4 EMERGENT EXPLANATION
Matei Candea and Thomas Yarrow
What is the relationship between ethnography and explanation? Is anthropological explanation necessarily explication?
Oriented by these questions, this slightly unusual chapter takes the form of a dialogue between two old friends. Its initial prompt was an exchange of book manuscripts in 2016–2017. Matt read a draft of Tom’s then-forthcoming Architects (2019), an ethnography of an architectural practice; Tom read a draft of Matt’s then-forthcoming Comparison in Anthropology: The Impossible Method (2018). Both were struck by the similarities and differences between the two texts.
Both books were animated by a shared sense of the nature and purposes of anthropological knowledge production. We both felt that a set of developments—including the increasing prominence of grant-based funding for anthropological research and university auditing of “research excellence”—have inflamed an existing anthropological passion for a certain kind of “pointiness.” The making of “take-home points”—preferably radically “new,” “groundbreaking,” and of international and interdisciplinary “significance”—is increasingly pushing out the slower, less easily transportable aspects of anthropological knowledge making.
Neither of us was surprised to find this shared ambivalence to singular, simplified argumentation. More striking was the fact that our accounts call for, and in their form exemplify, two precisely opposed explanatory orientations. While Tom’s is informed by an argument for the power and importance of the implicit in anthropological exposition, Matt’s book makes a case for the need to be more explicit about the purposes and limitations of our conceptual devices. These different concerns are mirrored in the form of the books themselves. Tom’s book is almost entirely ethnographic, an experiment in the backgrounding of theory. Matt’s book, by contrast, has no ethnography at all. In sum, the books are inside-out versions of each other.
Although neither of us took “explanation” as a key focus, the arguments just discussed bear directly on the question of what form anthropological explanation does and should take. Specifically, these relate to the ways in which anthropological explanations entail implication and explication of various kinds. Asking the question in these terms made us both wonder whether we had elaborated a shared sensibility in different directions or more profoundly disagreed. Is there an “explanation” within which both positions or strategies can sit? As we began to write this chapter, we were still not sure.
The form of the chapter tracks this emergent conversation. The first and second sections, written in our individual voices (the first by Tom and the second by Matt), outline the place where each of us started. The third section is organized as a turn-taking dialogue that sets off from these initial positions. It probes differences between our approaches, as a way to exemplify broader questions about the nature and value of anthropological explanation. The conclusion, like this introduction, is in a shared voice. Though we had not anticipated this at the outset, this dialogue led us to tease out a concept—that of “emergent explanation”—that we argue plays a particular role in anthropological thinking. Retrospectively, it became evident that the chapter is a recursive demonstration of the logic it helps to conceptualize: its form is also its finding.
Implicit Explanations
Architects is an ethnographic account of an architectural practice, focusing on the lives of ten architects and the work they undertake, mostly in the confines of a single office, based in the United Kingdom. Their comments on early drafts were not encouraging. “A bit dense,” as one of them put it, “my eyes slightly glazed over.” Another used architectural imagery to highlight a linked problem: “It’s as though you’ve constructed a building and left the scaffolding on,” he remarked. The “scaffolding,” by which he meant conceptual reflections and theorized arguments, seemed a distraction from the descriptive passages he found most engaging. The metaphor of the scaffold is drawn from his own professional practice and is also a reflection of the sensibilities that orient it. Architects, at least in this practice, spend a lot of time discussing “precedents,” drawing influences and inspiration from other designs, but in the final instance they are clear: a building cannot be explained; it has to speak for itself.
These responses made me question how and for whom I was writing, and provided the stimulus for an experiment in ethnographic form. The analogy is not precise but got me thinking: What would a description look like if conceptual engagements with other scholars were treated as “scaffolding”—enabling the construction of a descriptive object whose effectiveness depends on their ultimate removal? Perhaps the problem was not to explain their lives but to refrain from explaining them too much. In a sense there is nothing particularly new in this approach, good ethnography having long been recognized as a matter of “showing, not telling.”
Rewriting the manuscript, I aimed to downplay, footnote, or excise various elements of narrative scaffolding, including some that academic readers have routinely come to expect: broadly speaking, a theorized argument of a singular kind, explicating novelty against an already-existing set of conceptual positions. Other scholars have highlighted some of the linked changes associated with the reification of this academic form of writing: the rise of interdisciplinary research relates to a proliferation of perspective so that novelty must be more explicitly stated to stand out; processes of research audit, at least in the United Kingdom, are associated with definitions of “world-leading” research, more readily recognized through novelty staged argumentatively using established professional discourse; the rise of social media as a research tool likewise leads to a proliferation of voices, and the imperative to speak loudly in order to be heard.1 Writing is more often driven by argument, resolved as “points” with a singular focus that can be easily and quickly grasped by a readership with limited time and attention. Even as ethnography is often reified and romanticized in anthropological discussions, in practice the “showing” seems to be increasingly less valued than the “telling.”
With these thoughts in mind, the first plank of academic scaffolding I sought to remove was theory, in the specific sense of externally derived explanatory frameworks of a singular kind. “Writing is an exercise in humility,” writes Nigel Rapport: “Theory is proud in its claims at comprehension. But theory would nevertheless appear to be the principal means of misrecognition—not the reverse—in its making of the other into an object whose point is to prove that theory’s assumptions. Academia would seem prone to theoretical pride: trafficking in coherent stories and plausible interpretations. But … this is to bring an artificial order to a wild world” (2015, 681).
By implication, his target is “grand theory,” and its claims to what Dominic Boyer, James Faubion, and George Marcus (2015) elsewhere characterize as a “monopolizing epistemic authority,” an inherent asymmetry of knower and known. Rapport advocates the antidote to this, in writing that “eschews theory for a return to the everyday.” Arguably, he presents the relationship between academic theory and everyday life in overly binary terms: all descriptions must “tame” to some degree, simplifying even if only enough to bring particular forms of complexity into focus; all are oriented by more or less stated interpretive approaches, ideas drawn from other scholars or examples, that open up ways of seeing, even as they may close down others. Marilyn Strathern’s (particularly 1988, 1991) insistence that “theory” and “description” occupy the same conceptual plane highlights how good descriptions arise through the comparative lens of other people and places. “Theory,” from this perspective, is not a fixed set of ideas but the conceptual “remainder” of the descriptive act: how anthropological concepts are changed and extended in the act of describing particular circumstances. Still, the thrust of Rapport’s argument has particular resonance in the current moment: pulled toward the assumptions and expectations of fellow professionals, anthropologists, like other academics, are routinely drawn into explanation that often seems to move away from the concerns that animate the lives of those we seek to understand. Even those approaches emphasizing the interdependence of theory and description have more often emphasized the theoretical implications of descriptions than the descriptive implications of theory.2 My account was an attempt to move in the other direction, scaling back argument as a frame and focus of description. I hoped to amplify understanding of the complexity of architectural lived reality, to give more attention to those aspects that remain specific and inchoate, to dwell in architects’ own explanations of what they do and why. Whether or not successfully, I aimed to refuse the kinds of exegesis that would render these details as epiphenomena of my own explanatory theory.
