/ The Logic of the Chiasm in Merleau-Ponty's Early Philosophy


The trajectory of Merleau-Ponty’s career is often seen as a progressive development: he begins by analyzing scientific consciousness in The Structure of Behavior, complements that account with a phenomenological analysis of behavior as lived in Phenomenology of Perception, and then overcomes the “philosophy of consciousness” to which the earlier texts are committed in the turn toward an ontology of flesh in The Visible and the Invisible. Through close readings of Merleau-Ponty’s engagements with Gestalt psychology in The Structure of Behavior, I argue that the immanent critique of Gestalt theory in that text already anticipates the chiasmic logic of flesh. This challenges the idea of a turn in Merleau-Ponty’s thinking. I begin by outlining the elemental, carnal, and reversible status of flesh. With careful attention to his source materials, I then distinguish Merleau-Ponty’s appropriations of Gestalt theoretical insights from his critical adaptations, defending three claims: (1) The Structure of Behavior borrows insights from Gestalt theorists that are undermined by their own, realistic ontology; (2) it modifies those insights to explicitly acknowledge the elemental status of nature; and (3) those modifications enable Merleau-Ponty to re-interpret Gestalt psychologists’ empirical findings, outlining how consciousness must emerge from nature as both carnal and reversible.

1. Introduction

In his introduction to The Structure of Behavior, Merleau-Ponty tells us that his goal “is to understand the relation between consciousness and nature” (SB 3/1). “By nature,” he clarifies, “we understand . . . a multiplicity of events external to each other and bound together by relations of causality” (SB 3/1). This is, at first blush, the classical picture of a law-governed nature, and it is tempting to suppose that no critical gesture will be made in its direction. Barbaras reaches just this conclusion when he tells us that “until [the Nature courses of] 1956–57, Merleau-Ponty utilized [the concept] in a non-critical way and conferred upon it the current philosophical meaning” (2001: 23). On this reading, Merleau-Ponty’s early attempts to unseat the naturalistic explanatory paradigm in psychology leave that psychology’s implicit ontological presuppositions intact. The trajectory of Merleau-Ponty’s career then appears to be one of progressive development: while he moves from analyzing scientific consciousness (The Structure of Behavior) to a phenomenological analysis of behavior as lived (Phenomenology of Perception), the philosophy of consciousness to which that analysis is committed is not overcome until the period of The Visible and the Invisible, when Merleau-Ponty develops the ontology of “flesh.” During this period, Barbaras tells us, we find a distinctive picture of sense-generating nature that is not “foreshadowed by anything” (2001: 23).[1]

There is much, I think, to recommend this reading. In The Visible and the Invisible, for example, Merleau-Ponty states that he is motivated by a lingering problem, namely that Phenomenology of Perception “in part retained a philosophy of ‘consciousness’” (VI 183/234). And while he continues to employ the concepts of “form” and “Gestalt”—so crucial to the argument of The Structure of Behavior—throughout the 1950s, these appear in The Visible and the Invisible mostly in the Working Notes, and in a sense that would be unrecognizable to the Gestalt psychologists from whom they are borrowed.[2] Despite this evidence, the trajectory Barbaras traces is misleading.[3] Merleau-Ponty’s appropriation of the “current philosophical” understanding of nature in The Structure of Behavior represents a strategic choice, not one borne of any philosophical allegiance; by the end of the text, he had abandoned it. Moreover, the picture of consciousness we discover there is already “profoundly modified” (SB 172/186) with respect to the classical picture of objective consciousness; it is a naturally emerging, bodily consciousness that cannot be appropriately accommodated within the naturalistic framework of Gestalt theory. Last, when Merleau-Ponty proposes a “philosophy of form” in The Structure of Behavior, he invokes a “primordial” sense of nature positioned directly against the classical picture presumed by Gestalt theory. This “primordial” sense anticipates the structure of flesh. The statement of The Visible and the Invisible that “what one might consider to be ‘psychology’ . . . is in fact ontology” (VI 176/228) is already guiding his thought there. These points can be seen, however, only with a careful reconstruction of the specific nature of Merleau-Ponty’s disagreements with Gestalt psychology in The Structure of Behavior—a project that has earned little attention among Merleau-Ponty scholars, and even less among those who focus on the later work.[4]

My plan, in making my case here, is as follows: I begin in Section 2 with a discussion of “flesh” in Merleau-Ponty’s late work, introducing the scholarly debates over its meaning. I then turn to Merleau-Ponty’s appropriation and critical adaptation of Gestalt psychology in Section 3 and Section 4, respectively. In these sections, I focus as much as possible on Merleau-Ponty’s source materials, since, absent this strategy, it is difficult to separate Merleau-Ponty’s gestures of approval from his immanent critique. Out of his engagements with these texts, I then reconstruct what I take to be the first gestures at the structure of flesh. I hinted that this is a surprising discovery, and I defend it in Section 5. To be clear, however, the point is not to foist the mature ontology onto The Structure of Behavior. It is simply to demonstrate that we cannot endorse the view that it is not “foreshadowed by anything.” With this in mind, I conclude with a brief return to The Visible and the Invisible. The ‘turn’ to ontology, I show, is best expressed—as Morris puts it—as a “shift [in] emphasis” (2010: 144).

2. The Structure of Flesh: An Overview

After the publication of Phenomenology of Perception, a problem became decisive for Merleau-Ponty. As he tells us in a Working Note from 1959, “the problems posed in Ph.P. are insoluble because I start there from the ‘consciousness’-‘object’ distinction” (VI 200/250), which is to say, he starts ‘too late’. To begin again, he needs to locate the common ground from which the subject-object division emerges, accounting, at the same time, for the primitive divergence within this ground that allows the division to take shape. The tension between these demands gives rise to the concept Merleau-Ponty first makes use of in 1951: the concept of “flesh.” With “flesh,” Merleau-Ponty describes a basic ontological structure that hitherto “has no name in any philosophy” (VI 139, 147/181, 190) because philosophy has been caught on one side or another of the antinomy that flesh is meant to resolve. To name and to develop this concept is therefore to respond to the question of the relation of the human being and the world, the object, or nature without presupposing a specific account of their difference. This is only an instrumental definition, however, and in this section I will sharpen it. To do this, I’ll follow Hass’s typology: the senses of flesh that coordinate to guide Merleau-Ponty’s use of the term emphasize the carnal and elemental status of flesh and its structure of “chiasmic” reversibility (2008: 201–3).

I begin with the “carnal” sense of flesh, since this association coincides with the first appearance of the term in “Man and Adversity.”[5] In that essay, “the notion of flesh, that is, animate body” (S 227/287) is said to emerge from a humanism capable of overcoming the antitheses of empiricism and intellectualism. “Flesh” therefore calls to mind what Merleau-Ponty names the philosophical advancement of ‘‘our century”: that philosophy has succeeded in ‘‘wip[ing] out the dividing line between ‘body’ and ‘mind,’ and sees human life as through and through mental [spirituelle] and corporeal” (S 227/287). This is a familiar theme for Merleau-Ponty. However, what emerges at this point is a different, more evocative way of articulating it: whereas in Phenomenology of Perception Merleau-Ponty spoke of the body as a “fabric into which all objects are woven,” and as “the general instrument of my ‘understanding’” (PP 244/282), the emphasis on the flesh of the body brings out its living aspect. The body is not an instrument that my conscious processes could take hold of, it is an “intertwining of vision and movement” (EM 124/20) that is in and of the world. To capture this second aspect—the sense in which the body is in and of the worldMerleau-Ponty speaks of the carnal status of the world. He states explicitly, for example, that “the world is made of the very stuff of the body” (EM 125/21) and that this “fundamental homogen[eity]” (VI 114/150) of sensing body and sensible world escapes the alternatives of a philosophy “flattened to the sole plane of ideality or to the sole plane of existence” (VI 127/166).

If body and world are made of the same stuff, this is not to detract from the privileged status of the body as we live it. In fact, the lived body takes center stage in Merleau-Ponty’s thought precisely because it marks the space where the subject’s interwoven mode of existing overlaps or “crosses” with the mode of existence of “the world.” The language of crossing, of course, recalls the optic chiasm, which is the literal crossing of optical nerves at the base of the brain. However, Merleau-Ponty evokes this comparison for two reasons. First, the optic chiasm is the condition of possibility of vision. The crossing of the lived body and a world made of the same “flesh” is thus invoked as a condition for perception. At the same time, the optic chiasm produces a blind spot where the optic nerves coincide with the retinas. This is the second point. By associating carnality with a chiasmic structure, Merleau-Ponty signals that a moment of vision includes an implicit reference to the invisible. That is to say, he makes ‘invisibility’ a condition of perception. As he writes in The Visible and the Invisible, what consciousness does not see is “what in it prepares the vision of the rest, . . . is what makes it see, is its tie to Being . . . its corporeity” (VI 248/296).

Now, a version of this point appears as early as The Structure of Behavior, where Merleau-Ponty notes that “my body as given to me by sight is broken at the height of the shoulders . . . [though] I am told that an object is visible for others in this lacuna in which my head is located” (SB 213/231). In this context, Merleau-Ponty is underscoring that the body is only available to sight (it is only an object for itself) when it excludes from sight the bodily conditions of its seeing. The physiological or biological importance of invisibility is therefore an early discovery. By the time of The Visible and the Invisible, however, the chiasm has taken on a second, related sense—the rhetorical structure of the chiasmus. This second sense, in which an expression is juxtaposed to its own syntactic inversion, is nicely captured in Gallagher and Zahavi’s claim that “the notion of an embodied mind or a minded body” is central to Merleau-Ponty’s thought (2012: 153). Indeed, Merleau-Ponty himself makes use of this structure when he speaks, as Toadvine points out, of the “becoming-nature of man which is the becoming-man of nature” (VI 185/236, as cited in Toadvine 2011: 339) By invoking this double (rhetorical and biological) sense of a crossing, Merleau-Ponty implies that flesh is not simply a carnal “thickness” that is shared by myself and world; it is an overlapping or “folding back” of one on the other that functions as the pre-condition of thought (VI 152/197). When commentators speak of the reversible structure of the flesh, it is this sense of Merleau-Ponty’s picture they are drawing on.

I want to be careful on this point. Already in the 1930s and 1940s, Merleau-Ponty links the chiasmic structure of the minded body to the reciprocity of that body and world. For instance, Merleau-Ponty points to the synesthetic imbrication of vision and tactility—“we see the depth, the smoothness, the softness, the hardness of objects” (SNS 15/26)—in order to demonstrate the mutual involvement of perceivers and their environments. “The sensing being [le sentant] and the sensible,” as he tells us in Phenomenology of Perception, “are not opposite each other like two external terms,” in part because the sensible is only brought out by its being “synchronize[d]” with my body (PP 221–2/259). By the late work, however, the imbrication of vision and tactility, of body and ‘things’ is compensated by the carnal status of the world. The relation cannot be one of mutual exteriority, then, not because the sensible requires the sensing body for its completion, but because this relation of completion goes both ways. In The Visible and the Invisible, Merleau-Ponty speaks of this phenomenon as the “pre-possession of the visible”: “How does it happen,” he asks,

that I give to my hands, in particular, that degree, that rate, and that direction of movement that are capable of making me feel the texture of the sleek and the rough? Between the exploration and what it will teach me, between my movements and what I touch, there must exist some relationship by principle, some kinship, according to which they are not only like the pseudopods of the amoeba, vague and ephemeral deformations of corporeal space, but the initiation to and the opening upon a tactile world. This can happen only if my hand, while it is felt from within, is also accessible from without, itself tangible, for my other hand, for example, if it takes its place among the things it touches, is in a sense one of them, opens finally upon a tangible being of which it is also a part. Through this crisscrossing within it of the touching and the tangible, its own movements incorporate themselves into the universe they interrogate, are recorded on the same map as it; the two systems are applied upon one another, as the two halves of an orange. (VI 133/173–4, emphasis added)

It is now clear, I think, why Merleau-Ponty wants to retain both the optical and the rhetorical senses of the chiasm. While he had long spoken of the sense in which things “look back”[6] at me, the carnal co-belonging of self and world that allows for a relation of “intertwining,” “reversibility,” or “encroachment” between myself and things now threatens to lose the perceiver in the perceived. To show that, in perceiving, we do not fuse with things, and so do not lose ourselves in them, Merleau-Ponty will have to show that there is always a node of divergence (écart) between the perceiver and the perceived, a “thickness of flesh between the seer and the thing” (VI 135/176) that holds open the space ‘between’ them. We discover this gap already in the sentient-sensible structure of the body: as Merleau-Ponty notes, “if I should suddenly wish to apprehend with my right hand the work of my left hand as it touches, this reflection of the body upon itself always miscarries at the last moment” (VI 9/24). However, now we find this same miscarriage halting the reversibility of body and world. What miscarries in this case is the loss of the privilege of the self-sensing body: the “fundamental narcissism of all vision” is, therefore, that even when “the seer is caught up in what he sees, it is still himself he sees” (VI 139/181). The immanence of auto-affection in the body therefore functions like the blind spot between my hands and between my eyes: it is the condition of possibility of my bodily exploration of the world.

The complicated network of crisscrossing structures and senses I just described marks an advancement over Merleau-Ponty’s earlier treatments of these themes (i.e., embodied mind, the synesthetic intermingling of sense modalities, our embeddedness in the perceivable world). But it now seems that they call for a new language. Merleau-Ponty must find some way to speak of flesh at a sufficient level of generality that he can capture the carnality of the body and the flesh of the world of things, while showing that that generality can be “pregnant” with each of these—embodied subjects, things—as possibilities.[7] This is the last point. To express its generality, Merleau-Ponty begins to speak of flesh as an element. “Element” is meant here in the Greek sense of the term, so that “flesh” expresses a mode of ‘being’ that is not coextensive with mind, or with matter, or with any substance at all. That flesh is “an element of being” is the closest Merleau-Ponty comes to a positive definition.

The elemental status of flesh is vague but evocative. Precisely what Merleau-Ponty means by it is, in part, where the debate about the status of flesh opens up. Consider, for example, the view that flesh is an element of lived perceptual experience. We find versions of this position in Dillon (1997) (for whom the late ontology makes explicit the “implicit ontology” of Phenomenology of Perception) and perhaps most prominently, in Barbaras (who is critical of it). In this framework, the structure of the lived body is analogically extended to explain or structure my experience of the world. I perceive with my body, but I am also perceived; I both touch and am touched; and these ways of being a perceiver actualize or realize different ways in which the world can be perceived. On this view, lived experience always relates to the structures of the lived body. The dynamism of ‘things’ is therefore a function of their synchronization with the body. Nature, on this view, is always the lived body’s perceptual engagement with nature.

This account of flesh has much to recommend it. For one, it offers a clear way to link the late ontology to Merleau-Ponty’s ongoing concern with anchoring a philosophy in lived experience. It is buttressed, moreover, by Merleau-Ponty’s tendency to speak of flesh in terms of its visibility. However, this conception, to its critics, threatens to become an endorsement of anthropomorphism (Barbaras 2002) or risks denying difference (Stawarska 2006;[8] Irigaray 1993; see also Grosz 1993). A sympathetic account of Merleau-Ponty’s ontology will therefore have to reckon with these charges: how can we accommodate Merleau-Ponty’s clear emphasis on the sensible-sentient structures of the human body without undercutting the autonomy of nature? I will come back to this.

