Manuel DeLanda has long been one of the most interesting (indeed, provocative) thinkers to work with Gilles Deleuze’s ideas — and to not just repeat Deleuze’s vocabulary and slogans, or apply them to a close reading of some work or artifact, but actually to rethink those ideas, and to rethink questions of history, society, and physical science with the help of those ideas. DeLanda’s latest book is quite refreshingly short and lucid, although it (rather immodestly) purports to offer nothing less than A New Philosophy of Society. (I should probably also cite the subtitle: “Assemblage Theory and Social Complexity”).
I find the book (like all of DeLanda’s work) extremely useful and thought-provoking, although my overall reaction is quite mixed. DeLanda actually does two different things in this volume. In the first two chapters, he argues, on philosophical grounds, for what he calls Assemblage Theory. The term “assemblage,” and the ideas behind it, are drawn largely from Deleuze, and Deleuze and Guattari; though DeLanda does not merely repeat Deleuze, but reformulates his arguments (and terminology) in some very crucial ways.
The remaining three chapters of A New Philosophy of Society are more empirical; with the help of theorists ranging from Erving Goffman to Fernand Braudel, DeLanda draws on the principles developed in the first two chapters in order to give a schematic account of how societies work on several levels, from that of the individual person (and even the sub-personal) and the “networks” in which he/she is directly involved, up to the larger aggregates which are “organizations and governments” on the one hand, and “cities and nations” on the other.
I like DeLanda’s basic argument: which is to insist on the exteriority of relations. Traditionally, positivist, atomistic thought has pretty much denied the importance of relations between entities: the entities themselves are the absolutes, and all relations between them are merely accidental. Thus neoclassical economics adopts a “methodological individualism” according to which “all that matters are rational decisions made by individual persons in isolation from one another” (4). On the other hand, what DeLanda calls the “organismic metaphor” (8) asserts that entities are entirely defined by the totality to which they belong, entirely constituted by their relations: “the basic concept in this theory is what we may call relations of interiority: the component parts are constituted by the very relations they have to other parts in the whole” (9). Hegelian thought is the most powerful example of this tendency, thought Saussurean linguistics and the “structuralism” influenced by it could also be mentioned.
Now, the partisans of both these views usually claim that the two opposed positions are the only possible ones: there are no alternatives. Partisans of methodological individualism simply deny the existence of units larger (or smaller, for that matter) than that of the “individual” (or at most, the patriarchal nuclear family): they see such formations as being metaphysical abstractions with no objective validity. Hence Margaret Thatcher’s notorious statement: “There is no such thing as society. There are only individuals and families.” Of course such “methodological individualism” is absurd, since it is contradicted by everything in our minute-to-minute and day-by-day experience. We are never as isolated as methodological individualism assumes, and we probably couldn’t survive for very long if we were. The very fact that we use language, that we use tools and techniques that we didn’t invent from scratch ourselves, let alone that we use money and engage in acts of exchange, belies the thesis. It’s a curious paradox that the most rabid partisans of methodological individualism tend to be free-market economists and rational-choice political scientists and sociologists, since their entire logic depends upon denying the very factors that make their arguments possible in the first place. But if you press the more intelligent methodological individualists, they will admit that their presuppositions are, indeed, “methodological” rather than ontological, that they represent a kind of abstraction, and that such a methodology, and such an abstraction, are necessary in order to avoid getting stuck in top-down, totalizing theories (their aversion to which is often justified with citations from Karl Popper, Isaiah Berlin, or Friedrich Hayek on the dangers of totalitarianism).