Second, and relatedly, my approach involved the deliberate attenuation of explicit argument. Focusing on Godfrey Lienhardt’s ethnography of the Dinka, Michael Carrithers elucidates some of the elements that made the classic monographs of the middle of the twentieth century so compelling: “Lienhardt devotes his effort throughout to the knotty labor of finding the most felicitous way of characterizing the Dinka themselves, rather than adopting the established conceptual coinage of professional anthropology or engaging argumentatively with established professional opinions. He leaves us to infer his understanding of those other voices and how they might err” (2014, 136, emphasis added).
The vivid qualities of Lienhardt’s writing were as much a function of what he said as what he did not. Literary theorist Wolfgang Iser develops this point while discussing Virginia Woolf’s exposition of the role of the reader’s imagination in the work of Jane Austen: “[The reader] is drawn into the events and made to supply what is meant from what is not said. What is said only appears to take on significance as a reference to what is not said; it is the implications and not the statements that give shape and weight to the meaning. But as the unsaid comes to life in the reader’s imagination, so the said ‘expands’ to take on greater significance than might have been supposed; even trivial scenes can seem surprisingly profound” (1980, 111).
Imagination works through language as an interplay between explicit and implicit, revelation and concealment. Many good examples of this interplay exist in ethnographic writing, but the general shift to explication and argument has tended to leave less implicit.
A third and final form of textual attenuation relates to analysis. In her introduction to Marie Olive Reay’s book Wives and Wanderers, Strathern explains, “In the book analysis remains very largely off stage … and the pride of place is given to descriptions of people’s doings, as they apparently occurred, in story-like form.” Comparing this to ethnographic writing of a more conventional kind, she notes, “Much ethnography is seemingly written of the moment. Yet the moment in which the ethnographer writes is also turned to the ends of exposition, and conveying a sense of immediacy has to compete with that. The trade-off between immediacy and reflection, between what is observed and what is analyzed seems inevitable.” (2014, 46). If observation and analysis are inherently connected, Strathern highlights how minimization analysis has amplifying effects with respect to the capacity of observational description: without the framing post facto analysis of the observer, description captures quick changes from moment to moment, replicating the unpredictable qualities of social interactions: “An element in any ‘encounter’ is its unpredictability: people try to guess what will happen, watch how others behave, see how this or that person will react. The dynamic of the relationship makes everything for a moment unknown.” (2014: 48) Analysis, of course, is needed, among other reasons, to spell out what is meant from what is said (or not), the contexts through which words and actions acquire significance. The aim of minimizing explanatory analysis related to a desire to amplify those forms of explanation that are emergent within the ethnographic contexts described.
I am aware of the irony that my own explanation of the limits of singular argumentation itself takes a rather argumentative and singular form. I am also conscious of the contradiction of arguing for a particular kind of exemplification that I have not in fact exemplified. I explicate these explanatory orientations and aspirations in the knowledge I have often and perhaps always fallen short of them. I hope that the book itself goes some way to demonstrating what I have here hoped to explicate. But many other and perhaps better examples of this approach exist, including a number of ethnographies whose narrative forms and sensibilities have inspired my own (particularly Crawley 2021; Luhrmann 2012; Pandian 2015; Yaneva 2009). Beyond their obvious differences of focus and approach, what I take these to have in common is the productive sense in which an explanation can be implied through an ethnographic description. In all these accounts, description is its own tacit explanation in ways that reach beyond a conceptual explication.
On Explaining Ourselves
Comparison in Anthropology: The Impossible Method is an account of anthropological comparison. As in a grotesque mirror image, the book consists almost entirely of what Tom’s book seeks to leave out: theory, argument, and analysis. The book’s prime material—what it describes—are theories and arguments. Whereas other meditations on comparison have woven their arguments and exemplified them through a range of ethnographic and historical contexts (see, for instance, Van der Veer 2016), the gambit in this book is to take anthropologists’ own theories, arguments, and disagreements about comparison as the object of study and the source material. The book is, in that sense, a kind of historical ethnography of anthropology’s own conceptual practices—it is, as Strathern perceptively noted, “a report from the field” (2020, 118), the field here being anthropological theory.3 One reviewer at least was frustrated by the lack of ethnographic exemplification in the book (Gellner 2022)—but the point is precisely that the thickness of this particular style of ethnography is made out of what we normally think of as “theory.”
The book is a description of these theories and arguments, but it is also self-consciously and explicitly an analysis, leading up to an argument of its own—indeed an explanation. By analyzing the recurrent patterns that emerge from anthropological writings about comparison, the book seeks to explain why, despite seemingly endless amounts of methodological reflection on the subject, anthropologists appear to have little agreement about what comparison is and how one ought to do it; it also seeks to explain why, despite a recurrent self-definition of comparison as the very heart of the discipline, anthropologists so often seem to conclude it is de jure impossible and yet carry on doing it all the same.
Its first part traces in some detail the extended and often convoluted debates anthropologists have had for around 150 years about comparative methods, paying particular attention to the recurrent ways in which they have sought to compare different modes of comparison. Over and again, these comparisons of comparatisms have tended toward dichotomies, marking out one older and misconceived vision against a newer and better alternative. These paired alternatives (historical comparison vs. functional comparison, structural vs. typological, interpretive vs. positivist, etc.) are never quite the same, but the form of the argument recurs: there always seem to be two ways of doing comparison, and one (the new one we are proposing) is better than the other (the old one they have been attempting). The result of these constant theoretical-methodological revolutions is a space of argument that undervalues continuities, shared techniques, and heuristic moves that carry over from one form of comparison to the next. There is a tendency to reinvent the wheel.
