Abstract
The anthropology of personhood has often contrasted the bounded Western individual with non-Western persons conceived as relational or dividual. This article argues that the comparison is potentially asymmetrical, because it moves between different kinds of evidence and different levels of recursivity. Drawing on Living Value Theory, it distinguishes ongoing living coordination, at the levels where coordination proceeds or is felt as disturbed without explicit formulation, from symbolic articulation and from the stabilisation of distinctions into portable classifications, and it insists that the latter are never comprehensive representations of the former. Personhood classifications, which like all binaries are symbolic stabilisations and not structures of the living world, must therefore be investigated through their actual or possible consequences for what is vitally at stake. Because ethnographic questions can themselves generate symbolic articulations and stakes, research must also distinguish what was consequential before an inquiry from what became consequential through it. The article proposes an anthropology that investigates the circumstances in which personhood classifications become useful and consequential in living, rather than one that presupposes their centrality.
I. A Question Supplied in Advance
The lecture opened with a vote. On the screen were nine candidates: a newborn baby, a foetus, an eleven-year-old, a man in a persistent vegetative state, a company, a river, a Hindu deity, an elephant and a chatbot. A theatre of first-year anthropology students was asked, case by case, whether each was a person. Hands went up readily, and the disagreements were lively: the chatbot was dismissed, the company and the river provoked laughter and then hesitation, the foetus divided the room.
While preparing the lecture, and more sharply while delivering it, I found myself troubled by what the exercise had done before anyone voted. Nine heterogeneous cases had been brought together under a single symbolic distinction, person or nonperson, without anyone first establishing why that distinction mattered to the living coordination involved in each. The cases become far more interesting once that question is asked. For a newborn, what is at stake on its first day is feeding, warmth, sleep and the attention of those who care for it. For a river, what is at stake are flows, the beings that live in and from it, and the people whose fishing, travel and ancestry are bound up with it. For a company, the stakes belong to its creditors, employees and owners. For an elephant in a zoo, they concern space, movement and the company of other elephants. For a chatbot, it is unclear that anything is vitally at stake for the chatbot at all. A foetus, a corporation, a river, an animal, a newborn and a program do not generate the same question merely because each can be placed before the same classifier. The apparent commonality was supplied by the classification. The concrete stakes differ dramatically, and some of them were introduced by the exercise itself, which invited a roomful of eighteen-year-olds to care, for an hour, about whether a river is a person.
The question did more than invite different answers. It made one symbolically stabilised distinction the governing problem before establishing whether that distinction was consequential in any of the cases. This article argues that the same move runs through the anthropology of personhood, and that its most influential result, the contrast between Western individuals and non-Western dividuals, is a particularly revealing consequence. The anthropology of personhood has often mistaken differences between recursivity levels for differences between cultural conceptions of persons. It has set symbolically stabilised philosophical, legal and economic representations of the Western individual against ethnographic descriptions of non-Western living, interaction, exchange and elicited articulation, and it has read the resulting difference as a difference between cultural ontologies.
The inquiry behind this article began by following the familiar organisation of personhood into philosophical, legal, anthropological and experiential perspectives, and that organisation proved increasingly inadequate. Disciplinary divisions are historically stabilised ways of organising knowledge, not divisions within the phenomenon. A legal classification transforms everyday material possibilities; a philosophical concept shapes someone's ordinary self-understanding; an ethnographic question generates an articulation; a psychological account takes an institutional identity category as its starting point. None of these processes can be understood if disciplinary boundaries decide the analysis in advance. Only by releasing the inquiry from those received divisions did the distinction that organises the rest of this article become clear: the distinction between living coordination at the first two levels of recursivity and its symbolic articulation and stabilisation at the levels above.
II. Living Before Personhood
Living Value Theory (LVT) is a process ontology, and it is best introduced through the problem it resolves rather than as a checklist of components. Its starting point is the order in which things happen. Living beings do not first know, judge or choose and only afterwards act; before anything is named or decided it is already being lived, in the ease or difficulty with which a body moves through its day, a relationship holds, a place remains habitable (Ecks 2026c). LVT calls this ongoing process coordination, and it describes it as co-mediated through embodiment, being-with, dwelling, made things and symbols. Those mediations, and the levels and counterpart types the theory also distinguishes, are constitutive of living coordination wherever it occurs. Their presence, however, does not predetermine which processes or distinctions will become most consequential in a particular inquiry. LVT is in this sense a practice of attention: the explanatory priority of any of its concepts has to emerge from what is happening in the living under study. In the case of personhood, what emerges as decisive is the distinction between levels of recursivity.