But on the other side of the divide, Hegelians and other proponents of the “organismic metaphor” are just as insistent that their systematic (or “dialectical”) ways of doing things are the only alternatives to the absurdities of atomistic reductionism. I can’t tell you how many arguments I’ve had with Marxists, Zizekians, and others over the years, who insist that my Deleuze-inspired objections to the very notion of totalization, or to the idea that events occur only through a dialectic of negativity (usually up to and including the “negation of the negation”), is untenable: they claim that, to reject these “relations of interiority” is ipso facto to lapse into the absurdities of positivism and atomistic reductionism. The same is true for partisans of various sorts of systems theory (all the way from followers of Niklas Luhmann, to devotees of the Lacanian Symbolic order), who tell me that I cannot escape their system, because anything I say against it already presupposes it, and is already positioned somewhere within it. (Hardcore deconstructionists, despite their denial of the very possibility of totalization or a coherent system, are nonetheless also in this camp: as they argue — just like Lacanians — that one can never escape the presuppositions and aporias of Language. Deconstruction is entirely a theory of relations of interiority, even though it recognizes that such relations are never completed but always still in process).
What DeLanda says — which is indeed what Deleuze said before him — is that we need not accept either term of this binary (nor need we be stuck in the aporia of shuttling endlessly between them). What Deleuze and DeLanda offer instead is not the golden mean of a “Third Way,” but rather a move that is oblique to the very terms of the opposition. What does it mean to affirm the exteriority of relations? As DeLanda explains it, an entity is never fully defined by its relations; it is always possible to detach an entity from one particular set of relations, and insert it instead in a different set of relations, with different other entities. For every entity has certain “properties” that are not defined by the set of relations it finds itself in at a given moment; rather than being merely an empty signifier, the entity can take these properties with it, as it were, when it moves from one context (or one set of relations) to another. At the same time, an entity is never devoid of (some sort of) relations: the world is a plenum, indeed it is over-full, and solipsism or atomistic isolation is impossible.
Put differently, no entity can be absolutely isolated, because it is always involved in multiple relations of one sort or another, and these relations affect the entity, cause it to change. But this is not to say that the entity is entirely determined by these relations. On the one hand, the entity has an existence apart from these particular relations, and apart from the other “terms” of the relation (i.e. apart from the other entities with which it is in relation) precisely insofar as it is something that is able to affect, and to be affected by, other entities or other somethings. On the other hand, what the entity is is not just a function of its present relations, but of a whole history of relations which have affected it — or of “aleatory encounters” (as Althusser might say) with other entities, over the span (temporal and spatial) of its existence.
DeLanda distinguishes between the properties of an entity (which are what it takes with it to another context) and the capacities of that same entity (its potential to affect, and to be affected by, other entities). “These capacities do depend on a component’s properties but cannot be reduced to them since they involve reference to the properties of other interacting entities” (11). An entity’s capacities are as real as its properties; but we cannot deduce the capacities from the properties; nor can we know (entirely) what these capacities are, aside from how they come into play in particular cases, in particular relations, in particular interactions with other particular entities.
What this means is that entites of various sorts and scales — persons, but also (to use DeLanda’s own list) networks, organizations, governments, cities, nations — are all entirely real. (DeLanda resists putting “society” in this list, because he fears that such a term implies the logical topmost point of a hierarchy, a category that includes everything. He insists that entities always come in “populations” — taking the term in the sense it is used in “population genetics” — so that there can never be one, all-integrating topmost entity. Though I take his point, I also see no objection, on his own principles, to talking about societies in the same way we talk about any other level of entities. More on that in a moment).
To say that both individuals and wider social formations (and narrower, sub-personal formations as well) are real, is to be committed to what DeLanda calls “ontological realism.” This is in opposition both to the neoclassical economists who only recognize the reality of the individual, and think that anything of broader (logical or social) scope is just a linguistic fiction; and to the Hegelians (and Hegelian Marxists, and perhaps Durkheimian sociologists as well) for whom only the social is real, and the individual is regarded as a linguistic or ideological fiction. This means also to think entities non-essentialistically (every entity is historically contingent: its existence and its properties cannot be inferred, let alone be deduced logically; for the entity exists only as an effect of processes over time which could have gone otherwise). And further, it is to recognize that all entities (not just living things, but everything) are mortal — they have dates of coming-into-existence and passing-out-of-existence, they are not platonic forms but occupy a finite and bounded stretch of time and space).
