Stengers on emergence (BioSocieties)

BioSocieties (2014) 9, 99–104. doi:10.1057/biosoc.2013.43

Isabelle Stengers. Cosmopolitics. Minneapolis, London: University of Minnesota Press, 299 pp., US$25, £18.50, ISBN: 9780816656868; 9780816656875

Reviewed by Graham Harman

American University, Cairo, Egypt. E-mail:

Cosmopolitics, the major work of Belgian philosopher Isabelle Stengers, has been available in English since 2011 in a lucid two-volume translation by Robert Bononno. In the Anglophone world Stengers is already known as a formidable interpreter of Alfred North Whitehead, a thinker she has done so much to disseminate in the French-speaking world (Stengers, 2011). But in the present work we encounter Stengers’ own philosophical voice even more directly. Originally published in the 1990’s in seven slim French volumes, Cosmopolitics should be regarded as one of the most unique works of continental philosophy in the past several decades.

For many years, continental philosophy was attacked for its focus on purely literary and social science texts, far from the stunningly successful labors of the natural sciences. Cosmopolitics is one of several prominent recent works that have begun to reverse this trend. Karen Barad’s Meeting the Universe Halfway also comes to mind (Barad, 2007), as does Manuel DeLanda’s Intensive Science and Virtual Philosophy(DeLanda, 2002), along with several books on neuroplasticity by Cathérine Malabou (2008) All of these works have had considerable influence, and they may look in retrospect like a collective turning point. Yet Cosmopolitics differs from the others in at least two respects. First, Stengers gives us a long historical narrative filled with a roster of heroes barely familiar to her philosophy readership: Lagrange, Carnot, Hamilton, Duhem, Poincaré, and de Donder among them. Second, Stengers writes as someone personally invested in this history, since she worked as a close associate of Ilya Prigogine, the celebrated Russian-born Belgian chemist. Despite receiving a Nobel Prize in 1977, Prigogine ended his career as a somewhat marginalized figure, accused of ‘delusions of grandeur’ for reasons described in haunting fashion by Stengers herself.1

What will become of Stengers’ detailed history of dynamics among her philosophically minded readers? Gilles Deleuze launched an unexpected fashion for the Stoics and Duns Scotus, to name just two of his ‘minor’ favorites. Alain Badiou has spurred a generation of young readers to learn set theory and idolize the madman-genius Cantor. Will Stengers’ Cosmopolitics provoke a comparable wave of enthusiasm for the history of dynamics? Or will it remain an intriguing dark horse of a book, left by the wayside as different paths are followed? As Stengers demonstrates in her approach to the history of physics, there is no way to predict which human works will become events that produce a tangible line of heirs. Just as Prigogine’s scientific breakthrough in 1972 could have rewritten the history of physics,2 Stengers’ lengthy treatise could lead to a new style of continental philosophy: steeped in Deleuze and Whitehead, while closely tracking the shift from Lagrangians to Hamiltonians to Niels Bohr’s model of the atom. While this may sound unlikely in view of the meager past relations between science and continental philosophy, who would have expected Cantor to become a favorite of young French philosophers? It is at least conceivable that Stengers has opened a door that other talented thinkers will enter, and this gives Cosmopolitics the refreshing air of a possible future, no matter what eventually becomes of the book.

What must now be asked is whether the elegance and historical thoroughness of Cosmopolitics go hand-in-hand with a compelling philosophical position. In her remarkably calm and even-tempered book, Stengers nonetheless shows considerable impatience with philosophers of the old style, who brandish their arrogant certainties about how the world really is. She argues instead for what she calls an “ecology of practices.” Yet this ecological program turns out to be an ontology in its own right, as open to counter-argument as any other. To zero in on Stengers’ ontology, I will focus on her concept of emergence. This is both a central philosophical theme of Cosmpolitics and a topic where she differentiates her views equally from the ‘reductionists’ and the ‘holists,’ who are presented as sharing the same basic flaw. Though it may seem difficult to extract isolated themes from such a densely interwoven work, the brief format of the present review requires that we make the attempt. Ignoring for now her critique of the notion of physical ‘states,’ and her important passing salute to Gilbert Simondon’s dismissal of fully formed individuals, we can find the core of Stengers’ concept of emergence in Chapters 13 and 14 of Cosmopolitics II. These chapters are entitled ‘The Question of Emergence’ and “The Practices of Emergence,”3 and together total just under thirty pages.