At L1, living proceeds through ongoing, largely unproblematic recursive coordination: the infant feeds, a practised walker crosses a familiar street, and nothing needs to be noticed. At L2, coordination becomes salient through disturbance, difficulty, felt difference or emerging consequence, before and without any explicit formulation: hunger, unease, the sense that something is wrong. At L3, aspects of living become symbolically articulated, named for this occasion: "she is hungry", "I am not the same as I was". At L4, symbolic distinctions become stabilised, portable and repeatable, available for conventional or institutional application beyond the coordination that produced them. At L5, established articulations and classifications can themselves become objects of reconsideration (Ecks 2026c). These are not separate compartments, nor a developmental sequence through which every life passes in order. Symbolic articulation and stabilisation arise within living coordination, feed back into it and reshape it: a stabilised category becomes part of the ground on which later living proceeds, often unnoticed at L1. Nevertheless, what matters at L1 and L2 is not exhausted by what becomes articulated at L3 or stabilised at L4. That is the distinction the anthropology of personhood has repeatedly failed to maintain.
The gap between living and its articulation is easily underestimated, and multisensorial embodiment shows why it cannot be closed. Consider a single second of walking along a street: breathing, placing a foot, perceiving the distance to a kerb, adjusting posture on uneven paving, hearing a bicycle behind, feeling the cold on one side of the face, registering a dog on a lead and a puddle ahead, keeping balance while turning to look. An extraordinarily detailed verbal description of that second might take hours, and even then it would not reproduce the second. It would be another ongoing activity of living, with its own breathing, posture, materials and symbols, taking place somewhere else at a later time. The difference between living and its symbolic articulation is therefore more than a deficiency in descriptive detail that more care could remedy: symbolic articulation is one process within living, not a complete representation of it. A classification of persons can therefore be articulate, consequential and ethnographically fascinating while participating in only a small fraction of the coordination in which it occurs.
The first-person orientation of living must accordingly be distinguished from first-person reflection about personhood. Questions such as "Who am I?", "Have I changed?" or "Am I being treated as a person?" arise through symbolic articulation, often drawing on categories stabilised long before the speaker uses them, and they are relatively occasional activities within a life mostly occupied with other things. Someone who moves from Kolkata to Edinburgh finds within a year that their mornings, meals, sleep and greetings have all been reorganised, and may describe this, if asked, as a change of weather, food or job, never as a change in who they are. A life can change profoundly without its bearer articulating the change as a question of identity.
This has a direct bearing on the most influential tradition in the anthropology of meaning. Clifford Geertz held that "culture is public because meaning is" and that "the culture of a people is an ensemble of texts" which the anthropologist reads over the shoulders of those to whom they belong (Geertz 1973a: 12; 1973b: 452). Geertz never claimed that every experience is explicitly verbalised, and a critique that caricatured him in that way would be easy to dismiss. The stronger criticism concerns analytical privilege. Interpretive anthropology made publicly available symbolic meaning the primary route to what is significant in a form of life, and so assumed that the vitally consequential aspects of living are sufficiently accessible through public articulation. LVT rejects that assumption. What is meaningful in symbolic articulation is not coextensive with what is vitally consequential in the mesocosm, and the symbolic availability of an account must be distinguished from its coverage of living. People articulate reasons, identities, obligations and classifications, and the bodily, material, multispecies and spatial processes through which they live vastly exceed what they articulate.
Two opposite methodological errors follow from neglecting L1 and L2, and they are related. The first is to assume that what has not been symbolically articulated cannot have been vitally consequential. It overlooks embodied, material and multispecies coordination that becomes significant without explicit formulation, and it is the characteristic risk of an anthropology that reads culture as text. The second is to assume that what has been symbolically articulated must already have been vitally consequential. It mistakes an available or elicited articulation for evidence of salience in living, and it is the characteristic risk of the interview about personhood. Both errors treat L3 and L4 articulations as a comprehensive measure of L1 and L2 living. The first finds nothing where nothing was said; the second finds a settled concern wherever something was said.
The question that must therefore accompany every step of an inquiry into personhood is what is vitally at stake in the mesocosm, for whom, and how we know. Vital stakes may lie in embodied vulnerability, nourishment, care, safety, movement, livelihood, belonging, reproduction, exclusion or access to resources, and these are examples, not an inventory. Stakes need not be verbally articulated in order to exist, and much of what matters vitally at L1 and L2 never becomes an L3 formulation at all. The methodological rule is simple to state: at every point where personhood appears as a question, a classification or a comparison, investigate the vital stakes that make it consequential in living, and do not presume those stakes from the existence of the question itself.
