1When Whitehead wrote that the status of life in nature constituted the modern problem of both philosophy as well as science, he may well have had Bergson in mind. In this paper I unearth a line of development that runs from Bergson through Whitehead and into modern embryology.
2It is well known that life is to be thought, according to Bergson, as a kind of impetus, a drive or force irreducible to physical determination. But Bergson insisted that this was only an image borrowed from psychology, the psychology of effort in particular — the best one available, but an image no less. Life (that is organization, evolution) is supposed to exceed any of the conceptual frames through which it can be determined. An image is required by which it can be imagined, grasped, made sense of. This — among some other reasons — is why so many images populate and animate Creative Evolution. In this paper I discuss one in particular: the canal, or canalization. The fact that this image managed to find its way, through Whitehead’s Process and Reality, into modern embryology via the work of C. H. Waddington has still yet to receive the appreciation it deserves. Canalization continues to serve, even after Waddington, as an image for the way epigenetic landscapes operate in order to bring about, or canalize, a small number of end results from a large number of developmental potentialities. Bergson’s image of the canal should be of interest not only to Bergsonians, or Whiteheadians, but to both philosophers of biology and historians of science as well.
3Though any relatively attentive reading of Creative Evolution should yield the position that Bergson’s élan vital is not a vital principle of the traditional variety, headway nonetheless remains to be made in its interpretation. I think that the status of the élan vital as an image is too often overlooked. Life — organization and evolution — is supposed to elude the binary alternative between efficient and final causality — or mechanism and finalism — and so cannot be determined in the causal terms recognizable to either. Bergson suggests that it is better thought for this reason through what he calls an “image” instead, and of an impetus or an “élan” in particular (Bergson 1998, 257). He explains why in a letter to Floris Delattre:
The image intervenes most often because it is indispensable, none of the other existing concepts being able to express the author’s thought; the author is then obliged to present it suggestively. […] To give just one example: when I relate the phenomena of life and of evolution to an “élan vital,” it is in no way an ornament of style. It is even less meant to mask in images our ignorance of the deepest causes, as when the vitalist in general invokes a “vital principle.” [...] The truth is that philosophy only offers philosophers two principles of explanation in this matter: mechanism and finalism. [...] Now [...] the place to be is somewhere in between these two concepts. How should we determine that place? I have to point to it, to indicate it since no concept between mechanism and finality exists. The image of an élan is only this indication (Bergson 1936, 399 translation mine).
4There is of course a lot to be said about Bergson’s philosophy of images, especially as it is developed in Matter and Memory. What I want to note here is that images serve as tools for the expression and communication of thoughts that frustrate or escape the delimited boundaries of available concepts. Images point to something outside of those concepts, prompting thought to think beyond its habituated forms. Yet, as tools for the expression and communication of thought, images are also necessarily breaks, arrests, halts, or circumscriptions of thinking; and so, with them “you can no more reconstitute thinking than with positions you can make movement” (Bergson 1920, 55). While the need for an image arises when thinking outstrips the concepts that are available to it, any particular image, as a satisfaction of that need, is only produced “when thinking, instead of continuing its own train, makes a pause or is reflected back on itself” (Bergson 1920, 45). The image is an interruption of the movement of thought beyond the concept. Thought then returns to the domain of concepts in order to communicate itself, enlisting the image in order not to have to settle back into any of the concepts that preexisted its movement. The image is something like an unstable intermediary between the intuition of something beyond the domain of concepts and the concepts beyond which it is directed.
5To say that the élan vital is an image is to say that it is a way of pointing beyond any of the already-existing concepts through which it might be thought. It should not be understood, in other words, as a concept of its own, with an identifiable object — like a force — to which it would correspond. But I want to leave that discussion to one side for now in order to consider another image in Creative Evolution, a less conspicuous one, the image of the canal. This image has a kind of subterranean legacy that remains underappreciated.
6Bergson introduces the image of the digging of a canal only briefly before appearing to replace it with his better-known image of the hand thrust through iron filings, but the two images work in tandem and should not be treated independently of each other.