Before considering Stengers’ own views, we should place the question of emergence in philosophical context. How does something new emerge irreversibly from the old? This was a central concern of Henri Bergson a century ago, and Francophone thought since the 1960’s has often been nearly obsessed with the question of the new.4 Beyond this theme of the new arising from the old, emergence can also be framed as the question of how the large emerges from the small or the more composite from the more simple. If this never happens, if mid-sized entities are always mere aggregates of tinier authentic things, then we are left with a reductionist or even eliminativist approach to the world in which a privileged ultimate layer is treated as the sole authentic reality. Thinking would thereby become a permanent exercise isundermining: debunking illusory macroscopic beings in favor of underlying subcomponents or perhaps even a barely articulate, gelatinous mass (See Harman, 2011). Particle physics would become the chosen discipline of the gods; all others would have to accept a subordinate local status, ruled by the ultimate primacy of physics.5

Yet we must also beware of a kind of reduction that moves upwards rather than downwards – namely, the kind that I have called overmining. Instead of dissolving a thing into its constituents, we might dissolve it upward into appearances, effects, manifestations, or events, while mocking the ‘naïve realism’ that posits discrete individuals hiding behind their tangible manifestations. Or we might play the double game of doing both at once, duomining the world by appealing sometimes to ultimate particles or indeterminate flux as the bedrock of reality, and other times to an uppermost layer of the visible, the evental, or the determinate that float without a bedrock.6 What is lost through this two-faced process is the middle kingdom: the robust reality of specific things that are more than the inner relations of their parts, but less than their outer relations with their environment. Object-oriented philosophy insists on the rights of the middle kingdom, with objects protected from reduction in two directions rather than just one. From an object-oriented standpoint, how does Stengers look when it comes to the question of emergence? Is she an underminer, an overminer, a duominer, or an ally? Or does she somehow escape all of these categories, which through her act of escape would be exposed as artificial or at least non-exhaustive? Whatever the answer, this will not be an exercise in name-calling or empty labelling. Each of the terms just mentioned (undermining, overmining, duomining, object-oriented) has a precise meaning and definite philosophical consequences.

Stengers is aware that the problem of emergence has ancient metaphysical roots: “Aristotle’s disciples were already arguing about composite bodies endowed with new qualities that arose from the elements that composed them. How could these new qualitative properties be explained?” (p. 208) The standard reductive approach is exemplified by today’s most zealous materialists, who “signal a future in which, from psychology to the social sciences and therapeutic practices, all forms of knowledge concerning human behavior will be understood in terms of neuronal interactions.” (p. 209) Such materialists, of course, do not even treat neurons as the fundamental basis of the world, since they too can be explained through the workings of even tinier constituents. Neurons for them are simply a convenient, provisional, local supply base for reductive explanations of the human realm. Undermining is treated as the very business of thought, the sole legitimate method for assaulting all that is supposedly mysterious.