A different strategy—one that seeks to avoid entrapping nature within human perception—is to think flesh within the context of an ecological ethic. This interpretation presses on Merleau-Ponty’s statements that everything is “of the same flesh” and treats the peculiarly ‘subjective-objective’ status of the body as ranging over the human and the natural environments. As Abram (1996) argues, the emergence of human (and other) bodies within the shared space of a carnal nature should establish a continuity among beings that implicates them in relations of reciprocal responsibility. A worry, however, is that this undercuts the instrumental value of the concept of flesh. For example, if both we and nature are “of the same flesh” in Abram’s sense, there can be no place for the necessary divergence that makes perception possible. This is in part why Toadvine (2009), also in the context of environmental philosophy, argues that carnal nature must retain an original sense independent of human perceivers. Nature, he tells us, “bears a meaning as its own internal configuration” (2009: 131) and human perceptions are expressions of that meaning. In this case, the irreducibility of human perceptual expressions to the nature from which they are drawn preserves the écart that makes possible our “contact with nature.” Merleau-Ponty’s thought therefore becomes a properly “environmental thought” only when, in contrast to the perceptual view, the body’s reciprocal self-relation of touching and being touched is no longer restrictively seen as “an activity of the subject” but as itself “an event of nature” (Toadvine 2009: 75). Whether or not this casts us back into the arms of anthropomorphism is an open question.

There is a third way to understand flesh that I want to highlight, one that takes flesh as the ground of a “relational ontology” (Bannon 2001). According to this view, the point of Merleau-Ponty’s emphasis on carnality is not to make the carnal structure of the body the basis of an ontology. It is to discover the carnal structures of being itself of which the body is an example. Bannon develops this reading by suggesting that flesh names a “general relationality . . . [or] an open relationship of affection” of which perception is “a singular instance” (2001: 345). The point here is subtle: whereas on the perceptual reading, flesh is “grounded in” the human being’s lived perceptual experience, on Bannon’s alternative, it is grounded “in a body’s relational engagement with its milieu (i.e., a system of affects)” (2001: 329–30). The point of seeing flesh as a relation is to target the idea that flesh is a projection outward of the lived experience of the body (that is, a projection from the site of a carnal auto-affection) and so to emphasize the “constitutive internal relations” of things (2001: 344). “The flesh of things,” on this view, names the logic of constitutive relations that makes any unity—the unity of thing or the unity of the body—the unity that it is.

In developing this picture, Bannon draws Merleau-Ponty’s late work closer to Heidegger’s. He argues, for example, that Merleau-Ponty assigns an “existential role” (Bannon 2001: 345) to flesh that is similar to that enjoyed by attunement in Being and Time. He also prescribes a challenge:

If an interpretation of flesh wherein the things themselves will possess a dynamism without resorting to an animistic hylozoism is possible, then it must begin . . . by showing how flesh is a common denominator between the human body and other sensible beings without . . . situating that commonality in the perceiver itself or a monistic substance. (2001: 343)

This double demand, I will argue, is precisely what is at stake in Merleau-Ponty’s earliest gestures at an ontology of structure in The Structure of Behavior. Indeed, a full account of structure in that text requires attention to each of the three essential feature of flesh that I have outlined: carnality, reversibility, and elemental being. To show this, let me now turn to the text.

3. The Appropriation of Gestalt Theory

The Structure of Behavior (1942), which is based on research conducted at the École Normale in the second half of the 1930s, is the published version of Merleau-Ponty’s thèse complémentaire—the preliminary thesis submitted in progress toward the doctorate. Among the published works, it represents his most sustained engagement with empirical psychology: the first chapter takes up, in a largely uncritical manner, the Gestalt-theoretical response to behaviorist psychology; chapter two extends that critique to “higher behavior.” However, precisely where Merleau-Ponty shifts to criticism is a matter of dispute.[9] This is in part because little scholarly commentary engages closely with Merleau-Ponty’s source materials.[10] In order to fully appreciate the significance of this text for understanding the late ontology, it is necessary to reconstruct Merleau-Ponty’s arguments in a way that corrects that unfortunate tendency. To that end, the present section begins by tracking Merleau-Ponty’s arguments, referring, wherever possible, to the material from which he adapts them.

3.1. The Critique of Behaviorism and the Primacy of Value

It is useful to begin by situating Gestalt theory. In his 1943 introduction, Gestalt Psychology: Its Nature and Significance, Katz tells us that “the rise of Gestalt theory can be understood only as a reaction against atomistic psychology” (1950: 3).[11] By “atomistic psychology,” he means behaviorist psychology, where behaviorism posits a one-to-one correspondence between any (physical or chemical) stimulus and its local response. Behaviorism is targeted, not only because it implies that an organism will reliably respond to a stimulus in a way that is not borne out by empirical data, but also because it situates itself within a scientistic framework that opposes scientific to ‘naïve’ explanations of simple behaviors. Since ‘scientific’ explanations reject (whereas naïve accounts permit) descriptions of behavioral responses that make ultimate appeal to meaning or intention, the Gestaltists argue that even behaviorism’s most sophisticated theory will remain an “artificially simplified system of psychology” (Köhler 1947: 248).[12] The Gestaltist alternative—which replaces the atomism of reflex theory with a concept of melodic form—results not only in a modified theory, but also in a modified definition of behavior and consciousness.

This difference in starting point also opens The Structure of Behavior. As Merleau-Ponty introduces it, it runs roughly like this: consider a spot of light as it moves along the wall of a darkened room. In what way do I visually register this event? One answer is that the light attracts my gaze. Glimmering at the edge of my visual field, the light draws my attention, such that, in tracing the beam with my eyes it is as if its movements “‘pull[ ]’ my regard along with it” (SB 7/5). This is a direct, first-personal description of behavior; it reflects the kind that is offered by naïve consciousness. As such, it makes no appeal to sub-personal physiological or nervous processes; it suggests that the action is goal-oriented or intentional; and, finally, it accords with the moment as I experience it. But what role should such descriptions enjoy in a psychological theory? The behaviorist responds that “naïve” descriptions of this kind must be eliminated from any account that rises to the level of scientific certainty. Those descriptions appeal to events (the experience of attracted attention; the motivated feeling of the movement of my gaze) that, qua internal and immaterial, are inaccessible to the neutral observer; for psychology to remain an objective science, behaviorists argue, the psychologist must therefore eschew any data gleaned from “direct experience.”

The method of behaviorist psychology is guided by this rejection of the data of consciousness. First, the behaviorist learns to identify, behind each observable behavior, whatever real stimulus is causally (thus predictably) producing the response. She then decomposes that cause into its various atomic data, demonstrating, for example, how the “vibratory movement” of “real light” evokes a behavioral response by provoking a one-to-one “successive excitation of points” on the “discontinuous receptors” of the retina (SB 8/6). That description allows her to maintain the ‘objectivity’ of psychology because, as Merleau-Ponty will put it in a lecture of 1950, it privileges quantifiabilty over a “consideration of values” and “general laws” over “the singularity of the event” (HSP 338). That is, it treats the “molar” behavior described by naïve consciousness as nothing more than the appearance of atomic molecular responses to a consciousness that does not directly experience them. Notice, in this case, that behavior is a natural thing in the classical sense of “nature”: as an object of neutral observation, it is composed of real events (vibratory movement, the atomic elements of the stimulus) interacting under causal laws. The relation of nature and consciousness is simply the relation of ‘the real’ to its paired subjective appearings. These appearings are not themselves the object of science.

For Gestalt theorists, there are several problems with the behaviorist strategy, beginning with its aim to conserve the status of psychology as objective science. As Köhler points out in Gestalt Psychology—one of Merleau-Ponty’s source materials for The Structure of Behavior—behaviorism is founded on a too-narrow conception of scientific objectivity that in fact obscures the relationship of consciousness and nature. Whereas it rightly insists that science should concern itself with objective certainty, behaviorism overcorrects by conflating direct experience with introspection. What this misses, Köhler argues, is that direct (viz., subjective) experience is not only “personal” and “private,” and thus that it is perfectly coherent to say that behavior as we experience it can be objectively described: if I reach out to touch the surface of the table I am sitting at, for example, the experience I describe would be publically accessible. “The table is hard and solid; it is cool to the touch.” The description therefore retains an objective character, which is clear enough when we contrast it with the personal experiences that are prompted by the same table: I notice, for instance, that this table is the one from my childhood living room; I remember the pattern of our wallpaper. What these experiences—the public, tactile experience and the subjective event of remembering—share is a dependence upon certain processes in the experiencing organism. It is this genetic sense of subjectivity that science should not—cannot—do without. As Köhler puts it:

The simple truth is that some of the experiences which depend upon processes in my organism have the character of objectivity, whereas others which depend upon different processes in the same organism have the character of being subjective. This contrast has nothing to do with the genetic subjectivity of both types of experience, i.e., with the fact that both depend upon events within the organism. (1947: 24)

The point of Köhler’s argument is that the reintroduction of descriptions of naïve consciousness—that is to say, of reports of direct experience—does not have to consign psychology to “a world of imaginary ghosts” (1947: 24). Indeed, in claiming to reject “subjectivity,” the behaviorists are already self-deceived.

The Gestaltists’ insistence on “subjective-objectivity” is underlined in The Structure of Behavior. Merleau-Ponty suggests, for example, that the “cleavage between the subjective and the objective” that opposes the “universe of science” to the “universe of consciousness” might well have been “badly made” (SB 10/8). He will repeat this criticism in a lecture at the Sorbonne, calling “the ideal of objectivity as . . . the simple notation of external givens” a “chimera,” since “the external world is always grasped from a human situation” (HSP 345), and again in the Nature course at the Collège de France, where he argues that the strict opposition of subjective and objective must be overcome in order to arrive at a “valid conception of Nature” (N 70). In each case, he underscores the primacy of direct experience: there is no necessary conflict in something being both ‘subjective’ and objectively certain, just as no conflict must separate psychology from everyday experience. Yet precisely the classical antinomy of “objective” and “subjective” has long prevented productive conversation between psychology and philosophy. That Gestalt theory seeks to overcome this antinomy explains why Merleau-Ponty singles it out for praise: in refusing to “turn its back on experience”—an error Merleau-Ponty associates with both scientistic psychology and “abstract philosophy”—“psychology. . . has been lead to converge with phenomenology” (HSP 337).

The stakes of this move are not only methodological: that is, the reintroduction of “subjective-objectivity” does not only bring psychology back into conversation with philosophy by permitting attention to direct experience, but in doing so, it “neutralize[s]” (HSP 342) the antinomy of quantity and quality and makes considerations of value an “adequate basis for science” (Köhler 1947: 26). Starting from value, Gestaltists can then show why the behaviorist framework was not only inadequate, but self-defeating: while it sought “to place the subject back in contact with the world,” as Merleau-Ponty put it—and “the social world in particular”—that aim can be accomplished only on the basis of a certain conception of ‘significance’ that behaviorism “cannot generate” (HSP 242).

To grasp this second objection, consider the constancy hypothesis. In a classical expression of the hypothesis, the one-to-one correspondence between stimulus and response predicts that any alteration in peripheral stimulation should register analogous changes in a subject’s response; in the same vein, any changes to a subject’s observable behavior are explainable by antecedent alterations of the stimulus. The atomic framework of behaviorism’s realist psychology first requires that this correspondence be conceived such that “the excitation and the reaction [decompose] into a multitude of partial processes which are external to one another in time as well as in space”—processes that are linked by “pre-established correlations” (SB 8/6). Since behaviorism understands behavior in exclusively ‘natural’ terms, however, stimulus and response must be linked by relations of causality: the response is a reflex response. Against this background, the constancy hypothesis helps maintain the ‘objective’ status of psychology by reducing purportedly meaningful behaviors to the series of observable and reliable reflex-reactions to real external stimuli.

On its own, this hypothesis is obviously insufficient. The idea of an absolute correlation of constancy between stimulus and response would render inexplicable any variation in behavior among subjects in a single environment, yet two subjects rarely produce identical responses to identical stimuli: Köhler showed, for instance, that when two monkeys are placed in a cage, one may stand on a box to grasp a banana, while another may not—the same physical stimulus (the box) provoking a response in one subject but not in another. The task of the behaviorist, who must maintain the direct, causal link between the organism and the world, is therefore to explain how the box is reflexogenic for one organism but not for the other. To do this, they must appeal to a series of “auxiliary hypotheses,” arguing, for example, that the pre-established pathways that link a given stimulus to its reflex response are not singular, but multiple, such that behavior can be altered through conditioned inhibition of or preference for certain pathways over others. The box, they argue, will provoke a response in a subject for which it has acquired significance. This is supposed to allow the behaviorist to incorporate a sense of significance that does not appeal to subjective internal processes, the elemental layer of real atomic stimuli being overlaid with significance through a process of psycho-physiological conditioning. The realist foundation of psychology is conserved, in this case, since significance is recast as a physiological preference for the performance of a conditioned reflex behavior.

There are serious problems, the Gestalt theorists argue, with these hypotheses: first, it is easily seen that aspects of an external stimulus that do not reduce to real atomic properties—e.g., the rhythm or intensity of stimulation—nevertheless influence the organism’s responses. Katz notes, for example, that

if an object is moving at a constant speed but is seen from various distances, its apparent speed does not change to any great extent . . . in spite of the fact that corresponding speeds of the retinal image vary considerably. [Conversely,] [s]ounds heard from various distances seem to be of approximately equal intensity . . . although their objective intensity changes markedly. (1950: 11; cf. PP 8/30)

Since “velocity” and “intensity” have no real existence in the classical sense, any attempt to reconcile phenomena of perceptual constancy with the theory of correspondence would have to accept that the stimulus and response are both, in some sense, more than the sum of their parts. How else to explain reflexive responses to the global properties of the stimulus? Absent some other compensatory hypothesis, behaviorism is forced to accept what Katz calls the “basic assumption” of Gestalt theory: that “a form (Gestalt) . . . is a final and irreducible entity” (1950: 49). Of course, to say that form is irreducible is not yet to say any more than that it is partially composed of irreal elements. The stronger point is therefore that any decompositional explanation of the stimulus will be out of step with empirical data. Such an explanation could not salvage the theory of conditioning for the reason that forms by definition have their own internal unity. Form, in a better phrase, is other than the sum of its parts.[13]

To illustrate this point, assume that a complex stimulus acquires significance for a subject over time by exciting a formal pattern of response, much like a machine could be programmed to modify its outputs according to the spatial or temporal arrangements of some external impact. (This is the basic claim of our auxiliary hypothesis, now complemented by the irreducibility of form.) How could this thesis explain the results of Katz’s amputation experiments, which demonstrate a striking facility among animals to compensate for the loss of a limb? Even under extreme circumstances, the adjustment is immediate: “A guinea pig which is deprived of all four legs tries to move by rolling its body”; a legless beetle will propel itself forward with its mandibles (1950: 16). In each case, Katz notes, the animals adjust “without the painstaking practice required by the principle of trial and error” (1950: 16) despite the fact that these organs “have never before functioned in such a manner, either in the animal’s own lifetime or, it is reasonable to assume, in its entire phylogenic history” (1950: 15). What Katz calls “the astonishing plasticity of motor structures” suggests, against the behaviorist, that an apparent behavior such as “forward locomotion” cannot reduce to patterned reactions to patterned stimuli, even if we conceive those reactions in complex terms; the subtraction of any one of the subject’s “motor innervation center[s]”—e.g., the loss of a limb—disrupts the pattern but not the apparent meaning of the behavior (1950: 16). The organism, in other words, can still engage in the learned behavior of “forward locomotion” without recourse to the local site conditioned to produce that behavior. What this shows, Katz argues, is that the loss of some “motor innervation center has repercussions on all the others. Holism must replace the older atomistic view of the nervous system” (1950: 16).