It is also a space in which differences in purpose tend to get elided. As Lawrence Fisher and Oswald Werner wrote of a different set of debates, “Any brand of anthropology can be shown to be woefully deficient if the objectives of one program of explanation are substituted for those of another explanatory program” (1978, 195). Different modes of comparison are often aiming at different ends. To ignore this is to forget the fundamental distinction between critiquing another scholar’s goals and critiquing their devices. We too easily dismiss earlier visions of comparison on the grounds of what they couldn’t do, or of what they systematically did wrong, without due attention to what they were trying (and not trying) to do. In so doing, we are often missing or choosing to ignore all the caveats that earlier authors set up about the necessary limits of their comparative devices. Concomitantly, in proposing bright new alternatives, or in praising those of our friends or mentors, we too easily forget what our own cherished devices can’t do, or choose not to do.
Hence the book is, among other things, an argument for being explicit about the nature and crucially the limits of our conceptual devices. It is an argument for the value of explaining ourselves and our devices. The second part seeks to clarify this discussion by picking out some key formal properties of anthropological comparison, which orient the radically different uses to which comparative devices can be put. In so doing it focuses on—to borrow a central term in Tom’s ethnography—“the space between” these different purposes, the methodological space in which anthropologists’ comparative devices remain shared even as their purposes diverge.
This space can only be kept in view if we do away with the engrained mental habit, and scholarly convention, of taking things “with a pinch of salt.” This habit gives one key to the paradoxical way in which comparison seems to be simultaneously impossibly complicated and wholly self-evident. Most of us are more or less acutely aware of the heap of objections raised at some point or other against almost every aspect of anthropological comparison—from the problem of identifying units of comparison, to the possibility of commensuration, to the politics of comparative representation … And yet—there’s the paradox—we go on.
Thus we invoke cultural units, social groups or patterns of behavior, while all the time implying or stating that we are well aware that these are just convenient fictions and that reality is far more complex. We analogize entities while mentioning in passing that of course they are also, in other ways, profoundly different, or contrast them while gesturing to the fact that in many other ways they fade into one another. Some of us appeal to philosophically abstruse techniques for challenging the very grounds of what counts as an object or a relation, while all the while appealing to ethnographic particulars grounded in descriptions and local generalizations of the most conventional kind. At every turn, an implicit or explicit appeal to taking things “with a pinch of salt” keeps these contradictions out of view.
In one sense this is fine—such bracketing is unavoidable and productive. Comparisons can productively be imagined as bundles of heuristics that get jobs done, humble and unassuming techniques that churn away below the level of grand epistemological debates. These comparative moves, tricks, and fixes bracket extensively, they make no guarantees to absolute truth or exhaustiveness, and yet they keep the discipline going, keep it together, and produce exciting new work. It would be impossible to do any kind of intellectual work—or to live any kind of life—without bracketing. The vision of complete explicitness is a mirage.
There is a world of difference, however, between bracketing something and just forgetting about it. Heuristics are valuable primarily because we know when they fail (Wimsatt 2007). Or to put the point otherwise, in the language of politics rather than engineering, it is fine to exclude, black-box, and simplify as long as we have a path back to and remain responsible for what is being left out (Barad 2007).
In part because of the “pinch of salt,” anthropologists have too often taken the impossibility of comparison for granted and just “gotten on with the job” under cover of some vague caveats. The resulting landscape is one in which we seem to be forever saying things we don’t quite mean, to others who don’t quite mean them either, but often in different ways or for different reasons. It is this habit of taking things with a pinch of salt, as much as anything else, that contributes to the sense that if we really thought about it, comparison would be impossible—so best not think about it too much.
Being more explicit about the limits and exclusions of each of our comparative heuristics, by contrast, can lead us to layer and combine them into thicker and more intricate comparative arguments. This means taking a step back from the ultimate point or aim of any given comparison, to ask what another, differently constituted comparative device might add. When your attention is hooked by a difference, ask also about similarities (and conversely); when you find yourself looking at objects, ask about the processes and relations of which these might in another sense be the effects (and conversely); and so forth. This might seem like a counterintuitive procedure. Why not simply get to the point? What it speaks to is the sense that to be animated by a purpose, to set a course for a particular horizon, is not the same as just imagining one has reached it. In that latter belief lies the risk of dwelling in platitudes and truisms. By themselves, our horizons have little power either to convince or to illuminate. After all, we can already see them from here!
This normative valuation of intricacy echoes the normative principle of “robustness.” Robust combinations of heuristics are multiply redundant; they come at the same questions from different angles; they are interwoven in such a way that some can fail without sabotaging the entire enterprise. Robustness as articulated by engineering-minded philosophers of science such as William Wimsatt are implicitly wedded to one particular aim: the pursuit of the real, the objective, and the generalizable. The kind of robustness envisaged here, by contrast, could be deployed in pursuit of a broader range of aims. Indeed it would come in part from the way in which anthropologists aiming in radically different directions (toward generalization or critique, objective identifications or increasing self-doubt) work alongside one another and hold each other to account, not for their divergent aims but for their moves in the shared space of method.
Conversation
MC
In a way, the core of my argument in the previous section rejoins a claim by Fisher and Werner: “We take it to be essential and axiomatic in anthropology that one should ‘explain oneself’—by making explicit one’s objectives in explanation—before one advances explanatory statements” (1978, 195). Precisely because, as the introduction to this book argues, there is no single form of what might count as an “explanatory statement” in anthropology, it seems to me, intuitively, that being explicit becomes an essential component of any such statement. So is your argument for the value of the implicit an argument against explanation, or would you say there is such a thing—as your section title suggests—as an “implicit explanation”?
TY
I am not against explanation, nor am I opposed to explication. I want to highlight some of the routine elements of already-existing good ethnography that are elided or devalued where the emphasis is on explicated argument. This is partly a question of speed: various circumstances conspire to encourage us to write and to read quickly, to overlook subtlety, and to mistake the implicit for the absent.
