In asking about the lives of “structure” and “relation,” I do not take a head on approach. Rather, I sidle up to the concepts. John Dewey, whose early twentieth-century method in these matters I admire, called it “a flank approach” when he took on the concept of “The State” in his 1927 The Public and its Problems (Dewey, 1927). Such a military metaphor is not for me. I prefer to imagine a child on a playground asking, “Can I play with you?” Thus, I begin by playing a game with other concepts: “assemblage” and “polity.” I could propose this game by analogy to mathematical method, imagining it as a peculiar wordy calculus involving differentiation and provisional (re)integration. But I prefer analogy to chemistry. Telling it as the chemical method of flocculation, I take the anthropologically entwined pair of concepts, “structure” and “relation,” lyse them so that they float apart and, in reacting with other concepts as little wordy stubs, change their form, and rise to the top in a process of analytic flocculation. In subsequently being skimmed off, these novel concepts become useful in catalyzing divergences.
Around ten years ago, I claimed that it is useful to think of collective method in knowledge and culture work as a complex form of assemblage. In developing this proposition, I worked with two quite disparate exemplars of such work: a nineteenth-century British scientific expedition and a twenty-first-century Indigenous Australian digitization project. Attributing a precise meaning to the commonplace term, assemblage, one that went on from the STS sense of sociotechnical bricolage, but also quite different than post-structuralist “assemblage theory,” my 2009 concept of assemblage envisioned complex material-semiotic entwining of two distinct moments of generalizing—a one-to-many, inductive form, and a whole-to-parts abductive form of generalizing. In this Field Note on the concepts “structure” and “relation,” I pick up that proposition, this time focusing on the work that concepts do to mediate relations in the workings of these complex processes of structuring assemblage.
Whereas ten years ago, I used two rather modest exemplars of what knowledge and culture work generates—a novel botanical taxonomic scheme and a local intervention in identity politics around Australian Indigeneity—here my exemplar is a much grander project. I point to a polity as generated in infinite complex happenings in collective knowledge and culture work, imagining an overall process of assemblage understood as the entwining of the dual generalizing moments of assemblage. In beginning on foregrounding the concepts, “structure” and “relation,” I retell my account of assemblage through telling the workings of a polity.
Before I plunge into that, let me say something about the idea of studying the lives of concepts, which is what I see myself as discussing here. I take the phrase “studying the lives of concepts” from Henning Schmidgen (2014), who uses it to sum up the work of Georges Canguilhem, a twentieth-century French historian of science.
Georges Canguilhem only rarely expressed himself programmatically on the history of science. When he did so, he left scarcely any doubt that to him the history of science was above all a history of concepts. In 1963 […] he wrote: “History of science cannot be a simple collection of biographies and even less a chronological chart decorated by anecdotes. It also has to be a history of the formation, deformation and rectification of scientific concepts”.
However, there is a big difference between what I do in studying the lives of concepts and what Canguilhem did. I study how concepts participate in knowledge and culture work in the wild. The ways concepts participate in scientific texts and theories is of interest, but this is just one aspect of the study of the lives of concepts as I see it, and rather a minor one at that. For many years I have focused on the work of concepts in the wild through telling ethnographic stories.
As I tell it, concepts, as knowers’ companions, participate in happenings of knowledge and culture work. In much the same way cats, as totems of witches, are said to be the companions of witches, so concepts are the companions of analysts. In witch stories, cats participate in magic happenings as extensions of witches, in the same way concepts extend the epistemic and cultural agency of analysts. To put this in the context of this Special Focus Section, it is participations of “structure” and “relation” as conceptual companions of analysts that I attend to here.
I propose that in knowledge and culture work in the wild, and in anthropological theories, “structure” and “relation” are inseparable from each other. My proposition about the lives of “structure” and “relation,” as they participate in knowledge and culture work, is that structure implicates relations and relations embroil structure; the concepts’ lives are deeply intertwined.
Lysing the Clotted Conceptual Assemblage of “Polity” to Disentangle (and Re-entangle) Structure and Relation
In offering a glimpse of the workings of “structure” and “relation,” the analytic story of the concept of polity as assemblage, which follows, disaggregates and disassembles. The exercise interrupts the bond between the conceptual companions of “structure” and “relation.” I am seeking to flush their entanglements out into the open. It is a way of reciprocally naming the relation between the two concepts, and this is the preliminary work we need to do in order to get a clear view of these concepts at work in analysis. I ask what is inside the concept “polity” and conclude by briefly considering what more is packed inside the concept of “democratic polity.” This analytic story of polity as assemblage introduces a complexifying modification of both “structure” and “relation” as concepts. In this first step, the more complicated reciprocal concepts of “structural relationality” and “relational structurality” come into view. This analytic story of the concept “polity” is a serial lysing into the “ana-,” the entities inside polity, which collective practices of doing a polity clot. In discussing the clotted entity of “polity”, I bring “structure” and “relation” into the foreground and show the forms in which they participate in the conceptual assemblage “polity.”