To Gestalt theory, the discovery of form, conceived holistically, serves to illustrate three things: first, behavior is best understood in terms of a total performative orientation of the body, a “dynamic self-regulation” of the organism in relation to its meaningful milieu (Katz 1950: 76). The legless beetle can continue to walk, even without the ‘appropriate’ physical apparatus, because the normal and the pathological performances of that locomotive behavior are unified by their common sense for the organism’s functioning. Likewise, when the child withdraws her hand from the candle flame, she is not mechanically repeating a gesture that is built from accumulated movements, she performs the meaningful behavior of protection—a behavior that does not simply actualize a series of afferent and efferent pathways in the nervous system, but can manifest through any number of physical motions, so long as they vary around “the fundamental theme” of self-preservation (SB 99/108). This thematic responsiveness cannot arise from a process of hierarchical conditioning because it is not a reaction to a determinate stimulus. It is, in Merleau-Ponty’s terms, an “aptitude” to respond to “a whole category of stimuli” of which the particular flame is representative (SB 99/108).

Second, the principal relations of organism and environment that govern behaviors can no longer be causal: this challenges a key claim of behaviorist psychology, according to which behavior is reflexive or reactive and the body a mechanism. Consider, for instance, the following example from Merleau-Ponty:

When my hand follows each effort of a struggling animal while holding an instrument for capturing it, it is clear that each of my movements responds to an external stimulation; but it is also clear that these stimulations could not be received without the movements by which I expose my receptors to their influence . . . Since the [organism’s] movements. . . . are always conditioned by external influences, one can . . . readily treat behavior as an effect of the milieu. But since all the stimulations which the organism receives have in turn been possible only by its preceding movements which have culminated in exposing the receptor organ to the external influences, one could also say that the behavior is the first cause of all the stimulations. (SB 13/11)

In this example, we see that external stimulations (which are putatively the cause of responsive behavior) are also pre-conditioned by (and so are the effect of) preparations of the body. The organism does not simply respond, but contributes to the constitution of form. Inasmuch as the organism and environment must be bound together by relations of causality, it therefore could not be a linear causality that links them, but circular. A closer look at adaptive behaviors of this type therefore shows that the classical concepts of stimulus and response are ambiguous. It is relevant, in this case, that Merleau-Ponty’s appeal to circular causality is conditional: “one could say” that behavior is both effect and cause. In bringing out the ambiguous status of the stimulus and response concepts, in other words, Merleau-Ponty does not take a stance on causality. The point is simply that the relation of body and environment must be primarily conceived as a reciprocal coordination of meaning.

In a moment, I’ll return to the status of causality. Finally, however, the discovery that thematic responsiveness is chiefly a coordination of meaning changes where behavior takes place. That is, behavior is no longer properly conceived as something in the body, but as a unity between the body and the world. As Weizsäcker puts it—Merleau-Ponty quotes him on this point—the behavioral coordination is such that “the properties of the object and intentions of the subject . . . are not only intermingled; they also constitute a new whole” (Weizsäcker 1927: 45, as quoted in SB 13/11).[14] Descriptions of behavior are therefore impossible without direct appeal to elements explicitly foreclosed by behaviorist analysis. Moreover, when it is conceived as a structured whole in which ‘stimulus’ and ‘response’ are meaningful and mutually constitutive (and so not properly stimulus and response), behavior necessarily loses its status as a thing. Though behavior comprises a “multiplicity of events,” these form a single, indivisible unity, which is held together by laws that organize and stabilize the “whole” from the inside. This makes it necessary to underscore that behavior, for the Gestaltist, does not simply establish a relation between the subject and the object (or between a consciousness and the natural world); the whole of behavior ranges over both and is irreducible to either. Behavior has a structure, which is to say that behavior is itself an organized form (Gestalt). To the question, “Can a science of behavior even begin without the categories of order and meaning?” Gestaltism plainly answers: “No” (Koffka 1935: 308).[15]

This final discovery requires Gestalt theory to introduce a new vocabulary. For example, to resolve the purported ambiguity of stimulus and response, it is necessary to distinguish an organism’s thematic responsiveness from its physical movements as existing in-themselves (to distinguish the behavior of walking-toward from the molecular reactions that propel the body of the organism forward). This becomes the separation of the geographical and the phenomenal organism. Whereas the first term describes the site in which physical, chemical, or muscular events are actually executed, the second describes the agent of behavior. In making this distinction, Gestalt theory can then demonstrate why the behaviorist is unable to generate the sense of meaningful behavior necessary to resolve the ambiguities of the response concept: her commitment to a certain ideal of “objective science” entails that all behavior consist in molecular reactions in the geographical organism.

We can track the identical problem for the ambiguity of the stimulus. In this case, Gestaltists separate the organism’s environment into a geographical setting and a behavioral or phenomenal milieu. The former, Koffka tells us, comprises the real, topological features of the environment; the latter is that same environment for the observer. Again, the point of the distinction is to undercut the behaviorist: if behavioral environments include, whereas geographical environments exclude, precisely those aspects of ‘internal’ or ‘subjective’ life eliminated by the behaviorist framework, the Gestaltist is able to say—perfectly coherently, and without appeal to ad hoc auxiliary hypotheses—why one monkey will stand on a box to reach a banana but another may not. Whereas the box-itself exists in a real geographical environment shared by both monkeys, only one of their behavioral environments contains a box-perceived-as-a-stool. When the “present trend of action” of one monkey causes the box to become “functionally alive,” these elements—the box, the trend of action—will coordinate into a meaningful whole (1935: 30).

Now, since Merleau-Ponty’s guiding question at this point concerns the ‘relation’ of consciousness and nature, I want to underline that neither the reintroduction of the “intentional character” of the animal’s reaction, nor its “internal relation to the situation” serves to cut the analysis of behavior off from “the world” (SB 93/103). Gestalt theory is quick to argue that the dynamic nature of situated behavior impacts the geographical setting and vice versa: the monkey desires the banana, not simply because it looks good (a behavioral fact), but because he is hungry; his hunger highlights that banana because it is there; finally, the accomplishment of the monkey’s desire will alter his geographical environment—he eats the banana. Yet Gestalt theorists are happy to describe this circular impact in causal terms. It is now clear why, in his reconstruction of the image of capturing the animal, Merleau-Ponty employed conditional language. That language signals, on the one hand, that the Gestaltist conception of behavior gets something right: unlike behaviorism, it refuses to locate meaning at the second-order, as if significance were “attached by the transference of reflexogenic powers” to “a first layer of reactions which would correspond to the physical and chemical properties of the world” (SB 130/140). On the other hand, Gestaltism maintains that an indirectly (circular) causal relationship exists between the geographic organism and its geographic environment. This causality functions through the mediation of the behavioral milieu instead of within the phenomenal environment or organism, occurring at the outer boundaries that separate the geographical from phenomenal environment (or organism). A clear statement of the relation of behaviorism to Gestalt theory would therefore be to say that Gestalt theory has not so much rejected but inverted the order of explanation of behaviorist analysis. While behavior, in other words, loses its status as a thing, it remains, by the classical terms, a natural event.

3.2. Consciousness in Gestalt Theory

I have shown that Merleau-Ponty is sympathetic to much of Gestalt psychology’s critique of behaviorism: he appropriates many of their examples, quotes liberally from Gestalt theorists, and at times follows their texts nearly point-for-point.[16] Merleau-Ponty also approves of the Gestaltist vocabulary: in his lectures on “The Human Sciences and Phenomenology,” for example, he tells us the distinction between geographical and behavioral setting “is valuable no matter what one’s position on consciousness may be” (HSP 343). But this praise helps to highlight the point at which they diverge. Where does the Gestaltist picture of behavior leave us concerning the status of consciousness? And in what sense does Gestalt theory respond to the question of the relation of consciousness and nature on behalf of a psychology of form? What Merleau-Ponty will call their lingering “abuse of causal thinking” (SB 176/191) is instructive here. For Merleau-Ponty, Gestalt theory’s answer to these questions still appeals to the “realistic postulates which are those of any psychology” (SB 132/143), and this undoes the gains it has made over behaviorism. In making this case, he anticipates the relational ontology of flesh.

We can approach the problem through Gestalt theory’s positive conception of consciousness. To tease out this point, recall that one of the chief goals of behaviorism was to construct a psychology “without consciousness.” On the one hand, this is a goal with which Gestalt theory sympathizes. It strives to incorporate the new conception of form into a structural psychology without, as Koffka puts it, “smuggl[ing] consciousness”—conceived in a classical sense—“through a back door” (1935: 35). At the same time, while Gestaltists insist that consciousness could not exist as a separate power for the organization of matter, it is nevertheless impossible to “write a psychology” without it (1935: 35). To approach this challenge, Koffka introduces “consciousness” in his Principles of Gestalt Psychology through its colloquial meaning, as when we speak of a boxer ‘losing’ consciousness. He thus highlights two things. Consciousness—whatever it may be—is an aspect of ordered behavior, and it has its site in the nervous system. The loss of consciousness, in other words, represents a change of neurophysiological condition and not the loss of ‘life’. Since Koffka, like Katz, conceives of the nervous system in holistic terms, he then insists that “we can study the chemical composition of the nervous tissue and will find no component we have not found in inorganic nature; we can study the function of this tissue and . . . find that it has all the characteristics of living tissue; and finally [we can find that] there is this relation between the life function and consciousness” (1935: 10) even if these are not coextensive. Considered under these aspects, “the central nervous system becomes, as it were, the nodal point where mind, life and inanimate nature converge” (1935: 10).

Now, by arguing that consciousness “converges” with inorganic nature in the nervous system, Gestaltists do not take themselves to advocate a “crude materialism” (Koffka 1935: 11). Materialism, Koffka thinks—which for him turns on an arbitrary privileging of matter—is so plainly false that it “is not necessary to refute it” (1935: 12). But it will not do to propose a different arbitrary division. That would be to tread the path of Cartesian supernaturalism (which separates mind from life and nature); vitalism (which separates nature from mind and life); or the philosophy of Scheler (who sharply distinguishes all three). Koffka argues that each of these paths forecloses the truly integrative psychology that is required by the mind-body problem, since each commits itself to establishing essential difference. As he understands it, the real task is to identify some dominant concept characteristic of—but not exclusive to—each category (i.e., material nature, life, and mind). This will allow him to accommodate their differences (as against materialism) without making them unbridgeable gulfs. The concepts Koffka identifies are quantity (in matter), order (in life), and significance (in mind) (1935: 13). He then notes that each of these is a feature of form. For example, a basic form like the figure/ground structure can be quantified—“units and shapes,” Koffka tells us, “must have a formula which will express them quantitatively” (1935: 174); it is a phenomenon of order;[17] and, finally, the figure/ground structure is meaningful as a holistic unity. Matter, life, and mind can therefore be united within a psychology of form.

This discovery has consequences both for what we mean by “quantity” and “order,” and for what we mean by “consciousness.” In its formal sense, for instance, quantity is not the negation of quality. It is simply a precise way of articulating qualitative phenomena. Likewise, vital order is not neatly opposed to a blind nature: “order is a consequence of organization, [and] since we need no special agent to produce it . . . organization [is] the result of natural forces” (1935: 175). Finally, if the meaningfulness of form is the dominant characteristic of “mind,” we can no longer sharply separate mind from matter or life (1935: 13). The psychology of form is not “without consciousness,” then, but its conception of consciousness is more or less equivalent to what Köhler called “direct experience”: consciousness is the site of the genetic dependence of behavior in the experiencing organism, yet its expressions in behavior can be objectively defined (1935: 35).

If this is what we mean by consciousness, we must highlight three things. First, since quantity and quality are no longer mutually exclusive, what constitute ‘mental’ behaviors must not be an orientation toward meaning, but an orientation toward “the richest [forms]” (Koffka 1935: 22). In other words, “significance” (the dominant feature of mind) cannot express a uniquely cognitive property, but is rather a complex or richly structured functional value. In Koffka’s example, just as the food attracts me when I am hungry, the mailbox has a functional value for me (it is significant) only when I want to mail a letter. What makes the latter a ‘rich’ functional value is that behavior oriented toward mailboxes is free in the sense that it is almost entirely disengaged from the vital functions of the organism. Mental forms of this type are not categorically distinct from forms of physical systems, but they seem to be because they are less expressible in quantitative terms.[18] To “express in a formula” a complex figure/ground structure like the emergence of the mailbox from the background of our sender’s behavioral milieu is more difficult than to express the formula of a circle drawn on a page (1935: 174–5).

Second, if the forms to which behaviors are responsive are distributed along a scale of ‘richness’, there is no longer an exclusive association of consciousness with ‘higher’ primates (with mind in the robust sense, for instance, of cognitive ethology). Koffka accepts this consequence, alluding to the “conscious” behavior of dogs and monkeys (1935: 35, 66). In doing so, however, he emphasizes that because we cannot know precisely when behavior becomes consciousness, the conscious/instinctive distinction will lose its privilege. Instead, vital forms of behavior are separated from higher, ‘conscious’ ones by the extent to which the behavior is oriented toward (and resistant or open to deformations from) the outside. The demand-character of the apple when I’m hungry, for example, is made functionally valuable by a strong ‘internal’ vital need—e.g., a need for nourishment—that regulates my dynamic interaction with my environment. The demand-character of the mailbox, by contrast, is not regulated by any vital need: the phenomenal form of the mailboxes is enacted not by a biological condition, but by a social one.

This picture of consciousness suggests a final consequence: if we accept the Gestaltists’ account of dynamic behaviors, we must also account for the circular relationship of the physical world to the sphere of ‘the mental’. That is, we must secure, first, how something immaterial (an intention) could impact something material (the movements of my body) and second, how non-egoic immaterial impacts (the phenomenal mailbox, the phenomenal apple) could affect an organism’s real action.[19] The demand for a truly integrative psychology then requires that this resolution avoid consigning mind and nature, or mind and world, to “separate realms of existence” (Koffka 1935: 47). That would only reintroduce the arbitrary dividing lines that are the target of critique. Gestalt theory finds its solution in the conceptual apparatus of physics. The key, Gestaltists argue, is to grasp that geographical and behavioral reality “are both manifestations of the physical world” (HSP 352). Pushed to its philosophical conclusion, Gestalt theory therefore comes down on the side of physical realism.