I would say, by extension, that there is such a thing as an implicit explanation. This is at the heart of a lot of ethnographic writing and is both celebrated and overlooked in the disciplinary construction of ethnography as foundational. W. G. Runciman writes, “Primary understanding can itself, as always, be construed as explanation at another level.… [The statement], ‘He is doing a rain dance,’ answers, ‘Why is Mr Morley stamping about on the outskirts of Bournemouth with feathers on his head?’ ” (1983, 168). In this instance the description (reportage, in his terms) is a fairly straightforward explanation to a straightforward, if unusual, question. The account of Mr. Morley could be further developed by adding details that might help us to understand what he was doing and why, without being explicit with respect either to the question that is being asked or to how this gives an answer. It is in the evocation of these details and the nuance with which they are evoked that the nub of good ethnography lies—hidden, in plain and obvious sight.
Though he does not put it in quite these terms, Michael Carrithers helps us to see how the craft of ethnographic description is partly in what is kept back or withheld. He sees the aim of ethnography as “creating … imaginative leaps to reveal the dense habits, arrangements and reasonings, and the forces of mutual entanglement and necessity that motivate human beings” (2018, 225), and he highlights the importance of apparently trivial details in opening out these “other worlds” (compare Narayan 2012, chap. 1):
Much of the force of these minor appearances lies precisely in the resistance of these fleeting appearances to full understanding. They are somewhat explicable—this is a gourd, these are flies, this is a bird—but on the other hand their full meaning is withheld. Why are those flies there? What is that bird? How do these things have significance? This world gains its force in part through this resistance, a resistance anyone might meet when stepping into an unfamiliar scene, one which has some features that are understandable enough, but whose force of reality is amplified by those other features which are not, and which therefore challenge one’s certainties. (Carrithers 2018, 226)
After Writing Culture (Clifford & Marcus 1986), and quite possibly before, these kinds of “vivid touches” have sedimented as part of a routine way of writing: richly evocative writing is often a rather cursory and formulaic preface to the “substance” of accounts that spell things out in more singular, more explicit terms: at one level ontologizing these details to emphasize the singular and general connecting strands (explanations of general ways of thinking and acting over and above the details); at another, elaborating how these add up to a conceptual argument (explanations of why existing theories are deficient or in need of revision).
None of this is per se an argument against explanation. At the level of contextualization, of the micro-comparisons of “this and that” (in your terms), it is indispensable. And at this level my argument is for more of it: more and richer extrapolations of what is meant from what is said; of what is happening in the silences or the gaps between the words; of how one person might say one thing in one moment and appear to contradict themselves in the next. This is not straightforwardly an argument for detail or complexity. A lot of very detailed and complex ethnographies can be very boring, which is to say that they fail, in Carrithers’s (2018) terms, to “open a world,” and so fail to allow readers to reimagine their own.
Returning directly to your question, my problem is not with explanation as such. I am rather voicing some disquiet that certain forms seem to be privileged at the expense of others. I am resistant to the forms of exegesis where too much emphasis seems to be on using the particulars of others’ lives as illustrations of explanations of broader generalities (culture, ontology, society, and the like); I think there is a danger that we end up giving explanations that only make sense to those who are already part of a disciplinary conversation. This emphasis also leads away from the explanations that are there, complexly, more or less explicitly, in the words and deeds of those we seek to understand.
To turn this around as a question, I want to push you on your own formulation of the explicit. When you say that “the vision of complete explicitness is a mirage,” to what extent is that a lamentation in relation to a goal you nonetheless think we should be aiming for? Are there, in your view, ways in which the curtailment of some forms of explication can be productive? And if so, where would these limits lie?
MC
That’s a really good question—that caveat about the mirage of complete explicitness was in there precisely because I have a tendency to forget it. My weakness is the typical one of so many arguments in anthropology and beyond: one identifies a problem (in my case, “the pinch of salt”) and then tugs in the opposite direction with all one’s might, forgetting that one is pointing not toward an absolute good (as if it more explicitness were always better!) but toward a relative one: I think it would be nice to have a bit more explicitness, of a particular kind, in anthropology at the present time. But fundamentally, I think about explicitness what I think about comparison, or about explanation: to know how much is needed, and of what kind, one first needs to know what for, what the problem is. Explicitness is purpose-relative.
So to return to your question: When is the curtailment of explicitness good? My answer is, “It depends what you’re trying to do,” but that at least (what you’re trying to do) needs to be made explicit. In every description, some things will need to be left out, whether that be ethnographic description or an account of “the literature,” or even of one argument in one article. To say everything is to say nothing; even if it were possible, it would be boring! It might seem obvious—and a bit of a cop-out—to point out that the question of where and what to cut can only make sense in relation to an aim. What I would add is a plea for making the aim itself explicit.
And I say that, in turn, because my particular aim—here—is to point to the way in which anthropologists often talk (or shout) past each other because they are not being explicit about the extent to which they are just trying to do different things. But if I know roughly where they’re heading, what their purposes and explanatory commitments are, it gives me a handle on what might be left implicit. It also allows me to have a conversation with these alternative accounts—maybe even to put them to use to sharpen my own—rather than just envision the encounter as a fight to the death.
I think this rejoins your earlier concern, about the way anthropologists can end up talking only to those already part of the conversation. One symptom of that, for me, is the kind of bad writing we all know (and I, for one, certainly have been guilty of), which is full of gesturing to influential authors and complex concepts—whether it be to praise or to trash them—without explicating them. It’s another “pinch of salt” gesture—it’s the “you know what I mean when I say X” tone. The stuff that you don’t need to spell out is what you have to assume your reader already knows. The more of that stuff there is, the smaller your implied readership. Hence why I so love the advice Annemarie Mol once gave me: “Write for your students, not for your teachers.” I always took that advice to mean spelling things out (and in the process—no small benefit!—realizing whether you yourself understood them properly). But I realize in reading your text that one could also take it the other way, toward less framing—not allowing the account of the actual subject matter to be overshadowed by endless theory or meta-meta-meta-reflexivity.4
So from a similar concern, we’re stressing different things. You’re suggesting—I think—that a text is more open, less narrowly targeted, if it is more focused on the object than on the framework. I’m suggesting that a text is more open if it comes with a clear, accessible account of what it’s for and what it is not trying to do. I don’t think those are incompatible—indeed they might be mutually strengthening.5
That’s my answer to your question. But your previous answer interestingly throws a spanner in the works by introducing or perhaps implying the idea—which I intuitively really like—of explanation as a kind of emergent property of the relation between a text and its reader. If you describe something well enough, I might find an explanation in there of something that was a puzzle to me but that you weren’t particularly intending to explain. That’s how I understand the idea that “the account … could be further developed by adding details that might help us to understand what [X] was doing and why, without being explicit with respect either to the question that is being asked or to how this gives an answer.” This isn’t so much an implicit explanation, however (which suggests you already know where you’re trying to get to and have somehow produced that effect by cutting something out), and more like a “by-product explanation.” The richer your account, the more likely that it will spark something off in someone (or rather, hopefully, lots of different things in lots of different people) that you could not have predicted or intentionally engineered. That space between the author and the readers is where a certain kind of understanding emerges. And it’s true, fundamentally: good ethnography is supposed to give more than what you put in.
