in the direction of formalism. Even while asserting the coexistence of physical and mental structures, it does not grapple with their complex relation but confines itself to positing a general isomorphism. Where mechanism postulates immediacy, Gestalt psychology espouses the coincidence of structures. In other words, whenever there is a transparency of structures, the Gestalt appears.
Even as he uses the results of mechanism and Gestaltism to refute the classic notions of transparency, Merleau-Ponty maintains that each of them lacks, for reasons of its own, an authentic concept of perception, because the fundamental element is missing: the living body, which will be placed at the center of Phenomenology of Perception.
From the Milieu to the World
What is a body? Anything we will be able to say about a body, about any body, we will say by the measure of the first body that is given to us as experience: our own. Nevertheless, we have very diverse experiences of that body. We experience it “from the inside,” as a sensible, living, moving body, a body we manipulate and by which we act. But that same body can also turn inside out, can become “external” whenever we feel another looking at us. For an instant, we are no longer altogether “in ourselves”; we imagine our bodies perceived by others, an object exposed to an alien gaze. Our bodies become objectified in the experience of being-seen, but also whenever we gaze into a mirror: What we see there “from the outside” never entirely corresponds to what we live “from the inside.”
The inherent doubleness of the body was one of Husserl’s first intuitions. Since the German language possesses two words where the Latin languages have only one, Husserl distinguishes between the body as object (Körper) and the body as subject (Leib), living body, Leib stemming from the same root as Leben (life). Everything seems to indicate, however, that the distinction, fundamental in Phenomenology of Perception, became essential for Merleau-Ponty even before he had read Husserl, through contact with the writings of Gabriel Marcel. How to account for the fact that I am a body and that this body is not just any body but rather my body? The sentence “I am a body” remains void, so long as we have not realized that being a body amounts to having a body. That having, however, should not be confused with possessing. In Marcel’s 1935 book Being and Having (Merleau-Ponty would publish a book review of it the following year),15 the body thus plays a pivotal role, situated at the crossroads between being and having but also outside that dichotomy. The body “is not,” inasmuch as it is never given to me entirely: It will never become an “object.”16 Even though I have my body, even though it is the vehicle of my actions, it is never fully accessible to me. Phenomenology of Perception can be read as a response to that configuration, which is still all too abstract, since it does not sufficiently account for the fact that to have a body is to act not on it but through it.
The matter at hand, then, is to become conscious that, “for a living being, having a body means being united with a definite milieu, merging with certain projects, and being perpetually engaged therein” (PP, 97/84). The perceiving body is always a project(ion), one that is directed, in tension: That is the meaning of the Husserlian notion of intentionality, which Merleau-Ponty will eventually turn into the concept of “motor intentionality” (intentionnalité motrice). But what is the body’s project, what does it “project”? Not its objects or even really its “intentions”—rather its field, like a projector or even, as Merleau-Ponty explains, a milieu. “Insofar as it projects a certain ‘milieu’ around itself” (PP, 269/241), the body is always more than its actuality; it is also and equally in its virtualities.17 In that sense, the embodied being is not only a being “in a situation,” it is defined as a “possibility of situations” (PP, 467/431). These possibilities are not only thinkable but concretely realizable, a characteristic that distinguishes them from any transparent thought. “By saying that this intentionality is not a thought, we mean that it is not accomplished in the transparency of a consciousness, and that it takes up as acquired all of the latent knowledge that my body has of itself” (PP, 269/241). As a potentiality in a milieu, the moving body cannot be reduced to the full autonomy of a pure subject or to the heteronomy of an environment. To grasp its mediacy, therefore, is to allow it to maintain a certain thickness, which it loses in that strange alliance between scientific naturalism and spiritual ontology: “Thus, while the living body became an exterior without an interior, subjectivity became an interior without an exterior, that is, an impartial spectator” (PP, 68/56). Whereas The Structure of Behavior deliberately begins “from below,” Phenomenology of Perception situates itself from the start in the human experience of one’s own body, which already shifts the perspective. “Human life … comprehends itself because it is thrown into a natural world” (PP, 377/341, trans. modified). The question of the human thus arises as a problem of the world.18 Here again, Merleau-Ponty is using Uexküll’s studies: The French philosopher’s concept of milieu is quite simply the translation of the German term Umwelt, which the ethologist takes care to distinguish from Umgebung and Welt.