With the Gestaltists’ transformation of consciousness we now come up against the key dilemma: Gestalt psychology seeks both to be integrative and to meet the purported demands of objective science. Consciousness, it concludes, must in some way reduce to physical forms. Merleau-Ponty will not endorse this conclusion, and as I will now argue, his critique lays the groundwork for a new conception of nature. Since this new conception of nature is neither the law-governed ideal of the classical picture, nor material reality exclusive of value, but is partway between these, that conception establishes the elemental understanding of flesh.

4. The Negative Move: Merleau-Ponty’s Objections to Gestalt Ontology

Let me pause to retrace Merleau-Ponty’s positions in our trajectory so far. First, we have seen that the subjective-objective nature of direct experience as proposed in Gestalt theory is essential to the scientific understanding of behavior. Merleau-Ponty agrees, lamenting over and over the “false conception of scientific objectivity” that stymied behaviorist psychology (HSP 345) by wrongly excluding “consciousness” (HSP 345). We have just seen, however, that the task of a truly integrative psychology, if it is pursued within the Gestaltist framework, requires an account of how the geographical and behavioral fields (within which conscious processes unfold) are related to one another—an account that must lie, as Koffka puts it, within “one unitary universe of discourse” (1935: 52). Koffka is explicit about the identity of that discourse: the universe “in which all events can take place . . . must be the one about which physics teaches us” (Koffka 1935: 48). This brings us to the core of Merleau-Ponty’s objection: that “the integration of matter, life and mind are obtained [in Gestalt theory] by their reduction to the common denominator of physical forms” (SB 135/145–6). The realist conclusions toward which Gestaltism is pushed therefore walk back the major gains of its critique of behaviorism.

To show this, I want to focus on three things. First, in developing their conception of consciousness, Gestaltists propose a theory of structural isomorphism that reduces phenomenality to processes in the brain. This is, for Koffka, a positive result. If all complex conscious processes reduce to physical causes, psychology can be realigned with the epistemological and ontological frameworks of the natural sciences. But Merleau-Ponty argues that this move is contradictory. Thus, second, Gestalt ontology must be fundamentally rethought: the right account of nature, Merleau-Ponty argues, must include an account of the origin of the idea of nature. This does not lead Merleau-Ponty to endorse an idealist alternative. This is my third claim. While the apparent tendency toward idealism in The Structure of Behavior is of central concern for those scholars who identify a ‘turn’ in his thinking—Barbaras notes such a tendency explicitly (2004: 220), and Baldwin highlights what he calls “a sub-Kantian idealism” in Merleau-Ponty’s discussions of structure in physics (2013: 192)—Merleau-Ponty’s positive view makes sense only as the rejection of both the Gestaltist and idealist alternatives. It is this move that establishes the logic of the chiasm. What Merleau-Ponty later calls flesh emerges as the dialectical “truth” of the antinomy of realist and the idealist ontologies of nature.

4.1. The Form of Reality

Let me start with Gestalt theory’s ontology, since this is the principal target of the criticism. I said that Gestalt theory endorses the ontology of the physicist, but the first point we must note is that this does not represent a return to atomism. As Koffka tells us, the physicist’s universe (pace the behaviorist) is clearly not composed of determinate parts or substances. For example, it is “meaningless” to the physicist to say that “‘The world consists of protons and electrons’, since the reduction of a molecule to atomic or sub-atomic parts is destructive of the reality of the molecule” (1935: 57, translation modified). Instead, what “assume the chief reality” for the physicist are “organized fields of forces” (1935: 57). By appealing to the “field” concept for psychology, Koffka means that “real behavior” is “determined” by a general “system of stresses and strains” (1935: 42)—that is, organized fields—that governs the coming-to-existence and the relative appearance of forms. This appeal unifies psychology within one “universe of discourse” since Koffka holds that the same system of forces shapes both physical forms and the molar behaviors that are responsive to them.

A common way to illustrate this point is through the example of a soap bubble. Consider first the physical form of the bubble. As a natural phenomenon in geographical reality, the bubble results from an organized distribution of soap particles “held in equilibrium” by the forces of cohesion and energy conservation (1935: 14). The resulting form is irreducible, Koffka then reasons, since the pattern of distributed particles (the order) and the forces that distribute and hold those particles in place are as much aspects of the physical form as are the particles themselves; it is not just the particles, in other words, but the particles in that patterned distribution that give the bubble its shape. This discovery informs how the field of forces is expressible in a law. Koffka notes, for example, that the law that “the soap will distribute itself on a spherical surface” because “of all solids, the sphere is the one which has the greatest volume for a given surface or the smallest surface for a given volume” is both quantitative—it “says of each particle that it is here and not somewhere else”—and qualitative—“it assigns [to the distribution] a definite shape” (1935: 14). The laws of physics that ‘naturally’ organize the bubble thus avoid the quantitative/qualitative antinomy.

Now imagine that the physical bubble is an object of interest. Imagine, for example, that I notice it out of the corner of my eye and I feel my attention drawn toward it. This recalls the experience that opened The Structure of Behavior (the spot of light that “pulls along” my gaze). Here, with the introduction of conscious experience, we are in a behavioral setting. Yet, for Koffka, my attention must be understood as drawn to the light by a force of attraction. That is, Gestalt theory reasons that forces shape molar behaviors just as they affect the distribution of particles in forming a bubble. When I am drawn by my attraction to engage in the act of popping-a-bubble, I thus become one holistically integrated element of a behavioral structure that—like a physical structure—is governed by a field of forces and is irreducible to its parts. The conscious behavior of popping-a-bubble now looks to have a general affinity with the properties of physical form.

A mere “affinity” would, of course, be insufficient. What Gestalt theory needs is a unified way to account for the relation of physical (geographical) reality and conscious behavior. Koffka’s “bold assumption” resolves the dilemma (1935: 62). Inferring a “thoroughgoing” psycho-physical isomorphism from the general affinity (1935: 63),[20] he argues that physical structures in geographical reality cause physical reactions in a geographical organism through the mediation of the nervous system: ‘Real’ forms, in other words, produce identically patterned nervous reactions, and those nervous reactions then constitute consciousness. This is what he means when he insists that matter, life, and mind intertwine in the nervous system. Through the nervous system’s psycho-physical mediation, the conscious experience of a soap bubble can be linked to a physical structure in a geographical setting (to the real, geographical soap bubble, as an organized whole). And since, on Köhler’s gloss, the “‘motions of the atoms and molecules of the brain’ are not ‘fundamentally different from thoughts and feelings’ but in their molar aspects, considered as processes in extension, identical” (Koffka 1935: 62),[21] a more properly ‘cognitive’ experience—my thought of a soap bubble—can reduce to structural processes in my geographical organism. Positing a field of physical forms “behind” naïve experience, Gestalt theory thus takes itself to accomplish what Köhler calls the “enormous feat . . . already achieved by the physicist”: to “jump from the world of direct, but confused, experience into a world of clear and hard reality” (1947: 5).

I noted earlier that consciousness, for Gestaltism, is rethought as an orientation to the richest forms. We now see how this is complemented with a physicalist ontology. Consciousness, for Gestalt theory, may not readily admit of a formula, but it always reflects a quantifiable process in the physical brain. That this walks back Gestalt theory’s advances over behaviorism should be clear in three respects.

First, Gestaltism ‘solves’ the value-body problem while leaving the classical picture of nature intact: “physical nature,” it argues, is the ‘real’, causally governed substrate to which all ‘conscious’, ‘perceptual’, or ‘mental’ structures reduce. Consciousness is modified, but by being recast as a manifestation of physical nature. For Merleau-Ponty, this re-inverts the primacy of meaning. Whereas Gestalt theory initiated its critique with an appeal to the idea that direct experience can be objective just as we experience it, the view it ends in endorsing has little to do with experience at all. This is the thrust of Merleau-Ponty’s observation, in the final chapter of The Structure of Behavior, that, unlike the explanation that is ultimately endorsed by Gestalt theory,

if I adhere to what immediate consciousness tells me, the desk which I see in front of me and on which I am writing, the room in which I am and whose walls enclose me beyond the sensible field, the garden, the street, the city and, finally, the whole of my spatial horizon do not appear to me to be causes of the perception which I have of them, causes which would impress their mark on me and produce an image of themselves by a transitive action. It seems to me rather than my perception is like a beam of light which reveals the objects there where they are and manifests their presence, latent until then. (SB 185/200)

To put this another way, Merleau-Ponty objects to the fact that Gestalt theory’s purported return to “lived experience”—a return whose necessity he initially praised Köhler for recognizing—is abandoned at the last moment in favor of a renewed ideal of “objective science.”

Second, consciousness for objective psychology is not consciousness as lived. “Experience” thus seems to be of merely strategic value. Köhler himself underscores this point when, marshaling the idea of “subjective-objectivity” against the behaviorists, he nevertheless notes that “the chair of objective experience [that is, the chair as I ‘subjectively-objectively’ encounter it] cannot be identified with the chair as part of the physicists’ world” (1947: 24). Gestalt psychology therefore looks to endorse a Critical or “transcendental realism.”[22] Since the ideal object of science (the real world subtending its various appearances) is necessarily beyond the reach of any observational method, I must learn to use the “experienced world” as a kind of placeholder from which “the real” can be inferred.

Merleau-Ponty’s objections here are not borne of ontological prejudice. He argues, finally, that the inference from phenomenal to physical reality won’t work if we take seriously the holistic unity of form. Since whatever distinguishes lived experience or conscious thought from their underlying physical processes would have to be a holistically integrated aspect of their structure, the distinction between a physical and a mental form should, according to the Gestaltists, be transformative; conscious processes could not be identical with physical forms as “thoroughgoing isomorphism” proposes (SB 133/143). Thus, it should not be possible to deduce a picture of physical nature from appearances to the mind. We are either led back to the crude materialism that isomorphism was introduced to avoid, or we are forced to pose some force or power to explain the structural differences that separate matter, life, and mind. This directly contradicts the demand of an integrative psychology. To borrow a metaphor, the physical ontology of form seems to catch us in an interminable oscillation between the mentalism and materialism to which the philosophy of form should offer an alternative.

The point is not to abandon Gestalt psychology. In fact, Merleau-Ponty’s objections to isomorphism are already anticipated in the Gestaltists’ appeal to the “richness” of forms—to see that the mark of mental forms is that these are least translatable into quantifiable terms is already to signal the dilemma. Gestalt psychologists simply had not grasped the implications of their discoveries. The strategy deployed against behaviorism is now deployed against Gestalt theory. Just as behaviorism believed it could get ‘behind’ the direct experience on which it already relied, Gestalt theory believed it could infer the true form of reality on the basis of (apparent) phenomena. What they missed is that, according to their own concept of form, phenomenality is reality. “Nature,” Merleau-Ponty will argue, cannot (pace the Gestalt theorist) name a mind-independent transcendent reality of physical forms bound together by relations of causality. Rather, “nature” must be the expression of a general structural field that is as “subjective-objective” as experience. Nature will be objective, insofar as the existence of structures remains independent of the organism, yet it is also subjective since those structures’ order and ideal significance will depend upon dynamic interactions with the organism. This structural picture of nature, for Merleau-Ponty, is both elemental and carnal in a sense akin to flesh.

We are now at the threshold of Merleau-Ponty’s positive point: on the one hand, he has told us, Gestalt theory rightly recognized a prejudice in psychology. Whereas science is always conducted from a necessarily human perspective, this dependence on the subject does not threaten its status as ‘objective’. This point was central to Gestaltists’ objections to behaviorism. On the other hand, however, Gestalt theory has not gone so far as to question behaviorism’s classical ontology. That is, they were unwilling to extend the subjective-objective status they identified in the domain of experience to the domain of nature. They therefore treat the human perspective as merely “a point of departure, a προτέραν πρὸς ἡμᾶς” (SB 145/157). For this reason, “Gestalt theory does not recognize that psychological atomism is but a particular case of a more general prejudice: the unquestioned belief in determinate being and in the world” (PP 510 Endnote 60/62 Footnote 1). We are in need of a new ontology.

I have gestured at the shape of that ontology. Below, I’ll show that The Structure of Behavior foreshadows the “elemental” status of flesh by situating nature as the ground from which forms and “ideal significations” emerge (SB 221/237–8). This conception of nature is essentially linked to carnality, moreover, since the body, which Merleau-Ponty conceives as the paradigm case of structure, is the arena in which an “ideal’ form is realized. Already in The Structure of Behavior, naturalized consciousness and phenomenologized nature intertwine at the site of the body. Before I do this, however, it is necessary to say just what is at stake in a critical distinction Merleau-Ponty draws between “actual structure” and “ideal form.” To that end, I want to clarify how the contradictions of Gestalt theory, which undercut their physicalist ontology, do not lead Merleau-Ponty to authorize a critical idealism. I’ll approach this through a brief reconstruction of Merleau-Ponty’s discussion of structure in physics, since this, by my lights, is the point at which his divergence from Gestaltist commitments first comes into focus. I’ll return to the positive argument in Section 5, below.

4.2. Realism, Idealism, and Structure in Physics: A Transition to the Positive Ontology

In the previous section, I showed that the universe of ultimate explanation within which conscious behavior, for Gestalt psychology, is reductively explicable must be conceived as a real, structured field of physical forms whose appearance is governed by forces expressible as laws. On Merleau-Ponty’s reconstruction, for example, the force of gravity will explain why the geographical field is not wholly composed of freely floating forms and we can discover that force in the tendency of unsupported bodies to fall. However, he also notes a point whose importance was not fully appreciated in Gestalt theory: that “gravity” remains the same ‘law’ (“gravitational force” has the same pattern of formal expressions) only against a particular cosmological background (SB 138/149). Any change in the relative positions of celestial bodies or in the speed of the rotation of the earth will effect a change in the motions of bodies in that system. Whereas Gestalt theory treats ‘law’ as explicative of the structure of the geographical environment (that is, of the organization of our field), we can now see that the law also presupposes structure. It is the law that it is only within a particular structure of the cosmos; the law is an irreducible part of a system of laws.