To my mind, this overall ontological argument is what is important about DeLanda’s work, rather than the particular way he constructs a theory of “assemblages” — using terms from Deleuze and Guattari, but altering and simplifying them when necessary — in order to meet the requirements of his ontology. I think that other formulations that meet these requirements are possible — and indeed, that some alternative formulations are preferable. In particular, I would point to Whitehead’s metaphysics, which I think is better (more useful, more capacious, more cognizant of change, and more open to possibilities) than DeLanda’s. Whitehead also theorizes the externality, and non-totalizability of relations: his “actual entities,” or “actual occasions,” are stubbornly “atomic,” while at the same time relating to, and influenced or affected by, other entities. Whitehead insists, both that no entity can be what it is in isolation from all other entities, and that no entity is entirely determined by these other entities: this margin of indetermination, which is the “freedom” of the individual entity, but it is better described as a contingent “decision” than as a set of “properties.” Also, Whitehead distinguishes between these “actual entities” and what he calls “societies,” or aggregations of entities that possess spatial extent and temporal duration (whereas the actual entities themselves in a certain sense produce temporality and spaciality, rather than being located within them).
The distinction between “actual entities “and “societies” would seem to violate DeLanda’s dictum of a “flat ontology” (all entities at all scales have the same degrees of reality and sorts of properties) — though the “flat ontology” does apply for whatever we encounter in lived experience, since everything of this sort is a “society.” This seeming violation of the principle of flat ontology is something for which Whitehead has often been criticized. But what he gains by posing his ontology in this way is, among other things, that he is able to talk about change in a way that DeLanda is incapable of, and that he doesn’t need to share DeLanda’s phobia about extending his ontological realism to “society” itself. For DeLanda, saying that relations are external rather than internal means renouncing any sort of holism; Whitehead, however, is able to cheerfully embrace holism while at the same time posing the “whole” in such a way that it is irreducible to closure or totalization or full internal determination. For Whitehead’s actual entities are themselves events; whereas, for DeLanda, as much as he wants to proclaim the importance of (contingent) event over (fixed and closed) structure, events are still things that ‘happen to’ entities, rather than entities themselves. (For Whitehead, the things to which events happen are “societies” — which at the same time are composed of nothing more than these events, and the “routes of occasions” that link them together).
Note for further elaboration: a lot of this has to do with the way that DeLanda, through Deleuze, is ultimately channelling Spinoza, to whom the language of capacities to affect and be affected is originally due; and also Hume — again via Deleuze’s reading — in order to account for how the individual person exists as an “emergent property” of the assemblage of a quantity of impressions, ideas, and chains of association. Now, Whitehead writes a lot about Spinoza and particularly Hume, recognizing their importance but also their limitations, which have to do with the fact that neither of them think sufficiently in terms of events. Spinoza fails to think the event because of his absolute monism; Hume, because of his denial of “causal efficacy”, and development of a theory of mind entirely in terms of “presentational immediacy.” Where Deleuze uneasily juxtaposes Spinoza and Hume with Bergson, and DeLanda entirely ignores the Bergsonian side of Deleuze in favor of the Spinozian side, Whitehead is the one thinker who actually does — much better than Deleuze — integrate (using this term in the mathematical sense) Spinozian and Bergsonian imperatives. This needs to be explained further, in conjunction with Whitehead’s aphorism that “there is a becoming of continuity, but no continuity of becoming” (Process and Reality 35).
A lot more needs to be said about Whitehead — indeed, this is what I am hoping to write about in the months to come (for several talks that I have promised to give, and essays that I have promised to contribute to various anthologies). Hopefully my tentative formulation here of what he is doing, and how he both coheres with, and differs from, DeLanda, is not too cryptic.
