Testez le Grand Continent. Les idées et les analyses qu’on ne trouve nulle part. Mais que vous verrez bientôt partout. Dans votre boîte mail à partir de 8 euros par mois
Philippe Descola and Bruno Latour have profoundly renewed the way we think about the relations between humans and nature. At the heart of their dialogue, long overlooked, lies a shared notion: the “grand partage.” What does this notion encompass? And why is it so central to their critique of modernity?
One aspect of my work, which my latest book documents, is to trace the genealogical line of a vocabulary common to Descola and Latour, and to determine the moment when one introduces such a concept and how the other reinterprets it.
In this case, the expression “grand partage” first appears in Latour’s writing, at the very beginning of the 1980s. In 1983, he published in the Revue de synthèse a major article of general theory on the anthropology of science, titled “How to redistribute the grand partage?”
Latour borrows this expression from Jack Goody, who uses it at the start of The Domestication of the Savage Mind. At first, he employs it in a meaning quite close to Goody’s: essentially it designates the distinction between science and belief, which in turn covers the dichotomy between modern and traditional societies as wholes.
Goody’s gesture notably consists of reinserting a gradation between illiterate societies and scientific societies by studying the “intellectual technologies” that enabled this transformation of knowledge. The aim is to move beyond the dualism that pits a tradition mired in belief against a scientific modernity.
Gradually, and especially starting from We Have Never Been Modern, published eight years after that foundational article, the expression comes to denote another opposition: not the science-belief division, but the nature-society division. It is in this sense that Descola later takes it up in Beyond Nature and Culture. There is thus a mediation: Latour connects Goody and Descola.
However, it is interesting to note that, at least within Latour, this second sense of the grand partage retains the first. Indeed, he distinguishes an internal grand partage, between nature and society, which is then transposed outward as a division between “them” and “us.”
In other words, we present ourselves as the only societies capable of objectively knowing nature, since we have invented science. The two senses are thus closely linked.
In short, one can observe an evolution in Latour’s oeuvre, which reverberates in Descola’s use of the expression: the “grand partage” no longer designates merely, as in Goody, the opposition between science and belief, tradition and modernity, but the opposition between nature and society.
How do the critiques of Descola and Latour—one on comparative anthropology, the other on science and technology studies—allow us to question the traditional ways of producing knowledge within their respective disciplines?
Perhaps here we can find a common thread with Goody: to attack the “grand partage” is also to undertake a form of self-critique of the social sciences. This shared impulse explains why Descola and Latour have drawn closer around the idea of an end to the grand partage that they would bring about. This idea even lent its title to Pierre Charbonnier’s first book, The End of a Grand Divide. Nature and Society, from Durkheim to Descola (a work that can, however, be read as a challenge to some of Descola’s and Latour’s theses, since it shows that the French social sciences have never taken the nature out of view).
Descola and Latour alike agree that the grand partage between nature and society constitutes a kind of intellectual matrix of Western modernity, in the sense that it would be a distinction from which many other distinctions flowed, and which would have been recorded and reinforced by the dominant social sciences during their construction.
Obviously, the history of ideas offers other ways to critique the opposition between nature and culture. It would be unjust to present Descola and Latour as the only ones, or even the first, to question it.
One of the ideas that seems to be theirs is this: to critique the nature-culture dualism, one must also undertake an autocritique of the social sciences and show how their adoption of this foundational distinction has, even indirectly, served as epistemological justification for the modernization project of conquering nature.
You also point out that these critiques diverge…
A difference indeed emerges: for Descola, this autocritique primarily targets the history of anthropology. That is why he more often targets the opposition between nature and culture than the one between nature and society.
According to him, culture has become the anthropology’s plane of objectivity, since the discipline defines itself as the science of cultural diversity. Nature then becomes a backdrop that cannot be studied in its own right by the anthropology proper (hence the gradual marginalization of what used to be called “physical anthropology”). The discipline instead focuses on the set of cultures that overlay this nature, understood as a series of symbolic and technical systems.