Jakob von Uexküll defines the Umgebung in terms of its location in an isotropic space, comparable in that respect to Koffka’s “geographical environment.” Conversely, the Umwelt would be a qualitative space corresponding to the “behavioral environment.” In conceiving of the Umwelt as specific to the living thing, in contrast to the purely geographical Umgebung, Uexküll confers a positive meaning on the term; his approach “from below” distinguishes him from Heidegger’s privative zoology.19 Heidegger, starting from the human Welt, will in fact conceive of the animal Umwelt as necessarily deficient. Whereas for Uexküll, the surrounding world (Umwelt)—unlike the Umgebung—is already a world, for Heidegger the Umwelt is only an impoverished Welt, an “un-Welt.”20
On the question of animality, some have been too quick to see Merleau-Ponty’s notion of the world as being derived from Heidegger’s. Unlike Heidegger, who believed that recent efforts of zoology to understand the organism in its relation to its environment ultimately remain pointless for philosophy, Merleau-Ponty confirms and extends Kurt Goldstein’s and Frederik Buytendijk’s analyses of organic life. Contrary to the Heideggerian conceptualization, there would be not so much an encirclement (Eingenommensein) and captivation (Benommenheit) of the animal by the Umwelt as, on the contrary, an Auseinandersetzung, an agonal negotiation of an almost Spinozan cast.21 The organism, far from “always fitting snugly into a determined milieu,”22 invariably engages with it. There is hardly a fundamental caesura between the animal and human realms, merely a difference in the modality of that engagement. In short, in Merleau-Ponty’s conception, the world—in contrast to an inhibitory Umwelt—does not comprehend be-ings (étants) in the ontological sense of comprehension (understanding, Verstehen) but rather comprehends (in the sense of encompassing, enveloping) the potentiality of the milieu. Rather than being dependent on triggers that would be something like the “casing” or “rails” of behavior (N, 283/221, trans. modified),23 human beings maintain a relation of distance, which is always a creative distance. Hence human life “ ‘comprehends’ not only some definite milieu, but rather an infinity of possible milieus” (PP, 377/341, trans. modified).
It follows that the notion of milieu takes on a new philosophical value in Phenomenology of Perception, inasmuch as it is redefined in relation to the body: The body is no longer a “transparent envelope of Spirit” (PP, 187/163) but a “means” to make a world out of a milieu (PP, 144/125). Whereas Heidegger sees the Umwelt as closure and encirclement, Merleau-Ponty draws from the milieu his conception of openness. Without abandoning the possibility of distinguishing between the animal realm and the human, he makes that distinction not categorical but a matter of degree. Fundamentally, both animals and humans are at once situated in a milieu and open to it, though human beings potentiate that openness by creating their own worlds. Even while maintaining a division between the animal and the human, Merleau-Ponty explains the development of the latter on the basis of the former; human beings, in using the possibility offered by the condition of openness, liberate themselves from their objective determination.
In the chapter “The Spatiality of One’s Own Body and Motility,” Merleau-Ponty considers the case history of “Schn.” (Schneider is his real name), who cannot manage to visualize himself while he is carrying out an action. A leather worker, he continues to sew and cut leather but is unable to take any distance from himself, to objectify his activity, even less to project himself into different activities. Schneider has merged with his Umwelt, the milieu of the leather worker, whose possibilities are limited and univocal: “The normal person reckons with the possible, which thus acquires a sort of actuality without leaving behind its place as a possibility; for the patient, however, the field of the actual is limited to what is encountered in real contact or linked to these givens through an explicit deduction” (PP, 127/112). In order for a world to be, a distantiation, a “spacing,” must occur: The immediate givens must be broken up, the actual virtualized. For the nonpathological subject, a stimulus is not simply a lever for actualization; it can also give rise to a “virtual movement” anticipating “a certain power for action within the frame of the anatomical apparatus” (PP, 126/111). The body of the normal subject thus possesses a faculty of nonactualization, the faculty to “situate itself in the virtual” (ibid., trans. modified). Oddly, then, the act of turning away from the world (ibid.) truly constitutes the condition for a projection of a world. In the course notes collected in Nature, Merleau-Ponty will later write: “No longer the body as fusion with an Umwelt but rather the body as means or occasion of the projection of a Welt” (N, 284/222).