To be clear, Gestaltists recognized this dilemma. They ‘avoid’ it, however, by re-inscribing the system of laws in nature itself. That is, Gestalt theorists noted the necessary inclusion of law in structure and so reified the structure inclusive of laws, positing what Merleau-Ponty, following Cournot, calls a “cosmological given”[23]—a stable configuration of forces that functions as the universal milieu within which particular law-governed physical systems emerge. What we call “nature,” as Merleau-Ponty puts it, is thus the general “infrastructure of the physical world” (SB 141/152). According to Gestaltism’s ontology, however, this cannot be a stopgap. Recall that, for Gestalt theory, any holistic system must reduce to physical form. And so, if the system of physical laws (the infrastructure) is itself a physical form—if, as Brunschvicg puts it, “the law is an entity”[24]—some deeper “organized field of forces” would have to stabilize the form of the system of law, and so on, all the way down. Even a “cosmological” given, in other words, would have to receive its unity from dynamic interaction with a broader milieu. The inclusion of law in structure in physics therefore means that physical form cannot itself be “ultimate reality.” Far from being the universe of discourse within which all others could be, reductively, contained, it now appears that “the universe of naturalism has not been able to become self-enclosed” (SB 145/157). The explanatory framework of Gestalt theory comes undone.[25]

So, in what ultimate universe is the universe of naturalism enclosed? An alternative, we can now see, is to endorse a turn to idealism. Perhaps the milieu in which the total system of physical laws is situated is the milieu of absolute consciousness. In this case the “ultimate universe” is not a physical order but an order of ideal meanings. “Form” would therefore belong to “a universe of thought and not . . . a universe of realities” (SB 242 Endnote 35/154 Footnote). To go this route is to take up a Kantian mantle. It is to propose, for example, that the ideal of law-governed nature regulates the appearance of forms.[26] Indeed, this seems to make better sense of the problem. For if the laws of the physical universe make sense only against the background of a certain unfolding of cosmological history, then those laws seem to presuppose ‘ideas’ (ideal time, ideal space, “an infrastructure of eternal laws,” SB 242 Endnote 35/154 Footnote) of which particular forms are partial expressions. Whatever nature is, on this view, it would therefore depend on intentional consciousness. “What we call nature,” Merleau-Ponty now tells us, “is always consciousness of nature” (SB 184/199).

Now, Merleau-Ponty sometimes seems to endorse this alternative. Reflecting on his critical appraisal of Gestalt theory, for example, he tells us approvingly that “We have been moved from the idea of a nature as omnitudo realitatis to the idea of objects which could not be conceived in-themselves (en soi), partes extra partes, and which are defined only by an idea in which they participate” (SB 202/218). In the same spirit, he revises the Gestaltists’ line of progression from matter to life and mind to reconceive these not as “orders of reality” but as planes of signification.[27] In doing this, he underscores how a science conducted, as it must be, “from a human situation” (HSP 345) necessarily appeals to ideal structures or laws. Yet he cautions against mistaking this for a complete solution. He is clear, for example, that ideal forms are not ‘thinkable’ except in terms of their concrete expressions. Moreover, the holism of form means that each expression will determine the idea differently and only partially.[28] Whereas a regulative idea of nature as law-governed may inform and guide the study of natural objects, the instantiation of ‘law’ and ‘nature’ in those objects necessarily transforms them. To posit “absolute consciousness” in the place of the “cosmological given” therefore looks to compound the Gestaltists’ error. Merleau-Ponty puts this point more directly in the Nature course: “Idealism,” he writes there, “is only another form of objectivism. . . . It objectifies human representations” (N 96).

So, why does Merleau-Ponty introduce the idealist alternative, if he does not end by endorsing it? The answer is twofold. First, he means to show that even the physicists’ most putatively reductive explanations appeal to ideal forms. The Gestaltists’ position that, as Merleau-Ponty puts it, “structures can be found in a nature taken in-itself [en soi] and that mind can be constituted from them” (SB 141/152) depends upon an idea of Nature in-itself that is a product of that mind. The ideal cannot play the role of ultimate explanation. The appearance of any physical system, which emerges from the ideal as a “partial totality” (SB 139/149) will, qua physical and partial, holistically transform the idea. The “reciprocal inclusion” (SB 141/152) of physical structures and ideal laws thus undermines the antinomy of realism and idealism.

It is here, however, that we locate a positive suggestion. Merleau-Ponty notes that, even if we were to situate the system of physical law against the background of a cosmological or an ideal ‘given,’ each would require an appeal to the figure/ground structure—a function, Merleau-Ponty tells us, that “has a meaning only in the perceived world,” where “we learn what it is to be a figure and what it is to be a ground” (SB 92/101–2). Whether we take the appearance of physical forms to depend upon a physical nature in-itself or an idea in which they participate, the appeal to their ‘appearance’ is therefore illuminating: “Form,” Merleau-Ponty reasons, “is not a physical reality” (SB 143/155) as it is for Gestalt theory, not because forms depend upon a conceiving consciousness, but because “form” is a concept that is borrowed “from the universe of perceived things” and is “encountered in physics only to the extent that it refers us back to the perceived” (SB 144/156). Whereas the idealist gesture resolves the problem opened up by the internal contradictions of reductive naturalism (and so Merleau-Ponty provisionally embraces it), it now appears as if the framework that defines objects “by the idea in which [they] participate” must be compensated with a reflection on the natural development of those ideas. In just this way “the problem posed by physics approaches the problem of perception” (N 97).[29]

Notice we’ve returned to the question with which Merleau-Ponty began: what is the relation of consciousness to nature? It is evident now that Gestalt psychology’s association of consciousness with the nervous system—its reduction of consciousness to nature—is not the reply. This failure leads Merleau-Ponty to the problem of perception. The task of retrieving an objective picture of nature through the subjective-objective mechanism of human perception requires him to determine the relationship, both of nature to perception and of perceptual to intellectual consciousness. The question with which Merleau-Ponty initially confronted himself is thus fractured in two. He approaches this double task by way of the body. The Gestaltist critique of behaviorism had erred, Merleau-Ponty reasons, not in its emphasis on conscious, bodily behavior, but in positing the dependence of behavior on a purely physical ontology that is mediated by the nervous system. For all its intolerable conclusions, in other words, Gestalt theory had rightly shown that the body is the means for active consciousness to make contact with the world. In taking embodiment as his point of departure, Merleau-Ponty thus shows that “the situated and incarnated aspect of the physicist,” as he later puts it, “must succeed the universal ‘I’ of transcendental philosophy” (N 97).

4.3. From the Body to an Elemental Ontology of Nature

In moving beyond the realist ontology of Gestalt theory, Merleau-Ponty takes embodied consciousness as his point of departure. We can now see how this strategy adapts that of Gestaltism: Merleau-Ponty accepts Koffka’s claim that psychology necessarily marks the point at which “the three great provinces of our world intersect”—Koffka is speaking of “inanimate nature, life, and mind” (1935: 10). Yet he locates this point of intersection in the body rather than the nervous system. The body, Merleau-Ponty tells us, fulfills this function because it is the “acquired dialectical soil upon which a higher ‘formation’ [the idea of ‘mind’] is accomplished” (SB 210/227). The language of “acquired soil” is meant to demonstrate that the lived body is a unity that comprises both “the natural body” understood as what is “always already there” and the “cultural body as the sedimentation of its spontaneous acts” (SB 249 Endnote 50/227 Footnote 1) (the former aspect is thrust to the fore in cases of pathological breakdown, but these orders are for the most part indiscernible[30]). To articulate this conception of the lived body at this point in the text is therefore to say that it is from this dialectical soil that we must—in a kind of archaeological excavation—interrogate perceptual consciousness and, through it, the consciousness of nature.[31]

If I’m right on this point, our lived, bodily experience leads to a picture of both ‘consciousness’ and ‘nature’ that neither realism nor Critical idealism could accommodate. That is, each rejected framework presupposes a transcendental ‘reality’ that is present to us thanks to its coordination with some psychological function. But lived perception does not—in fact, cannot—distinguish these. The distinction of ‘form’ from ‘content’ comes “after the fact.”[32] Before I show how lived perception is, by consequence, both carnal and reversible, I want to say, very briefly, how this imbrication of form and content in The Structure of Behavior leads Merleau-Ponty to an elemental picture of nature.

Merleau-Ponty introduces the ‘late’ distinction of form and content through an example from child psychology. Research in child development suggests that infantile perception is directed first toward facial expressions and gestures rather than toward objects in the environment. If, following Gestalt theory, we can no longer make recourse to the false hypotheses of atomistic psychology in order to explain the ‘contents’ of this perception, it follows, he concludes, “that it is possible to perceive a smile, or even a sentiment in this smile, without the colors and the lines which ‘compose’ the face . . . being present to consciousness” (SB 166/181). A face, in other words, is not a ‘thing’, interpreted as smiling by a perceiver who joins its content beneath the form of a happy expression. It is grasped “as [an] experienced reality, rather than [as a] true object” (SB 166/180). Gestalt theory was already positioned to say this; they rightly identified the holism of perceptual meaning and the bodily activity of ‘consciousness’. Yet they lacked the framework for these discoveries to be worked out. Specifically, since they were locked within the classical picture of nature, their theory offered no conceptual resources to explain how ‘consciousness’ is passively conditioned to structures of sense. Merleau-Ponty makes this point explicitly in a footnote:

Psychology establishes with certitude the chronological and transcendental priority of the perception of others over the perception of objects in the sense in which the natural sciences understand them. But because it also takes the word Nature in the sense of the sciences of nature, it is not in a position to apperceive that primordial Nature, that pre-objective sensible field in which the behavior of other persons appears, which is prior . . . to the perception of other persons just as it is prior to the Nature of the sciences, and which transcendental reflection could discover. (SB 245 Endnote 82/180 Footnote, emphasis added)

Psychology, in other words, cannot see that the classical picture of nature as comprising “external events borne together by relations of causality” is a regulative ideal that must first be generated out of the sensible field before it can coordinate our meaningful perceptions within it. Merleau-Ponty thus reasons that the “primordial Nature” that could be materially present to us, that could give rise to idealizable perceptual forms and so could count as the origin of the idea, must be a fusion or inextricable intertwining of these. The classical picture of nature is not false. But it is an “ideal signification” that would have to be derived from a more basic ontological structure and, qua ideal, must depend upon embodied human perspectives. To say that Merleau-Ponty directs no critical gesture at the classical picture is misleading; it is in the spirit of critique that he turns his attention to the body.

It is important to add here that Merleau-Ponty does not confine his discussion to the footnote. While he does not use the term “primordial Nature” again, he clearly identifies what satisfies its double demand: the ontology of primordial Nature is an ontology of “actual structure.” By “actual structure,” he underlines, we mean “the joining of an idea and an existence which are indiscernible, the contingent arrangement by which materials begin to have meaning in our presence, intelligibility in its nascent state” (SB 206/223). In this way, the “actual structure” of primordial Nature shares in the elemental status of flesh. In The Visible and the Invisible, for example, Merleau-Ponty will argue that the ultimate reality within which behavior is explicable could not be a reality of “matter in the sense of corpuscules of being which would add up or continue on one another to form beings. Nor [could it be] . . . mind [or] substance, . . . [but must be] a general thing midway between spatio-temporal individual and the idea” (VI 139/181–2). That is, he will suggest that the universe of “ultimate explanation” sought by Gestalt theory can be neither material nor absolute consciousness, but must instead be a structure that could give rise to each (VI 250/298)—it must be flesh as a “pregnancy of possibles.” Since Merleau-Ponty concludes The Structure of Behavior by arguing that “actual structure” is “the philosophical truth of naturalism” (SB 224/241)—a truth dialectically generated from the failures of realism and idealism—we might reason that a philosophical naturalism is an elemental ontology of flesh.

Of course, The Structure of Behavior is not chiefly seeking out an ontology. What Merleau-Ponty is after in that text is a philosophically responsible integrative psychology. Thus, he tells us in the footnote that “a more complete explication [of primordial Nature] must be reserved for another work” (SB 245 Endnote 82/180 Footnote). Nevertheless, he has made two discoveries of value to the later pursuit: on the one hand, matter, life, and mind are unified in the acting body. (This point nods approvingly at a central tenet of Gestalt theory.) The “acquired soil” of the body—not an idea but also not ‘mere’ matter—then functions as our instrument for comprehending the world. On the other hand, the picture of the active body must be supplemented by a structural concept of nature. On Merleau-Ponty’s picture, material beings, living organisms, and conscious animals are integrated manifestations of the same general principle, the same general structure, which subtends and makes possible the organizations of the field within which material, living, and conscious beings are found. If it is true that behavior does not take place ‘in’ the body, then, but ‘in the world’ of which the body becomes a part—an insight again taken over from Gestalt theory—it’s not enough to say, with Gestaltists, that behavior has a structure. Behavior, Merleau-Ponty insists, is a structure: the body, in other words, is a “prototype” of the being that is also the being of the world. I take it that this is what Merleau-Ponty means when, in The Visible and the Invisible, he tells us that “it is through the flesh of the world that one can understand the lived body” (VI 250/299), that is, that “carnal being . . . is a prototype of being, of which our body . . . is a variant” (VI 136/177).

5. The Positive View: Carnal Structure and the Reversibility of Bodily Consciousness

In Section 2, I introduced three senses of flesh. Emphasizing a carnal sense, I then said that Merleau-Ponty broadens the phenomenological account of the lived body defended in Phenomenology of Perception so that embodiment is understood as an imbrication of self and world: it’s not enough to speak of the body as an “instrument” of my understanding, this must be compensated with a picture of the world as made of “the same stuff.” I just suggested that we find the seeds of this point in The Structure of Behavior. Chiefly, however, the argument has been negative. We have seen Merleau-Ponty uncover the need for a primordial Nature of “actual structure” only by pitting Gestalt theory’s physicalism against its idealist alternative.

We are now positioned to see the constructive move: the failures of Gestalt ontology require an ontology of flesh. I’ll reconstruct the positive argument, in this last section, in two steps. First, returning to Merleau-Ponty’s close engagements with experimental psychology I’ll show where these inform the idea of bodily consciousness as structure. I’ll then show how these findings impact Merleau-Ponty’s discussions of nature. Here, my point is that Merleau-Ponty’s empirical engagements help bind the carnal and reversible status of human perception to a critical revision of ‘classical’ nature. The chiasmic logic to which he appeals therefore reflects a “relational ontology.”

5.1. Embodiment and Animal Consciousness

The point I am urging is clearest in Merleau-Ponty’s engagements with Köhler, whose Mentality of Apes provides much of the material for the account of “vital structures” in Chapter 2 of The Structure of Behavior. In this discussion, Merleau-Ponty’s treatment of a pair of peculiar experimental findings emphasizes a role for the body that anticipates the reversibility thesis that shapes Merleau-Ponty’s later thought. Since this emphasis on the body directly contradicts Köhler’s interpretations of his own findings, the difference between Merleau-Ponty’s position and Gestalt ontology comes out quite sharply.

Briefly, The Mentality of Apes summarizes a series of experiments designed to determine what “degree of relationship” exists between human beings and apes (1925: 1). In each case, Köhler’s animal subjects are prevented from reaching a goal such as grasping a banana. Within the experimental parameters, they must therefore discover some way to achieve their goal either by means of planned detours around concrete obstacles or by producing instruments from materials within their environment. Since each of these calls for the involvement of capacities humans employ without effort, yet which Köhler’s subjects struggle to acquire, he interprets his experiments as revealing something about both human and animal intelligence. In particular, he concludes that a chimpanzee’s behavior is motivated by “his optical apprehension of the situation,” the structure of which is often “too much for his visual grasp” (1925: 267). Human beings do not suffer from the same “weakness of form perception” as do other primates and so can make use of problem-solving strategies that do not rely on what is perceived (1925: 267).[33]

Let me bring Merleau-Ponty’s argument out in an example: in the first experimental scenario, a banana is suspended from the ceiling. Köhler arranges this setting so the apes could access the banana by stacking available boxes together as a ladder. In practice, however, they are only able to place the first box without incident. Whereas a large box and the level ground, brought into contact, achieve a stable structure “by themselves,” the ape presented with the second box “meets a static problem” that requires his involvement to resolve (1925: 148). It appears that he cannot analogize the strategy he used with the first box, because the stability of the structure now depends upon the relative placement of the second box on top of it. Each of Köhler’s subjects, we see, ends up waving the second box haphazardly over the surface of the first one, dropping it suddenly, or abandoning it entirely, but none can immediately build the ladder. Confronted with “a limited body of a special shape [that] is to be brought into contact with a similar one in such a way that a particular result is [to be] obtained,” Köhler tells us that “the chimpanzee seems to reach the limit of his capacity” (1925: 148).