The fact that DeLanda doesn’t say enough about, or give the right space in his theory for, becoming and events (which I think, would require more of a Whiteheadian language and approach than he is willing to adopt) leads to the other problems I have with his work. There’s a sense in which DeLanda adopts an overly cut-and-dried schematicism (dare I even say scholasticism?). Every phenomenon he discusses is classified in terms of its material and expressive components, its potentials for territorialization and deterritorialization, and so on and so forth. It’s quite disappointing, after DeLanda announces the openness that comes from rejecting the organicist internality of relations, that everything fits so neatly into these little boxes. Of course, this is always the problem with schematicim (going bck to Kant); but DeLanda seems inordinately and disappointingly reductive in the way that he schematizes. Deleuze is a big schematizer himself, of course; and it is important in certain instances to emphasize his schematicism, against those who see him (for good or for ill) as an undisciplined philosophical wildman. But on the other hand, there is a certain delicious poetic quality to Deleuze, and this is something that his acolytes all too often miss; this poeticism is entirely lacking in DeLanda.
The further result of DeLanda’s schematicism, and his inability to think about becoming, is that his actual discussion of society, in the later chapters of A New Philosophy of Society, is disappointingly bland, and entirely devoid of any consideration of such things as power, domination, inequality, or the production, appropriation, and distribution of a social surplus (I use this latter formulation to encompass both Bataille and Marx). He simply dissolves such things into a general description of aggregations of various sorts; he mentions negotiations and disputes between groups over the allocation of resources, but ignores the fundamental dissymmetry (and thereby, the antagonism) that are crucial factors in all such disputes. And he simply leaves out of his account the ways that Deleuze and Guattari, and Foucault as well, are deeply concerned with these issues. Marxists and Zizekians would probably argue that DeLanda’s omissions in this regard are an inevitable consequence of his pluralism and non-totalism. But it seems to me, again, that DeLanda’s (often expressed) hostility to Marxist formulations is not an inevitable consequence of his ontology; and that adopting this ontology (or better, the Whiteheadian version of it that I have tried to point to here) actually has significant advantages for Marxist theory, as well as for much else. All this is something I can only assert for the moment — a lot of my effort these days is devoted to working it out. So stay tuned. In the meantime, and to summarize, I think that DeLanda’s book is enormously valuable for the way it works out, and states so clearly, its ontological argument — even if DeLanda’s more concrete development from his premises is enormously disappointing.
38 thoughts on “DeLanda, A New Philosophy of Society”
thanks for this. i was wondering whether I would have to read delanda’s new book before finishing the diss, but now I’ll let it slide until I have finished. I doubt it would have much relevance to car culture anyway, lol
on the concresence of prehensions as actual occassions versus the percipient arrangement of actual occasions as a society, whitehead he makes a distinction in the temporality of actual occassions versus societies:
“It is evident from this description of the notion of a ‘Society’, as here employed, that a set of mutually contemporary occasions cannot form a complete society. For the genetic condition cannot be satisfied by such a set of contemporaries. Of course a set of contemporaries may belong to a society. But the society, as such, must involve antecedents and subsequents. In other words, a society must exhibit the peculiar quality of endurance. The real actual things that endure are all societies. They are not actual occasions. It is the mistake that has thwarted European metaphysics from the time of the Greeks, namely, to confuse societies with the completely real things which are the actual occasions. A society has an essential character, whereby it is the society that it is, and it has also accidental qualities which vary as circumstances alter. Thus a society, as a complete existence and as retaining the same metaphysical status, enjoys a history expressing its changing reactions to changing circumstances. But an actual occasion has no such history. It never changes. It only becomes and perishes. Its perishing is its assumption of a new metaphysical function in the creative advance of the universe.” (Adventures, 204)
Relating to the ‘eventness’ of assemblages I thought the entry for Assemblage/Agencement by John Phillips in the recent theory, culture & Society ‘glossary issue’ or whatever it was called was very good on this (esp. compared to the other entries on ‘assemblage’).