It is this transition to a world that Gestalt psychology cannot manage to grasp because it remains prisoner to the idea of an objective world. Inasmuch as it places behavior—as one form among others—in the world, it does not take into account the fact that behavior indicates above all the possibility for the emergence of a world. In short, Gestalt theory would ultimately locate forms in a physical world and would confine itself to describing the structural correspondence of that world to the representations we make of it. In that sense, it would be a strange alliance between a physicalist presupposition and a critical philosophical procedure. Yet the world, Merleau-Ponty argues, is not, as Kantianism still understood it, a system of a priori relations, it is not “like a crystal cube” that shows “its hidden sides … in its present transparency” (PP, 378/342). On the contrary, it is an inhabited, invested, and worked space, to borrow a term from Hegel that recurs frequently in Merleau-Ponty’s writings. When the world’s genesis is linked to the milieu, the world cannot be idealized: Like the Umwelt, the world “is an intermediate reality between the world such as it exists for an absolute observer and a purely subjective domain” (N, 220/167).
In following the notion of milieu throughout Merleau-Ponty’s reflections on perception, we have been able to see how his thinking developed by incorporating technical terms to which he assigns new meaning. In placing ourselves outside the domain of philosophy, we are witness to that philosophical revision, the reworking of a term that, far from becoming a rigid concept, remains operative and multiple. The issue at hand is not to remove a notion to the rarefied spheres of the mind but rather to remain in the space between, as Merleau-Ponty concludes in the summary of his philosophical itinerary: “Perhaps these convergent research projects will ultimately bring to light a milieu common to philosophy and to positive science, and will reveal to us, on the near side of the division between subject and pure object, something like a third dimension where our activity and our passivity, our autonomy and our dependency, would cease to be contradictory” (P2, 13).
Merleau-Ponty’s first phase, then, already participates fully in the effort to avoid reductionism and to consider intermediaries. As a result, milieu is not merely the operative concept that traverses that phase as a common thread; it also points to his philosophical project as such.24 It is hardly by chance that Merleau-Ponty returns to the concept much later, dedicating many sessions to the milieu in the courses on nature he taught in the late 1950s. By virtue of the notion of Umwelt, “the view of the world is not reduced to a sum of exterior events or to a relation to the interior which is not taken in the world” (N, 232/177). To understand the role that the milieu plays for animals, Merleau-Ponty continues, one would have to compare it to our dream consciousness, directed toward something that is never seen for itself (N, 233/178). The dream could be seen as the prefiguration of a “new notion of the possible” that cannot be reduced to “another eventual occurrence” (RC, 137/98). Rather, it would be the place where, in a series of dream images lasting no more than a few seconds, the infinitude of possible links, always remaining in suspense, would take shape. In place of a direct and adequate vision, reveries offer us a vision that, though it hardly allows us to fix phenomena, does provide a glimpse of them. (So too does perception in a state of fatigue, as Peter Handke describes it in his memorable “Essay on Tiredness,” which it would be worthwhile to include among the classics of phenomenology.)25 In opposition to any epistemology of transparency, what is required is a return to behavior within the milieu, to the “Alltäglichkeit, [which] is always in the in-between world, always between the lines” (N, 268/207, trans. modified). Instead of “trying to squeeze [things] in the pincers of fragmentary events,” we must restore the opacity proper to them; we must understand “the organism or the species as mass reality” (N, 268–69/207, trans. modified). We would then arrive at an “envelopment-phenomenon … which is not to be sought behind, but rather between the elements” (N, 275/213).
That enveloping milieu is less a container than the gap, the distance, that sustains the elements on the inside. It will become the guiding idea not only of the ontology of the visible (as I endeavor to show in chapter 3), but already of the phenomenology of linguistic expression, the second phase in Merleau-Ponty’s thought. The hinge between that first inquiry on perception and the reflection on the phenomenon of expression is the critique and reinterpretation of the classic notion of transcendence.
The Problem of Transcendence
“It would be necessary to define transcendental philosophy anew.” The Structure of Behavior ends with that statement (SC, 241/224). It may come as a surprise, given that Merleau-Ponty’s approach seems so different from a transcendental enterprise of the Kantian variety. The long succession of examples, the incessant return to the results of the empirical sciences, and the insistence on the prereflexive are all factors that, at first sight, would qualify Merleau-Ponty as a “precritical” philosopher in the sense Kant conferred on that term. But the truth is that Merleau-Ponty not only had the ambition to move beyond critique, he even aspired to reformulate its very principle, namely, the transcendental. In The Structure of Behavior, that reworking of the status of the transcendental extends little further than a programmatic sketch, but in Phenomenology of Perception he explains the reasons for it even before turning to the analysis of the phenomenal field.