All resistance to such undermining must hold that something new emerges at levels higher than the physically ultimate. This approach need not come from outside the sciences. Stengers notes that the anti-mechanistic chemists of the eighteenth century “claimed there was a difference between composition, which was their problem, and the simple aggregation of physicists” (p. 209). The quarrel between reduction and anti-reduction thus became a disciplinary dispute between physicists and chemists. Here the chemists are joined by Leibniz, one of Aristotle’s greatest heirs, who “pointed out the foolishness of those who dreamed of explaining sensation, perception, and consciousness in terms of inert matter,” and in doing so “he seems to have been taking part in a quarrel that continues today with the unfortunately celebrated mind-body problem.” (p. 208) Of course, there are several different intellectual camps that might view the celebration of the mind-body problem as ‘unfortunate.’ One of these camps is that of the hardcore materialists mentioned in the previous paragraph. For them there is no mind-body problem simply because body is destined to win; mind will eventually cave in to advancing physical explanations of the brain. For a hardcore idealist such as Berkeley, by contrast, there can be no mind-body problem because mind has already won; to be is to be perceived, and there is no autonomous ‘body’ outside the configurations it displays for some human or divine mind. Against these two options, seekers in the realm of the mind-body problem at least mark a place of uncertainty, a temporary bastion against quick reductions in either direction. Given that Stengers views this bastion as ‘unfortunate’ (as I do, but for very different reasons) we will need to see how she hopes to outflank all three positions simultaneously.

A point of especial interest in Stengers’ story is the changing status of clocks, one of the most useful and earliest-perfected machines of the modern era. “The clock is a weapon against Aristotelian thought, for which matter is unintelligible as such but requires a form, with which are associated both the existence of individual beings, each of which is endowed with its own end, and the possibility of knowing them.” (p. 210) However, “in the case of the clock, matter and finality can be understood separately: consisting of inert parts, and as such subject to the laws of mechanics, it owes its clocklike existence to the genius of the maker, who has subjected those parts to their own ends, who has incorporated them into a coherent mechanism defined by a finality – telling time.” (p. 210) The question of emergence is thereby conflated with the question of purpose or final causation: “The question of finality designates the stronghold that must be defended or conquered.” (p. 210) This appeal to finality proves to be a bad move for the anti-mechanists, since Darwinian natural selection allows Richard Dawkins to replace the purposeful timekeeper with the “blind watchmaker” whose living creatures have no internal finality, but simply survive or fail to survive in the environment they happen to confront. (p. 210) By the same token, it allows Jacques Monod to dismiss final causes and say that living creatures are merely “teleonomic,” meaning that we can describe them on the basis of their aim of self-reproduction, but without metaphysical commitment to an actual finality inherent in these beings themselves. (p. 210)

And here we encounter Stengers’ impatience with both the reductionist and vitalist sides of the dispute. Locked in mutual polemic, each adopts self-defeating strategies that open up vulnerable paths to their mortal enemy. “What I want to emphasize here is that understanding the challenge to which the living being exposes the biologist is barred to the vitalist biologist just as it is to the believer in neo-Darwinism. In both cases, the polemical position is expressed by the production of an identity that is substituted for practical requirements and obligations the way a solution is substituted for a problem.” (p. 211) In other words, the neo-Darwinist defends inert mechanical matter and the vitalist defends non-mechanistic purposes, but these are both ‘identities’ that ought to give way to ‘practical requirements and obligations.’ This is the pragmatist gist of Stengers’ call for an “ecology of practices”: disputes over the nature of reality are pointless polemics that ought to be re-inscribed in the practical soil that enables the two opposite positions in the first place. We are led not to an ambiguous real world in which everything is both mechanistic and purposive, but to an ultimate human practical context in which things are neither mechanistic nor purposive, apart from the ‘requirements and obligations’ following from how the problem is posed at any given time.

Stengers cites the cases of Pasteur demonstrating the autonomy of the microorganism, Körner displaying the hexagonal structure of benzene, and Nirenberg using an artificial DNA molecule to synthesize a protein. (p. 213) Stengers’ ontological conclusions about these events might be called ‘deflationary,’ since they neither add real autonomous microorganisms, benzene molecules, and proteins to the world, nor do they shatter these things reductively into tinier components. As she puts it, “events of this kind mark the creation of new laboratory beings and the new laboratories that correspond to them. But they do not pose the problem of emergence and do not allow any reduction to occur. They mark the success of an operation of delegation.” (p. 213) The understated tone of the passage cannot mask its radical philosophical claim. For it is not just that Pasteur, Körner, and Nirnberg happened not to brush against the philosophical question of reduction and emergence. Instead, for Stengers, reduction and emergence are not legitimate philosophical problems at all. They are pseudo-problems that ought to be replaced by the true problem of how successful and unsuccessful ‘delegations’ are made. Instead of disputing over the criteria for what would or would not count as an ‘emergent’ being immune to mechanistic reduction, “it is much more interesting to point out how the operations of experimental delegation that have treated bacteria as targets or actors have been possible.” (p. 213) Is Stengers’ theory of delegation simply ‘much more interesting’ than ontological disputes over emergence, or are there more convincing grounds for dissolving those disputes into her own pragmatic theory? I for one do not share her lack of ‘interest’ in emergence, nor can I accept the concluding lesson of her Section 13: “all the confrontations that serve as ecology in the modern sciences converge around the question of emergence. Therefore, it is from this field of battle that we must escape… a practical, constructivist sense must be given to the issues covered by [the term ‘emergence’].” (p. 218; emphasis added)