Also, a last comment or question actually, do you think it is worthwhile thinking about Deleuze’s “baroque house” from _The Fold_ as something between the notion of the assemblage and Whitehead’s ‘societies’?
Ok, the very last comment, the obvious difference between Deleuze and Whitehead is that Whitehead doesn’t have an equivalent of the virtual, not in the sense it is developed in DeLanda from (the ‘late’) Deleuze as a kind of mathematical/complexity theoryfication of the virtual. The rough equivalent would be the neoplatonic eternal objects of whitehead to the ‘pure events’ of (the ‘early’) Deleuze. In Whitehead, the problem of integration is of conceptual prehensions (within which eternal objects are ‘ingressed’) with sensual prehensions (within which actual objects are ‘ingressed’). Isn’t there a tension in Deleuze’s work between a version of this integration in _TLoS_ and the differenciation of _D&R_ (which could be read in Whiteheadian terms as a kind of the exhaustion of actual occassions within societies)? (Which I think is a much more productive way to read the movements in Deleuze’s work compared to Badiou/Zizek’s reading that takes _TLoS_ as the final word.)
Thanks for all this, Glen.
I hadn’t remembered the passage you cite from Adventures of Ideas; I will have to look at all that more closely. Whitehead has been criticized for setting up a (supposed) dualism between actual occasions and entities that can be experienced directly; it is said that this reintroduces the bifurcation of nature — which is a little like saying that it violates DeLanda’s “flat ontology.” But I am inclined to think that the distinction is necessary, just as the atomization or quantization of reality is necessary (this was an argument I had with Sha Xin Wei, who argues that Whitehead’s mathematics would work better if a continuum were assumed), in order to avoid the sort of undifferentiation (if everything is “different”, nothing is) that Bergson perhaps lapses into, and also the ultimate blandness of DeLanda’s categories and descriptions.
Thanks too for mentioning the baroque house in The Fold; I have to think more about that as well (especially given that the chapter in The Fold is the only place where Deleuze writes at any length about Whitehead).
And the question of how to relate Whitehead on potentiality via eternal objects to Deleuze’s sense of the virtual is something I have been thinking about a lot lately. I will have to think more too about your suggestion re: integration and differenciation, because this is something I don’t really understand. I’m inclined to think that the doctrine of eternal objects (as pure potentials) actually may obviate some of the difficulties I have with the Deleuzian virtual…
(As I write this, it is too late at night for me to have any more focused comments…)
Very cool post. Dense, but ultimately clear and new to me. Thanks so much. You perhaps won’t be surprised to hear that your anticipated response from “Marxist/Zizekians” to Delanda very much mirrors the actual response (of late) of Marxist Zizekians to Badiou.
curious about this resistance to ‘doing philosophy’ – do you think this is it? the failure to engage whitehead as an effort to stave off metaphysics as such
This is a useful review – thanks.
But – belatedly – I have a question/comment about deconstruction in relation to assemblages. DeLanda doesn’t mention Derrida in his book, and deconstruction is invoked here only in the claim that it “is entirely a theory of relations of interiority, even though it recognizes that such relations are never completed but always still in process”.
So – not ‘entirely’ a theory of interiority…
More to the point, though, the point made here – that “an an entity is never fully defined by its relations; it is always possible to detach an entity from one particular set of relations, and insert it instead in a different set of relations” – is ‘pure’ deconstruction. In fact it is its starting point – for example, in the idea of iterability that Derrida elaborates in ‘Signature Event Context’, where “a written syntagma can always be detached from the chain in which it is inserted or given”, and so can always be ‘grafted’ onto other chains.
