Pure variation and organic stratification

Progress in Biophysics and Molecular Biology xxx (2012) 1e13 Contents lists available at SciVerse ScienceDirect Progress in Biophysics and Molecular Biology journal homepage: www.elsevier.com/locate/pbiomolbio Original research Pure variation and organic stratificationq Jérôme Rosanvallon Université Paris 7 Diderot, France a r t i c l e i n f o a b s t r a c t Article history: Available online xxx The fundamental problem posed by Darwin distinguishes his theory from any transformism of the past as well as any evolutionism to come: since variation is inherent to the living, it is a question of explaining, not at all why the living varies, but instead why the living does not vary in all directions to the point of constituting a continuum of forms varying ad infinitum. What limits and stabilizes this intrinsically unlimited variation, allowing certain forms to subsist and multiply to the detriment of others, is natural selection. This double principle of intrinsic variation/extrinsic selection constitutes a vector for the unification of reality that underlies Jean-Jacques Kupiec’s ontophylogenesis as well as Deleuze and Guattari’s global philosophy of Nature. Therefore, everything would potentially tend to incessantly vary. The work of Kupiec and others identifies an intrinsic random variation within ontogenesis itself. For Deleuze and Guattari, it is nothing but the figure, already selected by the organic stratum, of a more fundamental or pure variation. But, in fact, nothing really varies incessantly: everything undergoes a selective pressure according to which nothing subsists as such except what manages to endure through invariance (physical stratum) or reproduction (organic stratum). Thus, organic stratification only retains from variation what ensures and augments this reproduction. In this sense, every organism stratifies, i.e. submits to its imperative of subsistence and reproduction, a body without organs that varies in itself and always tends to escape the organism, for better (intensifications of life) or worse (cancerous pathologies). Ó 2012 Elsevier Ltd. All rights reserved. Keywords: Variation Stratification Cosmogony Symbiogenesis Ontophylogenesis Body without organs 1. Darwin: intrinsic variation and extrinsic selection 1.1. The two meanings of variation in the Origin of Species In what sense can we really speak of a Darwinian revolution? What is the comparative advantage of the Darwinian theory, which has today still allowed it to extend and impose itself to the detriment of all other theories, including those affiliated with it? Darwin does not simply mean to break with creationism and to affirm definitively that “each species had not been independently created, but had descended, like varieties, from other species1”. More subtly and with a certain prescience, he tends to break with theories of transformism, whether past or contemporary (Lamarck, Owen, etc.), while also breaking with their future evolutionist variants that nevertheless ally with him (Th. Huxley, Spencer, etc.). In a sense, all these theories are looking to explain the variation of the living and are therefore led to suppose that the living tends to adapt itself to its milieu, to complexify itself, to individualize itself, etc., q Translated from the French by Taylor Adkins. E-mail address: jrosanva@orange.fr. 1 On the Origin of Species [OS], 1st ed., 1859; “Introduction”, p. 3. supposing in this way some tendencies within individuals, species or life itself that would constitute determined directions of variation. But Darwin simply posits axiomatically or in principle the fact that there is, within both domesticated and natural living beings, variation.2 Just as the order of the presentation of the Origin of Species (OS) seems to suggest, every theory of the living must take this fact as its point of departure and not its destination: every theory must found itself entirely upon this fact rather than seeking to found it, i.e. to completely explain it. Although Darwin did not necessarily grasp its overall revolutionary capacity, this principle’s position has been fully brought to light by one of his most faithful correspondents, the botanist Joseph Dalton Hooker, who correctly reproaches him for not having emphasized it in a more compelling way in the OS and afterwards for too often neglecting e in his ulterior hesitations e “to dwell on the facts of infinite incessant variation3”. Thus all the stakes of natural selection are deduced from this primary fundamental fact: to explain that everything within the 2 Ibid., chapters 1 and 2. Letter from Hooker to Darwin 26 November 1862, Correspondence of Charles Darwin [CCD], X, p. 570. 3 0079-6107/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 2 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 living does not vary incessantly, that there are invariances e there is stability, which is itself provisional, and there is reproduction of identity, which is itself partial, i.e. ultimately that there are groups of living beings that are distinct from and cross-fertile with others and are thus reckoned to constitute a certain number of species e second constitutive fact of the living and first observation. In this sense, natural selection would designate nothing but the existence of a condition of non-variation, the existence of a constraint that limits this potentially infinite and incessant variation by conserving only the favourable variations, which are then led to impose themselves to the detriment of others, thus laying the foundations for specific divergences. We understand in a similar way that Hooker could consider that a rigorous, axiomatic and pedagogical presentation of the Darwinian theory sets out to “disentangle the two phenomena of variation and Natural selection4”, the first and always disguised fact of incessant variation and the second and always partial fact of reproductive invariance. But the same text of the OS seems to doubly resist this disentanglement that Darwin never fully unravels. On the one hand, through several reiterations, chapter four presents natural selection as having the capacity to produce variations, to “act on and modify organic beings5”, which is what Hooker explicitly bemoans: “Natural Selection is supposed to make the varieties as well as to fix them”, whereas it is nothing but the stabilizer or “fixer of these” varieties into species.6 On the other hand, chapter five seems to claim to partially explain variation by exposing its “laws”. Darwin’s difficulties in completely leaving behind this pattern advocated by his predecessors and a number of his successors, who believed it was necessary to explain the variations of the living and not its invariances, are further confirmed by the way in which he responds to his interlocutor: “I wish I had done what you suggest, started on the fundamental principle of variation being an innate principle; and afterwards made a few remarks, showing that hereafter perhaps this principle would be explicable7”. The resistances Darwin seems to have do not at all contradict the axiomatic framework redefined by Hooker but simply require for new distinctions to operate within it. Concerning the “production” of variation through natural selection, it is a question of an abbreviated formula that does not support any ambiguity if one takes care to distinguish, according to the celebrated “queer diagram” presented in chapter four, between variation envisioned “horizontally”, the spontaneous branching out with each generation, and variation envisioned “vertically”, the progressive modification of the same lineage. In the latter case, variation is nothing but that of a subject, i.e. of a certain species taken as an arbitrary point of departure: it simply measures the divergence between this point of departure, “common descent”, and a destination that is just as arbitrary, its so-called “modified” descent; this divergence in the end is actually an effect induced by natural selection, which tends to eliminate all the less advantageous, intermediary variations. Nevertheless, in the first case variation is subjectless or has no subject other than the living itself: it returns to a pure variability of the living, “an inherent tendency to vary8” with each generation; this variation, which remains without determined or perhaps determinable cause, is instead intrinsic in the sense that natural selection only intervenes in it secondarily, and thus extrinsically (even if e we shall return to this e it also itself tends to conserve and amplify this intrinsic variability of the living). If he often passes from one to the other unexpectedly, Darwin acknowledges the 4 5 6 7 8 Letter from Hooker to Bates 2 February 1862, CCD, X, p. 128. OS, 1st ed., 1859, IV, p. 86. Letter from Hooker to Bates 2 February 1862, CCD, X, p. 129. Letter from Darwin to Hooker 26 March 1862, CCD, X, p. 135 (my emphasis). The variation of animals and plants under domestication, 2nd ed., 1875, p. 2. importance of distinguishing between these two meanings of variation, most notably when, in the final edited version of the last chapter, he once again casts his theory as that of “descent with modification through variation and natural selection9”. It should be noted that the “laws of variation” are not the translation of some theory of variation that would attempt to explain the primordial fact of variation by revealing its origin, reason or causes. Concerned with the how and not the why like any other law, they only seek to comparatively determine the “relative degrees of variability10”, i.e. the greater or lesser variability of a certain part of the organism (according to its more or less developed character, the loss of its usage rendering its elimination advantageous, its structural or hereditary correlation with other parts, etc.) and of a certain population of living beings (according to whether it lives domestically or in the wild, in an extensive or restrained habitat, etc.). Darwin’s question is therefore never e for what reasons or according to which causes does it vary? e but always e according to which differential speeds and rates does it vary? e being given the greater or lesser selective constraint that secondarily exerts itself upon this intrinsic variation. 1.2. The “Darwin-Hooker principle” Having clarified this axiomatically, we can now call this principle e according to which, on the one hand, what varies is a primordial fact that is itself given without having to be explained and, on the other hand, only what does not vary must be explained as the effect of a constraint that extrinsically limits variation e “Darwin’s principle”, or more precisely, in order to do justice to his most prescient reader and interlocutor on this point, the “DarwinHooker principle”. But could this principle be restricted to the sphere of the living, since it is not simply valid within what we call the “organic stratum”? Does it not necessarily call for a metaphysical systematization that would take variation alone as the point of departure and simply give itself the task of explaining on its basis everything that does not vary, namely biological but also cosmological and physico-chemical or even anthropological and socio-historical invariants? The entire program of naturalist metaphysics elaborated for a span of about twenty years by Deleuze and Guattari could be taken back up in this way (from Anti-Oedipus in 1972 up to What Is Philosophy? in 1991): this program implicitly relies upon the “Darwin-Hooker principle”, which is the same as reversing Leibniz’s fundamental ontological question by no longer examining why something happens rather than nothing, but on the contrary why sometimes nothing happens, i.e. why everything does not incessantly vary. This fundamental reversal thus calls for a radical generation of all forms of invariance, starting with those that constitute the framework of the physical world (space-time, constants, matter, things, etc.), which then cannot be an irreducible given of nature but simply a secondary effect resulting from a constraint exerted upon a primary variation. The “Darwin-Hooker principle” seems to constitute a new formulation, which is potentially valid for all the sciences and their studied domains of reality, of the “principle of reason”: in effect, this formulation is substituted point by point for the preceding formulation that masterfully reigned over modern science and philosophy and completely reverses it. This preceding formulation can identically be called the “Newton-Galileo principle” and can be defined in this way: every system tends to conserve its state and only changes the latter if an external force is exerted upon it. This fundamental principle of modern physics actually at the same time 9 OS, 6th ed., 1872, XV, p. 404 (my emphasis: “through variation” was only added in the 6th and final edition). 10 OS, 1st ed., 1859, V, p. 151. Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 underlies the kinematics of Galileo’s supposed inertial movements and the dynamics of Newton’s supposed external forces: it is not only translated by the principle of inertia but also by the principle of relativity (which posits the equivalence of “immobile”, inertial and e with its Einsteinian generalization e accelerated systems of reference) and the various laws of conservation (energy, quantity of movement, angular momentum, electric charge, etc.). Does it not almost immediately become the object itself of a rigorous metaphysical systematization through Spinoza? The entire Ethics in a sense finds a logic of the conatus, i.e. the inertial perseverance of every (physical, organic and human) thing in being, a perseverance that only varies (augmenting or diminishing its capacity of being) and only stops functioning due to external encounters alone. In this sense, we can more easily grasp the fundamental reversal operated by the “Darwin-Hooker principle”, which on the contrary affirms that every system tends to vary in itself and only conserves its state in accordance with an external constraint. As foundational for modern biology, this principle subtends both phylogenesis e following the Darwinian revolution e and ontogenesis e following the extension of this revolution initiated by Kupiec: it combines intrinsic variation (whose nature, as we shall see, is fundamentally random) and extrinsic selection (which, in the case of the living, does not with each generation restrain the variation and varieties it generates without tending to conserve this variability that is inherent11). It therefore remains for us to show how Deleuze and Guattari’s philosophy constitutes, with the double idea of pure variation and stratification, the most ambitious metaphysical systematization of what remains after a century and a half a new and revolutionary formulation of the “principle of reason”. 2. Everything potentially tends to vary incessantly: random variation and pure variation 2.1. Decodings and deterritorializations: Neo-Darwinism according to Deleuze and Guattari On the Origin of Species therefore does not contain a theory of variation but a theory of selection that limits variation, whatever the scale, causes and nature of the latter might be. Darwinian theory in this sense assigns a place to variation, taking it as a primary matter without needing to determine it except by a variability upstream that produces it and a variety downstream that it produces. Darwinian theory must singularly determine, on one side, degrees of variability, which are envisioned as relative to one another in accordance with conditions of existence and, on the other, varieties of population, which are envisioned as an array of nascent species. As for the incessant variations produced in this way, they cannot receive any supplementary determination e the theory does not require it: this is not so merely by default, because, as we said, on the contrary this is what focuses its entire force. Based on Darwin’s own quite classic judgement, it therefore does not matter much that the seemingly random character of all these variations is nothing but the simple effect of our ignorance12: his theory will constitute no less than the guiding thread for all those whose point of departure is an intrinsic random variation upon which an 11 A recent reformulation of this principle e in contrast with that of NewtonGalileo e has also been proposed by Brandon and McShea (2010). But is not diversification a secondary consequence of intrinsic variation, and “complexification” a useless evolutionary residue? 12 “I have hitherto sometimes spoken as if variations e so common and multiform in organic beings under domestication, and to a lesser degree in those in a state of nature e had been due to chance. This, of course, is a wholly incorrect expression, but it serves to acknowledge plainly our ignorance of the cause of each particular variation”, OS, 1st ed., 1859, V, p. 106. 3 extrinsic pressure of selection can be exerted, whatever the source of variety and type of “population” concerned might be, whether within the organic strata or beyond. Thus we have been able to speak of molecular, cellular, immunitary and neuronal, as well as social, economic and even mineral or cosmological Darwinism.13 The “synthetic theory of evolution” labelled as “NeoDarwinism” is obviously the first among all the others that has given to the double Darwinian principle an irrefutable support, that of the genetics of populations: on the one hand, genetic mutations constitute the single a priori material support for transmissible and thus selectable variations and, on the other hand, selection is carried out a posteriori by the milieu, which eliminates unfavourable mutations and conserves favourable ones insofar as the latter allow the organisms to which they belong to acquire new resources or utilize available resources better than other organisms. The synthetic theory constitutes a primary authentic re-foundation for Darwinism, insofar as it accentuates the divergence and establishes a single direction between the genotypic source of variability and its selected phenotypic effects, thus confirming that variations depend not on some internal finality of development or external finality of evolution e whatever it might be e but solely on what arrives in a stochastic fashion in the genetic code, which is exposed to the constant bombardment of the milieu (oxidations, radiations) and above all constrained to ongoing replications and recombinations (cellular division and sexual reproduction). This fact alone, however, does not prevent us from envisioning the replication of the DNA complementary double helix as the most reliable process imaginable, just as much as it does not prevent us from consequently assimilating all these random variations with “accidents”, “errors”, simple “noise” e which are all certainly decisive, since all varieties and, in the end, all specific divergences result from them. Such an interpretation illustrates the current attitude of the biologists who, lacking the ability to account for a source of chance deterministically, at least tend to cast it in a simply perturbative role. But a properly Darwinian perspective always implies taking not invariance but variation as the norm and thus implies never giving up on asking, including within DNA, not “how do variations appear, but the other way around, how is it that they do not appear more often!14”. In truth, a genome owes its stability to various mechanisms of control and correction (notably the DNA mismatch repair system15), which are the precocious products of selection and which have been identified with multiple variants in (almost) all uni- or multicellular prokaryotic and eukaryotic organisms. Despite this reinforced fidelity of genomic replication, the stability of genotypic forms on the geological scale is as provisional as the species that express them phenotypically. These forms therefore seem like “so many islands consolidated here and there in the ocean of becoming16”, a scattering of little islands emerging in time through selection and necessarily drifting within a perpetual genetic whirlpool. Describing the organic stratum’s unity of composition and logic of constitution in the third plateau of A Thousand Plateaus titled “The Geology of Morals”, Deleuze and Guattari specifically point out the fact that “a code is inseparable from a process of decoding that is inherent to it”, a process of decoding signifying for them neither the transduction nor the translation of the genetic code alone on the scale of the individual, 13 For a complete panorama of all these Darwinisms, see Heams, 2011b; the work of Gerald Edelman in neurobiology (see specifically Edelman, 1987) and that of Lee Smolin in cosmology (see specifically Smolin, 1999) should also be included. We shall take up the question of cosmological selection in our third part. 14 Thomas Heams, 2011a, p. 54e55. 15 See, historically, Meselson and Wildenberg, 1975, and, for a recent review, Galles et al., 2009. 16 Bergson, The Creative Mind: An Introduction to Metaphysics, p. 64. Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 4 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 but the transformations and drifts of genomes in their entirety on the scale of populations. No genetics without ‘genetic drift’. The modern theory of mutations has clearly demonstrated that a code, which necessarily relates to population, has an essential margin of decoding: not only does every code have supplements capable of free variation, but a single segment may be copied twice, the second left free for variation. In addition, fragments of code may be transferred from the cells of one species to those of another [.] by viruses or through other procedures.17 Deleuze and Guattari here offer a short and somewhat faithful, general idea of the multiple mechanisms of stochastic genetic variation discovered with the synthetic theory. These mechanisms can be divided into two basic types: mutations that create variants of code (alleles) and the recombinations that can combine favourable variants. But everything happens as if the bacterial prokaryotes and the domain of uni- or multicellular eukaryotes form nothing but two distinct exploratory paths of these two types of variation or margins of decoding. On one side, prokaryotes slightly experience more elevated and above all variable rates of mutation (SOS response which thwarts the effects of the mismatch repair system in critical moments18), but their mutations are essentially punctual (single base substitutions) and quite rarely favourable (lesser size of genomes). On the other side, the mutations of eukaryotes are slower and more constant, but they essentially are produced by the duplication of longer or shorter fragments of DNA, most notably genes, including all the chromosomes during the course of reproduction (polyploidy); these duplications are such that they ballast the genomes of an increasing non-coding part and therefore augment the probability of neutral or favourable variations that have been able to open the most diversified path, for then we shall witness the “deterritorializable” path of multicellularity. From that moment on, where recombinations are made simply in bacteria through the horizontal transfer of one species to another (via plasmids, circular fragments of DNA or viruses), eukaryotes have invented another path of genetic recombination, the sexual path, which assures a greater variability and thus a greater adaptability within the same species but also induces an inevitable genetic drift, i.e. a random and non-optimizing selection of these variations. Must we conclude that genotypic variation is the only source of random variation of the living? It is important never to forget that this perpetual genetic shuffling wouldn’t have more value than a simple molecular disorder if there were not carried out in parallel, at the other end of the spectrum, a systematic selection of favoured phenotypes and thus a constant optimization of mutations: “milieus always act through selection on entire organisms, whose forms depend on the codes that these milieus indirectly sanction19”. The reproductive success of each organism therefore will depend on its adaptation to the milieu, i.e. on its capacity to effectively utilize the available nutritive and energetic resources compared to other organisms. The fact that this adaptation is the mechanical outcome of selection does not prevent it from also bringing with it an important margin of indetermination. The external milieu of organisms, including their “organic and inorganic conditions of life20”, in fact itself always tends to vary, whether concomitantly with or independent of the actions of the latter: climatic variations, the arrival of new predators, evolution of prey, scarcity of resources, etc. All these variations intervene in the reproductive success of this or that individual or population, and therefore no adaptation is acquired once and for all. Adapting in fact always returns, according to Deleuze and Guattari, to “being deterritorialized”: “it is populations that are deterritorialized and reterritorialized” while “they are coded and decoded”, and if “deterritorializations and reterritorializations do not bring about the modifications” of code, they “strictly determine their selection21”. Ecosystems can therefore be seen as the continually changing territorial spinoffs of this great vector of deterritorialization that has borne all the phylogenetic lineages and specifically assured the maximal phenotypic diversification of multicellular organisms. To deterrorialize consists in conquering, first and above all else, new sources of energy through the extension of types of respiration (this is specifically the main success of Archaea e between bacteria and eukaryotes e which can only live in the most extreme milieus), but also and correlatively new territories through the acquisition of a partial or total mobility (this is the great success of multicellular eukaryotes), since it is a question of the passive and aleatory dissemination of only the reproductive elements (case of vegetal life) or the active, but also aleatory, displacement of the entire organism (case of animal life): “[t] he more internal milieus an organism has on its own stratum, assuring its autonomy and bringing it into a set of aleatory relations with the exterior, the more deterritorialized it is22”. Thus multicellular organisms attain this maximal point of deterritorialization/ reterritorialization by becoming capable, solely based on chance disseminations and migrations, of traversing, i.e. of occupying and/or travelling throughout all the territories of the Earth favourable to life. 2.2. The third source of intrinsic variation: ontophylogenetic unification Between the margin of decoding inherent to the code and the vector of deterritorialization inherent to organisms, is there not, however, any other source of random variation within the organic stratum? Wouldn’t variation simply intervene upstream from the replicated codes or downstream from the produced forms? Wouldn’t a phenotypic form be the programmed product e i.e. simultaneously determined and finalized e of a given genotype, through the intermediary of the proteins that the latter codes and that compose and allow all the cells making up the organism to function? Ontogenesis would then be the only stage of the living that would result from a supposed genetic program and not from the double Darwinian principle of intrinsic variation/extrinsic selection. But, as Jean-Jacques Kupiec has already insisted, sources of random variation have been discovered on every stage of ontogenesis23: above all else, genes are expressed in a stochastic way within each cell according to the frequency of encounters with regulative proteins, the topological and temporal dynamics of chromatin, etc.; moreover, molecular and proteinic interactions form a veritable non-specific rhizomatic network24 where each molecule can potentially participate in any transformation whatsoever; consequently, each cell follows paths of differentiation, which are themselves stochastic, without obeying any inductive molecular signal; and ultimately, all the cells of multicellular organisms conserve a stochastic possibility of autonomous variation within the organism in its growth, maintenance and 21 A Thousand Plateaus, p. 54. Ibid., p. 53e54. 