Descola’s idea is that ethnology and anthropology failed in their mission by projecting onto non-Western terrains a distinction that is itself a pure product of the West.
That is why Descola often adopts a form of epistemic modesty, arguing that his work has simply deepened the constitutive dynamic of anthropology, namely the critique of ethnocentrism. The distinction between nature and culture, which lies at the heart of the discipline, remains for him among the most ethnocentric aspects of our intellectual heritage. It must be shed if we want to continue this self-critique.
Latour, for his part, certainly identifies more with anthropology than sociology, but he remains a specialist in modern societies, i.e., those traditionally studied by sociology. He thus attacks the history of sociology.
Like Descola, he defends the idea that at the moment of their institutional and epistemological emergence, at the end of the 19th century, the social sciences needed to design a domain of objectivity of their own in order to justify their existence. This is true for sociology, where this domain is generally named “society,” and for anthropology, where it takes more the name of “culture,” as we have just seen.
According to Latour and Descola, the question was then to define a domain that obeys its own causal regime, not unilaterally conditioned by physical or biological determinants. This is what justified the creation of new sciences charged with studying this dimension of reality.
The grand partage was thus necessary, for a time, to found the social sciences. But once these were established, one could discard it to allow them to reach a higher stage, where they would be capable of studying ecological relations that, according to Descola and Latour, they struggled to describe because they held a dualist cosmology. In this sense, these two anthropologists are among the many authors who, since the late 20th century, have sketched the theoretical outlines of what we now call the “environmental humanities.”
You explain that, for Descola, the difference between animist, totemist, or analogist peoples and naturalists is not that the former would be more respectful of nature, but that their ontological framework does not permit the rise of productivism. This also ties into the question of the emergence of capitalism and its consequences.
Indeed, and this is another point of convergence between them: the critique of the grand partage is, for Descola as for Latour, inextricably epistemological and political.
The critique they level at the social sciences is epistemological, in the sense that it aims to make them better at describing situations and the kinds of relations between humans and non-humans with which they contend. But it is also political, because it shows that these sciences have, in a sense, been the objective allies of a modern project of mastering the world.
By relegating nature to the background of social relations, they would have provided an indirect theoretical support (and probably involuntary) to the modernization project, particularly in its productive dimension, based on the exploitation of what was regarded as nature: which is merely a “material backdrop” to social life, to borrow a phrase from Latour, and which can be seen as a simple ensemble of resources.
This critique, both epistemological and political, is accompanied by a desire to foster ecological reform within the social sciences to respond to this diagnosis.
Here again, the two authors converge. Descola acknowledges, in an interview, that Latour showed him this path by making him recognize the ecological reach of his own work.
Can you return to the notion of collective in Latour and Descola, a concept shared by the two thinkers but whose definition seems to vary at times? It seems characteristic of this co-construction of Latourian and Descolian concepts.
Descola and Latour share a critical project, which invites us to bend the historical trajectory of the social sciences, but they also share an ambitious constructive aim: to replace the flawed categories of naturalism with a new set of fundamental concepts. The notion of “non-humans” provides an important example. In Latour’s usage, it temporarily replaces that of nature, before the author ultimately judges it to be overly anthropocentric. In Descola’s usage, it takes hold more durably.
This concept is, in some respects, counter-intuitive. It testifies to a kind of pure anthropocentrism: it is as anthropocentric as one can get, since it defines beings by their difference from humans. Yet it also enables us to view beings formerly lumped together under the indistinct banner of nature as a set of actors with whom social relations are formed (hence the importance of the plural “non-humans,” opposed to the singular “nature”). This is the common ground in how the two authors use it.