The idea of a by-product or, better perhaps, an “emergent” explanation does make my insistence on stating one’s purposes seem a bit clunky—who cares what the author intended if the text works? And yet, I still think that would be letting the author off the hook a bit too easily. I still want to know what they were aiming for, even if in the end their account takes me elsewhere. “The author is dead” is convincing as a sort of abstract perspective on literature in general, but it’s no way to live an academic life.
My questions to you: Have we cracked—some? all?—of anthropological “explanation” if we say that it happens, not within a single text, but somehow between an author and a reader (see Reed, this volume)? And a crucial reader here, as your own example illustrates, would be those whom in an older language one thought of as the “informants” themselves.6 And if there’s value in that thought, what, if any, are its limits?
TY
I agree this is an important aspect of anthropological explanation, though, for the kinds of reasons you so well set out in your book, I would be resistant to any sense that anthropological explanation could ever really be “cracked.”
Your answer prompts the reflection that there are two ways in which an explanation can be clear: either in showing your “workings,” allowing the reader to trace the steps you have taken; or through focusing on what those workings lead you to—in this case the descriptive object. I would agree that both of these can be effective (or ineffective) strategies, and the question of effectiveness is relative both to the writer’s aims and to the readers’. Knowing how a painting was made (using what techniques, during what period, by what kind of painter, under what kind of influences) might help to explain the meaning of that painting, though it may also undermine the capacity of the painting to “speak for itself.” Hence a lot of artists’ resistance to these kinds of contextualizing moves.7 I take it that this is the distinction you highlight between the object and the framework. I am not saying anthropologists should concern themselves with the object instead of the framework but rather, to reprise an earlier point, that the current political economy of higher education seems to lead to a distorting preoccupation with the framework.
I agree that a lot of ethnographies have the “by-product” explanatory function you describe. The reader gets more than the writer intends and the more so, the richer and more multistranded the description. From the writer’s perspective, this quality might be amplified by a kind of underdetermination, dwelling in the specifics in a manner that allows the generalities, comparisons, and lessons to be drawn in different directions. The less the analogy is made in one explanatory direction, the more the reader is free to make it in their own. Perhaps this by-product explanatory function also explains why Bronislaw Malinowski’s ethnography continues to speak to us, even as his theories are now of mostly historical interest. We can return to those circumstances and elaborate them in endlessly new directions. And perhaps there is also something in these classic ethnographies that we have lost, or at least abridged. The details of the case far exceeded the explanation they were intended to support. A lot of more recent ethnographic writing is more centrally oriented to illustrate argument.
Ethnography can also work as an implicit explanation, in the sense of a concept or idea that is there without being explicit. Most ethnographic writing does this to some degree, making a description a demonstration of an idea that is less than fully spelled out. Maybe this is more like ethnography as allegory: there is a message, which is more powerful because the reader finds it themselves. I would see these as two distinct ways of “saying more” by, in certain respects, “saying less” in others. In both instances the conceptual framework disappears in order to foreground the descriptive object. A world is opened for a reader more directly and immediately, because the descriptive language is transparent—not in the sense of revealing itself, but rather to the extent the reader sees through it.
Your discussion of comparison helpfully highlights how these are routinely invoked, in clarifying the forms these comparisons can take, and in reminding us how frequently we are drawn back to the same heuristic devices. I am in full agreement with your diagnosis that valorizations of “frontal comparison” (how understanding “them” helps us to understand “us”) have tended to elide or devalue the central role of “lateral comparison”—the way in which we understand the “here” of a particular case in relation to the “there” of others. Am I right to infer from your account that you are saying that comparisons help us to explain, and the more, and more explicit, the better? Or could you imagine such a thing as an “implicit comparison”? This is really an extension of my thinking laid out earlier in this chapter, and the argument that there can be a virtue to not fully articulating the comparative relations through which a text is built. For instance, “here and now” comparisons are intensified by curtailing “there and then” comparisons. Might ethnography open a world more effectively, or anyway differently, without the constant deferral and relation to others? Even to the extent that ethnographic understanding is pieced together through these various comparative understandings, might there, on the foregoing logic, be a narrative rationale for reining them in? Of many good examples, Kath Weston’s Traveling Light (2008) illustrates the amplification and intensification that occurs when the focus is squarely on the ethnographic “here and now.”
MC
Before answering your question, I just wanted to mention that one thing you wrote in your previous answer clears up for me how we end up in different positions while aiming at the same thing—the point where, as it were, our sensibilities go out of synch. You write, “There are two ways in which an explanation can be clear: either in showing your ‘workings,’ allowing the reader to trace the steps you have taken; or through focusing on what those workings lead you to—in this case the descriptive object.”
That formulation made me realize that we share a contrast between object and framework, but we apply it differently. In your contrast, the “object” is the description (or the reality behind it, perhaps?) and the framework is the additional layer of explicit commentary on the object or description.
By contrast, in my usage, ethnography (thick description, profusely lateral comparison, etc.) is the framework, the device, and the setup, and the “object”—what this is all leading toward—is the “point,” the argument, the conclusion, the end, or the purpose of the description. This is why, in arguing for essentially the same thing—the value of “staying with” slow ethnographic and descriptive richness—you call for less framing, and I call for more.
But to come back to your question. The short answer is, yes, I agree that simply outlining one case and letting the reader compare with other cases they already know can be productive. The text itself is not explicitly comparative, but it becomes or rather “affords” a comparison in a relational way, like the “emergent explanations” introduced earlier. Indeed this could be a classic instance of what might be meant by a relational or emergent explanation: your case makes me think of a comparison that explains something to me about something else. We could apply our (emergent!) distinction between “implicit” and “by-product” here: an implicit comparison would be one in which the author already has a further point of application in mind that he or she artfully conceals, hoping that the reader will tease it out for themselves. By-product comparisons would be all of the possible other ways in which readers could build comparisons out of the one case presented. And since we never know what our readers already know—indeed, that’s the beauty of it—the range of by-product comparisons or explanations could be limitless, and surprising.