It should be noted, writes Merleau-Ponty in the introduction to his 1945 book, that classic transcendental philosophies first postulate the necessity of experience, then seek to establish its conditions of possibility, but at no moment do they reflect the actual event of experiencing (PP, 74/61–62). Within such a system, there can be no real exteriority, either as a facticity of perception or in the experience of an Other irreducible to myself. Exteriority is assumed, but it never takes place: The transcendental “I” is anonymous; it is neither in me nor in others. But despite what seems to be an irrevocable condemnation, Merleau-Ponty seeks to save the Kantian project. What he criticizes in Kant is less his transcendental approach than, more precisely, the fact that he did not follow his program—namely, to define cognition in terms of the factual condition of the knowing subject—to its conclusion (PP, 255/266). Although Kant clearly saw that there could be no cognition without the intuition of the senses, he introduced the stratum of the a priori, which, though not chronologically antecedent, is logically so.26 Inasmuch as “the a priori maintains its character in Kant’s philosophy of that which ought to be the case, in opposition to what exists in fact and as an anthropological determination,” he introduces a hierarchy between “what the world ought to be and what the world actually is” (PP, 255/266). And that hierarchy leads him to fall short of his own objective. “If a world is to be possible …”: In that expression, which recurs several times in Kant’s writings, the rational subject is not only placed in a position that is in some sense prior to the world, the subject also becomes its lawmaker as it were, since the subject posits the conditions for the world’s genesis (PrP, 50/16–17). The heterogeneity between the transcendental aesthetic on the one hand and the transcendental analytic on the other cannot be resolved—still according to Merleau-Ponty—except by conferring on the subject the status of a God who posits the world, not that of a human who “is attached to it” (PP, 254/228). In the last instance, that amounts to reintroducing dogmatism into a critical philosophy that purports to be free of it. Only in appearance would such a philosophy of the constituting subject be the antithesis of the naturalistic view that the world is populated by constituted objects: Together they are only the twofold aspect that the prejudice “of a universe perfectly explicit in itself” can assume (PP, 51/44). The only difference is that, in the transcendental version, the de facto condition is combined with a de jure condition: In short, the explanation must be possible and consequently must already be “completed somewhere” (PP, 74/62).
That line of argument, which had a considerable impact on French phenomenology,27 remains somewhat hasty, however. To grasp its import, we need to take a detour through Merleau-Ponty’s sources: Husserl and Sartre, of course, but also and especially Eugen Fink. Before Merleau-Ponty, Husserl had criticized Kant for the idea of a law inscribed before the fact. Transcendental philosophy is thereby reduced to a “regressive inquiry” that wonders “in what conceptual and legal forms [Begriffs- und Gesetzformen] an objective world in general (a world of nature) must present itself.”28 In his effort to demarcate his own transcendental philosophy from that of his predecessor, Husserl judges that, for Kant, the idea of the transcendental is still confined to a “scientistic” framework consisting solely of the laws of nature. Husserl, by contrast, believes that this framework must be “broadened” to take into account “the multiple forms of human societies and the cultural products originating in the life of the community,” inasmuch as they too configure the experiences possible.29 Transcendental phenomenology is thus the description of intentional consciousness. That consciousness, much more than a mere condition a priori, would be something quite real, directed toward life and its “concrete plenitude.”30
And yet in The Transcendence of the Ego (1934), which we know Merleau-Ponty had read attentively, Sartre warned all who, on the pretext of wanting to correct Kant’s critical philosophy, sought to “realize”—that is, to give factual reality to—what was in Kantian thought merely a simple form of understanding. In his “transcendental turn,” therefore, Husserl would have reified the transcendental into an ego in which consciousness is situated: “Thus transcendental consciousness becomes thoroughly personal.”31 In that way, Husserl supposedly “weighed down” consciousness and—Sartre is playing on words here—in making it “ponderous,” he also made it “ponderable.”32
According to Sartre, however, the transcendental consciousness must be entirely purified of the ego, which would still be an alien element installed within it. Consciousness must be conceived as “clear and lucid” through and through: “The object with its characteristic opacity is before consciousness, but consciousness is purely and simply consciousness of being consciousness of that object.”33 There is thus no place for an “I” that would inevitably become an “occupant” surpassing the limits of mere consciousness, an object for that consciousness, which for that reason will never be able to be a subject of that consciousness. As a result, the transcendental field is inevitably a “pre-personal” agency and the precondition for any empirical “I.” Purified of any egological structure, consciousness recovers its initial “limpid density”: In order to be able to welcome the whole, it must be conceived as nothingness.34 With that interpretation of transcendental phenomenology, Sartre is notably in conflict with the reading Eugen Fink gives of it in his article “The Phenomenological Philosophy of Edmund Husserl and Contemporary Criticism,” which Sartre twice quotes.