Latour, however, will abandon it precisely because of the paradoxes inherent in the notion I just highlighted. Descola, on the other hand, almost certainly keeps it because it functions well within a structural combinatorics that rests on simple oppositions: continuity/discontinuity, interiority/physicality, humans/non-humans…
The notion of the collective thus stands in counterpoint to that of non-humans. It replaces the idea of society, but it also opposes nature. What it takes on, in my view, is the dimension of a common enveloping or totalizing field that unites these two notions.
It is therefore no longer a matter of opposing a nature composed of material beings to a society formed by human individuals, but to consider a single collective of actors (humans and non-humans) situated on the same plane of reality. This innovation belongs first to Latour, who proposes to use the term “collective” to designate ontologically more heterogeneous ensembles than societies or cultures, which, in his view, grant primacy to humans.
How are these collectives formed? What role do non-humans play in this constitution?
We can answer at several levels, but one thing is certain: in Latour’s view, the construction of the collective passes notably through science and technology.
In modern societies, these practices stand at the forefront of the gradual integration of an increasing number of non-humans into social life. Pasteur, for example, is not merely a scientist uncovering an external reality for Latour; he is the agent of a “socialization” of microbes, which, through him, become actors in the common world.
That does not mean Latour denies the independent existence of microbes, as some have accused him of saying. He has made provocative remarks on this question, but never questioned that existence. Pasteur does not create microbes ex nihilo: only those who oppose science studies have thought that constructivism would go as far as such absurdities. There is indeed a scientific discovery in the sense that Pasteur produces new facts involving preexisting entities.
Yet, simply by the fact that Pasteur discovers them, microbes are not exactly the same, since we now know they exist and we must live with them. It is in this sense that Latour can say that science consists of the socialization of new entities into the collective.
Michael Fœssel delivers a critique of Latour and Descola concerning the finitude of the world and the ontological indeterminacy. This echoes the question of the constitution of collectives, especially when you write that we must “return to the work of world-making the ontological indeterminacy with which naturalism partly deprived it.”
This critique is ultimately Kantian. According to Fœssel, by returning us to the world’s finitude, these authors would make us lose the world’s infinity as an horizon of historical projection into a future that we would be free to construct collectively. This horizon would constitute one of modern political humanity’s grand achievements.
Fœssel plays here with Alexandre Koyré’s famous title: From the Closed World to the Infinite Universe. He essentially reproaches Descola, and even more so Latour, for performing the inverse movement: we would now be returning to a closed world, prior to political modernity. There could therefore be a tension between the injunction to respect planetary limits, on the one hand, and the ideal of autonomy and self-determination, on the other.
Latour has contributed to a volume edited by Émilie Hache, aptly titled From the Closed Universe to the Infinite World. Fœssel’s critique is certainly not without foundation. It should be noted, however, that returning to Earth, a stem more Latourian than Descolan, also implies rediscovering the surprising and, legally, infinite character of what we make with the world: ecological crises are also a form of ongoing relearning of the attachments necessary for the possibility of the future.
The aim is therefore to preserve a new form of indeterminacy: to relearn what the world is made of. This indeterminacy perhaps resides less in projecting into the future than in the rediscovery of a common space that we sometimes do not know with whom and with what we share.
Hence the idea, frequently advanced by Latour, of a shift in primacy from time to space. We would thus have moved from a modernity whose priority was to project itself into the future to the necessity of being contemporaneous with our own space.
This leads us to a first fundamental difference, which concerns their relation to naturalism: Latour seeks to “dissolve” it, while Descola aims to “provincialize” it.
So far, we have indeed described a set of convergences that explain why Descola and Latour are often associated in the reception of their works. But their paths diverge on the following question: to what extent does the grand partage really govern our societies?
Par-delà nature et culture (Beyond Nature and Culture) (2005), published fourteen years after We Have Never Been Modern, can be read as a response to the latter, even if that is just one of the many theoretical operations of the book. Significantly, one of the texts Descola dedicated to Latour is titled How We Became Modern (2016). It seems to me that much of the work can be understood in light of this reversal of Latour’s initial proposition. It is by no means a book written against Latour, but viewing it through this prism helps reveal its singularity in relation to the gesture of We Have Never Been Modern.