23 Kupiec, 2008, pp. 47e66. 24 Rhizome is a concept that Deleuze and Guattari take up from botany mainly to describe multiplicities that are not reducible to unity and more specifically to describe non-arborescent networks, see “Introduction: Rhizome” in: A Thousand Plateaus, pp. 3e25. 22 17 Deleuze and Guattari, A Thousand Plateaus, p. 53. See the whole work of Miroslav Radman (notably Radman, 1975; Radman et al., 1995, 2003). 19 A Thousand Plateaus, p. 52. 20 OS, 1st ed., 1859, IV, p. 84. 18 Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 reproduction in which they contribute to the unique fact that its internal milieu itself contributes to their subsistence. Another fact must be added to these variational facts that are irreducible to simple “noise”: the apparent reproducible stability of the products of ontogenesis. Kupiec concludes from this that only the Darwinian paradigm can account for this double aspect and that it is thus applied to both ontogenesis and phylogenesis, including all the stages and scales of the living. Before analysing the concrete consequences of this ontophylogenetic unification and showing that the logic of the living’s stabilization that it envisions unites with the logic of organic stratification described by Deleuze and Guattari, let us attempt to further define the nature of random variation beyond and throughout all the figures that it takes within the organic stratum. As we mentioned, the “Darwin-Hooker principle” primarily affirms the intrinsic nature of variation. But it is actually upon this intrinsic nature that its random character depends, and not the other way around. In effect, intrinsic variation must not simply be understood in a negative sense, i.e. as everything that is not determined extrinsically. The force of the Darwinian paradigm is precisely to offer a positive definition of this mode of determination: an intrinsic variation is not in itself without determination, it is simply determined or caused by another mechanism than that which determines or limits it; in truth it is not undetermined but doubly determined. The limiting mechanism e selection e will therefore be called extrinsic in relation to the causing mechanism e the source of variability. It is this disproportion between cause and effects, this irreducible gap between two determining orders e for example, the molecular transformations of the genetic code and the adaptation to the external milieu that its phenotypic expression requires, or even the adaptability with which an organism is equipped and the effective variations of the milieu into which it is plunged e that leads us to qualify an intrinsic variation as random and makes us wrongly believe that it would thus neither have causes nor conditions. If random variation and causal indetermination are often confused de facto, it is because of another major confusion that manages to graft itself onto it and sustain it. Envisioned a priori or abstractly, the variation intrinsic to the living is necessarily unlimited and undetermined e or the “Darwin-Hooker principle” would be violated and external limits or internal determinations to the variation reintroduced, tending to explain it and restrain the explanatory power of natural selection.25 But when we concretely envision it within its various fields of intervention (genetic shuffling, ontogenetic variability, phylogenetic deterritorialization), intrinsic variation always seems determined by conditions (genomes, cells and territories) and limited in its possibilities e which fully open themselves for each new drawing according to the preceding drawing, as Deleuze and Guattari describe them starting with Anti-Oedipus by utilizing the image of Markov chains as “partially dependent, aleatory phenomena” or a “succession of partial re-linkings26”. 25 Darwin thus refuses any a priori limitation of variation: “the ordinary belief that the amount of possible variation is a strictly limited quantity is likewise a simple assumption”, OS, 3rd ed., 1861, IV, p. 89. Hoquet (2009) has clearly revealed the stakes of this non-limitation/indetermination in order to preserve the integrity of the explanatory power of natural selection: “On the one hand, if variation is neither infinite nor unlimited, this reduces the scope of natural selection: it is revealed to be confined within certain boundaries. On the other hand, if variation is not random but follows determined directions, this reduces the latitude of natural selection and seems to channel it along certain directions”, pp. 183e184. 26 Deleuze and Guattari, Anti-Oedipus, p. 289 and 343 and Deleuze, Foucault, p. 71 and 96. Markov chains are used today specifically in bio-informatics in order to model the relations between the successive symbols of the genetic code and thus to distinguish between coding and non-coding genes, specifically in eukaryotes. 5 2.3. Random variation, contingent variation and pure variation However, taken as a guiding thread and a theoretical imperative, the “Darwin-Hooker principle” prohibits us from remaining with this sole intrinsic random determination of variation and obliges us to extract it from the figure that it takes within the living or the organic stratum: in effect, is it not, in this case, always already constrained and selected, thus always already conditioned and limited, in short always already partially explainable? Doesn’t affirming that everything that is varies in itself, before receiving some invariance or identity e whatever it be e lead to thinking a variation outside all constraint and condition, i.e. absolutely outside every stratum, a variation that precedes every real experience that there can be of it, so to speak a pure variation? Quentin Meillassoux has recently shown the necessity for every speculative enterprise seeking the absolute or the in-itself of things to take as its point of departure a “principle of factuality” somewhat equivalent to what we have been calling the “Darwin-Hooker principle”. This principle of Humean obedience affirms that nothing can be necessary except the single fact that there is contingency.27 The fact of contingency therefore plays for Meillassoux the same absolute, inexplicable and irreducible role as the fact of variation for Darwin and Hooker or Deleuze and Guattari. The absolute primacy implicitly granted by Darwinian theory to variation led us in a sense to the threshold of such a speculative conclusion by inviting us to consider as always necessary both the fact that it varies and the fact that it does not vary in a necessary manner (i.e. explicably through any sort of determinism or finality). But the fact that it is simply necessary that there be the advent of variation without necessity could signify either that the latter is random or that it is, more simply and more radically, contingent. A random variation is in effect nothing but a contingent variation already submitted to conditions, i.e. able to be assimilated into a simple drawing among a defined set e be it infinite e of combinational and spatio-temporal possibilities that are equivalent or even into a fortuitous linkage of causal series or orders of independent determinants. On the contrary, a radically contingent variation governs every condition, establishing a regime where “[t]here is no reason for anything to be or to remain the way it is; everything must, without reason, be able not to be and/or be able to be other than it is28”, no longer a multi-causal regime of equivalent possibilities and fortuitous encounters, but instead an acausal regime of unforeseeable possibilities and incessant ruptures, which Meillassoux calls “hyper-chaos”. Purifying variation of all its necessarily limiting empirical baggage under which it always already presents itself leads us therefore to think a more radical contingency, even more unconditional than chance alone. In their last work written together What Is Philosophy?, Deleuze and Guattari carry out a complementary purification of the character e this time intrinsic e of variation. The undoing of every extrinsic determinant boils down to characterizing it according to three correlated components. Primarily, a purely intrinsic variation does not predicate something; it is neither the variation of a subject nor a substance, for it is the sole subject and sole substantiality,29 the variation without beginning or end of which all invariances and finite things are merely ever provisional ends. Secondly, a purely intrinsic variation does not unfurl itself into something; it is not reducible to a movement in space or even to a duration in time, for it subtends all space-time, 27 Quentin Meillassoux, 2006, pp. 82e111. Ibid., p. 60. 29 Bergson was the first to defend this unique substantiality of variation in “The Perception of Change”, conference lecture reprinted in: The Creative Mind, pp. 107e132. 28 Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 6 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 a variation in itself whose variable number of dimensions or degrees of freedom, including the continuity of succession and distance of separation, are nothing but secondary consequences30. Thirdly e and this is the central point of the Deleuzo-Guattarian perspective e a purely intrinsic variation is nothing but a single rhythm; it is emphasized by no other regularity than the fact that it never stops, an incessant variation comprised of all the rhythms or intensive degrees of speed of variation, from the infinite speed through which the authors define “chaos”31 in their own way, down to the slowest speeds through which “stratification” variously operates. “Stratification is like the creation of the world from chaos, a continual, renewed creation32”: we must now fully deploy the logic of this ongoing creation that is differentiated by limitation or constraint. 3. Nothing really varies incessantly: physical subsistence and organic reproduction 3.1. Physical stratification: constants, invariances and material durations Governed by the principle of unreason e which rejects any principle of reason, whether reformulated or not e (but also governed by an inexplicable presupposed time of successive contingency), hyper-chaos as defined by Meillassoux forces us to ask de facto why everything and anything whatsoever does not take place each instant. Characterized only by the infinite speed of variation (in which each time of succession is merely one of the derivations, a continuous and uniform derivative), chaos as defined by Deleuze and Guattari first forces us to ask de facto how anything whatsoever can subsist and have a more or less stable duration without incessantly varying in all directions and thus disappearing immediately as such. In fact, the infinite speed of variation, the “infinite speed of birth and disappearance” of everything that can be, can really only give rise to nothing or the infinite void, i.e. an absolute instability or inconsistency: “chaos makes chaotic and undoes every consistency in the infinite33”. The fact that at least one universe, namely ours, would be created on the basis of this “nothing” thus signifies that a relative consistency is provisionally subtracted from this absolute inconsistency. This initial means of taking on consistency within chaos that defines the object of physics, i.e. also the physical stratum, calls for a selectionnist or expanded Darwinian view: both what determines the genesis of the universe itself (cosmogony) and that of which it is composed (cosmology) in effect benefit by being seen as the result of a series of constraints exerted upon and on behalf of pure variation. But since nothing escapes pure 30 The variation or movement of the infinite “does not refer to spatiotemporal coordinates that define the successive positions of a moving object and the fixed reference points in relation to which these positions vary”, Deleuze and Guattari, What Is Philosophy?, p. 37. For a rigorous reconstruction of the mathematical models (intrinsic and n-dimensional Riemannian geometry, phase space, etc.) that underlie or at least inspire this aspect of their fundamental ontology already present in the early Deleuze, see the now classic study by DeLanda (2002). 