In that essay, which appeared in Kant-Studien in 1933 and which influenced Merleau-Ponty very early on,35 Husserl’s last assistant tried to separate the use of the word “transcendental” in phenomenology from the sense neo-Kantianism was giving it at the time. The error committed by Husserlian critical hermeneutics would supposedly consist of applying to Husserl its own conceptual grids. Although beginning in 1906 Husserl used a somewhat Kantian terminology, he took care to distinguish the meaning he was giving to the terms he borrowed. According to Fink, one has to understand that, unlike in Kantian thought, for Husserl the transcendental ego is an individual and existing ego. Instead of contrasting an empirical self to a nonempirical transcendental subject, one would have to distinguish between an empirical ego, the object of a worldly apperception, and a transcendental ego, which is deprived of that apperception but is nonetheless a be-ing.36 That “constituting entity” (être constituant) is both less and more than the Kantian subject: less, because, in the correlation with the world, the form of subjectivity has no priority over the be-ings of the world qua possible objects of cognition; more, because consciousness is not only that which contains within itself the form of any possible cognition but also the very site of the constitution of the world. But the specific question of the being of that constituting entity is left hanging. Fink, even while presenting himself as the defender of Husserlian philosophy closest to its author’s will, has undeniably already added on his own thinking, strongly influenced by that of Heidegger. Like Heidegger (and unlike Sartre), Fink believes that what is at stake in the transcendental must be formulated not on the basis of a pure consciousness but on the basis of the facticity of the being engaged in the world. Although Husserl, with respect to Kantian epistemological laws, truly brought about a revolution by asking the question of constitution, “the question of the mode of being [Seinsart] of what does the constituting”—as Heidegger insinuates in his famous 1927 letter to Husserl.37
These issues, which are echoed in Eugen Fink’s and Ludwig Landgrebe’s subsequent texts of 1939, which Merleau-Ponty read immediately after his visit to Leuven,38 did not fail to give rise to reflections on his part. Husserl, of course, could do no more than dismiss flat out the Heideggerian criticisms, since in his eyes the operation of the transcendental reduction brought to light only the being of the constituting entity. In Phenomenology of Perception, Merleau-Ponty indicates, however, that phenomenology so conceived is barely distinguishable from intellectualism, or worse, psychologism: “This new ‘reduction’ could thus know but one true subject, namely the meditating Ego. This passage from the created [nature] to the creating [naturant], or from the constituted to the constituting, would complete the thematization begun by psychology and would no longer leave anything implicit or implied in my knowledge” (PP, 73/61). And he adds that “such is the standard perspective of a transcendental philosophy as well as, at least in appearance, the program of a transcendental phenomenology” (PP, 73/61). But, asks Merleau-Ponty, what remains of the phenomenological in such a philosophy? If the description of the lifeworld (Lebenswelt) is only a propaedeutic, destined to be replaced by a transcendental reflection “where all of the obscurities of the world would be clarified,” “it would not be clear why reflection would have to pass through the lifeworld” (PP, 419n1/553n14, trans. modified). But then what is an authentically phenomenological notion of the transcendental? Without a transcendental subject and on the basis of a description of intentional life experiences, can one still develop a philosophy that deserves the name “transcendental”? The Sartrian thesis of consciousness as a “prepersonal” agency would in any case provide an alternative to Husserlian subjectivism. But, Merleau-Ponty believes (PP, 320n/544n60), if one displaces agency away from the “I,” in order to make it a “relative and pre-personal” “I,” one pretends to resolve the aporia when in reality it only becomes more solidly entrenched. In his endeavor to abandon the idea of an ego that would be in front of the world, Sartre’s “nothingness” of consciousness is reduced to a container-milieu from which one can then draw an inventory, at will.