Let me take one example, to try to tease out this “implicit” versus “by-product” distinction further. It starts from Alexei Yurchak’s argument in Everything Was Forever (2006), about semantic shift in late Soviet socialism—the way official language came to operate as a pure form, dissociated from its earlier meaning. Yurchak’s account was and continues to be an incredibly rich source of by-product comparisons for me, and I will come to one of those in a minute. But Yurchak and Boyer later explicitly drew out one key comparison, between the hollowing out of late socialist discourse and the hollowing out of “late capitalist” or “neoliberal” discourse (see, for instance, Boyer 2013; Boyer and Howe 2015; Boyer and Yurchak 2010). This explains interesting echoes in terms of parodic humor, for instance. That’s a very interesting analogy (although it might be intriguing to excavate further some of its limits).
By contrast, the by-product comparison that struck me most directly upon reading Yurchak’s book was grounded in his description of the way the emptying out of official language affords the formation of an “us” (svoi) community: a majority of people who bond around the knowledge that they say things without quite meaning them. To be svoi is to plot a course between two ways of taking Soviet language seriously: svoi are neither true believers (extremists, apparatchiks) nor committed or outspoken critics (dissidents). They are “normal people,” living in the middle, in a space characterized by a particular kind of pragmatism, humor, and everyday ethics. Many believe in the original ideals of socialism, the onetime reference of a now-empty language, while recognizing that the systemic way of pursuing them is broken—they thus do their best to act well in an untenable situation. And they also make some occasionally unsavory compromises. Indeed it is precisely the refusal of such compromises that marks out dissidents, from the svoi perspective, as unsympathetic characters who refuse to accept a kind of moral community of compromise and in the process make waves for their svoi counterparts. To paraphrase an archetypal joke told by Yurchak, everyone is standing in a pool of shit up to their necks, but the dissidents are making it worse by waving their arms about in indignation.
Now as it happens—and I have no reason to think Yurchak specifically intended this—that description made me think (on a completely different scale of seriousness, of course, but that is how comparisons often work) of the particular ambivalence created among (some) academics today by the managerial language increasingly imposed on our practice by university administrations. I’m picking this example precisely because it nicely takes us back to the common gripe from which we both started. Excellence indicators, feedback forms, and quality assurances have devolved into meaningless formalities associated with direct power. That situation has fostered explicit critique, certainly, but it has also led to the constitution of more ambivalent everyday intimations of community among working academics, svoi, “normal people” as it were, who neither believe in the meaning and value of these forms nor stand up against them explicitly (which would make waves and trouble for everyone else). Like svoi in Yurchak’s account, many of us hold dear the actual ideals that this managerial language is purportedly trying to point to. Indeed many of us struggle to make space for actually good teaching and actually rigorous and thought-provoking research, often against the grain of these managerial requirements themselves.
Whatever the value of this rough-and-ready comparison, it exemplifies the sort of dynamic we’ve been talking about: through the comparison that I drew out of it, Yurchak’s account explained things about a situation familiar to me, without ever explicitly setting out to do so. And I mean “explained” in a number of different senses here, which range across the epistemological spectrum of visions of explanation. Reading Yurchak with my own academic life in mind outlined a structure, linking discursive, affective, practical considerations into a pattern; it suggested some complex bundles of causal or quasi-causal relations (the emptiness of language, added to a clear power structure, can lead to a particular set of moral and relational options); it helped me get an interpretive handle on why self-proclaimed “dissidents” against academic managerialism might occasionally seem unlovable even to those who might fundamentally agree with them—and so forth.
Is that a by-product comparison or explanation, or was that implicitly “there” in Yurchak’s account? I don’t know, but if it was the latter—that is, if Yurchak had intended it but then kept it hidden or held it back—I don’t see what the added value of that move would have been. Conversely, this by-product comparison or explanation sprang to my mind without being hampered by the fact that Yurchak himself has a very clear and explicit set of arguments in that book—the book is, among other things, an explanation of how state socialism could seem unshakable and yet, as soon as it had fallen away, could seem so obviously to have been teetering. Nor was this hampered, either, by the fact that I encountered the text in relation to specific further comparisons by Yurchak and Boyer, which made it seem like the key comparative point of relevance was to a particular genre of political comedy or political performance. In other words, I don’t think that the author’s telling me what they intended is likely to limit me or throw me off the scent of other potential comparisons. Being told what lessons (the author thinks) are to be drawn doesn’t limit my own ability to draw other lessons.
That being said, I agree that a text in which the description is crowded out by theoretical apparatus will be less effective at doing that elicitative work. Equally, I don’t think it is productive or necessary to actively erase or submerge one’s explanations, arguments, or comparisons. When you say, “Most ethnographic writing does this to some degree, making a description a demonstration of an idea that is less than fully spelled out,” I agree—but I would add that this is because it can’t be fully spelled out, not because of a careful decision to withhold full exposition. There is always more than one point to any good description. That’s why you can’t reduce a description to a point. To give the point and not the description is a radical loss, including of the potential for further (unexpected) comparisons or by-product explanations. But to give one point—to add it, rather than substitute it to the description—doesn’t curtail further ones. To come back to your metaphor, I don’t think that knowing the techniques of painting, or the intentions or historical context of the painter, dulls the effect of any given painting—not in the way in which, for instance, knowing how a magic trick is done destroys the magic trick, or being explicit about how much a meal cost destroys the hospitable effect of inviting someone to partake of it.
Where I think we’re in full agreement—and here we’re back to these values of thickness and slowness—is that the author’s point or purpose can’t ever replace the actual description. This is true whether this is a description of one case or a description of lots of lateral comparisons of this and this and this and that. Indeed one key argument of my work on comparison is that those reduce to each other: if you zoom into the texture of ethnography, what it is actually made of, then you see that every “single-case” ethnography is already built out of lateral comparisons of moments, instances, individual people, particular statements, and so on. The thickness and richness of the description is already—in my terms—a matter of the multiplication of lateral comparisons. And it’s true that, by contrast, frontal comparisons (comparisons not of “this and that” but of “us and them”) are often used as ways of drawing things to a point. They are very good at marking out, explicitly, what matters, where the key contrast or similarity lies. As you noted earlier, frontal comparisons are a classic device for making “pointy” theoretical value out of thick ethnographic description (“Here, precisely, is how their conceptual world transforms ours!”). My concern—that the excitement of frontal comparison can tend to crowd out the value of lateral comparison—thus maps very closely onto yours. And of course, just as you’re not suggesting that we do away with theory, I’m not suggesting that we do away with frontal comparison—just that we replace it within its proper role as one among other anthropological heuristics.