7The canal is an image for the idea that “the vision of a living being is an effective vision, limited to objects on which the being can act” (Bergson 1998, 93). In Matter and Memory, Bergson theorized the way living beings cut, select, and distribute a field of perception images on the basis of their sensorimotor capacities as an operation by which the total aggregate of material images is narrowed into a navigable and actionable world of experience. In Creative Evolution, he describes the pragmatically limited capacity to see as “a vision that is canalized” (Bergson 1998, 93). “Vision,” he claims, “is a power which should reach in principle an infinity of things inaccessible to our eyes” (Bergson 1998, 93). Such a vision would be impractical for the vital necessities of living organisms. A vision that is canalized is a vision that is trained on and contoured by the objects that are relevant for it, in view of the ability to act on them. Canalization denotes the tapering down of a possibility space for some effective purpose.
8Before referring the “canalization” of “effective vision” to the image of the digging of a canal, Bergson adds that “the visual apparatus simply symbolizes the work of canalization” (Bergson 1998, 93 translation mine). By “visual apparatus” he means the structural complexity of the organ. Vision should in principle be able to register otherwise invisible objects, but is canalized (or narrowed) according to what is relevant for it. The visual apparatus is a product (or symbol) of the work, activity, or process of that canalization. Far from generating the function of vision as a capacity of its structural organization, the visual apparatus is supposed to be an effect of the function instead.
[...T]he creation of the visual apparatus is no more explained by the assembling of its anatomic elements than the digging of a canal could be explained by the heaping-up of the earth which might have formed its banks. A mechanistic theory would maintain that the earth had been brought cart-load by cart-load; finalism would add that it had not been dumped down at random, that the carters had followed a plan. But both theories would be mistaken, for the canal has been made in another way (Bergson 1998, 93-94).
- 1 This temptation is evidenced in the fact that despite the widespread existence of discussions of th (...)
9This is all Bergson has to say about the image. He turns afterward to the hand, submitting that “the process by which nature constructs an eye” — presumably analogous to that by which “the canal has been made” — can be fixed “with greater precision” in the image of a hand thrust through iron filings (1998, 94). It may seem tempting to leave the canal to one side and focus upon the image of the hand instead.1 This move is a mistake on four counts. First, the term “canalization” was introduced to explain the relationship between the function of vision and the structure of the visual apparatus before the image of the canal was produced and displaced (Bergson 1998, 93); second, it insists throughout the description of the hand, whose movement is said to be canalized by the iron filings that outline it (Bergson 1998, 95); third, the term reappears in later, pivotal discussions of the thermodynamic accumulation and release of free energy in the living being (Bergson 1998, 110, 126, 256); and fourth, there exists a line running from Bergson’s use of the term through Whitehead’s process metaphysics to the embryological theory of C. H. Waddington and from there into mainstream embryology today.
10Bergson says of the canal that the process by which it was formed cannot be explained by the accumulated dirt that makes up its banks. Canals are formed in three ways: (1) a channel is dug into dry land through excavation and supplied with water from elsewhere; (2) a channel is dredged into the bottom of an already existing body of water that is subsequently drained down to the canal; and (3) a stretch of raised banks are constructed in parallel in order to create a channel between them by outlining it. As Bergson mentions both “digging” as well as the “heaping-up” of earth in banks, we can imagine a canal of both the first and the third kind. This canal seems to have had its channel dug out and its banks formed out of the dirt displaced. As a water supply is introduced into the canal, it flows through the negative space of the channel and adopts a trajectory contoured by the material structure that contains and directs it. As a moving reality, the flow can be seen to stand in relation to the structure of the canal in a manner analogous to the relation between the banks and the channel. Whereas the negative space of the channel corresponds to the visible structure of the floor and banks that define it, the moving reality of the canalized flow corresponds to the sedimented and static material composition of the canal that directs it.
11The process through which the canal was formed stands in the same relation to the constituted material of its structure. It is on this point that both mechanist and finalist explanations falter: “mechanism would maintain that the earth had been brought cart-load by cart-load; finalism would add that it had not been dumped down at random, that the carters had followed a plan” (Bergson 1998, 93-94). Both agree in attending only to the finished material of the canal, and in so doing they miss the process responsible for its formation. While that process — the digging out of its floor — consists primarily in the generation of the negative space of the channel, its only visible material effects are embodied in the floor and banks around it. Both positions mistake what are the visible products of an invisible process of formation for its real component parts.