We can now move to Stengers’ attempted coup de grâce in Section 14. “It is not often,” she reports, “that I have the opportunity to speak well of the work of philosophers of science.” (p. 219) Yet she now sees opportunity for praise when referring to the three-tiered model of emergence proclaimed by J.K. Feibleman. Though he starts with “a conventional definition of emergence, which associates the relation between a whole and its parts to the relation between ends and means,” (p. 219) he seems to add an extra layer to the problem. In Feibleman’s own words: “For an organization at any given level, its mechanism lies at the level below, and its purpose at the level above. This law states that for the analysis of any organization three levels are required: its own, the one below, and the one above.”7 For Stengers, the value of this model lies in its implication that “the purpose of an organization is not found in itself but is always seen from the point of view of something else.” (p. 219) This gives ammunition to her claim that the identities of whole and part must be determined in terms of “the practices that allowed those identities to be defined.” (p. 220) In the case of water, for instance, we can actually speak of two waters: “one of its identities corresponds to the chemist’s purpose in understanding it as a molecule that will interact with other molecules; the other corresponds to the purpose of understanding it as a solvent that is a liquid.” (p. 220; emphasis added) The purposes of the understanding are always what is central, hence my added italics in the passage. But even more surprising is Stengers’ brazen rewriting of “emergence” so that it dwells within the understanding itself. As she puts it, ‘ “water” had to emerge twice: as a molecule composed of ‘parts’ and as a liquid with specific properties, composed of molecules.’ (p. 220) The scare-quotes around ‘water’ and ‘parts’ in this passage should not distract us from what is happening to the non-scare-quoted “emergence.” For Stengers, the term “emergence” no longer pertains to levels of reality where something new happens independently of our understanding; instead, it is produced by that very understanding.

It certainly looks as though Stengers is simply replacing the part/whole dualism of classical disputes over emergence with a new and unimproved twofold in which a non-articulate or semi-articulate world is confronted by human scientists whose practical purposes serve to cut the world into neatly defined sections for the first time.8 Here, Stengers might answer that she does not advocate a two-leveled theory of emergence, but something more like Feibleman’s three-leveled model. Let’s consider how such a model might operate in the framework ofComsopolitics.

Stengers briefly develops her own three-level approach with the example of chemical elements. “Ever since Mendeleev,” she recounts, “the element has been a part of the chemical definition of molecules and reactions, but it presents no problem for emergence.” (p. 220) From there, Stengers goes on to describe an asymmetry between elements/molecules on the one hand and atoms on the other; I will treat ‘elements/molecules’ as a pair only because Stengers does not distinguish between them in this passage. As concerns elements/ molecules: “The chemical element, like matter in the Aristotelian sense, has no properties that could be used to define it ‘in itself.’ Its definition entails the definitions of simple and compound bodies and their reactions.” (p. 220) We will discuss this again shortly. But the case of the atom is apparently quite different: “On the other hand, the atom claims to explain the molecule the way the part explains the whole. It owes its scientific existence to practices of a very different kind, which do not address it as a chemical actor; therefore it can, unlike the element, claim a separable identity.” (p. 220) Along with elements/molecules and atoms, we also find the anticipated third level: “element and atom came to designate the same being only after a series of complicated negotiations in which data from various practices had been articulated and coadapted. And in this process of negotiation, the ‘purpose’ is found ‘above,’ on the level of the practice of negotiation itself.” (p. 220)