Köhler links this, as I said, to a visual deficiency. He notes, first, that the task of building a stable structure requires a kind of ‘folk physics,’ or a sensitivity to the pull of gravity, on the one hand, and to “visual forms in space,” on the other. He then reasons that “at least one of the components must be in a very undeveloped state in the chimpanzee” (1925: 149). Since the apes have the tendency, when confronted with “static problems,” to grope around “blindly,” he concludes that the deficiency is visual (1925: 149). Merleau-Ponty disagrees. He points out that whenever Köhler’s chimps are perched atop a precarious structure that has already been built, they can skillfully adjust their posture to counterbalance the wobbling. While Köhler called attention to the “most masterly fashion” in which they do this, suggesting that, “in this respect, the chimpanzee . . . is superior to man” (1925: 152), this puts the point too strongly for Merleau-Ponty. No special ‘intelligence’ is required to keep the chimp’s body stable, since the “scaffolding” necessary for “balancing the body proper” (SB 119/129) is already there in his own body. That the animal is only able to balance the ladder through “the struggle for not wobbling” (Köhler 1925: 151, as cited in SB 116/126), however, and not by constructing the ladder to be stable, suggests that the animal cannot analogize this bodily experience in order to locate the same scaffolding in the object. His body has a kind of privileged stability, in other words, and he cannot relinquish this privilege. If this is right, his deficiency is bodily, not visual. Merleau-Ponty therefore signals two things: through Köhler’s experiments, we discover (1) that there is a kind of bodily pre-condition for so-called ‘higher’-order reasoning; and (2) ‘higher’ reasoning remains linked to those bodily conditions.

For my purposes, the explanation is not complete. The second experiment secures the point. In this scenario, instead of suspending the banana from the ceiling, Köhler separates it from the animal by the barrier of the cage. He then notes that an ape permitted to exit the cage will happily go out of his way to access the banana. However, if the task requires him to move the banana (if he must push it with a stick, for example, toward the open end of his cage), he only seems to succeed by chance. The animal will thus vary his ways of approaching an objective, but he will not start by moving a goal into his projected path. On Merleau-Ponty’s gloss, the goal consistently remains a “fixed point and the organism the mobile point[;] they cannot exchange their functions” (SB 117/127). Again Merleau-Ponty’s explanation appeals to the body. He reasons that in order to back-chain from a goal, one must be able to “put oneself in the place of the moveable thing” (SB 118/128); one must be able to move as if one were the mobile goal. But to do this, he continues, one must first recognize that an external object is “a concrete unity capable of entering into a multiplicity of relations without losing itself” (SB 118/128)—a concrete unity like the carnal unity of the body. It is this capacity Köhler’s apes were shown to lack. Thus, just as they were unable to analogize from their own unity to the unity of the object, they could not see their goal as a unity into whose place they might be transposed.

We can put Merleau-Ponty’s crucial insight like this: a felt sense of bodily stability clearly structures the ape’s experience (they evidence this when balancing on the unstable boxes). But what is missing from their experience—to borrow from the language of The Visible and the Invisible—is a sense of “this perceived that we call my body applying itself to the rest of the perceived” (VI 250/299). More concretely, the apes lack the capacity to discover a carnal unity outside themselves—a capacity integral to the task of ‘transposing’ our bodies into the position of external objects so as to ‘see’ ourselves from their position (SB 117/127). That they are incapable of this perceptual reciprocity then explains why an ape who has learned to stand on a box to reach a banana will not borrow one another ape is sitting on: if the ‘thing’ (the box) does not have a fleshly unity like the unity of his body, it cannot be perceived as the bearer of properties at all—it cannot be perceived, in other words, as a body. A merely contingent and ephemeral vector of value, the box “remains for [the ape] a means of support or rest [and so] cannot become an instrument” (SB 114/124).

A crucial Romantic thought is lurking beneath the surface here—a point Merleau-Ponty develops in Phenomenology of Perception. Central to human perception is the capacity to grasp the world, not just as meaningful for me but as meaningful like me.[34] This is not the pragmatic, “functional value” of Gestalt theory. I am able to imaginatively transpose my body into the place of the external object not only because I perceive it as a source of value, but because I see it as another “pole of activity” (PP 117/146). It is only in recognizing the carnal co-belonging of myself and the world (hence recognizing the persistence of unities like the unity of my body) that I become, qua body, an instrument of understanding.[35] This is precisely how a “pure consciousness of self” (SB 224/241) can emerge from a lived, perceived reality. In touching and being touched, in looking and being looked at, perceptual experience prefigures the reflexive act of thought turned back on itself. Phenomenality therefore becomes, not simply the appearance of objective, physical reality, but “the always presupposed foundation of all rationality, all value and all existence” (P 13).

There are two consequences I would like to disentangle here. (1) Through Merleau-Ponty’s engagements with Köhler, we find carnal unity in the natural percept in addition to the perceiver. A resonance to that unity is a trait of human perception. I’m able to forge and take up instruments to alter and appropriate my environment, for example, because my percepts signal unities that persist beneath a change in aspect—that is, unities like me. Human perception is thus linked to capacities to extract ideal significations (a box that becomes an instrument without ceasing to be a means-of-support) from actual structure. (2) This capacity establishes the reciprocal relation of body and world. Since this is a bodily relation, percepts gain their status as “things” neither independently nor as an act of mental construction. We find this point made again in The Visible and the Invisible. The middle path between naturalism and idealism, Merleau-Ponty tells us, requires us to conceive of “things [not as] first identical with themselves”—that is, as objects “which would then offer themselves to the seer”—but to see the world around us as “something to which we could not be closer than by palpating it with our look, things we could not dream of seeing ‘all naked’ because the gaze itself envelops them, clothes them with its own flesh” (VI 131/171). What Merleau-Ponty describes here, by my lights, is a bodily attunement to what Bannon called the “constitutive internal relations” of things (2001: 344).

These consequences are related in the following way: first, Merleau-Ponty’s critical re-reading of Köhler re-situates the human/animal distinction as a difference in scope that becomes a difference in kind. The human resolves the static problem of constructing a ladder by feeling the promised stability of that structure like the scaffolding of her body; she does not have to discover it, as the chimps do, by accident. This becomes a difference in kind, however, since she cannot “get behind” the accomplishment of her body. The living body is itself a holistic structure; the arrival of ‘mind’ in the ‘matter’ of the natural body—an arrival evidenced in the capacity to extract ideal significations from actual structure—is by definition transformative. This warns against appealing to the body as mere instrument: “The mind,” Merleau-Ponty tells us, “does not use the body, but realizes itself through it while at the same time transferring the body outside of physical space” (SB 209/226). In “returning to this structure as the fundamental reality,” he continues, “we are rendering comprehensible both the distinction and the union of the soul and the body” (SB 209/226).

This threatens to trap us in an anthropomorphizing dilemma. The second point gives a way out. What I’m calling the ascription of carnal unity is not an act of consciousness. What Merleau-Ponty shows in his reading of Köhler is that I do not make things ‘things’ in the active sense of subsuming them under an objective category. They ‘become’ things for me because I recognize within them the kind of persistent unity I am. The felt unity of my own lived body is thus linked inextricably to the asymmetric reversibility of perception: the accomplishment of the “thing-structure” requires that we see ourselves in things, thus that we can see the things look back. This speaks to the carnal and reversible pre-history of human consciousness: as Merleau-Ponty writes in a Working Note, “to grasp what makes the leaving of oneself . . . a return to oneself, . . . to grasp this intertwining (chiasma), this reversal; that is the mind” (VI 202/249, translation modified). The Structure of Behavior therefore demonstrates empirically that the body-world relation mimics the body’s own structure of auto-affection without simply being a projection of the structure of perceptual consciousness onto the world. This is a discovery, through the experience of the body, of the general structures of being according to which a mind can “come into the world” (SB 209/225).

It is here that our detour through idealism proves instructive. Recall that Merleau-Ponty accepted, following Köhler, that the objective description of any observable state (whether it is a conscious behavior, a living organism, or a physical object in the natural world) is genetically dependent upon a perceiving organism. The apparent endorsement of idealism is recast in light of that view: whereas ideal significations inform human perceptions, those significations are themselves drawn from “actual structure” (that is, are genetically dependent upon its perception). Merleau-Ponty ends his text by cashing out the significance of this argument for the concept of nature: the human being’s contact with ‘nature’, he argues, is shaped by an ideal form of nature that is never a pure idea; it retains its rootedness in perceptions of actual structure. Köhler’s experiments then bear this out: for a human being to perceive a world of ‘things’—stable unities that share in the stability of the body—she must presume that those things, qua stable, exist when she’s no longer perceiving them.[36] This idea of persistence refers to embodiment and perception. “In affirming that [objects] continue to exist” when I’m not perceiving them, for instance, “I mean that a properly placed, psycho-physical subject would see this or that sensible sight, articulated in this or that way” (SB 212/229). The ‘idea’ of an independent nature has been wrested from a sensible world, on this view, and maintains a constant reference to it, at the same time it structures and regulates human perceptions. The ideal signification of nature—the independent nature of the classical picture—is a logical necessity and not itself a measure of reality.

In defending this view, Merleau-Ponty does not equate the idea with perception. This last point returns us to the logic of the chiasm. “For there to be perception,” he argues, “that is, [for there to be] apprehension of an existence, it is absolutely necessary that the object not be completely given to the look which rests on it, that aspects intended but not possessed in the present perception be kept in reserve” (SB 213/230). The ‘idea’, wrested from actual structure by the structure of the body, thus functions to hold these other possible perceptions “in reserve.” The suggestion that the human order is a “plane of significations” is therefore the idea that perceptual experience includes an aspect of invisibility. The ‘idea’ is the blind spot of human perception. What fulfills the transcendental position cannot therefore be the classical picture of nature: that picture presupposes the idea of which it would have to be the source. Instead, what must occupy the transcendental position is whatever could function as the origin of both idea and appearance. If we take seriously Merleau-Ponty’s gesture at the “pre-objective sensible” structure of “primordial Nature,” we begin to uncover there what Heinämaa calls Merleau-Ponty’s “transcendentalism of sensuous flesh” (1999: 57).

5.2. Back to The Visible and the Invisible

I opened my discussion by showing that Merleau-Ponty confronts a dilemma. Tasked with uncovering the point at which the subject-object dualism emerges, yet without thereby fusing the subject and object in a way that threatens the possibility of perception, he proposes a new ontology of flesh as a “pre-objective” “pregnancy of possibles.” A widely accepted view in Merleau-Ponty scholarship associates this project with the overcoming of a limit: prior to The Visible and the Invisible, Merleau-Ponty is trapped in a philosophy of consciousness that will not allow him to discover its natural origins. The turn to ontology therefore frees him for another attempt. I argued that this narrative misstates the sensitivity with which Merleau-Ponty treats the first emergence of consciousness from nature in The Structure of Behavior; that is, it misses the sense in which the picture he develops there is already “profoundly modified.” What Merleau-Ponty says of the argument of The Visible and the Invisible is also true of his own thinking: “The end of a philosophy is the account of its beginning” (VI 177/229).

If I’m right, Merleau-Ponty’s close but critical relationship to Gestalt theory in the early, explicitly psychological texts helps us navigate the enigmatic concept of flesh that grows out of their demand. What are we to make of the vast, sometimes contradictory literature on the status of flesh if we accept the argument that the concept is anticipated? Let me now note two things.

First, it seems clear that, inasmuch as Merleau-Ponty contributes to an ecological philosophy, this cannot be meant in Abram’s sense of an ethic. If we accept that Merleau-Ponty anchors the reversibility of flesh in the privileged unity of the body, as attested by his examination of Köhler’s primate research, this makes it difficult to see how we could locate in it the grounds for reciprocal responsibility. Toadvine (2009) offers a similar objection when he argues that the normative demands of an environmental ethic would outstrip what is offered in Merleau-Ponty’s ontology. We do not find compensation in the psychology. But while Toadvine rightly draws a connection between the earliest and latest writings on nature, he implies that it is only in the later Nature course that embodiment is explicitly thematized in that context. I argued, by contrast, that the connection between the body and primordial nature (of which our bodies are the prototype) is essential for Merleau-Ponty’s positioning against Gestalt theory in The Structure of Behavior.

Second, my reading highlights a shortcoming of the perceptual conception of flesh. According to that view, the dynamism of the perceived environment is lent to it by that environment’s involvement in activities of the body. When I introduced this “perceptual” account, I promised to return to the question of anthropomorphism: the worry was that an ontology founded in the auto-affective structure of the lived body would entail the dependence of Nature on its appearance to the body. Because The Structure of Behavior emphasizes a developmental ontology, however, the connection of flesh to that text shows us the sense in which the body is a point of contact with nature: the body attests to its belongingness to nature. As an “acquired dialectical soil” that retains the residue of ‘prior’ dialectics (the physical dialectic, the vital dialectic), it is its mediation through the body that entitles Merleau-Ponty to say, in the Nature course, that “perception is not an artificial construction of nature” (N 100). The ideal significations that guide my perceptions get “the index of real existence” from the “valid grounds” of “the concrete articulations of the perceived field” (SB 218/235). Again, this view treads quite closely to Toadvine’s, for whom “nature has its own meaningful configuration to which we are oriented . . . at the level of our bodily engagement with the perceived” and yet for whom “Merleau-Ponty does not simply reduce human existence to a moment of nature’s meaning” (2009: 131). However, the re-situation of this picture in the general context of a critique of empirical psychology helps highlight why the problems of child development and psychopathology so consistently preoccupy Merleau-Ponty. They represent an interest in the dialectical progression from the vital to the human order with attention both to what the dialectic acquires and to what it leaves behind—with attention to what binds us to nature and what keeps us apart. The disintegration of that dialectic in cases of psychopathology disentangles the first of these threads. The method through which Merleau-Ponty later tries to hold the entire framework in view bears the name, in The Visible and the Invisible, of “hyper-dialectic.”

This last point shows why I’ve endorsed Bannon’s language of “relational ontology.” The idea that flesh describes a general logic of “constitutive internal relations” neatly highlights its indebtedness to the language of experimental psychology. What’s more, in pointing to the embodied, reciprocal dimension of human perception, Bannon argues that “flesh is a relation between bodies, the connection between them that isolates each as a separate body and yet holds them all together in one world” (2001: 345)—precisely the relation that Merleau-Ponty describes, at the conclusion of The Structure of Behavior, as “rendering comprehensible both the distinction and the union of the soul and the body” (SB 209/226). Merleau-Ponty himself will make the connection to Gestaltist language in a Working Note. “[T]o be conscious,” he tells us, is “to have a figure on a ground.” It is this figure-ground structure that “introduces a third term between the ‘subject’ and the ‘object’”—that fulfills the need announced in an ontology of flesh. “It is [therefore] that separation [écart] [of the figure/ground distinction] first of all that is the perceptual meaning” (VI 197/247). On this view, flesh is not coextensive with nature, it is the actual structure of nature.