31 “Chaos is defined not so much by its disorder as by the infinite speed with which every form taking shape in it vanishes. It is a void that is not a nothingness but a virtual, containing all possible particles and drawing out all possible forms, which spring up only to disappear immediately, without consistency or reference, without consequence. Chaos is an infinite speed of birth and disappearance”, What Is Philosophy?, p. 118. For a detailed analysis of this latter fundamental ontological aspect, of which the principles of superposition and indetermination in quantum physics give a general scientific idea, cf. our work, Rosanvallon, 2009, pp. 93e105. 32 A Thousand Plateaus, p. 502. 33 What Is Philosophy?, p. 42. variation, or in other words since nothing is exterior to it, how could it initially be constrained by something besides itself? These constraints are therefore produced solely from the fact that pure variation, which is in itself infinitely infinite, can only intersect itself: the infinite cutting the infinite creating some breaks that will play the role of cutoffs or limits.34 These constraints in fact produce three distinct limitations that tend to cumulate and conjugate together: the infinity of variables or degrees of freedom of variation, the infinity of values taken by these variables or ultimately the infinity of speed according to which these variables vary could successively be limited. This triple limitation then makes possible the fact that the “finite”, as partial as it may be, i.e. for whatever the infinity it conserves (infinite space-time, perpetual chaotic inflation, superposition of quantum states, space of mathematical possibilities35), manages to take on consistency within chaos or pure variation, which is in itself neither exhaustible nor limitable and never stops remaining infinitely infinite.36 But such constraints do not carry out a selection of universe (“cosmogonic” selection) without first carrying out a selection of these constraints themselves (“physical” selection). Since pure variation varies in a purely contingent way, the breaks, which are its effect, are just as contingent; but what is not contingent is the effect of these breaks, namely what, with them, acquires not simply a form of consistency but also subsistence amidst everything that comes to exist. This is how the breaks or limits that produce a “physically” subsistent and not merely “mathematically” consistent reality will be selected: in the same way that what the organic stratum selects (and what forms it) gauges its reproductive success, above all what the physical stratum selects (and what forms it) gauges its degree of subsistence. But which forms of relative subsistence does this process of the self-limitation of pure variation specifically produce on the basis of its absolute inconsistency? And how to gauge this subsistence without intervening with a prior time, which would come more or less provisionally to occupy or fulfil a certain kind of consistency (any sort of multiverse, universe or physical process whatsoever)? From a Deleuzo-Guattarian perspective, time envisioned formally does not at all constitute the prior form of any subsistence but simply the belated product of a multiplicity of subsistences, i.e. the simple offspring of a correlation of variables establishing both a continuous and uniform variable in step with all the other rhythms or durations of variation.37 The primordial forms of subsistence therefore in no way constitute temporal forms but simply forms of infinite non-variation, the partial cessation of infinitely infinite variation. These partial cessations can be regrouped into three broad overlapping categories. First, a definitive subsistence is acquired by what no longer varies, i.e. the breaks or limits productive of physical reality that therefore become constants not only of a universe, but also inevitably of an infinite 34 The so-called technique of renormalization in quantum field theory, which introduces an arbitrary cutoff between the infinity of field oscillations in order to be able to subtract the infinity of the excited field (with particles) from the thus determined infinity of the empty field (without particles, unless they are virtual) so as to end up with finite values, can offer a general idea of the genesis and functioning of what such an intersection and break of the infinite by the infinite would be. 35 These products of pure variation are the four levels or sources of staggered multiverses proposed by Tegmark (2007). 36 “It is not the limited thing that sets a limit to the infinite but the limit that makes possible a limited thing”, What Is Philosophy?, op. cit., p. 120. 37 Rovelli (2008) has thus shown the necessity and possibility of formulating a classical and quantum mechanics without time, which is nothing but the dependent product of a correlation of variables and not the independent frame for the evolution of variables. Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 population of distinct universes.38 Second, a relative subsistence is acquired by what varies within these limits that border the infinity of variables and/or values, thus forming, when these variations can all be constituted and invert themselves, groups of transformations (whether discrete or continuous and/or global or local) that compose a number of symmetry groups or groups of invariances. In effect, each group defines a double invariance in relation to all its variations or transformations: on the one hand, it renders them equivalent either as descriptions or as possible states of that which varies, following the example of the principle of relativity which in truth is only the translation of a spatio-temporal or gravitational symmetry39; on the other hand, it implies that they conserve or at least leave invariant a quantity or an aspect of that which varies (energy, charge, isospin, etc.). Third, a more or less provisional subsistence is ultimately acquired by everything that varies more slowly than at infinite speed, thus participating in this “primordial slowing down” from whence all the superposed material durations that compose a universe like ours arise: the expansion of space, interaction of elementary particles and atoms, galaxies, stars, planets and molecules that are basically compounds of the elementary, stabilized and variable at the same time40. The Deleuzo-Guattarian perspective prompts us to consider each of these durations as the outcome of a stratification of continually renewed chaotic variation. The entirety of what has already been selected (constants, interactions and other persistent physical processes) in fact plays the role of stratum that will constrain, capture and give durable consistency to what never stops varying: the variable duration of stars results, for example, only from the combinational play of the four fundamental interactions. The (so-called spontaneous) symmetry breakings described by contemporary physics can be better understood as the crucial results of such a logic of stratification. A primary symmetry breaking would be the breaking of the perfect cosmological principle into a simple cosmological principle that conserves the 38 The three so-called fundamental constants, since they are neither the product of any theory nor any other constants (the Planck constant h or minimum of variation, the constant c or maximal speed and the constant of gravitation G), therefore do not determine the physics of our universe and the physical theories that describe it without also being at the heart of an infinity of other universes as a consequence of these same theories (for example, the theory of eternal chaotic inflation specifically has as a preliminary condition the structuration of space-time operated by c e this limit-speed is in fact in no way violated by the inflation that does not define a speed of propagation in space but a speed of expansion of space itself, i.e. a speed of intrinsic variation which would indeed also be initially always already infinite). 39 The generalized covariance induced by general relativity can in fact be interpreted in terms of a group of diffeomorphisms or the “supple group” (see the work of Jean-Marie Souriau specifically Souriau, 2002) that includes the subset of the Poincaré group of special relativity and determines the gravitational field, i.e. spacetime itself. Three other continuous symmetries e which are called gauge invariances, this time internal e correspond with three fields of interaction or fundamental types of variation other than gravitation (weak, strong and electromagnetic interaction): they do not make equivalent the spatiotemporal descriptions or becomings of physical objects, which are already determined, but determine the presence and the becoming itself of fundamental physical objects (bosons or particles of interaction) as conditions of their invariance. Ultimately these symmetries include the discrete symmetries of charge, parity and time (CPT) from which, on the one hand, anti-matter and, on the other hand, the “perfect cosmological principle” (isotropy of space and time) are deduced. 40 “It is by slowing down that matter.actualizes itself”, What Is Philosophy?, p. 118 (“primordial slowing down” is employed by Guattari, 1992, p. 112). If cosmogenesis is conveniently presented as the history of a linear chronology, it in fact more profoundly constitutes the history of a temporal slowing down (and not simply a spatial inflation), i.e. from a production and superposition of durations on the order of 109 years (duration of the universe’s current life) and even 1030 years (duration of the proton’s minimal life) to the primordial durations characteristic of Planck time (10 43 s), which marks the frontier of contemporary physics and the domain of quantum gravity still to be elaborated. 7 isotropy of space but implies the anisotropy of time, i.e. the entropic rupture of symmetry between past and future (the arrow of time): rather than deploying our universe on the basis of a singularity and a time zero (Big Bang theory), the expansion of observable space would instead be nothing but one anisotropic episode e among an infinity of others e of an eternal, i.e. perfectly isotropic, chaotic inflation.41 This dynamic of space expansion was at first inflationary before slowing down and then accelerating again after several billion years: described by the standard model of cosmology (Lambda-Cold Dark Matter model) but not theoretically understood, it also seems to result from a play of symmetrical forces between an attractive force of gravitation (matter and dark matter) and a repulsive force of antigravitation (cosmological constant Lambda or dark energy) slowly broken in favour of the latter42 in a way such that the universe has a duration of expansion sufficiently long enough to allow for the appearance of other durations, i.e. other processes of stratification or instances of its slowing down. Besides the primordial quantum fluctuations that explain the anisotropies of the cosmic microwave background and their post-inflationary residue e which are the great structures of the universe (galaxies clusters) e and besides the violation of CP within CPT symmetry that explains the disappearance of antimatter,43 this first breaking entails a second series of symmetry breakings e this time internal e that describe, albeit still without being able to explain, the standard model of particles. If inflation allows for the appearance of particles which are no longer virtual but now actual, i.e. susceptible to interaction and composition,44 the expansion that has slowed down, by cooling down the temperature of the universe, in other words, by reducing the amplitudes and 41 See Aguirre and Gratton, 2008; the model of eternal chaotic inflation has been furnished by Andréi Linde, who was the first to think inflation not as one episode of the Big Bang among others but the Big Bang as one episode of inflation among others (theory of bubble universes). 42 The symmetry between past and future and/or its breaking indeed seems to play a central role in this regard. Two types of completely different works, which are trying to deduce in a natural way e without the creation of an ad hoc field e the effect or value of the cosmological constant, attest to it: on the one hand, when temporal symmetry is fully integrated into general relativity, although Frédéric Henry-Couannier proposes to reformulate it albeit sadly in a way that is not background independent (Henry-Couannier, 2005), it naturally generates a flat and accelerating universe (Henry-Couannier et al., 2006); on the other hand and more profoundly, when on the contrary temporal dissymmetry introduced by material durations, i.e. by the necessarily cumulative character of past cones of light or causality, is formalized in a theory like that of causal sets (candidate for quantum gravity), the cosmological constant is equipped with a non-null value that always fluctuates in the midst of the value of material density, thus explaining this “cosmic coincidence” in a natural way (Sorkin et al., 2002). 