For Descola, there is thus a provincialization of the grand partage: our cosmology is relativized and placed among the set of cosmologies documented by ethnological knowledge. It aims first to show that “nature” does not exist everywhere or at all times, and that the modern separation is only one of the possible ways of distributing the fundamental properties of human and non-human entities.
Nevertheless, in Western societies, it is quite pertinent to speak of nature, insofar as this category guides the way we have built our knowledge and how we organize our relations with the other beings that populate the world.
Naturalism thus possesses an undeniable efficacy, which Descola’s comparative gesture helps to unveil: by placing our cosmology in contrast with all others, it reaffirms its substance. We can then consider that we have truly “been modern,” in the sense that this cosmology has genuinely shaped our societies.
Latour, for his part, adopts a different stance toward this cosmology.
He likewise acknowledges the existence of this cosmology, but goes so far as to dissolve it. The central argument of We Have Never Been Modern is that we have never practiced what we preach.
The paradox at the heart of this work is that the separation between nature and society which our discourse claims to uphold, in practice and covertly, multiplies interdependencies between humans and non-humans. The grand partage in its official form thus enables, under the surface, its own opposite: a growing mobilization of non-human entities within societies. In other words, for Descola, the separation between nature and culture is effective, in the sense that it deeply structures our categorical topology and our collective organization, whereas for Latour, it is indeed effective, but only insofar as it constitutes the condition of possibility for its own practical denial.
Latour even reverses the exceptionalist tale of the grand partage. If there are societies that have not implemented the separation between nature and culture, these are modern societies themselves. We did not merely refrain from carrying it out: we have performed it less than others.
There lies a fundamental divergence between Descola and Latour, which has occupied them a great deal and seems never to have been resolved.
From the 2000s onward, the texts Descola and Latour dedicate to each other increasingly take the form of a critical dialogue. In the beginning, they cite each other essentially as allies, which they have nevertheless remained. But this disagreement fed, for about twenty years, an open debate.
To continue on this point, why does Latour ultimately defend the idea that we have remained analogists? It is interesting that he acknowledges and uses Descola’s categories, while reinterpreting them.
As I mentioned, the divergences between Descola and Latour are less well known than their convergences. It is even less known that their discussion evolved progressively and that their disagreement about the effectiveness of naturalism manifested itself in the terms of Descola’s own system.
Recall that Descola distinguishes four major ontologies. Naturalism opposes animism, which constitutes its perfect inversion: unlike naturalists, animists attribute interiority to non-humans, while locating the discontinuity with humans at the level of physicality.
Yet Descola also describes another ontology, analogism, characterized by a double discontinuity of physicalities and interiorities. In such an “ocean of singularities,” the order of things depends solely on humans’ work of analogy-making.
The Latour–Descola debate gradually shifted. The question is no longer merely whether we have truly been modern, but whether we have remained analogists.
There are at least two reasons for this shift.
The first lies in the internal logic of Descola’s system, according to which the societies in which naturalism first appeared were previously analogist. In this framework, refuting the effectiveness of naturalism necessarily leads to the conclusion that we are ultimately still analogists. It is a demand of his combinatorics, insofar as it unfolds historically.
The second reason is that this shift presupposed Latour’s acceptance of the validity of the analogism concept, notably in describing premodern Europe, which he seems to have done.
An ambiguity remains, of course: does Latour agree to enter Descola’s categories for the sake of a debate, or does he subscribe more deeply to the very idea of analogism?
I would lean toward the second option, because certain structural elements of analogism recur in Latour’s ontology itself, such that, viewed through Descola’s typology, his work resembles, in many respects, a form of Neo-Analogism.