For Derrida, it is this possibility which means, not only that no context can ‘totally’ determine the sign/element of the ‘chain’, but also that – in its detachability, in its iterability – the sign never fully belongs to the chain or the context in which it appears. (That is: if the sign is to remain detachable, it must always ‘stand out’ from, and so not fully belong to, that context.) For Derrida, then, there is no context in which ‘the’ sign/element ‘itself’ appears. Hence, there can be no privileging of any ‘proper’ – or original – context; such that the sign/element always – and even ‘originally’ – appears only as a repetition of ‘itself’.
There is always context; but it is never determinant: “there are only contexts without any absolute center or absolute anchoring”. So, for Derrida too, it is in this respect that the sign/element evinces a relative autonomy. And Derrida calls it ‘resistance’ in a sense that may perhaps be compared to Whitehead’s ‘â€œfreedomâ€ of the individual entity’.
So, compared to DeLanda, what we got from Derrida is, it seems to me, a more rigorous thinking of how ‘assemblages’, if they are thought through this analysis, are not totalities (they are not, for example, ‘individuals’, or ‘wholes’, as DeLanda often says). And, along these lines, perhaps deconstruction might also provide the basis for a more specific – if not strategic – approach to how assemblages are constituted (though never fully) in ways that have to manage/contain (but can never fully) the resistance/surplus of their elements….
Might such a richer account of assemblages, then, provide one way in which DeLanda’s later “description(s) of aggregations of various sorts” could avoid ‘blandness’, and perhaps come to attend to issues of power – for example, as these are raised as or around the question of the ‘center’ or the ‘anchoring’ upon which (despite their ‘assembled’ character) the very idea and reality assemblages must still depend?
thank you for your review,
I think your thoughts are spot on for the most part. after reading de landa’s other works I was really looking forward to his take on a host of social issues that are normally left to post-positivists (subjectivity, socialization, etc.) unfortuanately, you are right, the first few chpts. of the book give an excellent ontological description of what the social world is, but doesn’t tell us much about why it is so. I was greatly dissapointed with the end of the book, which seemed to leave most of the crucial questions about social change unanswered. however, de landa’s description of social assemblages is an excellent contribution to our knowlege base, and more theories can be built upon his work.
Thanks you for your review and to the commentariat for their reponses. Like many here I thought that De Landa’s construction of the Assemblages was very useful yet, on the other hand and also like the many here, I thought it lacked any useful embodiment of power into the contruct. A worker is an assemblage as is her employer, but when they are pitted one against the other the social construction of their intercation reverberates with the sound of power. My finaly conclusion was that, perhaps, realists like De Landa think that power is not part of the assemblage.
I agree with much of your review. I have to say that I’m less bothered by DeLanda’s schematism than you, largely because I think it needed to be strict in order to, dare I say, formalize Deleuze’s concept of assemblages.
As a social scientist, I turn to philosophy for useful concepts. The first time I tried to read ‘A thousand plateaus’ I set it aside after two pages as poetry at best, gobbledy-gook at best. A philosopher friend thankfully encouraged me to stick with it. Like DeLanda, I found ‘Empiricism & subjectivity’ the most useful text, as well as incredibly readable. In an interview DeLanda comments on Deleuze’s style. He says that Deleuze believes that it took 400 years to really understand Spinoza, so it’s fine if it take 400 years to finally understand him. However, DeLanda remarks, we don’t have 400 years, we need Deleuze’s thought now. I agree that the three chapters where DeLanda
applies his schemata are quite dry, but they are also incredibly unique. The use of the Deleuzian concepts he refines need not be as he did it, but the development of such a schemata was incredibly important.
Finally, as for the lack of talk of ‘surplus’ I see that as a mark in DeLanda’s favour and it of course fits with his empiricism bordering on positivism (in the sense that Hacking refers to himself as a ‘positivist’). The surplus is a fiction in the way it has been conceptualized and employed to death since Marx.
Must check when Stengers’ ‘Thinking with Whitehead’ will be finally translated and published (Harvard)….