On Emergent Explanations
This conversation doesn’t have a natural endpoint. For the purposes of this chapter, however, one might round things off here by pointing to one central concept that has come together in the foregoing pages of the chapter—namely, the concept of an “emergent explanation.” Retrospectively we might define this in general terms as the ways in which an explanation emerges between and across different explanatory contexts that relate without being commensurate. Our account has foregrounded the productive ways in which such explanatory differences may be reconciled without resolving into a singular frame. Emergent explanations keep different explanatory aims and assumptions productively in view and actively related.
We offer the concept of emergent explanation as a critical alternative to those explanatory forms that seek to collapse or resolve difference, most obviously in the form of singularizing arguments, monocausal explanations, and those that generalize—for instance in ontological, cultural, or sociological terms—as truths over and above the contexts they relate. We hope this formulation adds conceptual precision to the more instinctive explanatory orientations that framed our account: emergent explanations necessarily work in slow and concrete ways that do not lend themselves to totalization or generalization. They are not incompatible with summary but remind us that the sum is always productively more and less than the parts.
Our shared investments in this concept are located in relation to specific aims and assumptions, which the dialogue has helped us to understand and formulate. In hindsight, we might recognize how our dialogue was framed by a normative question, which can now be recast more descriptively as two distinctive versions of how best to encourage explanatory emergence. Each of these anticipates and frames two specific kinds of readerly response.
Through Tom’s contributions, the idea of an implicit explanation foregrounds the explanatory potential of description: how descriptions of specific ethnographic contexts open explanatory possibilities, conceptual affordances, and imaginative possibilities in ways that are unanticipated and open-ended. In these cases, it was suggested that description involves tacit explanation that is distinct from the straightforward absence of explanation. The reader infers or imagines concepts in a way that depends on the absence of explication.
Matt’s discussion, by contrast, foregrounds the productive effect of being explicit about what particular accounts and analytical devices (such as particular forms of comparison) were intended to do, and what they were not intended to do, in order to leave the reader free to go somewhere else with the material, to do something else with the analytical devices, beyond the horizon and interests of the author. Here explanatory emergence is facilitated by the way in which the author’s explanatory frameworks are laid out explicitly, so that they can be bypassed or borrowed and put to work in other contexts.
In specific ways, our contributions have highlighted distinctive forms of explanatory emergence through which anthropological texts are routinely built. Matt has made explicit how explanations are built intertextually as relationships between concepts and contexts that are in some sense analytically reconciled. Tom foregrounds how ethnographic writing has an emergent quality involving the juxtaposition and comparison of explanatory difference within a given field. In both these senses, the explanatory work of anthropology often has an interstitial quality, residing between and across other explanatory concepts and contexts. Comparisons juxtapose contexts that help to explain each other. “This” illuminates “that” and vice versa. However, the insights that emerge from these ethnographic and analytic relations cannot be subsumed in positions over or beyond these elements.
Our dialogue led to a concept that neither had anticipated (emergent explanation) and is therefore an instance of what this concept purports to explain: through our chapter, explanation emerges as an unfolding relation between positions. At least for the authors, the result is not a collapsing or resolution of those differences but a better and more reconciled sense of where those differences lie. From both of our perspectives, the chapter helps to explain something that we hadn’t understood as we set out to write it. Our explanation is emergent in the dialogue, in the sense of being led toward an unanticipated understanding through a process. It remains epistemologically emergent in the sense that it relates our differences of orientation without resolving these. We hope this ultimate irresolution creates a space in which readers can draw their own conclusions.
NOTES
1. This point has been made by a number of commentators from various disciplinary perspectives, including anthropologist Marilyn Strathern (2000), sociologist Frank Furedi (2004), science studies scholar Isabelle Stengers (2018), and literary critic Stefan Collini (2016).
2. Godfrey Lienhardt’s approach to ethnography has some resonances with recent accounts, in their insistence on starting from understandings of the ontological basis of others’ categorical distinctions (I am thinking particularly of Viveiros de Castro’s “Perspectival Anthropology and the Method of Controlled Equivocation” [2004] and Holbraad’s Truth in Motion [2012]). My own account draws inspiration from these approaches: in their insistence that “theory” and “description” occupy a single plane of explanaton; and in the methodological orientation that engenders commitment to the effort to understand others’ lives, as the necessary corollary to a skepticism toward anthropology’s own concepts and theories. Description of the particularities of others’ lives requires that we—professional anthropologists—reconfigure our categories in the act of bending them to circumstances for which no encompassing explanation exists. By the same token, ethnographic description is the means by which new concepts are generated, as old ones are extended or found to be wanting. The rationale is compelling (Englund and Yarrow 2013), and many of the resulting descriptions are insightful. However, asymmetries are reintroduced where the point of this equivocation is less the production of faithful descriptions than the novel conceptual points that derive from these. Despite a number of notable examples to the contrary, proponents of this approach seem more often oriented by the aim of unfolding theory (ethnographically derived concepts) from description (the circumstances of other people’s lives) than to the production of accounts in which description is the point. In this respect Lienhardt exemplifies a distinct approach, from which I draw inspiration.
3. In respect of that strategy, the book has a far more distinguished precursor in Strathern’s own Partial Connections (1991; updated 2005). As one reviewer just as perceptively noted, this is the sort of book a former journal editor would write “in recovery mode” (Shryock 2019, 414).
4. I have some sympathy with Latour (1988) when he sarcastically points at the towering layers of reflexivity piled on in some postmodern accounts: here is me thinking about me thinking about me thinking about me writing this thing about me thinking that … It’s as if we thought our readers were too naïve, he says, and so we had to diminish the power of our writing, by stepping outside the text to caveat it. Equally, though, I am no longer convinced by Latour’s converse proposal, that we should just live with the fact that all we do is tell stories (even when they are stories about us telling stories …), and so that instead of caveating and diminishing their power to convince, we just try to make our stories as convincing as possible, using every rhetorical and stylistic trick in the book. That’s in essence his argument against explanation too—a description that needs an explanation is not a good enough description—and I no longer find that convincing.