6. Conclusion

In the course of my argument, I brought into contact two purportedly different lines in Merleau-Ponty’s thought: the close engagement with empirical psychology that occupies him from his thèse complémentaire through his lectures at the Sorbonne in the early 1950s, and the ontology of flesh developed in the posthumously published The Visible and the Invisible. A common line identifies a ‘turn’ between these: Merleau-Ponty’s growing interest in the concept of nature while lecturing at the Collège de France thus coincides with a reflection on the genesis of sense that betrays an earlier failure to break free from a philosophy of consciousness. I tried to push back against this view. By reconstructing the precise relationship between Merleau-Ponty and the Gestalt psychologists on whose research he draws, I argued that while “flesh” and “nature” receive no independent consideration until the 1950s, the late ontology does not represent a turn so much as a shift in emphasis. Already in The Structure of Behavior Merleau-Ponty is aware of the need to articulate the pre-objective Nature from which consciousness and its object emerge and he does so in ways that align quite closely with the elemental, carnal, and reversible structure of flesh. My position is that this connection has long been obscured because of the difficulty in distinguishing Merleau-Ponty’s appropriation from his critical adaptation of Gestalt theory. Thus, whenever possible, I made my argument by engaging directly with Merleau-Ponty’s source materials. This enables us to see that, far from ensnaring himself in a philosophy of consciousness, Merleau-Ponty attempts to ground empirical science on a relational ontology of “actual structure” that explains how consciousness and mind emerge from nature. This attempt to “see behind the back of the [scientist] what the [scientist] himself does not see” (N 86) is what later becomes worked through in the concept of flesh.

I want to conclude by noting two things. First, if The Structure of Behavior foreshadows the argument of The Visible and the Invisible, what are we to make of Merleau-Ponty’s own admission that the problems of his earlier work—he is speaking in particular of Phenomenology of Perception—are insoluble because they depart from the consciousness-object distinction? Much has been made of this admission that I think is illuminating. To offer one example, Morris argues that

if we look from the outside at what Merleau-Ponty is doing in Phenomenology of Perception (as opposed to looking at what he is saying, or what he, in looking back from the inside, thought he was doing), we find a thinker who is trying to install himself in the sway of phenomena in order to return to an encounter prior to the subject-object distinction. (2004: 59)

That is, we do not have to take Merleau-Ponty at his self-critical word. On this reading, we can also detect in Phenomenology of Perception a concern for the pre-objective ground that I argued is already guiding his research by 1935. I am sympathetic, of course, to Morris’s line of argument. But it is helpful on this point to recall the final insights of The Structure of Behavior. As I showed, Merleau-Ponty takes himself to reveal the “truth” of dualism by discovering the “acquired dialectical soil” of the body in which nature and culture interweave and, through a dual consideration of the shortcomings of idealism and scientific realism, to reveal the dialectical truth of naturalism as actual structure. It is through this process that he poses the problem of perception that Phenomenology of Perception interrogates head-on. There’s no inconsistency in suggesting, then, that The Visible and the Invisible returns to the earlier insights in order to deepen them—and in a way he could not do in the framework of Phenomenology of Perception. The position I defend is well complemented by reading Phenomenology of Perception ‘against’ Merleau-Ponty’s later criticisms, as Morris does, but I don’t think it commits us to adopting that approach.

Second, I have been concerned with reading The Structure of Behavior in order to undercut the idea of a turn to ontology, that is, to locate the seeds of the late concept of flesh in the earlier work. For this reason, I chiefly made reference to commentators whose interest touches on the meaning of that concept. I’ve only been able to present this vast literature in outline. It’s worth noting, however, that there is also a strand of scholarship that seeks to bring Merleau-Ponty into conversation with analytic philosophy of mind and with empirical work in embodied cognition; this work also reads Merleau-Ponty’s early writings as distinct in aim and strategy from what follows the “ontological turn.” Among such commentators (see, for example, Dreyfus 2005a; 2005b; Noë 2004),[37] attention to The Visible and the Invisible is scant, if entirely absent. If I’m right, however, one upshot of my argument is that Merleau-Ponty’s otherwise enigmatic Working Notes can be brought into conversation with contemporary philosophies of mind. This is consistent both with Merleau-Ponty’s own aims and with the project of reading him against himself, as Morris proposes. In fact, Morris himself suggests that, in uncovering Merleau-Ponty’s encounters with the pre-objective in Phenomenology of Perception, we should note where the “effort to bed the personal in pre-personal habits and movements”—precisely the kind of effort to which commentators like Dreyfus and Noë appeal—“prolongs The Structure of Behavior’s efforts to conceptualize a discontinuous continuity of the natural, vital and human orders” (Morris 2010: 143 Footnote 11). Morris thus concludes, though he does not develop the point, that the “account of the body subject” in Phenomenology of Perception “already overlaps with an ontology of nature, albeit this is not yet explicit” (2010: 143 Footnote 11). It is in this spirit that I’ve endeavored to make Merleau-Ponty’s position explicit. For while I argued that the image of flesh we discover running, in various expressions, from Merleau-Ponty’s first published texts to his untimely death, does not reduce to an environmental or a perceptual sense, I have shown where it can nevertheless form the basis for an ecological philosophy, a phenomenology of perception, and—if I am right—a general psychology.


Portions of this paper were presented, in earlier versions, at the Society for Phenomenology and Existential Phenomenology in Salt Lake City and at the Southwest Seminar in Continental Philosophy at Northern Arizona University. I would like to thank the audiences of those conferences for their helpful comments. I am especially grateful to Anthony Vincent Fernandez and to Ergo’s anonymous reviewers for their constructive criticisms on an earlier version of this paper. This paper was completed with the generous support of a College of Humanities Faculty Fellowship from California State University, Northridge.


  • EM = Maurice Merleau-Ponty (1994/1964). Eye and Mind. In Galen Johnson (Ed.) and Michael Smith (Trans.), The Merleau-Ponty Aesthetics Reader (35–56). Northwestern University Press / L’Œil et L’Esprit. Gallimard.
  • HSP = Maurice Merleau-Ponty (2010). Human Sciences and Phenomenology: 1950–52. In Child Psychology and Pedagogy: The Sorbonne Lectures (1949–1952) (Talia Welsh, Trans.) (316–372). Northwestern University Press.
  • N = Maurice Merleau-Ponty (2002). Nature: Course Notes from the Collège de France (Robert Vallier, Trans.). Northwestern University Press.
  • P = Maurice Merleau-Ponty (1964). The Primacy of Perception (William Cobb, Trans.). Northwestern University Press.
  • PP = Maurice Merleau-Ponty (2012/1945). Phenomenology of Perception (Donald A. Landes, Trans). Routledge. / Phénomenologie de la Perception. Gallimard.
  • S = Maurice Merleau-Ponty (1964/1960). Signs (Richard C. McCleary, Trans.). Northwestern University Press. / Signes. Gallimard.
  • SNS = Maurice Merleau-Ponty (1964/1966). Sense and Nonsense (Hubert L. Dreyfus and Patricia Allen Dreyfus, Trans.). Northwestern University Press. / Sens et Non-sens. Les Éditions Nagel.
  • SB = Maurice Merleau-Ponty (1963/1942). The Structure of Behavior (Alden Fishcher, Trans.). Duquesne University Press. / Structure du Comportment. Les Presses Universitaires de France.
  • VI = Maurice Merleau-Ponty (1968/1964). The Visible and the Invisible (Alphonso Lingis, Trans.). / Le Visible et L’Invisible. Gallimard.


  • Abram, David (1996). The Spell of the Sensuous. Vintage.
  • Baldwin, Thomas (2013). Merleau-Ponty’s Phenomenological Critique of Natural Science. Royal Institute of Philosophy Supplement, 72(1), 189–219. https://doi.org/10.1017/S1358246113000118
  • Bannan, John F. (1967). The Philosophy of Merleau-Ponty. Harcourt, Brace & World.
  • Bannon, Bryan E. (2001). Flesh and Nature: Understanding Merleau-Ponty’s Relational Ontology. Research in Phenomenology, 41(1), 327–357.
  • Barbaras, Renaud (2001). Merleau-Ponty and Nature. Research in Phenomenology, 21(1), 22–38. https://doi.org/10.1163/15691640160048540
  • Barbaras, Renaud (2002). The Ambiguity of the Flesh. Chiasmi International, 4(1), 19–26. https://doi.org/10.5840/chiasmi200247
  • Barbaras, Renaud (2004). A Phenomenology of Life. In Taylor Carman and Mark B. N. Hansen (Eds.), The Cambridge Companion to Merleau-Ponty (206–230). Cambridge University Press. https://doi.org/10.1017/ccol0521809894.009
  • Brunschvicg, Léon (1922). L’Expérience humaine et la causalité physique. Alcan.
  • Čapek, Jakub (2010). Percevoir c’est Sentir: Merleau-Ponty et Herder. In Karel Novotny, Taylor S. Hammer, Anne Gléonec, and Petr Špecian (Eds.), Thinking in Dialogue with the Humanities (43–58). Zeta Books. https://doi.org/10.7761/9789731997971_3
  • Carmen, Taylor (2008). Merleau-Ponty. Routledge.
  • Cournot, Augustin (1861). Traité de l'enchaînement des idées fondamentales dans les sciences et dans l'histoire. In Oeuvres Complete (Vol. 3). Vrin.
  • Dillon, Martin C. (1997). Merleau-Ponty’s Ontology (2nd ed.). Northwestern University Press.
  • Dreyfus, Hubert L. and Stuart E. Dreyfus (1999). The Challenge of Merleau-Ponty's Phenomenology of Embodiment for Cognitive Science. In Gail Weiss and Honi Fern Haber (Eds.), Perspectives on Embodiment: The Intersections of Nature and Culture (103–120). Routledge.
  • Dreyfus, Hubert L. (2005a). Overcoming the Myth of the Mental: How Philosophers Can Profit from a Phenomenology of Everyday Experience. Proceedings and Addresses of the American Philosophical Association, 79(2), 47–65.
  • Dreyfus, Hubert L. (2005b). Merleau-Ponty and Recent Cognitive Science. In Taylor Carman and Mark B. N. Hansen (Eds.), The Cambridge Companion to Merleau-Ponty (129–150). Cambridge University Press.
  • Embree, Lester (1980). Merleau-Ponty’s Examination of Gestalt Psychology. Research in Phenomenology, 10(1), 89–121. https://doi.org/10.1163/156916480X00082
  • Evans, Fred and Leonard Lawlor (2000). Introduction: The Value of Flesh: Merleau-Ponty’s Philosophy and the Modernism/Postmodernism Debate. In Fred Evans and Leonard Lawlor (Eds.), Chiasms: Merleau-Ponty’s Notion of Flesh (1–22). SUNY Press.
  • Fóti, Véronique (2013). Tracing Expression in Merleau-Ponty. Northwestern University Press.
  • Gallagher, Shaun and Dan Zahavi (2012). The Phenomenological Mind (2nd. ed.). Routledge.
  • Gardner, Sebastian (2015). Merleau-Ponty's Transcendental Theory of Perception. In Sebastian Gardner and Matthew Grist (Eds.), The Transcendental Turn (294–323). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780198724872.003.0015
  • Grene, Marjorie (1976). Merleau-Ponty and the Renewal of Ontology. Review of Metaphysics 29(4), 605–625.
  • Grosz, Elizabeth (1993). Merleau-Ponty and Irigaray in the Flesh. Thesis Eleven, 36(1), 37–59. https://doi.org/10.1177/072551369303600103
  • Hass, Lawrence (1999). Sense and Alterity: Rereading Merleau-Ponty’s Reversibility Thesis. In Dorothea Olkowski and James Morley (Eds.), Merleau-Ponty, Interiority and Exteriority, Psychic Life and the World (91–106). SUNY Press.
  • Hass, Lawrence (2008). Merleau-Ponty’s Philosophy. Indiana University Press.
  • Heinämaa, Sara (1999). Merleau-Ponty’s Modification of Phenomenology: Cognition, Passion and Philosophy. Synthese, 118(1), 49–68. https://doi.org/10.1023/A:1005140809802
  • Heinämaa, Sara (2009). Phenomenological Responses to Gestalt Psychology. In Sara Heinämaa and Martina Reuter (Eds.), Psychology and Philosophy (263–284). Springer.
  • Irigaray, Luce (1993). The Invisible of the Flesh: A Reading of Merleau-Ponty, The Visible and the Invisible, “The Intertwining—The Chiasm”. In An Ethics of Sexual Difference (Carolyn Burke and Gillian C. Gill, Trans.) (127–153). Cornell University Press.
  • Katz, David (1950). Gestalt Psychology: Its Nature and Significance (Robert Tyson, Trans.). The Ronald Press Company.
  • Kockelmans, Joseph J. (1993). On the Function of Psychology in Merleau-Ponty’s Early Works. In Keith Hoeller (Ed.), Merleau-Ponty and Psychology (119–142). Humanities Press.
  • Koffka, Kurt (1935). Principles of Gestalt Psychology. Harbinger.
  • Köhler, Wolfgang (1925). The Mentality of Apes (Ella Winter, Trans.). Liveright.
  • Köhler, Wolfgang (1947). Gestalt Psychology: An Introduction to New Concepts in Modern Psychology. Liveright.
  • Kruks, Sonia (2006). Merleau-Ponty and the Problem of Difference in Feminism. In Dorothea Olkowski and Gail Weiss (Eds.), Feminist Interpretations of Merleau-Ponty (25–48). Penn State University Press.
  • Landes, Donald A. (2013). Merleau-Ponty and the Paradoxes of Expression. Bloomsbury.
  • Marratto, Scott L. (2012). The Intercorporeal Self. SUNY Press.
  • Morris, David (2004). The Sense of Space. SUNY Press.
  • Morris, David (2010). The Enigma of Reversibility and The Genesis of Sense in Merleau-Ponty. Continental Philosophy Review 43(2), 141–165. https://doi.org/10.1007/s11007-010-9144-7
  • Nöe, Alva (2004). Action in Perception. The MIT Press.
  • Penn, Derek C., Keith J. Holyoak, and Daniel J. Povinelli (2008). Darwin’s Mistake: Explaining the Discontinuity between Human and Nonhuman Minds. Behavioral and Brain Sciences, 31(1), 109–130. https://doi.org/10.1017/s0140525x08003543
  • Povinelli, Daniel J. (2000). Folk Physics for Apes: The Chimpanzee’s Theory of How the World Works. Oxford University Press.
  • Sanders, John T. (1993). Merleau-Ponty, Gibson, and the Materiality of Meaning. Man and World26(3), 287–302. https://doi.org/10.1007/BF01273397
  • Sanders, John T. (1994). Merleau-Ponty on Meaning, Materiality, and Structure. Journal of the British Society for Phenomenology25(1), 96–100. https://doi.org/10.1080/00071773.1994.11007052
  • Schenck, David (1984). Meaning and/or Materiality: Merleau-Ponty’s Notions of Structure. Journal of the British Society for Phenomenology, 15(1), 34–50. https://doi.org/10.1080/00071773.1984.11007644
  • Sheredos, Benjamin. Merleau-Ponty’s Immanent Critique of Gestalt Theory. Human Studies 40:2 (2017), pp. 191-215.
  • Smith, Barry (1988). Gestalt Theory: An Essay in Philosophy. In Barry Smith (Ed.), Foundations of Gestalt Theory (11–81). Philosophia.
  • Stawarska, Beata (2006). From the Body Proper to Flesh: Merleau-Ponty on Intersubjectivity. In Dorothea Olkowski and Gail Weiss (Eds.), Feminist Interpretations of Maurice Merleau-Ponty (91–106). Penn State University Press.
  • Toadvine, Ted (2009). Merleau-Ponty’s Philosophy of Nature. Northwestern University Press.
  • Toadvine, Ted (2011). The Chiasm. In Sebastian Luft and Søren Overgaard (Eds.), Routledge Handbook to Phenomenology (336–347). Routledge.
  • von Weizsäcker, V.F (1927). Reflexgesetze. In A. Bethc (Ed.), Handbuch der normalen und pathologischen Physiologie (Vol. 10). Spring. https://doi.org/10.1007/978-3-642-48586-2_3
  • Welsh, Talia (2006). From Gestalt to Structure: Maurice Merleau-Ponty’s Early Analysis of the Human Sciences. Theory & Psychology, 16(4), 527–551. https://doi.org/10.1177/0959354306066205
  • Welsh, Talia (2013). The Child as Natural Phenomenologist. Northwestern University Press.