12The canal is explained fully when its material structure is also referred to the process of canalizing that it facilitates. The sedimented materiality of the canal figures both as the material product of the process of its formation and as the material substrate for the canalizing of a flow. The image of the canal depicts the distinction between process (formation) and product (material structure), illustrating the irreducibility of the former to the latter; and it is an image that depicts the reincorporation of product (the canal) into process in turn (canalization).
13After asserting that both the canal and the eye have been made in a way that eludes the explanatory resources of mechanism and finalism alike, Bergson replaces the image of the canal with the image of a hand. He introduces the second image by adding the idea of resistance to the example of raising one’s hand: “Let us now imagine,” he says, “that, instead of moving in air, the hand has to pass through iron filings which are compressed and offer resistance to it in proportion as it goes forward” (Bergson 1998, 94). By offering resistance, the filings “canalize” the trajectory of the hand by limiting its motion and eventually exhausting the effort required to push any further. We are told to imagine that the outline of the hand remains preserved in the arrangement of filings wherever it stops; and, finally, to imagine that the hand and arm are invisible. The result is a mass of iron filings coordinated around the silhouette of an absent hand. This image is analogous to the canal in the bifurcation between its material configuration and the invisible process responsible for its formation.
14The coordinated filings should not be thought as a causally inert representation of the hand’s final position. The reason why the image involves iron filings and not a more plastic medium is because the filings resist the hand plunged into them, canalizing its movement and exerting a causal influence on the trajectory taken by it. The filings can be conceived both as the product of a canalized process and as the material substrate for the canalization of that process. The distinction between process and product is at least partially bidirectional. The filings are a spatialized expression of the trajectory taken through them as well as ingredient factors in the calibration of that trajectory itself.
15Bergson proposes that “the relation of vision to the visual apparatus would be [on his hypothesis] very nearly that of the hand to the iron filings that follow, canalize and limit its motion” (Bergson 1998, 95). If it is the act of vision that is canalized, then it is the physiology of the visual system that canalizes it. The structure of the visual system, as the material substrate for the canalization of vision, is what explains the pragmatic efficacy of vision in the living being. The living being sees in its environment only what is meaningful for it as an effect of its body’s material constraints on and delimitations of what it would be otherwise possible to see. Canalization is both limitation and facilitation at the same time. Limitation is facilitation’s material means.
16If vision is canalized by the visual apparatus, then any attempt to explain the composition of the apparatus on the basis of its structure alone would be akin to mistaking the constituted banks of a canal for the process through which the canal was dug. When Bergson calls the materiality of the organ “a negation rather than a positive reality,” he means that the organ is the spatialized outline of the process of its formation, reflecting the positive reality of its genesis in inverse, by silhouetting it (1998, 93). The greater the hand’s effort, the further will be its trajectory through the filings that resist it. This is Bergson’s model for vision as well, for the visual organ’s sophistication — its degree of complexity and the coordination among its parts — stands in direct proportion to the effort or advancement of “the undivided act constituting vision” (1998, 95). The tendency toward vision canalizes the visual apparatus. “The materiality of the organ [of the eye] is made of a more or less considerable number of mutually coordinated elements” in direct proportion to how far “the undivided act constituting vision” was able to realize itself in a given medium (Bergson 1998, 95). The structure of the visual apparatus corresponds in its complexity to the intensity of the function embodied in it. While structures are always concrete and individual, differing from each other in the coordination of their component parts, the function of vision is an undivided act whose advance in intensity is realized across structures of increasing complexity in evolutionary time. The specific form of each visual apparatus “only expresses the extent to which the exercise of the [evolutionary] function has been obtained” (Bergson 1998, 96). Behind all the visual apparatuses of the animal kingdom — from gastropods and spiders to cephalopods and mammals — there is one tendency, one movement, one élan.
17Evolution is canalized — patterned or trended — in the direction of vision just as it is canalized in the direction of indetermination. Both ends are canalized by the structures that facilitate them. In Bergson’s discussion of thermodynamics, he suggests that the tendency towards indetermination (which is how he defines the work of the élan vital, as the introduction of indetermination into matter) can be understood on the basis of its physiological canalization through the metabolic processes that afford the animal its latitude of possible actions in response to external stimuli. This tendency is canalized subjectively as well, via the sensorial presentation of choice to consciousness, contoured by the material means by which visual information is captured, processed, and converted into the visual images in which possibilities for action are instantiated. For any of evolution’s ends, any of its tendencies, there is a process of canalization that works to constrain and facilitate their realization.