5. It’s no surprise, perhaps, that Mol’s own The Body Multiple (2002) is written in two layers—I wonder if you’re arguing for the top layer and I’m arguing for the bottom layer?
6. This idea gets us close to the vision of explanation that, as we note in the introduction, philosophers of science have described as “pragmatic”—a vision of explanation as relative to the interests and perspectives of those who receive it.
7. See also Rapport (this volume).
REFERENCES
- Barad, K. M. 2007. Meeting the Universe Halfway: Quantum Physics and the Entanglement of Matter and Meaning. Durham: Duke University Press.
- Boyer, D. 2013. “Simply the Best: Parody and Political Sincerity in Iceland.” American Ethnologist 40:276–287.
- Boyer, D., J. D. Faubion, and G. E. Marcus, eds. 2015. Theory Can Be More than It Used to Be: Learning Anthropology’s Method in a Time of Transition. Ithaca, NY: Cornell University Press.
- Boyer, D., and C. Howe. 2015. “Portable Analytics and Lateral Theory.” In Theory Can Be More than It Used to Be: Learning Anthropology’s Method in a Time of Transition, edited by D. Boyer, J. D. Faubion, and G. E. Marcus, 15–38. Ithaca, NY: Cornell University Press.
- Boyer, D., and A. Yurchak. 2010. “American Stiob: Or, What Late-Socialist Aesthetics of Parody Reveal about Contemporary Political Culture in the West.” Cultural Anthropology 25:179–221.
- Candea, M. 2018. Comparison in Anthropology: The Impossible Method. Cambridge: Cambridge University Press.
- Carrithers, M. 2014. “Anthropology as Irony and Philosophy, or the Knots in Simple Ethnographic Projects.” HAU: Journal of Ethnographic Theory 4:117–142.
- ______. 2018. “How to Open a World 1: Humanism as Method.” In Anthropology as Homage: Festschrift for Ivo Strecker, edited by F. Girke, S. Thubauville, and W. Smidt, 225–250. Cologne: Rudiger Koppe Verlag.
- Clifford, J., and G. Marcus, eds. 1986. Writing Culture: The Poetics and Politics of Ethnography. Berkeley: University of California Press.
- Collini, S. 2016. “Who Are the Spongers Now?” London Review of Books 38:33–37.
- Crawley, M. 2021. Out of Thin Air: Running Wisdom and Magic from above the Clouds in Ethiopia. London: Bloomsbury.
- Englund, H., and T. Yarrow. 2013. “The Place of Theory: Rights, Networks and Ethnographic Comparison.” Social Analysis 57(2): 132–159.
- Fisher, L. E., and O. Werner. 1978. “Explaining Explanation: Tension in American Anthropology.” Journal of Anthropological Research 34:194–218.
- Furedi, F. 2004. Where Have All the Intellectuals Gone? London: Continuum.
- Gellner, D. 2022. Review of Comparison in Anthropology: The Impossible Method, by M. Candea. Journal of the Royal Anthropological Institute 28:378–379.
- Holbraad, M. 2012. Truth in Motion: The Recursive Anthropology of Cuban Divination. Chicago: University of Chicago Press.
- Iser, W. 1980. “Interaction between Text and Reader.” In The Reader in the Text: Essays on Audience and Interpretation, edited by S. Suleiman, 106–119 Princeton, NJ: Princeton University Press.
- Latour, B. 1988. “The Politics of Explanation: An Alternative.” In Knowledge and Reflexivity: New Frontiers in the Sociology of Knowledge, edited by S. Woolgar, 155–176. London: Sage.
- Luhrmann, T. M. 2012. When God Talks Back: Understanding the American Evangelical Relationship with God. New York: Vintage.
- Mol, A. 2002. The Body Multiple: Ontology in Medical Practice. Durham, NC: Duke University Press.
- Narayan, K. 2012. Alive in the Writing: Crafting Ethnography in the Company of Chekhov. Chicago: University of Chicago Press.
- Pandian, A. 2015. Reel World: An Anthropology of Creation. Durham, NC: Duke University Press.
- Rapport, N. 2015. “The Consequences of Anthropological Writing.” Journal of the Royal Anthropological Institute 21:680–683.
- Runciman, W. G. 1983. A Treatise on Social Theory. Vol. 1, The Methodology of Social Theory. Cambridge: Cambridge University Press.
- Shryock, A. 2019. “The New Basics: Pushing ‘Refresh’ on the Comparative Method.” Review of Comparison in Anthropology: The Impossible Method, by M. Candea. History and Anthropology 31 (3): 410–416.
- Stengers, I. 2018. Another Science Is Possible: A Manifesto for Slow Science, translated by Stephen Muecke. London: Wiley.
- Strathern, M. 1988. The Gender of the Gift: Problems with Women and Problems with Society in Melanesia. Berkeley: University of California Press.
- ______. 1991. Partial Connections. Savage, MD: Rowman and Littlefield.
- ______. 2014. “Introduction.” In Wives and Wanderers in a New Guinea Highlands Society: Women’s Lives in the Wahgi Valley, edited by Marie Olive Reay, 43–56. Acton, Australia: ANU Press.
- ______, ed. 2000. Audit Cultures: Anthropological Studies of Accountability, Ethics and the Academy. London: Routledge.
- ______. 2020. Relations: An Anthropological Account. Durham, NC: Duke University Press.
- Van der Veer, P. 2016. The Value of Comparison. Durham, NC: Duke University Press.
- Viveiros de Castro, E. 2004. “Perspectival Anthropology and the Method of Controlled Equivocation.” Tipiti 2:3–22.
- Weston, K. 2008. Traveling Light: On the Road with America’s Poor. Boston: Beacon.
- Wimsatt, W. C. 2007. Re-engineering Philosophy for Limited Beings: Piecewise Approximations to Reality. Cambridge, MA: Harvard University Press.
- Yaneva, A. 2009. Made by the Office of Metropolitan Architecture: An Ethnography of Design. Rotterdam: 010 Publishers.
- Yarrow, T. 2019. Architects: Portraits of a Practice. Ithaca, NY: Cornell University Press.
- Yurchak, A. 2006. Everything Was Forever, Until It Was No More: The Last Soviet Generation. Princeton, NJ: Princeton University Press.