    1. It is also, Barbaras notes, “contemporaneous with the Merleau-Pontian turning point that leads him toward ontology” (2001: 23). The argument is not unique, of course, to Barbaras. See, for example, Evans and Lawlor, who argue that Merleau-Ponty expresses a “grave dissatisfaction with the earlier work” and “attempt[s] to go beyond it” by “substituting” a “new notion” (flesh) for the earlier consciousness-world distinction (2000: 9–10). See also Fóti (2013), who explicitly aligns herself with Barbaras’s trajectory. Finally, see Gardener (2015) for a reading of Phenomenology of Perception that helps highlight the problematic of that text’s relation to the earlier and later thought.return to text

    2. See, e.g., VI 205/255, where Merleau-Ponty refers to the Gestalt as “a pivot of a system of equivalencies.”return to text

    3. In pushing back against Barbaras’s thesis, I align myself, e.g., with Landes (2013), Marratto (2012), Hass (1999; 2008), and Morris (2010). However, most of these commentators devote comparatively little attention to the late themes as they appear before Phenomenology of Perception. Marratto, for example, reads Phenomenology of Perception as an early call for what Merleau-Ponty later names “an ontological rehabilitation of the sensible” (S 166–7/271, as cited in Marratto 2012: 6), but mentions The Structure of Behavior only by way of introduction. Hass (1999) and Morris (2010) root the “reversibility” thesis in the earlier works, but center their attention on Merleau-Ponty’s treatment of sense experience in Phenomenology of Perception. The closest to my project here is Landes (2013), who devotes a chapter to The Structure of Behavior, though he is concerned with the concept of “expression” rather than “flesh.”return to text

    4. Some attention is paid in this direction, but much of it, on my view, is mistaken. Moreover, there is little scholarly consensus. Carman, for example, refers to “the standard Gestalt theory Merleau-Ponty has been evoking and elaborating for years” before identifying a “major departure” in The Visible and the Invisible (2008: 121). Compare this to Bannan’s remarks that in The Structure of Behavior, Merleau-Ponty “makes his entire case [against Gestalt theory] in every chapter” (1967: 31).return to text

    5. “Man and Adversity,” published in Signs, is a lecture delivered in Geneva in 1951; Hass (2008) does not make note of this appearance, focusing his typology on the “late” writings.return to text

    6. See, for example, PP 71/96–7: “When I see the lamp on my table, I attribute to it not merely the qualities that are visible from my location, but also those that the fireplace, the walls, and the table can ‘see.’ The back of my lamp is merely the face that it ‘shows’ the fireplace.”return to text

    7. The language of “pregnancy” comes as early as The Primacy of Perception, where human perception is said to be “‘pregnant’ with its form” (P 11), and at VI 250/298, Merleau-Ponty speaks of flesh as a “pregnancy of possibles, Weltmöglichkeit.” Bannon (2001: 333 Footnote 26) argues that “pregnancy” should be heard in the “strict psychological sense” of prägnanz—a basic principle of Gestalt perception—and not in the biological sense of reproduction. While it’s true that the French term (prégnance) does not convey the biological sense, the emphasis in the law of prägnanz on the tendency toward order does not seem to me to fully capture Merleau-Ponty’s very rich conception of the “contingent arrangement” or “nascent” sense of the perceptual field, which he cashes out in The Structure of Behavior in a way that, I argue, is in tension with Gestaltism.return to text

    8. For some pushback to this idea, still internal to feminist theory, see Kruks (2006).return to text

    9. My own view is that the shifting point in The Structure of Behavior comes at the transition from the Gestalt critique of reflex theories to the sections on “The Structures of Behavior” that constitute the last portion of Chapter 2. In those sections, Merleau-Ponty takes up a number of Gestaltist discoveries (drawing on Köhler’s experiments, for example, with chickens and apes), but interprets those in terms that are often different from (and sometimes in direct contradiction to) the Gestaltists’ explanations. I offer examples of these differences in Section 5.1, below.return to text

    10. A few commentators have tried to reconstruct the precise relationship of Merleau-Ponty to Gestalt theory. See, for example, Embree, (1980), Kockelmans (1993), and Heinämaa (2009). None of these focuses exclusively on The Structure of Behavior, however (though see Sheredos 2017), and none reconstructs the relationship through the lens of the source materials. An exception has been Welsh (2006; 2013), whose interest in Merleau-Ponty’s relationship to Gestalt theory concerns Merleau-Ponty’s contributions to child psychology.return to text

    11. The first German edition of Katz’s text was published in 1943, and so does not appear in the bibliography of The Structure of Behavior. However, Merleau-Ponty’s familiarity with Katz’s work at that time is attested by a brief mention (without citation) in the discussion of “The Reflex and Gestalt Theory” (SB 39/39). Katz himself was a student of Husserl’s at the University of Göttingen, where Husserl served as an examiner for Katz’s doctoral thesis in 1907.return to text

    12. This text appears in the bibliography of The Structure of Behavior.return to text

    13. This conception of form distinguishes the Berlin School (Wertheimer, Köhler, Koffka) from the earlier Gestalt psychology of von Ehrenfels, for whom forms are more than (and not other than) the sum of their parts. For some discussion of this development in Gestalt theory, see Smith (1988); for the link between that development and the phenomenologies of Husserl and Merleau-Ponty, see Heimämaa (2009).return to text

    14. Tellingly, Weizsäcker is not himself a Gestalt theorist and is critical of Gestaltism’s reliance on causal mechanisms.return to text

    15. This text appears in Merleau-Ponty’s bibliography to The Structure of Behavior.return to text

    16. Compare, for example, Merleau-Ponty’s introduction to Chapter 3 of The Structure of Behavior (especially SB 129–37/139–46) and the introduction to Koffka’s Principles of Gestalt Psychology (1935: 10–21).return to text

    17. The unit is a “real organization,” the consequence of which is what we call “order” (Koffka 1935: 175).return to text

    18. Note that, for Gestaltists, this is why their account avoids a relapse into “mentalism.” Mind, they argue (as one extreme of the spectrum of consciousness), is not discontinuous with matter (and so is not a separate power); it only seems so because the relation it describes does not fit as neatly into a formula as relations among material forms. Of course, exactly this fact makes it urgent to understand the status of form. For if the properly ‘mental’ forms we associate with ‘higher’ conscious behaviors are not easily captured in the language of physics, then what is it they share with the physical forms?return to text

    19. As circular, the impact goes both ways: we saw the other direction of impact already in Köhler’s apes, where a behavior (eating) transformed the geographical environment (i.e., by ‘erasing’ the banana).return to text

    20. Katz calls the thesis of isomorphism the most important discovery of Gestalt theory. See his discussion in “Physical Forms and Isomorphism” (1950: 54–8).return to text

    21. Citing the 1920 German edition of Köhler’s Gestalt Psychology, 193.return to text

    22. Cf. Merleau-Ponty’s arguments at SB 185–8/200–3. It’s common to suggest that Merleau-Ponty objects to Gestalt theory’s “naïve realism” (e.g., Embree 1980: 111) but this doesn’t seem to me to be right: the target of his objection is precisely the positing of an in-itself imperceivable substrate of physical form.return to text

    23. Merleau-Ponty’s discussion on this point, as well as the example of gravity, is drawn from Cournot’s Traité de l'enchaînement des idées fondamentales dans les sciences et dans l'histoire (1861: Section 181), which Merleau-Ponty quotes in the 1956–7 course on “The Concept of Nature” (N 3–122, the relevant quote appears on page 30); the portion of the course called “The Concept of Causality” (N 29–35) repeats much of the content of the discussion of physics in The Structure of Behavior.return to text

    24. Brunschvicg (1922: 516), quoted at N 31.return to text

    25. Merleau-Ponty’s argument here is admittedly difficult to track; the example of gravitational forces is, by my lights, the clearest expression of it. The discussion is significant for my purposes because it is where many commentators begin (wrongly, on my view) to detect an endorsement of idealism on Merleau-Ponty’s part. When Baldwin, for example, identifies the “sub-Kantian idealism” in Merleau-Ponty’s critique of structure in physics, he writes that it “does not do justice to the abstract complexity of physical law.” His explanation then betrays the misreading: he urges us to consider “the tensors which have a central role in general relativity,” writing that “they are certainly not ‘forms’ that are manifest in the perceived world” (2013: 192). That is not a claim Merleau-Ponty (or Gestalt theory) would hold. The claim isn’t that physical laws are perceivable forms, but that a description and analysis of physical laws requires an appeal to Gestalt structures (like the figure/ground structure) that are ‘learned’ in perception.return to text

    26. For Merleau-Ponty’s discussion of this view in Kant, see, e.g., N 21–6, especially 23.return to text

    27. Grene (1976) identifies two directions of natural development (which she calls pluralism A and pluralism B) and associates these with The Structure of Behavior and The Visible and the Invisible, respectively. “Pluralism A,” she tells us, “envisage[s] a universe that rises level upon level in a hierarchy of emergent realities, [and] Pluralism B, stress[es] not so much orderly and encompassing ascent as proliferation of structures, a multiplicity of forms, indeed, but not a scala naturae in either the classical or the modern, ‘emergent’ mode” (1976: 609). While she goes on to qualify that the ‘pluralism’ of The Structure of Behavior is only a pluralism “in a certain sense,” since it differs from a standard hierarchical view in several respects, her argument is structured around the supposed “dichotomy of directions” that distinguishes the early and the late ontologies (1976: 609). By my lights, her argument is important for its consideration of these two texts in parallel. However, I think it misstates the argument of The Structure of Behavior. While Grene defines Pluralism A as “admitting at the start some hierarchical organization, some ‘levels of reality’” (1976: 608), Merleau-Ponty says directly that when he speaks of “planes of signification” in that text we are not to think of these as “orders of reality” (SB 201/217) and that we should move beyond the language of higher and lower. That the image of planes of signification is his preferred image underscores, I think, where that text does not sit in such a neat opposition to the Pluralism B of The Visible and the Invisible as Grene supposes.return to text

    28. To show this, Merleau-Ponty offers an archaeological example, adapted from Brunschvicg:

      Each step of progress in Egyptology modifies the history of Egypt. But the reconstituted structures function to complete a “time” of the universe, the idea of which they presuppose. . . . But Egypt, as an economic, political and social structure, remains an object of thought distinct from the multiple facts which have constituted and brought it into existence. It is an idea, a signification, common to an ensemble of molecular facts, which expressed by all the facts and which is not contained completely in any one of them. (SB 143/154–5)

      That this comes from Brunschvicg (Merleau-Ponty’s exemplar of neo-Kantian Critical thought) suggests the shortcoming of this reasoning: it is right, but incomplete—right in the sense that we can interpret the history of Egypt only through the lens of the ideal toward which it tends, incomplete if it does not compensate this with an account of the development of the ideal and with attention to how the interpretive framework is learned in perceiving.return to text

    29. This is the point I argued Baldwin misses (Footnote 25).return to text

    30. At SB 209/226 Merleau-Ponty suggests that the pathological “disintegration” of soul and body, where these ‘appear’ to be distinct, gives us “the truth of dualism.” His point is that this is an abstract division of the natural and the cultural body that presupposes their original intertwining; the actual accomplishment of the apparent dualism (in which case the body has no ‘meaning’ of its own) implies biological death.return to text

    31. This image of ‘soil’, of course, becomes a central thought of the late ontology.return to text

    32. This is a subtle point, and commentators who miss it (e.g., Schenck 1984) accuse Merleau-Ponty of reauthorizing the form/content dualism in adapting the psychology of form. For a direct rebuttal of this point, see Sanders (1994). See also Sanders (1993).return to text

    33. More recently, Povinelli (2000) makes a similar point. He argues that primate cognition is limited because apes lack insight into unobservable properties of their environments. He clarifies elsewhere that, confined to perceptual problem-solving strategies, even higher primates cannot “reason about the higher-order relation between causal relations in an analogical or theory-like fashion” and so—he concludes—are only able “to solve . . . problems” based “on evolved, domain-specific expectations about what perceptual features are likely to be most salient . . . and a general ability to reason about the causal relation between observable contingencies” (Penn, Holyoak, & Povinelli 2008: 119). This confinement to perceptual strategies is supposed to explain the fact—already on evidence in Köhler—that a monkey will appropriate a tree branch to reach a distant banana, but only when the branch is positioned in front of him just so.return to text

    34. I’m thinking of Merleau-Ponty’s discussion of the “active qualities” proposed in Herder’s “rich notion of sense experience” (see PP 60/78). The distinction between pragmatic value (value for me) and Romantic value (value like me) comes from Čapek (2010).return to text

    35. This is the sense in which those stable structures give my body “the power of organizing the visual spectacle” (SB 117/127–8). Merleau-Ponty makes this point directly in Nature: “I organize with my body an understanding of the world, and the relation with my body is not that of a pure I, which would successively have two objects, my body and the thing, but rather I live in my body, and by means of it I live in the things. The thing appears to me in this way as a moment of the carnal unity of my body, as enclosed in its functioning” (N 74).return to text

    36. Just as, Merleau-Ponty tells us, “truths do not cease to be truths when I am not thinking about them” (SB 212/229).return to text

    37. For further discussion of these readings, especially with regard to Phenomenology of Perception, see Gardner (2015).return to text