43 Weak interaction in fact violates CP symmetry and compensates this violation with that of the symmetry or invariance through T in order to conserve CPT symmetry. This violation of T is translated by a dissymmetry between the duration of the life of certain particles (like neutral kaons, beautiful mesons, but also neutrinos) and the shortest life of their corresponding antiparticles. This dissymmetry creates a surplus of matter over antimatter that will thus be selected or, in other words, the complete primordial annihilation of matter and antimatter will be allowed to subsist as the residue from which all the matter of our world stems. 44 We call “virtual particles” the creations/annihilations of particles (or production of pairs of particle/antiparticle or fluctuations) of which the quantum void is the incessant site due to the relation of Heisenberg’s time/energy indetermination that renders the conservation of energy violable (and thus, due to the Einsteinian equivalence of mass and energy, creatable particles) on condition that this violation lasts a sufficiently brief amount of time, in such a manner that the product of this energetic indetermination through the interval of time must not be inferior to the reduced Planck constant (or quantum of action). This specifically signifies that there always subsist infinite energies in infinitely brief times. To make this energy durably consistent, to transform it into an actual particle, external energy must manage to repay this loan granted by nature: inflation would precisely constitute such a source of external energy; concerning this point, see Gunzig, 2008, pp. 176e217. Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 8 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 possibilities of variation, crosses certain thresholds of energy that tend to break and constrain symmetries, i.e. to successively disjoin and render autonomous strong, weak and electromagnetic interactions and thereby render particles of interaction e other than photons e and particles of matter massive themselves (chiral symmetry breaking and Higgs mechanism). Particles of matter thus become able to be composed into neutral atoms (primordial nucleosynthesis) which will themselves then be able to be fused into heavy atoms (stellar nucleosynthesis) and composed into metals or molecules (chemical bonds). Thus symmetry breakings do not result necessarily from a deterministic hypothetical theory but contingently from the pressure exerted by the “environmental” constraints that constitute the residues, which are also contingent and productive, of pure variation. What results from these breakings then in turn constrains, i.e. conditions and selects the ever persistent possibilities of variation. Therefore the expansion will have selected a certain physics of low energies which has itself selected the possibility of chemical bonds and reactions, opening them, among other reactions, to the immense combinational and reactive possibilities of carbon compounds (organic chemistry). 3.2. Organic stratification: dynamic kinetic stability, membranous selection and natural selection The phenomenon of the living is naturally grasped within this frame of thought: it does not emerge mysteriously from the physical world (theories of emergence), nor is there an ultimate adjustment of its parameters that makes it miraculously possible or at least “anthropically” probable (multiverse theories); it simply marks the capture and continuation of pure variation by other means. In effect, everything happens as if a new stratification would then lodge itself within physical subsistences and would manage to capture or subtract pure variation more than the cascade effects of the causal subsistences that constitute physical stratification. Everything happens as if stratification would give variation a more direct and more immediate consistency than simple subsistence, albeit without using other materials, other forces or energies than those resulting from physical stratification and the conditions that it has randomly established on a planet like ours. The constraints of the physical world are also necessarily those of the living world, but the latter has its own additional constraints that paradoxically allow it to break free from the first constraints or rather to provisionally suspend them by inventing another mode of organization capturing intrinsic variation more directly. Before analysing in detail this new mode of organization that alone legitimizes speaking of a new stratum, let us insist upon the specificity of the Deleuzo-Guattarian approach e a faithful generalization of the Darwinian schema e in relation to the “emergent” and/or “anthropic” frameworks currently utilized e specifically due to the lack of other existing theoretical frameworks. In the first place, guarding against “any ridiculous cosmic evolutionism” that would like to make one stratum emerge from another, for example the organic stratum from the physical stratum, or that would like to define a progression in complexity between these strata, Deleuze and Guattari ask us to never “question how something escapes from the strata, but instead how things enter it45”. As a pure differential process of selection without any directed evolution or presupposed progress, stratification consists in sifting a variation that is presented as external to it: each stratum captures, retains or subtracts whatever it can, thus 45 A Thousand Plateaus, p. 49 and 57. transforming stratification into a vast array of relative movements or limited variations within it; each stratum is distinguished by the sieve that it operates in such a way that, from one to another, there is a “change of organization, not augmentation46”; each stratum is therefore intrinsic, not to one another from which it would then not be distinguished, but to pure variation alone as the inexhaustible ground exhausting all that there is. In this way A Thousand Plateaus envisions the strata as “spinoffs, thickenings on a plane of consistency that is everywhere, always primary and always immanent47”. Stratification thus is resumed by a constraint, a selection or a sieve carried out, secondarily, as extrinsic and polyrhythmic, on an intrinsic variation e which is always primary e by nothing other than its products or “residues”. But, in the second place, the “anthropic” selection of variation allowed by the multiverse depends upon a calculation of probabilities which is, despite certain appearances, the exact opposite of such a Darwinian schema: the source of variability is then conceived as exclusively extrinsic to our universe, whereas the effect of selection by contrast would be purely intrinsic to it due to the “anthropic principle” as the imperative of observability of such a universe presumed to determine its conditions facing downstream.48 Only “natural cosmological selection” responds to the double Darwinian requirement; this type of selection has been proposed by Lee Smolin for whom the actual values of parameters would simply be those that maximize the production of black holes, i.e. “baby universes”, since each birth within a black hole would be brought about after a slight stochastic variation of values already selected from the parent universe.49 But this selection also supposes that variation would have taken place only upstream from a cosmogenesis that remains deterministic in the same way as ontogenesis within NeoDarwinian optics e as if the distinct universe or individual organism were both completely determined by a legal or genetic program whose sole preliminary definition would be partially stochastic. A theory of “grand Darwinian unification” can really be nothing but a “cosmo-onto-phylogenesis” that introduces, on all the stages of formation of the physical world or living world, simultaneously intrinsic and stochastic possibilities of variation and extrinsic and necessary constraints of optimization. If the physical stratum is constituted by eliminating possibilities that lack a durable causal effect and by optimizing cascade effects of subsistence (the value of each subsistence being gauged by the number of variations and 46 Ibid., p. 49. Ibid., p. 70. “Plane of consistency” (which, as these two authors point out, more properly speaking constitutes a “plane of inconsistency”) in A Thousand Plateaus would be equivalent to the idea of pure variation, which will only be implicitly theorized as such in What Is Philosophy? correlative with the idea of chaos. 48 Aguirre (2007) perfectly shows that the “standard” framework of the multiverse has no other scientific stakes than that of calculating a distribution of probabilities such that it “would explain”, i.e. would make less improbable, the known values of the parameters of standard models, a number of which would seem so suitably adjusted that their variation in the slightest would have produced a universe unfit for life: so as to be calculated, these probabilities should therefore be balanced or conditioned by an “anthropic principle”, and this conditioning should be made typical or not improbable by a complementary “principle of mediocrity”. Aguirre demonstrates that reflection on the unavoidable steps of such a calculation (for example, the determination of the more or less fundamental character of the parameters and their order of precedence) is often neglected on behalf of “shortcuts” that often lead to making a masked evolutionist and even finalist usage of the anthropic principle, which is supposed in order to then determine the direction taken by our universe from its origin. 49 See specifically Smolin, 1999, along with his article (Smolin, 2007). Nevertheless, this theory has the major flaw of projecting onto the physical stratum the principle of reproduction proper to the organic stratum without being able to translate this modifiable memory into a hereditary material equivalent to the genetic code. 47 Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 thus of potential subsistences that it conserves)50, how is the organic stratum constituted, what will it eliminate and therefore tend to optimize? Although not without certain theoretical shortcuts, we can consider that, within the physical stratum, potential energy is always minimized, i.e. minimally “acquired” and maximally “expended” (principle of least action): every system tends to attain its state of maximal stability or, in not so rigorous terms derived from statistical physics, of maximal kinetic “disorder” (increasing entropy or second principle of thermodynamics). But biochemical processes are called “far from equilibrium” because they doubly manage to free themselves from this thermodynamic stability. Primarily, it seems that they simply suspend this stability: the carbon compounds that all these processes make intervene are specifically characterized by their metastability, i.e. their intrinsically delayed expenditure of potential energy. It is in this sense that Gilbert Simondon proposed to define the vital as “the physical in suspense, slowed down in its process and indefinitely expanded51”. This does not at all signify however that the living varies more slowly because this slowing down together is specifically obtained only due to the partial acceleration of the variations or transformations within it, i.e. due to the energetic degradations (catabolisms) and material syntheses (anabolisms) that the living carries out in order to conserve itself in time and grow in space: the organic metabolism of the cell, the elementary biochemical system for all other systems, is in fact itself founded upon the action of catalysts (enzymes) that accelerate chemical reactions by reducing their activation energy. Thus uni- or multicellular organisms constitute fundamentally open systems that use the energy of solar rays or terrestrial chemical reactions (phototrophs or chemotrophs) and establish themselves on the basis of inorganic or organic matter (autotrophs or heterotrophs) in order to conserve a local negentropy within the global entropy of the Sun-Earth system. But if the organic stratum then remains far from thermodynamic equilibrium, this would be secondarily, because it responds first to another equilibrium that would be induced by the foundational element around which it is organized, including the chemically acquired capacity e on the molecular scale e no longer simply to subsist but to be reproduced in all senses of the word and in every way possible. What in this case tends toward equilibrium is the difference between the rates of reproduction of what is produced and the elimination rates of what is thus variably reproduced, in such a manner that a dynamic kinetic stability e according to what the chemist Addy Pross calls it e is created, namely a stability of the population of entities (first molecular, then cellular and multicellular) with the capacity to 50 Within this “cosmo-onto-phylogenetic” optics, the presumed goal of adjustment of the parameters becomes tautological: since the parameters’ values have been duly selected amongst all possibilities in virtue of the causal effects of subsistence and variation in their core which they successively or correlatively produce, there is no surprise that they seem so optimal for producing such effects! In this sense, a parameter would be confused with its concrete instantiation, with all of its effects, without preceding them anywhere else but in theory. 