18Trends, patterns, and convergences are the results of canalization. Canalization is everywhere — that is my idea. By the end of Creative Evolution, it is trendedness that comes to comprise the canalization’s primary sense (Bergson 1998, 110, 126, 256). With some qualifications, it is this sense that is retained by the use of the term “canalization” in the embryology of today. It was most likely Whitehead who first discerned in Bergson’s image the value of this aspect in particular.
19The question of the extent to which Bergson played an influential role in the development of Whitehead’s philosophy of organism has been the subject of some debate since F. S. C. Northrop declared Bergson the chief resource for “the basic concept and doctrine of Whitehead’s entire philosophical and scientific outlook” (Northrop 1941, 169). Northrop identified Wildon Carr’s 1912 monograph, Henri Bergson: The Philosophy of Change, as the most likely intermediary between the two. The critical response to this suggestion is exemplified by Victor Lowe, who argues that Bergson exerted little to no positive influence on Whitehead, seeing the two thinkers as having developed some broadly similar general commitments independently (Lowe 1949, 271, 272, 278). In his more recent study on the topic, Didier Debaise concedes an initial convergence, but argues for a basic difference in the way each develops his system (Debaise 2009). Neither Lowe nor Debaise addresses the idea of canalization. Neither do any of the other critics of the idea of Bergson’s influence on Whitehead.
20Whitehead introduced the term “canalization” in The Principle of Relativity to emphasize the productive dimensions of limitation, attributing to Bergson both the word and the insight as well (Whitehead 1922, 16). Though Whitehead did not elaborate its details initially, the term reappeared throughout Process and Reality, serving there as an essential factor in the explanation of the emergence of order (Whitehead 1978, 107-108, 129, 178). According to Whitehead’s use of it, canalization is a primarily positive process. The material of a canalized process should be understood not as the process’s negative or spatialized outline, but as the means by which it was progressively tapered, becoming increasingly irreversible over time. Whitehead thus conceived of canalization as a general metaphysical principle capable of explaining everything from gene expression to the order and continuity required for the maturation of personality. It might not be that surprising that Whitehead saw something powerful in Bergson’s image of the canal. What he saw was the insight that creatively can only realize itself within the confines of material bodies. Without material “brakes” on its tendency to differentiate, evolution would be able to produce nothing stable at all, nothing on which natural selection could subsequently act. That would not be a victory won on behalf of creativity, but a failure, for totally unrestricted variation wouldn’t be productive as much as chaotic — all experimentation and no artwork. It is this generative sense to canalization that Whitehead affirms and develops — it supplies order, which is required for creativity and an increase in what he calls “intensity.”
21Whitehead emphasizes the way Bergson utilizes the image at the end of Creative Evolution, as a mechanism for the concentration and intensification of an initially diffuse and vague activity. Whitehead recodes the idea of productive concentration in the terms of his own system. In Whitehead’s formulation, canalization refers to the appropriation — or “prehension” — of an inherited past of consolidated acts and relations. As the organism develops, it integrates more and more of its relations and transmits those integrations along a temporal line. The organism doesn’t have to constantly decide how to relate to the world; in some sense, the outlines of its relation are increasingly rigidified for it — which implies a certain degree of irreversibility, just as it implies a developmental tapering in the space of possibilities.