Though Stengers does not do all the work for us of mapping her threefold schema onto Feibleman’s triad, it is not difficult to see how she proposes to do so. Stengers’ Feiblemanian analysis runs as follows:

  1. We must consider the element/molecule on its own level. According to Stengers, this level is reminiscent of Aristotelian matter, having no properties in its own right but serving as a kind of amorphous receptacle that gains its qualities only from the levels below and above it. We should note in passing that this first level is both dubious and surprisingly innovative. It is dubious because it is by no means clear that the properties of a chemical element can be reduced either to the properties of its atoms or the uses one makes of the element. Indeed, this is one of the chief recurrent arguments of partisans of real emergence. Yet Stengers simply declares their argument irrelevant by her fiat of comparing chemical elements to ‘Aristotelian matter’ lacking intrinsic properties of their own. Yet in another sense her model is also quite innovative, since normally the defenders of matter-without-qualities place it at the very bottom of the cosmos, rather than at an intermediate level as Stengers does.
  2. Following Feibleman’s threefold method (which Stengers endorses), we must now consider the element/molecule at the level below it. In the present example, atoms are the level just below molecules. “Unlike the element,” Stengers already told us, atoms ‘[can] claim a separable identity.’ Obviously Stengers does not take this to be a permanent special feature of atoms, which (as the scientist Stengers knows even better than the chemical layman) can be analyzed downward into quarks and electrons just as easily as molecules can be analyzed into atoms. What she evidently means is that, given our momentary interest in the element/molecule as a chemical agent, and given the sub-chemical status of atoms, we can treat atoms for the moment as explanatory agents or ‘black boxes’ lacking internal articulations of their own. Certainly, we could always change our question and focus on the composition of atoms instead. But the ‘practices’ relevant to our current question allows us to treat the atom (for now) as an explainer that does not need to be explained in turn.
  3. Finally, we must consider the element/molecule at the level above it. For Stengers (there is no evidence that Feibleman would see it this way) this third level is the most important, since it is not just one among equals, but governs the very production of the difference between the other two. For as we saw, “element and atom came to designate the same being only after a series of complicated negotiations in which data from various practices had been articulated and coadapted. And in this process of negotiation, the ‘purpose’ is found ‘above,’ on the level of the practice of negotiation itself.” (p. 220) There may be three layers, but practice is the layer that rules them all.

In short, Stengers does not argue for a three-leveled theory at all, but for precisely the sort of twofold theory of which we were complaining a few pages ago. First, given that Stengers shows no traces of frank Berkeleyan idealism, she seems to concede that there is a world out there that resists our conceptions and allows for some negotiations to succeed and others to fail. That’s the first level: a world that is not just an image in our minds. And second, there is the dominant layer of praxis and negotiation that allows for the very articulation between parts and wholes in the first place. And what of the additional level that Feibleman requires – the consideration of the element/molecule (or anything else) ‘on its own level’? We recall Stengers’ rather noncommittal description of this level: “The chemical element, like matter in the Aristotelian sense, has no properties that could be used to define it ‘in itself.’ Its definition entails the definitions of simple and compound bodies and their reactions.” (p. 220) The upshot is that nothing has any qualities in its own right (here we are speaking of elements/molecules, but the same would hold for atoms, horses, balloons, persons, nations – for anything at all). A thing gains its properties either from the explanations provided by its own parts, or the ‘purposes’ that articulate it in one way rather than another.