That is a point I emphasize in the book: we link the two authors to what has been called the ontological turn in anthropology, but, in Latour’s case, there is at least as much an ontological turn in anthropology as an anthropological turn in ontology. One might try to place Latour’s stance within Descola’s ontological chart, while Descola does not intend to produce an ontology of his own, but rather a cross-cultural anthropology of existing ontologies. For Latour, anthropology remains a detour for reforming philosophy.
So what is this Latourian ontology?
It’s a broad question, but we can try to sketch the traits that bring it close to what Descola calls analogism. Notably, there is a simple principle: in the analogical world, continuity between elements is never given in advance. It must always be produced by a work of relating beings through arts, through knowledge…
The analogical world is thus fragile: continuity is always achieved against a backdrop of discontinuity, and order against background disorder.
This is an axiom that runs through Latour’s entire oeuvre. His thought has a metaphysical dimension, which I will not develop here, but which is distilled, notably in two major works: Irréductions and An Inquiry into the Modes of Existence.
Throughout this trajectory remains the idea that the world is first filled with discontinuities, which we must cross by producing continuity. This aligns with a structuring motif in the Inquiry into the Modes of Existence: everything that lasts lasts because something that does not last lasts—that is, it endures through processes that transiently come and go.
Thus Descola’s analogism has a deep affinity with Latour’s ontology. This probably explains why Latour accepts debating whether we have remained analogists.
If Latour can be associated with what Descola calls analogism, what then is their debate on this point about?
The difference lies in the fact that Descola maintains that a transition occurred in Western modernity, moving gradually from analogism to naturalism. Latour, by contrast, reframes the thesis of We Have Never Been Modern by arguing that we have never been naturalists, and that we have therefore remained analogists.
In the book, I show that moving the discussion toward the question of analogism also shifts it from a synchronous to a diachronic plane: it now concerns the stability and longevity of the corresponding ontologies.
One of Latour’s essential arguments rests on an internal contradiction within Descola’s ontological scheme, which claims to be built on a group of structural transformations.
According to this taxonomy, there are only four possible ontologies. In a structural frame, these provide as many solutions to the same logical problem: the continuity and discontinuity between humans and non-humans. Since there are only four ways to distribute these continuities and discontinuities, each ontology should actualize itself in a relatively regular fashion throughout long histories of societies.
That is precisely what allows Descola to find, across the whole body of ethnological knowledge, analogist, animist, and totemist collectives in cultural areas as diverse as they are. He acknowledges, however, that naturalism constitutes an exception: it would have only a single historical cradle of emergence, namely Europe’s modernity. If one then observes this ontology in other parts of the world, it would be due to diffusion rather than independent appearances.
Latour notes, however, that this situation poses a structural problem. If naturalism is a logically possible option just as much as the other ontologies, it should normally actualize in several foci. It actualizes only once and lasts only a few centuries.
According to Latour, does naturalism thus occupy a special place among the ontologies described by Descola?
Latour’s argument is indeed that naturalism is not an ontology like the others. It would rather be a historical aside, lasting only a few centuries, while other societies remained analogist or animist for millennia—a passing that, moreover, is already crumbling.
This argument, among others, seems to attest to the value of reading Descola and Latour against each other, and sometimes against one another.
These two authors have been extremely influential and remain the subject of many criticisms today, sometimes uncharitable and more destructive than understanding. Their dialogue, by contrast, is stimulating because they take one another very seriously: each often seeks to challenge the other from the very premises of that other’s thinking.
Latour strives to produce a critique of Descola from Descola’s own categories (pushing to the end the logic of structural combinatorics), while Descola takes aim at Latour with Latour’s own weapons.
In Latour, one finds what could be called a structural critique of Descola’s post-structural anthropology. He confronts the exceptional status of naturalism with the demands of a structural combinatorics and highlights the contradictions this creates within the Descolan system.
What about Descola’s own immanent critique of Latour?