22Canalization affords the organism an increase in order, which is just as necessary for the intensity of experience as is originality. Originality always moves away from — breaks the strictures of—the order out of which it emerges and into which it settles, even if after expanding or reshaping it. In biological terms, Whitehead might say that canalization allows the organism to take one relatively stable route through a massively chaotic environment — whether interior or external. It allows for the relatively reliable expression of phenotype, just as it ensures that different organisms of the same species behave in similar ways and so are able to survive in the same niche. Creativity, originality — or we might say mutation, variation — requires that reliability as its starting point. “Thus life is a passage,” on Whitehead’s account, “from physical order to pure mental originality, and from pure mental originality to canalized mental originality” (Whitehead 1978, 107-108). Creativity necessitates some means by which societies of “actual subjects,” or existent organisms, can interact with what has yet to come to pass. This is what is meant by “mental originality”: the prehension or appropriation of what exists over and above the actual — that is, the conceptual. If prehensions were simply physical, if they were related only to their actual pasts, then novelty would be made impossible from the outset. The universe would be caught in a cycle of repetition. Its creative advance is made possible on the basis of the creative decisions of actual subjects capable of prehending the physical past as in some sense other than it was; prehending the past conceptually — in the indeterminacy afforded to it by what Whitehead calls “eternal objects.” If physical feeling relates to the settled facts of the past, then conceptual feeling concerns the future, the formal possibilities left open by those facts. Every process of actualization therefore concretizes what is initially (or potentially) indeterminate by prehending the conceptual (that is, the eternal objects) along with the actual, thereby adding another layer of realized actuality to the ongoing creative advance of the cosmos. Unity and continuity are achievements — not starting points or givens (Whitehead 1978, 108). Canalization supplies the mechanism of their explanation.
- 2 Peterson (2011, 306-307); Gilbert (1991); Fox-Keller (2000, 117-119); Alseekh et al. (2017); and Wa (...)
23C. H. Waddington — a developmental biologist responsible for initiating the conceptual revolution in the life sciences known now as “systems biology” — elaborated his theory of the epigenetic landscape and its branching pathways of development while reading Whitehead’s Process and Reality.2 Waddington drew on Whitehead’s reformulation of the image of the canal as the means by which to explain organismic order in the midst of environmental variability. He sought an explanation for the reliable production of similar phenotypes in a highly variable population scattered across significantly different environments. Waddington imagined a plane into which several divergent canals have taken shape. What flows through them, on his account, is not a tendency in the process of realizing itself, but — according to a certain secularization of Bergson — the cell instead. Before becoming canalized along one line of development, the cell’s fate is plastic; it can follow any of a number of pathways, each of which is contoured by the interactions of various genes. But once the cell begins to develop along certain trajectories, it gains in expression — what Whitehead called “intensity” — what it loses in plasticity.
24Developmental irreversibility is the key to Waddington’s formulation of the image. Becoming canalized means falling into a groove, taking one pathway of development at the expense of initially possible others. “Developmental reactions,” in Waddington’s own words, “are in general canalized […] so as to bring about one end result regardless of minor variations in conditions during the course of the reaction” (Waddington 1941, 563). It was on the basis of Whitehead’s redeployment of Bergson’s image of the canal that Waddington developed a theory of epigenetics that, as Adam Wilkins has suggested, was nothing short of “a premature discovery” (Wilkins 1997, 257). Waddington encoded Whitehead’s insight into material facilitation in his formulation of cellular development. It remains important to the study of epigenetic stability today.
25It is important to note that Waddington’s image does not correspond to anything concrete. There are no canals in the cell. It is an abstraction with purely heuristic value. To what extent is it the same image, the same idea, that is behind the use of the same abstraction in each thinker? No doubt the idea of canalization evolved as it passed from Bergson through Whitehead and into Waddington. Bergson began by employing canalization as an image for the way structures are organized by their corresponding functions, taking vision and the visual apparatus as a case. Visual structures of varying material complexity are in some sense the downward causal results of the intensity of the function of vision that they embody and facilitate. This is part of what Bergson meant when he claimed that the visual apparatus was canalized by vision. He extended the image further than this too, generalizing canalization across evolutionary history, employing it as an image for the way simple tendencies are channeled and embodied through a series of complex forms, material organs, over planetary time. The image in this more general form is what Whitehead redeployed as a principle for the production of novelty on the basis of ordered, limited, concentrated creativity. Whitehead emphasized the facilitation effect of material constraint, but it is Bergson’s concept of canalization that he is employing nonetheless. It is only a small step from there to Waddington’s idea of canalization in cellular development. He reimagines the material structure of constraint in the form of the genetic landscape, drawing from structuration the idea that stability is achieved as a function of narrowing possibilities over the course of development. If we replace the genetic landscape with the banks of the canal, substituting for the cell in the course of development the idea of irreversible process, or tendency, in the course of materializing itself, then it is not so difficult to see the affinity.