51 Simondon, 1995, p. 150. This intuition is taken up again by Deleuze and Guattari, for whom the plastic properties of carbon carry out a “bifurcation” that will “search the infinite chaos of the virtual for new forms to actualize by carrying out a sort of potentialization of matter” What Is Philosophy?, p. 123 (my emphasis). But the first to inspire such an intuition is actually Bergson, for whom, after encountering the caricatured interpretation which is often foisted upon the concept of élan vital, life “has not the power to reverse the direction of physical changes” determined by the increase in entropy: “Incapable of stopping the course of material changes downwards, it succeeds in retarding it”, Creative Evolution, pp. 245e246 (my emphasis). 9 multiply, specifically due to the perpetual replenishment and variation of these same entities that make up the population.52 This equilibrium translates a new stratification, a fundamental change of organization according to which what subsists is not only what tends to no longer react or react less (thermodynamic equilibrium of the physico-chemical stratum with which every dead organism also tends to unite), but what is all the more capable of reacting and participating in reactions as well as reproducing it and, in this sense, being reproduced (dynamic kinetic stability of the chemico-organic stratum). What does “to be reproduced” mean? Life fundamentally derives its origin from a process of the reduction and canalization of a physico-chemical variation that takes form both in the (essentially cosmic) prebiotic soup of elementary organic building blocks and in the (essentially geochemical) energetic cycles based on the transfers of proton and electron: on the basis of only a part of all these possible building blocks and cycles, little by little organic stratification has constructed and developed the constitutive materials and energetic transfers that belong only to the living. If the imperative of reproduction has indeed played a role of selective constraint carrying out this productive canalization, this imperative is realizable in three irreducible and thus originally distinct ways: to be reproduced and to multiply passes through the fact either of being replicated (from RNA to DNA), or of reproducing its conditions of production (from “proto-metabolism” to metabolism53), or of growing to the point of being divided (from the permeable passive membrane to active cellular division). These three forms of organic reproducibility each concern a type of population whose mode of operation has entailed the rise of a specific molecular path: the populations of “replicators” (only ones able to be replicated identically) that function through the pairing of the bases of a linear chain have opened the nucleic path; populations of “catalyzers” (only ones able to reproduce conditions of production) that function through the foldings of a linear chain of amino acids have opened the proteinic path; and populations of “selectors” (only ones able to distinguish an outside filtered from a retained inside and divide themselves) that function through the closure of a dual layer of amphiphilic molecules have opened the phospholipidic path.54 It is therefore not a question of asking which of these three populations e which are by right reproducible and even reproductive e would have initiated organic stratification by preceding, or indeed generating the others, but to which encounters, correlative operations and durable couplings, leading each to their current molecular form, the imperative of reproduction, upon which this stratification depends, has constrained all three e thus designating a primary 52 Pross, 2011 (http://www.jsystchem.com/content/2/1/1#B50). As we shall see, the capacity for molecular replication upon which Addy Pross explicitly applies his reasoning seems to us a certainly fundamental, but merely particular case of a more general logic of reproduction which is constitutive of the organic stratum. 53 The idea of “proto-metabolism” has been proposed and explored in detail by de Duve, 2005, pp. 15e24. The reproduction of the conditions of production does not signify that the living completely auto-produces itself (according to the idea of “autopoiesis” developed by Maturana and Varela (1979)) since it cannot do so without external materials or energy, but signifies that, in their presence, what it produces manages to never stop reproducing itself. 54 “In actual cells, membranes are never born de novo; they grow by accretion, i.e. by the insertion of new molecules in a pre-existent network. Thus membranes arise from pre-existent membranes linked by an uninterrupted filiation with an ancestral membrane which could harken back to the first days of life on the Earth. This genre of development leaves to the imagination an evolutionary process in which the original components of membranes would have disappeared after a long time so as to be progressively replaced by more elaborate molecules to ultimately arrive at the phospholipids and other complex constituents of today”, de Duve, 2005, pp. 131e132. Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 10 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 series of (molecular) synergies or symbiogenesis of which unicellular living beings are the ultimate product.55 The renowned coupling between replicators and catalyzers appears in all current living beings as inevitably circular: in a sense, it gives rise to genetic code, i.e. the coded production in DNA of the proteins that reproduce the conditions of production for the living, and, in an opposite sense, to proteinic enzymes as necessary vectors of the activation, acceleration and precision both of the transcription and translation of coding replicators and already of their replication itself. On the basis of chemically and thus randomly produced chains of amino acids and nucleotides entering into cross-catalytic56 networks, the optimization of such a reproductive coupling has undoubtedly taken millions of years: if RNA is evidently the precursor of DNA, through which the dissociation of the function of replication and functions of transcription becomes possible, it remains to identify the precursors of RNA itself. The fact that the residues of the vectors of organic metabolism, i.e. its fundamental energetic currency (ATP but also GTP, CTP, UTP) are also the basic constituents of RNA (AMP, GMP, CMP, UMP) in each case proves the ancient status of this first coupling. The coupling, which is also completely fundamental and circular, between productive catalyzers and membranous selectors, in a sense, for actual translation has the metabolic synthesis of the lipidic constituents of cellular membranes and, in an opposite sense, the proton-motive force based on the electrochemical gradients maintained by the membrane that allows for the permanent recreation of ATP. Also, the extremely long selective history of this coupling could be described as the passage from a passively obtained reproduction of the necessarily selective conditions of production of a reproducible cycle (proto-cellular protometabolism) to the randomly active and progressively acquired reproduction of each of these conditions, up to the function of membranous selection itself, whose growth no longer becomes possible but produced (cellular metabolism).57 However, this passage can only be accomplished through the coupling e which is itself optimized e of the first two, from which therefore the just as 55 Beyond the badly posed debate between advocates of the primacy of replication, metabolism or the membrane which thus obey the three logics of reproduction irreducible to one another, scientists are forced to identify entities that would all alone from the start have been able to fulfil at least two of these three functions. Thus, for example, is grasped the whole interest of the theory of mineral surfaces (proposed by Günter Wächterhäuser in 1988, see Wächterhäuser, 1988) that simultaneously combine selective functions (due to their bi-dimensionality) and catalytic functions (due to their electric charge) or the idea of an RNA world (proposed by Walter Gilbert in 1986, see Gilbert, 1986) that endorses replicative and catalytic functions (according to the current model of ribozymes). However, the idea that irreducibly distinct logics of reproduction would be secondarily linked and that their functioning in symbiosis would have been itself selected among other possibilities, and thus constantly optimized, is imposed by the inevitable coupling of the three forms of reproduction that no molecular entity can assume alone. 56 On the random chemical production of peptides or “multimers”, see de Duve, 2005, pp. 15e24; on the cross-catalytic network or “hypercycle”, see Eigen and Schuster, 1979 and Pross, 2011. 57 Through this movement of internalization of its conditions of production (the first cellular membranes no doubt having survived only by inventing autotrophy and no longer simply heterotrophy, i.e. the direct degradation of inorganic matter once the organic components of the cosmic prebiotic soup have been exhausted), the organic stratum then globally seems to auto-maintain itself, the majority of conditions apparently external to membranes in truth remaining internal to it, for they are completely transformed by its increasing reproduction: “the materials furnished by the substrata are an exterior medium constituting the famous prebiotic soup, and catalysts play the role of seed in the formation of interior substantial elements or even compounds. These elements and compounds both appropriate materials and exteriorize themselves through replication, even in the conditions of the primordial soup itself. Once again, interior and exterior exchange places, and both are interior to the organic stratum. The limit between them is the membrane that regulates the exchanges and transformations in organization”, A Thousand Plateaus, pp. 49e50 (my emphasis). decisive coupling between replicators and selectors also results. The importance of this latter coupling is translated by the fact that a third of the genome of today’s uni- and multicellular organisms would code for the proteins inserted into the dual lipidic layer that makes the membrane multifunctional,58 and in an opposite sense, that another membrane in eukaryotes surrounds DNA (cellular nucleus), this time dissociating the site of transcription and the sites of translation. Even so, during the lengthy course of primordial phylogenesis, only the membranous containers that each time receive a selective advantage for their replicative and productive contents have been able to grow and divide in order to attain the first great bifurcation between bacteria, Archaea and eukaryotes. 3.3. The organism and the body without organs Once cellular reproduction takes on dynamic kinetic stability in a selectively optimized way, two other great ensembles correlative with selectable symbiotic possibilities arose during nearly three billion years: “If evolution includes any veritable becomings, it is in the domain of symbioses that bring into play beings of totally different scales and kingdoms59”. A second series of (unicellular) symbiogenesis traces the decisive dividing line between prokaryotes and eukaryotes. In the first case, living in symbiosis is limited to molecular diffusions (“quorum sensing”) and the random, instantaneous and genetic (and perhaps even proteinic) transfers between growing populations of cells that increasingly reinforce their autonomy: their phospholipidic membrane is doubled by a peptidic cellular barrier and the characteristics of this membranous foliation end up diverging, thus giving unicellular entities a mesophilic tendency (bacteria) or extremophilic tendency (Archaea). On the contrary, in the second case, living in symbiosis opens into an “endosymbiosis in series60” between Archaean and bacterial cells, of which eukaryotic cells would be the product that is ultimately selected and thus optimized: methanogenic Archaea would have benefited from the hydrogen and carbon dioxide produced by bacteria to the point of completely surrounding them to become their single nourishing milieu and to link up with their genetic material and their “respiratory” capacities so as to transform them into simple organelles delegated to the production and supply of energy for the “host” cell (mitochondria and chloroplasts, the latter opening the royal phototropic path for vegetal life).61 Even a unicellular eukaryote (protists) de facto already constitutes in itself a veritable organism with differentiated functions, beginning with the function of nutrition or molecular transfer made possible by the membrane’s vesiculation (endocytosis and exocytosis). A third series of (multicellular) symbiogenesis, which relies upon these various forms of unicellular reproduction, inevitably follows from this in a wide variety of eukaryotes (and several rare bacteria) e since multicellular reproduction in a sense has been invented on various occasions within distinct cellular lineages. But this new reproduction cannot occur without somewhat reactivating and transforming each of these three logics of molecular reproduction that will become the three inseparable poles for all unicellular reproduction. First, the replicative pole gives rise, already within unicellular eukaryotes, to a new path of reproduction e sexual reproduction, where a cell no longer divides simply by replicating its genetic material identically (mitosis and asexual reproduction) but by also randomly dividing this material (meiosis and formation of gametes) so as to then have the ability to recombine it with other 58 59 60 61 Arkin and Stevens, 2000. A Thousand Plateaus, p. 238. Margulis, 1981. Martin and Müller, 1998. Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 material (fertilization and formation of a zygote). Second, the membranous pole opens what could be called a new path of cellular adhesion, where, due to reasons simultaneously internal (genetic mutations) and/or external (environmental constraints), cells that are nevertheless divided aggregate together in a terrestrial milieu or remain non-separated in an aquatic milieu so as to form a single organism that can grow and reproduce e at least partly and ultimately fully e and consequently be submitted to the optimizing action of natural selection.62 Lastly, the metabolic pole on this basis authorizes and even favours a new path of division of cellular labor that involves the differentiation of the organism into multiple cellular types that can ultimately form tissues and distinct organs. It is with these two last poles that ontophylogenesis intervenes, conceiving the differentiated production of organisms (ontogenesis) and the divergent production of phyla and species (phylogenesis) as the same unique process governed by natural selection. In effect, no multicellular organism develops and grows by itself without exterior nutritive supplies e whether they first be partially stocked, notably in the zygote phase (seed, embryo, etc.). Flows of matter and energy are indispensable in order for each organism to develop and endure throughout the same process of division, differentiation and thus cellular renewal on the basis of (animal) stem cells or (vegetal) meristematic cells that themselves issue from zygotes. Then how could these flows constitutive of the exterior milieu not have a selective effect before the development of the fully grown organism? How could the phenotype of an organism be submitted to natural selection without the phenotype of each of the cells that compose it and on the basis of which it develops not be submitted as well? In fact, natural selection or organic stratification presents itself as a two-faced Janus that operates, upon an intrinsic variability, a double constraint that produces each cell of the organism in a tendentially stabilized and optimized way. On the one hand, a certain cellular phenotype is selected, i.e. favoured, by the immediate environment to which each cell must adapt itself. This cellular environment is primarily determined by cells nearby, including the extra-cellular, tissual milieu and ultimately the rest of the organism, forming so many metabolic filters of the exterior milieu, which tend to produce variously concentrated gradients of factors of growth or regulation, i.e. “signal-foods63”. Whether the organism, i.e. all the cellular microenvironments, remains e notably in plants e dependent upon variations of the exterior milieu or it breaks free from them by continually compensating them for enabling constant conditions of cellular life, thus constituting e in warm blooded animals e a veritable “interior milieu64”, it no less offers environments that are variable according to its phases of life and its constitutive parts and therefore tends to favour differentiation, specialization and consequently the growing interdependence of cellular tissues and organic systems (digestive, respiratory, circu- 62 Bonner, 1998. 63 Formulated by Atamas (1996), the idea of “signal-food” has been taken back up by Jean-Jacques Kupiec in order to think the fact that, even present in small quantities, all intercellular signals (including hormones) function through the metabolic model, i.e. serve to transform or reproduce the conditions of production of cellular life (see Kupiec, 2008; pp. 137e142). 64 The idea of “interior milieu” has been developed by Claude Bernard concerning living beings that acquire a life that is no longer “latent” or “oscillating” but “constant or free” (see Bernard, 1878; pp. 65e124). Despite conceiving the interior milieu as a selective milieu, Bernard however never integrated this idea into the Darwinian schema, although Wilhelm Roux will propose this shortly afterward (see Roux, 1881) and today ontophylogenesis theorizes it (see Kupiec, 2008; pp. 118e123). 11 latory, nervous, etc.).65 But on the other hand and simultaneously, a certain cellular phenotype is selected, i.e. rendered more probable, by the distant history in which each multicellular being is inscribed and which is inscribed in the genetic material of its cells. In a sense, the genome resumes the reproductive success of an evolutionary lineage. It constitutes a memory that does not serve a deterministic program, but a probabilistic reservoir, factory and architecture. The fact that the first multicellular lineages had no genetic material of their own e which would not already be present in their unicellular ancestors even in a different degree or quantity66 e proves the original inadequation of the concept of genetic program. Do not homeogenes, so-called “architect” or “developmental” genes, then determine, through their stabilized order in the genome, the plane of organization of all bilateria? Ordered distributions of coding sequences within DNA, variable distributions of sequences expressible within chromatin and probabilistic diffusions of regulators within the nucleus veritably constitute so many historically acquired and selectively optimized factors that simply come to constrain and stabilize the frequency of expression e which is by right purely stochastic e of genes and, consequently, the cellular maturation and differentiation in all multicellular beings.67 Nevertheless, how do we understand this intrinsic cellular variability which is only doubly constrained secondarily? Whence does this objective margin of flexibility come? How do we pass from simple cellular indifferentiation, without differentiated tissues, of the sponge or algae, to the veritable cellular dedifferentiation of organisms with multiple organs and interlocking systems, such as bilateria or triploblastic organisms (including the majority of the contemporary animal world) and angiosperms or flowering plants (including the majority of the contemporary vegetal world)? The evolutionary success of the former group, which marks the Cambrian explosion, like that of the second group, dominating dry land from the end of the Cretaceous, would precisely seem to be tied in both cases to an accrued capacity of dedifferentiation (and consequently at the same time a capacity of ontogenetic differentiation and phylogenetic divergence): in effect, bilateria are characterized by the intercalation e between the other two embryonic layers directly exposed to the exterior milieu and destined to protect them from it and sense it (ectoderm) or provide respiration and nourishment for it (endoderm) e of a third layer (mesoderm) on the basis of which internal organs (locomotive, reproductive, circulatory, etc.) of the body arise; in the same way, angiosperms tend to accrue the dedifferentiation characteristic of the whole vegetal world by intercalating a new meristematic layer that specifically allows for a wide variability in the formation of tissues (heteroxylated and no longer homoxylated wood) and organs (leaves, flowers and fruits). Thus everything happens as if the preexistent cellular layers of the organism would serve as a membrane making possible dedifferentiation or random variation accrued from a more internal layer. Everything happens as if embryonic 65 “In order to make possible and more rigorously regulate cellular life, organs join together with organs and systems with systems. The task imposed upon them is to qualitatively and quantitatively unite the conditions of cellular life”. Thus there exists an organic perfectibility that consists in a “more thoroughly marked differentiation of preparatory labor in the constitution of the interior milieu”, Bernard, 1878, pp. 358e59. In this sense, Deleuze and Guattari already fully situate themselves within a Bernardian lineage: “Neither is the organic stratum separable from socalled interior milieus that are interior elements in relation to exterior materials but also exterior elements in relation to interior substances. These internal organic milieus are known to regulate the degree of complexity or differentiation of the parts of an organism” (A Thousand Plateaus, p. 50), i.e. primarily, in cellular differentiation, according to Kupiec’s model (Kupiec, 1997). 66 Rokas, 2008. 67 See Kupiec, 2008, pp. 162e174. Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 12 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 tissues would tend to be duplicated in a logic of dedifferentiation analogous to the logic of decoding inherent to the genetic code whose duplications offer so many possibilities of variation. Having been present in the living on every scale, membranes veritably owe their importance to the fact that they do not carry out a membranous selection without at the same time being the site of an intrinsic random variation that also submits them to natural selection. The reduction of the physico-chemical possibilities of variation ultimately leads to the creation of a series of variations internal to the organic stratum which, in the first place, is thus translated by the decoding inherent to the code. This decoding or this variation is simultaneously minimized to allow for the conservation of every selective advantage (with the selected appearance of enzymes that augment the precision of replication, correcting errors and locating mutations), but is also reinforced and maintained to perpetuate genetic variation, the indispensable source of every adaptive advantage (notably with the selected appearance of sexual reproduction in animals and plants that submits genetic material to a shuffling no less passively undergone but actively produced). The double membranous redoubling of uniand multicellular eukaryotes (cellular nucleus and organism in its entirety playing the role of membrane) then leads to the entry into this selective play of two other sources of intrinsic random variation which we catalogued earlier: the ontogenetic dedifferentiation of cells and the phylogenetic deterritorialization of organisms. By systematically eliminating the possibilities with null, meager or lesser rates of reproduction, the organic stratum thus tends to optimize its capacities for reproduction and tends to conserve all the variations that favour it, to begin with variability in all its forms: in this way the organic stratum manages to capture intrinsic variation more than the physico-chemical stratum by directly, and no longer indirectly, selecting it. As an epistemic consequence, this fact explains that the ontological primacy of variation has indeed been discovered and identified within its inner core (Darwin-Hooker principle) before being able to be generalized for all the strata. The organic stratum thus only subsists by being reproduced, i.e. by submitting to a dynamic kinetic stability that governs all uni- or multicellular living beings and each multicellular organism that is constrained to maintain its gradients of concentration and reproduce itself as such. If every organism is therefore fundamentally subjugated by this imperative of reproduction, it is at the same time intrinsically submitted to the force of variation from which the organic stratum retains or captures whatever it can in order to subsist: this is what both translates the idea of ontophylogenesis as well as the idea of body without organs (BwO) that Deleuze and Guattari take up from Artaud and meticulously conceptualize.68 Should we understand the strange concept of body without organs as an echo of the undifferentiated elementary cell of the unicellular living being (even if, as we have seen, every eukaryotic cell already contains within itself organelles or functions that are specialized on behalf of its reproduction) or rather as an echo of the fertilized egg cell or zygote e which is not yet divided and organized into differentiated tissues and organs e of the multicellular living being? 