Descola proceeds in a similar fashion. He addresses Latour with a critique that could be described as pragmatic, given that Latour is one of the historical figures associated with what has sometimes been called “pragmatic sociology.”
Latour sought to distance himself from Bourdieu’s sociology by giving more credit to actors’ sociological capacities. It was a move to shift, following Luc Boltanski’s distinction (another central figure in this intellectual space), from a “critical sociology” that reveals what actors do without realizing it, to a “sociology of critique” that takes their own interpretations more seriously.
That is precisely where Descola detects a tension in Latour. On the one hand, Latour claims to adopt an approach that takes actors seriously; on the other hand, he does not seem to give much weight to the grand partage as something that structures their practices. He treats it as an artificial construction that enables him to do the opposite of what it proclaims.
Descola thus suggests that his own system is more understanding and more forgiving, because it takes into account the effects of the grand partage on how modern societies have built their regimes of knowledge and their political organization. He so addresses Latour with a critique that almost approaches what Latour directed at Bourdieu.
To sketch it bluntly, one could say that Descola offers a pragmatic critique of Latour’s pragmatic sociology, while Latour, conversely, offers a structural critique of Descola, who is indeed an heir of Lévi-Strauss’s structural gesture.
As I noted earlier, this is what makes their confrontation particularly rich: each tries to situate themselves within the intellectual coordinates of the other. This approach distinguishes itself from critiques that merely claim Latour is wrong because he neglects capitalism, for example (which would, by the way, oversimplify his thought). Make no mistake: it is entirely fruitful, and perhaps even necessary, to produce eco-Marxist critiques of Descola and Latour, but they do not exhaust the full range of tensions that can be discerned in these two anthropologies of nature.
A key point in your book is what you call the “complementarity of critique” between the two authors. How then to think Descola through Latour, and vice versa, Latour through Descola?
The two authors are complementary precisely because they critique each other and each highlights what remains in the shade in the other. In the book, I try to show that, in the final analysis, their anthropological projects are very different, insofar as they occupy, in a certain sense, opposite poles of the comparative gesture.
Descola embarks on a search for maximum alterity. He goes through the study of societies that, when he encountered them in the 1970s, were still relatively insulated from exchanges with Western modernity and globalization.
He deliberately sought to study collectives as distant as possible from our economic development and cosmology. It is thus that he encounters peoples whom he ends up calling animists, whose worldview constitutes almost a total inversion of our way of conceiving relations between humans and non-humans.
Descola thus embodies a fairly classic figure of the anthropologist of the “faraway”: the aim is to compare oneself with an otherness that is radically different. In Latour, the project is different: he intends to bring the anthropology of the tropics back to critique and reorient its gaze toward our own societies.
In We Have Never Been Modern, Latour revisits elements of Descola’s approach but argues that we must draw lessons from the study of other societies to understand our own. We must observe them as anthropologists, treating them as strange as those we have learned to describe. This strangeness is particularly situated in science, and this already in Latour’s first book, co-authored with Steve Woolgar: Laboratory Life (1979).
For Latour, if we want access to our own anthropological singularity, we must examine in the most prosaic and detached way what seems self-evident to us: science, considered as the culmination of our rationality. We must return it to its ordinary character and place it within the spectrum of ways to produce relationships between humans and non-humans.
Latour is therefore an anthropologist who does not take radical alterity as his object, but rather ourselves. His project is not so much to study the gap between us and others as to probe the difference that separates what we do from what we say about our practices.
Both approaches seem complementary precisely because of this methodological difference.
In Descola’s case, one grasps what is singular in our cosmology, but also its persistence and its capacity to impede our struggle against the ongoing climate transition. In Latour’s case, one discovers that this cosmology does not permanently liberate us from the common lot of humanity. The ecological crises into which we are only just entering show that our supposed detachment from nature was but a fantasy, whose deleterious effects become more visible each year.