26By way of conclusion, I want to return to what I said about images at the outset. On the one hand, images are spatializations or halts. They cannot capture the complexity of the phenomena for which they are images, and they do not purport to. On the other hand, they are heuristics, helpful simplifications that point beyond themselves, using their own limits as resources for communicating more of whatever it is that they depict.
27There is a kind of structural analogy between the image and what it images, and life (or that for which the élan vital is an image) and individual organic forms. Life itself, so to speak, might be thought according to the model of the image. The evolutionary movement is spatialized across a series of individual organic bodies, which outline in material form a process that is always underway, canalizing it. There is a relationship that obtains between the logic of life and the inability of the image to exhaust its explanandum; or yet again, between the relation of the evolutionary movement to individual organisms. Scientific analysis is also structurally analogous to perceptual experience, as each operates by isolating a thing from its relations and dividing a whole into parts.
28Understood in this context, the image of the canal might occupy a particularly interesting position in Bergson’s understanding of life. The canal, as an image for the relation between life and matter, is an image, a halt or spatialization, deployed as a means by which to explain precisely that same limitation imposed on life by matter. Canalization is supposed to show that what is essential about biological structures cannot be captured spatially without recourse to the moving reality or indivisible tendency that is at work in organizing them. Yet it is supposed to accomplish this as an image, characterized by the same spatializing operation that it is implemented in order to address. The image of the canal is a spatialization of thought intended to demonstrate the inadequacy of spatialized thinking. Like the filings, it outlines its object in negative relief. That may not be coincidental. It might even reveal something about Bergson’s view of the scientific study of living things. Just as life — or the élan vital — needs the matter that limits it in order to realize itself, perhaps it is the case that we make use of images in order to think — even if through them we always necessarily fall short of our objects of thought. Understood in this way, images of life are far from necessary evils. They are also potentially productive ingredients in the development of scientific theory in their own right. Science makes use of spatialization, mereological disintegration, analytic and experimental stabilization, isolation, delimitation, and so on. These are features of imagistic cognition. Life cannot be studied apart from its images, but these images may work best when they can be turned back on themselves, made to reveal how they are functioning as images, what their imagistic limitations are.
29I want to conclude with a distinction made by Peter Godfrey-Smith between philosophies of science and nature, and to suggest that both Bergson and Whitehead are situating the former within the latter, delineating its boundaries in accordance with the limits of imagistic thought, and granting to the philosophy of nature a meta-scientific purview. While the philosopher of science attends, on Godfrey-Smith’s account, to the specificities of some domain of research, its requisite boundaries, presuppositions, and operative images, the philosopher of nature “comments on the overall picture of the natural world that science, and perhaps other types of inquiry, seem to be giving us” (Godfrey-Smith 2001, 284). Where the scientific thinker makes use of helpful images, expedient categorizations, and ontologically suspect distinctions in the service of some research program, the philosopher of nature feels no such obligation. This philosopher comes, in a certain sense, after empirical research, and tries to synthesize its findings with insights generated by other investigations in other domains in order to do as much justice as possible to the global complexity of the situation. Often this is unhelpful, even paralyzing for scientific research. Science has to delimit, cordon, isolate. But the philosopher of nature has concerns over and above research feasibility and experimental purchase.
30Bergson and Whitehead should be understood as philosophers of nature in this sense. Their philosophies function both as commentaries on the findings generated by the sciences of their time as well as attempts to supplement those findings with theories that remained foreclosed to the methodologies of the sciences. If Bergson’s élan vital and Whitehead’s primordial creativity are part of a philosophy of nature, then their use of the image of the canal might signal the role played by a philosophy of science. The life sciences proceed by decomposing the living world into graspable parts that are conducive to analytic manipulation. That is also how images work. They isolate and delimit for some purpose, in service of some end. So too for science, which means that it should be located within a philosophy of nature whose sphere is wider — and, indeed, metaphysical at its bottom. By tracing embryology’s concept of canalization back to the image of canal as it is situated, in both Whitehead and Bergson, within a larger field of activity — of temporality, creativity, life — I tried to indicate one way in which these thinkers recommend a philosophy of science concomitant with a philosophy of nature that will always nonetheless necessarily outstrip.