We can conceptualize a polity as embedding two core entities: the citizen and the sovereign. “Citizen” and “sovereign” are the structural concepts of polity. So, what is their relation? The sovereign rules many citizens, and to know those citizens, a census might be carried out. In the logic of the practices of a census, a series of concepts are invoked: child/family/a household/a quarter/a city, and so on. Through a series of enumerative practices, a precisely defined one—the individual embodied human—becomes a precisely definable many—a generalized unit, the sovereign (these days usually a nation state rather than a monarch). All children, families, households, and so on are the same in precisely being children, families, and households. They are some of the structures of a polity. Of course, within those samenesses many interesting differences can empirically be found—between children, between families, and between households—and much anthropological work has been, and remains, dedicated to articulating those differences. In the end, such differences are interesting to the extent that they will not go away; they cannot be explained away by reifying the structural concepts. In knowing a polity through a census, structural concepts, or the samenesses, are carried to the fore, and the reasoning process is induction. Structure and relation mediate a one-to-many form.
But, there is another way of knowing a polity where relational concepts and differences are foregrounded and the reasoning process is abduction. If instead of considering citizens, we look at what a sovereign must accomplish in order to stay a sovereign, the polity becomes evident as a vague whole thing to be (re)accomplished from many parts. Many different functional entities (re)generate citizens as many different governable entities: territories, militaries, tax-collecting organizations, order-keeping groups, hospitals. The conceptualized functional units differ, all work with particular practices, although there are enough samenesses between the groups to connect them as functional units of a particular polity, even if it is only that the officers of these functional entities wear the same token embroidered onto their sleeves. This sameness, explicitly made within difference, announces a relation. Again, much anthropological work details the working of such functional units of this or that sort of polity as well as the generation of their relations. What is usually of interest here is how myriad sorts of sameness might be contrived. Here “relation” and “structure” as concepts mediate a parts-to-vague-whole form.
Polity, then, is a conceptual assemblage that embeds a precise unified concept of sovereign, where citizens are inside the sovereign, along with a vague emergent concept of sovereign as a whole, where citizens are variously accomplished as the sovereign’s outside. If, with this understanding of polity, we wish to go on to define a parliamentary democratic polity, we might imagine precise sovereignty and vague sovereignty as set within the structural relation between the electorate and the political parties of a parliament, the voter citizens and the voted-for political parties. I could adumbrate the relational structurality and the structural relationality of the concepts “electorate” and “parliament,” in just the same way that I have for polity with its “citizen” and “sovereign,” but will refrain from doing that. You get the idea.
My account of a polity as a relational processual entity is far from original, yet my adumbrating it here is not with the intent to make a claim in political philosophy. Before me, Rancière, Dewey, Hobbes, and Aristotle have explained what a polity is in these terms. My point in telling this story about polity as assemblage is to show what is inside the concept assemblage. I foreground dual relational concepts, “relational structurality” and “structural relationality,” each one being completed in their entanglement in polity’s assemblage.
Pausing for a moment here, I feel I need to make a couple of comments on this process that I’ve glossed as conceptual lysing. First, let me comment on the relation of these two derivative concepts. We might be tempted to think of the relation between these concepts as one being the inverse of the other, but that is a misnomer signalling misunderstanding. The core of the concept of inverse is transposition—like sewing the left sleeve of a blouse into the right armhole. These two concepts (relational structurality and structural relationality) are reciprocals. Each can be everted into the other. Relational structurality is inside structural relationality and vice-versa; imagine lava and volcanic mountains, or more precisely, mobius strips.
Next, noting the proliferation of metaphors, I offer an aside about the complicated grammatical acrobatics I have deployed here in naming these two concepts. How is “a relationality” different from “a relation”? Or “a structurality” different from “a structure”? Starting with the noun “relation,” a fragment of an old word is added. Relation plus –al (from an old Latin word meaning “pertaining to”) turns the naming term “relation” into a qualifying adjective—something can be said to be relational; a shared mode of existence of two entities is commented on.
Then, in a second complicating grammatical step, -ity (a short form of a Latin word used to express a state or condition) is added; relational becomes relationality, naming the condition, or form of being that all relational things have in common. To give an example by discussing fingers: we might talk about a finger on one of our hands as having a particular relation to each of our other fingers (usually nine others). So, we can say that each and every finger is relational. Thus, relationality is a condition that all fingers share, although, as we know, every finger is different from every other finger. Albeit semantically complicated, relationality is just “a thing” that is proposed as being or existing in the world; a relationality is an entity just like each finger is an entity. As a thing, in English, relationality is said to be “abstract” because, unlike a finger, you cannot actually suck it.
Doing Things With Such Contrived Concepts
In my peculiar exercise in playing with words I have written about concepts—polities, assemblages, fingers, and grammar—and I have recklessly mixed my metaphors along the way. In a process I named conceptual flocculation, two derived concepts have floated to the top: relational structurality and structural relationality. In introducing conceptual flocculation, I suggested that the purpose and value of this analytic process lies in capacities that the derived concepts have for catalyzing divergences: not divergences from anything, but rather divergences valued for the reason that they (might) make a difference.