Along with Fink, Merleau-Ponty thinks that, on the contrary, we must conceive of the transcendental subject as a be-ing among be-ings, as intra-ontic being. Although transcendental reduction is imaginable, it can hardly be conceived in terms of the model of an ultimate ground that would have to be reached. Rather, it is to be seen as an interminable exercise that, through the modification of being—and thus of be-ings—makes it possible to glimpse a common ground that is never possessed. While he was reading Fink, Merleau-Ponty also discovered in Leuven Husserl’s fragment “The Earth Doesn’t Move” for The Crisis of European Sciences, a fragment that gave new coherence to that idea. Husserl says that, in response to the Copernican and Galilean revolutions, against which Kantian thought can be measured, he is starting a counterrevolution. In this text, which is also known as “The Overthrow of Copernican Theory” (“Umsturz der kopernikanischen Lehre”),39 Merleau-Ponty believes he can detect the archetype for another “transcendental ground,” at the opposite extreme of the early Husserl’s logicism. The Earth, he paraphrases, “is not in motion like objective bodies, but not at rest either, since we cannot see what it could be ‘tacked on’ to” (SG, 227/180). As “soil” or “stem,” it would prefigure a new idea of the sensible foundation “of our thought as it is of our life” (ibid.). Even before there can be a transcendental philosophy, empirical experience is necessary: “As Kant himself said profoundly,” Merleau-Ponty writes in “The Primacy of Perception,” “we can only think the world because we have already experienced it” (PrP, 50/17). Kant’s error was to have wanted to seek a ground more solid than the uncertain ground of the world of the senses, which, however, is the only one available to us. To rethink the transcendental, the issue raised at the end of The Structure of Behavior, is therefore simply to become conscious that, after all, the world may be the transcendental condition of the transcendental itself—if, that is, that hyperbole still has any meaning.
“Along with the natural world and the social world,” Merleau-Ponty concludes, “we have discovered that which is truly transcendental, which is not the collection of constitutive operations through which a transparent world, without shadows and without opacity, is spread out in front of an impartial spectator, but rather the ambiguous life where the Ursprung of transcendences takes place, which, through a fundamental contradiction, puts me into communication and on this basis makes knowledge possible” (PP, 418–19/382). As such, that transcendental is literally a condition of possibility. Merleau-Ponty insists on that last term: The field, insofar as it is limited but nevertheless unenclosable, allows for variation; as “principle of indetermination” (PP, 197/172), it opens the possibility of another oriented attitude (the Husserlian concept of Einstellung) against the backdrop of a world characterized as horizontality.40
What interests Merleau-Ponty in the notion of transcendence is not so much a new notion of foundation (like Husserl in his transcendental idealist phase) as the act of transcending. (In that way, Merleau-Ponty paradoxically returns to the near side of the rigorous distinction, dating back at least to Kant’s Critique of Pure Reason, between the transcendent as a precritical dogmatic principle and the transcendental as an epistemological condition of critical philosophy.) Yet the matter at hand is not to move out of the world but to move permanently toward what is not yet possessed within the world, as Eugen Fink’s 1933 article suggested.41 The opacity of the world is correlated to and inseparable from an “active transcendence” (PP, 431/395), an “ek-stase” of the subject “oriented or polarized toward what he is not” (PP, 491/454), an “act of transcendence by which the subject opens himself to the natural world and carries himself along” (PP, 180/191, trans. modified), rather reminiscent of Ernst Bloch’s “transcending without transcendence.”
We would therefore have to see that reflection on apriority less as a critique actually directed at Kant (Merleau-Ponty grafts his own inquiries onto Kantian terminology) than as an indicator of a general movement. Undeniably—and despite the emphasis on inherence in a world-milieu—Phenomenology of Perception truly constitutes an effort to conceptualize the activity of the subjective pole. When the subject is collapsed to its corporeal condition alone, there is no longer any possibility of explaining how one moves beyond oneself—a point to which The Visible and the Invisible will return. The notion of a transcendental movement is still present in the uncompleted work, though Merleau-Ponty explains that that movement beyond cannot be conceived as a movement “by oneself,”42 only as a “mute transcendence” accompanying phenomenality itself. The field in which Merleau-Ponty finds, first, the marker for an active transcendence—which he attempts to describe as “ex-pression”—and then the milieu for an anonymization that dissolves the residue of a philosophy of consciousness and leads toward an ontology of the sensible, is the field of language.