In a word, from the standpoint of object-oriented philosophy, Stengers is a classic duominer who reduces entities simultaneously both to lower-level atoms and higher-level scientific purposes, while reserving for entities themselves nothing but the amorphous status of inarticulate Arisotelian matter, fit only to be shaped by our ‘ecology of practices.’ Reality becomes a hot potato, passed either downward to tiny pieces or upward to all-encompassing practices, but is never stationed wherever we happen to be searching for it. This is the philosophical pitfall of duomining, and I hope that the unfamiliarity of the term does not overshadow the seriousness with which I use it. For all her claims to surpass all the stale old dualistic polemics, Stengers simply shows us the most classic reflex of Western philosophy: a simultaneous reduction of the world in two separate directions rather than one, with each reduction providing an alibi for the other.


1 For Stengers’ treatment of Prigogine see Cosmopolitics II, Chapter V, Life and Artifice, pp. 105–204. For a more detailed earlier collaboration between the two, (see Prigogine and Stengers, 1984).

2 Stengers makes this claim about Prigogine in Cosmopolitics II, p. 151.

3 Stengers, Cosmopolitics II, pp. 207–233.

4 For an intriguing account of this phenomenon, see the treatment of Badiou and structuralism in Bryant (2011), p. 243 ff.

5 I have criticized this tendency in the work of James Ladyman and Ross, among others. See Harman (2010).

6 I borrow the term ‘duomining’ from the credit card industry, where it refers to the simultaneous use of data and text mining. See Harman (2013).

7 Stengers is quoting here from page 61 of Feibleman (1954).

8 The ‘human scientists’ part is slightly unfair, of course, since Stengers like Latour tries to reinterpret words such as ‘negotiate’ in non-anthropocentric terms (see her remarks on the body’s twofold treatment of water in Cosmopolitics II, p. 221). But the same ontological problems occur even if we allow non-humans to join humans in using their own purposes to carve a largely inarticulate world into pieces. See my remarks about how a global ‘relationism’ is only marginally better than a human-centered ‘correlationism’ in Harman (2009).


  1. Barad, K. (2007) Meeting the Universe Halfway: Quantum Physics and the Entanglement of Matter and Meaning. Durham, NC: Duke University Press.
  2. Bryant, L.R. (2011) The Democracy of Objects. Ann Arbor, MI: Open Humanities Press.
  3. DeLanda, M. (2002) Intensive Science and Virtual Philosophy. London: Continuum.
  4. Feibleman, J.K. (1954) Theory of integrative levels. British Journal for the Philosophy of Science 17: 59–66. | Article |
  5. Harman, G. (2009) Prince of Networks: Bruno Latour and Metaphysics. Melbourne:
  6. Harman, G. (2010) I am also of the opinion that materialism must be destroyed. Environment and Planning D: Society and Space 28(5): 772–790. | Article |
  7. Harman, G. (2011) On the undermining of objects: Grant, Bruno, and radical philosophy. In: L.R. Bryant, N. Srnicek and G. Harman (eds.)The Speculative Turn: Continental Materialism and Realism. Melbourne:
  8. Harman, G. (2013) Undermining, overmining, and duomining: A critique. In: J. Sutela (ed.) ADD Metaphysics. Aalto Finland: Aalto University Design Research Laboratory, pp. 40–51.
  9. Malabou, C. (2008) What Should We Do With Our Brain? Trans. S. Rand. New York: Fordham Univ. Press.
  10. Prigogine, I. and Stengers, I. (1984) Order Out of Chaos. New York: Bantam.
  11. Stengers, I. (2011) Thinking with Whitehead: A Free and Wild Creation of Concepts. Trans. M. Chase. Cambridge, MA: Harvard Univ. Press.

About the Author

Graham Harman is Distinguished University Professor at the American University in Cairo. He is the author of numerous books, most recently Bells and Whistles: More Speculative Realism (2013).

Deixe um comentário

Preencha os seus dados abaixo ou clique em um ícone para log in:

Logotipo do

Você está comentando utilizando sua conta Sair /  Alterar )

Foto do Google

Você está comentando utilizando sua conta Google. Sair /  Alterar )

Imagem do Twitter

Você está comentando utilizando sua conta Twitter. Sair /  Alterar )

Foto do Facebook

Você está comentando utilizando sua conta Facebook. Sair /  Alterar )

Conectando a %s