68 In the transmission of his radio play from 1947, To Have Done with the Judgment of God, Antonin Artaud thus asserts: “Man is sick because he is badly constructed/ We must make up our minds to strip him bare in order to scrape off that animalcule that itches him mortally,/god,/and with god/his organs./For you can tie me up if you wish,/but there is nothing more useless than an organ./When you will have made him a body without organs,/then you will have delivered him from all his automatic reactions/and restored him to his true freedom./Then you will teach him again to dance wrong side out/as in the frenzy of dance halls/and this wrong side out will be his real place.”, pp. 571e572. [W]e treat the BwO as the full egg before the extension of the organism and the organization of the organs, before the formation of the strata; as the intense egg defined by axes and vectors, gradients and thresholds, by dynamic tendencies involving energy transformation and kinematic movements involving group displacement [.] the egg always designates this intensive reality, which is not undifferentiated, but is where things and organs are distinguished solely by gradients, migrations, zones of proximity. The egg is the BwO. The BwO is not ‘before’ the organism; it is adjacent to it and is continually in the process of constructing itself.69 The BwO therefore naturally returns to the multiple sources of the intrinsic variation of the organic stratum, most notably to the intrinsically undifferentiated or rather dedifferentiated cellular multiplicity of every organism. In its theoretical as well as practical usage, it primarily designates our capacity for corporeal being, such that it escapes from the organization of the organs that constitute the organism, but also, secondarily, it designates the variations intrinsic to the organism, such that they spontaneously escape from its control. The relation between the BwO and the organism is consequently the same as that between variation not yet stratified or destratified and the organic stratum: For the BwO already exists in the strata as well as on the destratified plane of consistency, but in a totally different manner. Take the organism as a stratum: there is indeed a BwO that opposes the organization of the organs we call the organism, but there is also a BwO of the organism that belongs to the stratum. Cancerous tissue: each instant, each second, a cell becomes cancerous, mad, proliferates and loses its configuration, takes over everything; the organism must resubmit it to its rule or restratify it, not only for its own survival, but also to make possible an escape from the organism, the fabrication of the “other” BwO on the plane of consistency70. On the one hand, the BwO can be viewed as inherent to the organism and as incessantly wanting to escape from it. The first relation between the two passes through instruments of control responsible for resubmitting the body without organs to the rule of the organism. The concept of BwO could thus anticipate the ontophylogenetic vision which considers that the organism is neither a programmed nor autotelic entity but an arbitrary scale of the living and that cells are not as such at the beck and call of the organism but must always be constrained from the exterior by the selective pressure of the intercellular and tissual milieu in relation to the available supports. When this external control loosens its grip, their own reproduction is no longer subordinated to that of the organism, and they inevitably become cancerous.71 But, on the other hand, the BwO also designates this possibility of blending with the plane of consistency. The second relation between organism and body without organs thus passes through (practical or conceptual) instruments of destratification, through crossing the strata and grappling with pure variation.72 This path of destratification, which includes the greatest dangers, can only be followed with the most extreme caution: 69 A Thousand Plateaus, p. 153 and 164. Ibid., pp. 162e163. See Kupiec, 2008, pp. 155e158 and the work of Jean-Pascal Capp (notably Capp, 2012). 72 See “How Do You Make Yourself a Body without Organs?” in: A Thousand Plateaus, pp. 149e166, which specifically describes in detail masochistic practices in their corporeal, sometimes extreme, aspect as an example of the destratification or construction of a body without organs. 70 71 Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002 J. Rosanvallon / Progress in Biophysics and Molecular Biology xxx (2012) 1e13 This is because the BwO is always swinging between the surfaces that stratify it and the plane that sets it free. If you free it with too violent an action, if you blow apart the strata without taking precautions, then instead of drawing the plane you will be killed, plunged into a black hole, or even dragged toward catastrophe [.] Dismantling the organism has never meant killing yourself, but rather opening the body to connections that presuppose an entire assemblage, circuits, conjunctions, levels and thresholds, passages and distributions of intensity, and territories and deterritorializations measured with the craft of a surveyor.73 In sum, pure variation therefore constitutes the fundamental given that does not require any explanation but is the basis upon which all aspects of reality must be explained. Stratification is the process through which this pure variation acquires a subsistence. The strata tend to thicken pure variation and give it consistency, i.e. forms but also and above all duration: it is due to the latter that they incarnate finite degrees and various rhythms of variation that are characterized by various degrees of slowness. They produce the reality of things by limiting and constraining variation from the exterior. The physical world is nothing but the first of the solutions discovered by pure variation to the problem of durable subsistence, i.e. its endurance as variation: if physicochemical stratification only conserves the subsistences or invariances of variation, organic stratification transforms the former into a genuine conservatory by maintaining and reproducing variation itself, specifically its intrinsic and stochastic components. Thus pure variation simultaneously constitutes a “nonorganic life” wider than the material world and a body without organs wider than the organism, and we can coincide with pure variation by loosening the stranglehold of the organism (more intense “non-organic life”) without ever for all that dismantling it completely (risk of organic death). Acknowledgements I would like to extend my most sincere thanks to Jean-Jacques Kupiec and Olivier Gandrillon for their cordial invitation to present these ideas in this special volume dedicated to the colloquium «Chance at the heart of the cell» as well as for their meticulous rereading and their infinite patience. I would also like to warmly thank Quentin Meillassoux for his fundamentally friendly and fruitful remarks. Lastly, I want to extend special thanks to Taylor Adkins for his elegant and precise labour of translation, his unwavering support and his consistent availability. References Aguirre, A., 2007. Making Predictions in a Multiverse: Conundrums, Dangers, Coincidences. In: Carr, B. (Ed.), Universe or Multiverse? Cambridge University Press, Cambridge, pp. 367e386. Aguirre, A., Gratton, S., 2008. Inflation without a Beginning: A Null Boundary Proposal. gr-qc/0301042v2. Arkin, I., Stevens, T., 2000. Do more complex organisms have a greater proportion of membrane proteins? Proteins: Struct. Funct. Gen. 39, 417e420. Artaud, A., 1947. English translation 1988. To have done with the judgment of god. In: Sontag, S., Weaver, H. (Eds.), Selected Writings. University of California Press, Los Angeles, pp. 555e572. Trans. Atamas, S.P., 1996. Self-organization in computer simulated selective systems. Biosystems 39, 143e151. Bergson, H., 1907. In: Mitchell, A. (Ed.), English Translation 1944. Creative Evolution. Random House, New York. Trans. 73 Ibid., p. 161 and 160. 13 Bergson, H., 1938. In: Andison, M.L. (Ed.), English Translation 1946. The Creative Mind: An Introduction to Metaphysics. The Philosophical Library, New York. Trans. Bernard, C., 1878. English Translation 1974. Lectures on the Phenomena of Life Common to Animals and Plants. Charles C. Thomas, New York. Bonner, J.T., 1998. The origins of multicellularity. Integr. Biol. 1 (1), 27e36. Brandon, R.N., McShea, D., 2010. Biology’s First Law: The Tendency for Diversity and Complexity to Increase in Evolutionary System. University of Chicago Press, Chicago. Capp, J.-P., 2012. Stochastic gene expression stabilization as a new therapeutic strategy for cancer. Bioessays 34, 170e173. Darwin, C., 1859. On the Origin of the Species, first ed. John Murray, London. Darwin, C., 1861. On the Origin of the Species, third ed. John Murray, London. Darwin, C., 1862. The Correspondence of Charles Darwin, vol. 10. Cambridge University Press, Cambridge. Darwin, C., 1872. On the Origin of the Species, sixth ed. John Murray, London. Darwin, C., 1875. The Variation of Animals and Plants under Domestication, second ed. John Murray, London. de Duve, C., 2005. Singularities: Landmarks on the Pathways of Life. Cambridge University Press, New York. DeLanda, M., 2002. Intensive Science and Virtual Philosophy. Continuum, New York. Deleuze, G., 1986. In: Hand, S. (Ed.), English Translation 1988. Foucault. University of Minnesota Press, Minneapolis. Trans. Deleuze, G., Guattari, F., 1972. In: Hurley, R., Mark, S., Lane, H.R. (Eds.), English Translation 1983. Anti-Oedipus. University of Minnesota Press, Minneapolis. Trans. Deleuze, G., Guattari, F., 1980. In: Massumi, B. (Ed.), English Translation 1987. A Thousand Plateaus. University of Minnesota Press, Minneapolis. Trans. Deleuze, G., Guattari, F., 1991. In: Tomlinson, H., Burchell, G. (Eds.), English Translation 1994. What Is Philosophy?. Columbia University Press, New York Trans. Edelman, G., 1987. Neural Darwinism: Theory of Neuronal Group Selection. Wiley, New York. Eigen, M., Schuster, P., 1979. The Hypercycle: A Principle of Natural Self Organization. Springer, New York-Berlin-Heidelberg. Galles, C., et al., 2009. From bacteria to plants: a compendium of mismatch repair Assays. Mutat. Res. 682 (2e3), 110e128. Gilbert, W., 1986. Origin of life: the RNA world. Nature 319, 618. Guattari, F., 1992. In: Bains, P., Pefanis, J. (Eds.), English Translation 1995. Chaosmosis. Indiana University Press, Bloomington. Trans. Gunzig, E., 2008. Que faisiez-vous avant le Big Bang? Odile Jacob, Paris. Heams, T. (Ed.), 2011a. Les mondes darwiniens. éditions Matériologiques, Paris. Heams, T., 2011b. Variation. In: Heams, T. (Ed.), Les mondes darwiniens, pp. 51e67. Henry-Couannier, F., 2005. The Dark Side of Gravity. gr-qc/0610079v55. Henry-Couannier, F., et al., 2006. Negative Energies and a Constantly Accelerating Flat Universe. gr-qc/0507065v2. Hoquet, T., 2009. Darwin contre Darwin. Seuil, Paris. Kupiec, J.-J., 1997. A Darwinian theory for the origin of cellular differentiation. Mol. Gen. Genet. 255, 201e208. Kupiec, J.-J., 2008. In: Hutchings, M., Hutchings, J. (Eds.), English Translation 2009. The Origin of Individuals. World Scientific Publishing, London. Trans. Margulis, L., 1981. Symbiosis in Cell Evolution. Freeman, New York. Martin, W., Müller, M., 1998. The hydrogen hypothesis for the first eukaryote. Nature 392 (6671), 37e41. Maturana, H.R., Varela, F., 1979. Autopoiesis and Cognition: The Realization of the Living. D. Reidel Publishing Company, Dordrecht. Meselson, M., Wildenberg, J., 1975. Mismatch repair in heteroduplex DNA. Proc. Natl. Acad. Sci. USA 72 (6), 2202e2206. Meillassoux, Q., 2006. In: Brassier, R. (Ed.), English Translation 2008. After Finitude: An Essay on the Necessity of Contingency. Continuum, New York. Trans. Pross, A., 2011. Toward a general theory of evolution: extending Darwinian theory to inanimate matter. J. Syst. Chem. 2, 1. Radman, M., 1975. SOS repair hypothesis: phenomenology of an inducible DNA repair which is accompanied by mutagenesis. Basic Life Sci. 5A, 355e367. Radman, M., et al., 1995. Interspecies gene exchange in bacteria: the role of SOS and mismatch repair systems in evolution of species. Cell 80 (3), 507e515. Radman, M., et al., 2003. Stress induced mutagenesis in bacteria. Science 300, 1404e1409. Rokas, A., 2008. The molecular origins of multicellular transitions. Curr. Opin. Genet. Dev. 18, 472e478. Rosanvallon, J., 2009. Deleuze et Guattari à vitesse infinie, vol. 1. Ollendorff & desseins, Paris. Roux, W., 1881. Der Kampf der Teile im Organismus. VDM, Saarbrücken, Germany. Rovelli, C., 2008. Forget Time. http://arxiv.org/abs/0903.3832v3. Simondon, G., 1995. L’individu et sa genèse physico-biologique. Millon, Paris. Smolin, L., 1999. The Life of the Cosmos. Oxford University Press, New York. Smolin, L., 2007 . Scientific Alternatives to the Anthropic Principle. In: Carr, B. (Ed.), Universe or Multiverse? Cambridge University Press, Cambridge, pp. 323e366. Sorkin, R., et al., 2002. Everpresent Lambda. astro-ph/0209274v1. Souriau, J.-M., 2002. Les groupes comme universaux. http://semioweb.msh-paris.fr/ f2ds/docs/geo_2002/Document02_Souriau1.pdf. Tegmark, M., 2007. The Multiverse Hierarchy. In: Carr, B. (Ed.), Universe or Multiverse?, Cambridge University Press; Cambridge, pp. 99e149. Wächterhäuser, G., 1988. Before enzymes and templates: theory of surface metabolism. Microbiol. Rev. 52 (4), 452. Please cite this article in press as: Rosanvallon, J., Pure variation and organic stratification, Progress in Biophysics and Molecular Biology (2012), http://dx.doi.org/10.1016/j.pbiomolbio.2012.06.002