So now, in concluding my Field Note, I ask about using the conceptual tools that have come out of my odd little exercise in opening up the possibility of divergences in the entangled lives of structure and relation in anthropology. If this text were to become something more than a Field Note, then with the rather oddly paired concepts of “relational structurality” and “structural relationality” as my companions, I could make a reading of current anthropological theory.
What if, say, I took the work of Marilyn Strathern (linking her theory back to Radcliffe-Brown) and that of Philippe Descola (pursuing his connections back to Lévi-Strauss)? I would propose that since Strathern foregrounds relational structurality and backgrounds structural relationality, she is a relationist; Descola reverses this foregrounding/backgrounding relation, so he is a structuralist. When it comes to their epistemic practices, these anthropology theorists are very different. Of course, pointing to that is not news, albeit that there is a certain elegance to seeing the one as the reciprocal of the other. We might contrast the methodologies in this way: Strathernian theory enacts empirical relationism in contrast to Descolian theory, which performs empirical structuralism. We can accept that each in its own way is constructivist, yet also recognize the effect of the differing metaphysical commitments of these analysts: the former to a metaphysical dualism; the latter to monism.
But, to compare and contrast Strathernian and Descolian theory is not all I want to achieve with this exercise. My main purpose is to ask what an anthropology that abstained from backgrounding and foregrounding relational structurality and structural relationality would look like. Can we imagine an anthropology that is neither and both empirical relationism and empirical structuralism? Here we are feeling our way towards an analytic that has learned to keep both these derived concepts in the foreground, in generative tension. When we can do that, I propose that the name of the process would be relational experientialism. But why would we want to diverge towards epistemic practices that might properly be called relational experientialism?
As anthropological theories, Strathernianism and Descolianism offer readings of the past, with one emphasizing emergent relations and the other emergent structures. Both offer possibilities to imagine futures different than pasts. The problem with that, however, is that we are currently marooned in a regime of historicity that has rendered such imagining redundant. As French historian François Hartog (2015) points out, both futuring the present and historicizing it are currently beset with uncertainty. The world now seems lodged between two impossibilities: the impossible past and an impossible future. He proposes this aporia as a pause or a gap in time. What we need today is an a-theoretical anthropology of the present that might offer a basis to intervene in a knowingful doing of the present.
Here is Adam Tooze, a political historian, on our present.
Since its inception, neoliberalism has sought not to demolish the state but to create an international order strong enough to override democracy in the service of private property…
… As long as it remains at the level of abstract gestures…the impulse of resistance mirrors what it opposes. We are still not engaging with the actual mechanisms…What we need to revive is the impulse to know. The will to intervene. The freedom to choose not privately but as a political body…
…[We must address ourselves to what the world order of neoliberalism] seeks to obscure: namely the engines both large and small through which social and economic reality is made and remade, its tools of power and knowledge ranging from cost-of-living indicators, to carbon budgets, diesel emission tests and school evaluations. It is here that we meet real, actually existing neoliberalism—and perhaps may hope to counter it.(Tooze, 2018)
“Cost-of-living indicators,” “carbon budgets,” “diesel emission tests,” and “school evaluations” are “tools of knowledge and power” that are expressions of diverse, instituted epistemics. They name “clotted” sets of meaningful physical arrangements and routines that effect certain conditions on-the-ground. As expressions of epistemics, these “tools” are among the myriad objects of contemporary governance—concepts. They act both as mediators and intermediates in the purposeful work of contemporary governance. They govern partially as themselves as they simultaneously, and partially, transmit the intentions of a polity in conducting the conduct of a polity’s collective life—shifted today to the benefit of capital. To engage Tooze’s diagnosis, we need an anthropology of the present.
Relational experientialism, as I imagine it, is an analytic platform supporting ethnographic research and analysis of collective knowledge-making-and-doing in the wild. Knowledge generated with the support of this platform offers partial insight into and possibilities for analytically telling, the happening of a “here-and-now” as knowingful doing. Research projects mobilizing relational experientialism can be imagined as applications inquiring into various aspects of collective action in the present. Happenings of times and places in ethnographic research, understood as knowingful doing in going-on together, are researched as collective epistemics, with the aim of designing careful interventions. Of course, doing this with certain concepts as our familiar companions can be done in better rather than worse ways, so knowing how to discern and agree on the criteria for worse and better, is a crucial aspect of this work.
Dewey, John. The Public and Its Problems. New York: Henry Holt, 1927.
Canguilhem, Georges. Etudes d’histoire et de philosophie des sciences. 7th edition, 2nd printing. Paris: Vrin, 2002. Pages 226-273.
Hartog, François. 2015. Regimes of Historicity: Presentism and Experiences of Time. Translated by Saskia Brown. New York: Columbia University Press.
Schmidgen, Henning. 2014. “Concepts have a life of their own: Biophilosophy, History and Structure in Georges Canguilhem.” Inflexions 7, “Animating Biophilosophy” 62-97.
Tooze, Adam. 2018. “Neoliberalism’s World Order.” Dissent Magazine, Summer 2018.
 Schmidgen cites Canguilhem (2002), specifically the chapter entitled “La constitution de la physiologie comme science,” pp. 226-273.