III. Discerning Counterparts Without Classifying Persons
Living from within is always also living with others. LVT calls the activity through which living beings engage what they encounter recursive discernment: the continual determination of which aspects of an encountered entity are relevant to the coordination at hand, whether those aspects answer back and how, and whether further discernment is warranted. It operates primarily at L1, in bodily attunement, before any symbolic elaboration, and it is aspect-specific (Ecks 2026c). A horse adjusts to a rider's hand while its weight in the saddle answers to nothing, and the rider engages both aspects at once and differently. A person notices a colleague's exhaustion and slows a conversation before either says a word about it. An infant quietens at a caregiver's voice and becomes agitated when the voice changes. In each case what matters is discerned within coordination, and no articulated concept of what kind of being the counterpart is plays any part.
Discernment differentiates continually. Living is not an undifferentiated flow, and a newborn distinguishes its own hand touching its face from being touched by another, as a dog distinguishes its own limb from the stick it chews. What discernment does not require is a classification of the beings it differentiates, still less a prior world of isolated individuals subsequently joined by relations. Recursive discernment is a living accomplishment with no need of personhood as its destination, and no primitive or incomplete form of personhood awaiting a concept to complete it. Particular capacities for discernment and self-recognition vary across species and remain empirical questions; the general point concerns living coordination, not a uniform cognitive endowment.
Three things that often travel together in human life must be kept apart. Discerning that a counterpart is responsive is one; granting it moral standing, so that what is done to it requires justification, is another; calling it a person is a third. Each can occur without the others. Rane Willerslev's account of Yukaghir elk hunters shows the first in its purest form: the hunter sways in the rolling gait of an elk, on skis covered with elk-leg skin so that they sound like an animal moving through snow, until the elk approaches him as one of its own (Willerslev 2007). The hunter discerns the elk's responsiveness with extraordinary precision, the elk discerns the hunter wrongly as a conspecific, and the coordination between them is asymmetrical, deceptive and lethal. What is vitally at stake is plain: for the elk its life, for the hunter his household's food. Willerslev frames the material through personhood and perspectivism. The scene itself needs no classification of persons to be described, and whether the hunter regards the elk as a person is a further question whose answer adds a claim to the description rather than following from it.
LVT's distinction between recursive attribution and recursive relevance sharpens the point. Attribution is the settled sense of what kind of counterpart a being is engaged as; relevance is how much ongoing coordination actually depends on it; and the two vary independently (Ecks 2026c). An ancestor may be addressed and fed while carrying little relevance in a household where veneration has lapsed. A newborn may carry overwhelming relevance, the household's sleep and labour reorganised around it, while the attribution of what kind of being it is remains unsettled for weeks. Relevance, in this sense, is the form in which vital stakes appear in discernment: it tracks what depends on whom, whether or not anyone names it.
The consequences of discernment are often greater than the consequences of any classification. A newborn cries in the night. Whether the cry is discerned as hunger, as discomfort, or as nothing that requires attention changes the next hour of both the infant's and the caregiver's lives, and households and generations differ in how they teach adults to discern such cries. Nothing in the infant's classification as a person changes between these possibilities; what changes is what is discerned, and therefore what is done. If an anthropology of infancy wants to know how a newborn comes to be treated as a particular kind of being, the practices through which its cries, movements and appetites are discerned are at least as likely to carry the vital stakes as any explicit statement about what an infant is.
IV. Symbolic Articulation and Stabilisation
Aspects of living and discernment can become symbolically articulated. Someone says "I am no longer the same person", or "That one is one of us", or "This is not a person". Some formulations narrate lived change; some classify counterparts; others invoke an established category to protest treatment or to claim standing. The level of a particular act and the stabilisation of the resource it uses need not coincide: an L3 utterance can employ a category stabilised at L4 long before the speaker was born, as when a child says "I'm a person too" in an argument with a sibling. Personhood, insofar as it operates as a repeatable category applicable across situations, is characteristically an L4 stabilisation, while particular uses and challenges of it occur within continuing living coordination and may involve L3 articulation and L5 contestation.
LVT's ontological principle concerning binaries must be stated without qualification. There are no binaries in the mesocosm. Binaries are invariably L4 symbolic stabilisations: they take a continuous or multidimensional coordination and compress it into a two-valued distinction (Ecks 2026c). They can be deeply consequential for living coordination, directing attention, supporting decisions, organising institutions and shaping how beings live, but they are not structures inherent in the mesocosm. Individual/dividual and person/nonperson are L4 distinctions in exactly this sense. They do not reveal that the mesocosm is divided into two corresponding kinds of existence, and no rival binary or third category would do so either. Nothing is gained by attacking a binary because it is a binary.
The appropriate inquiry concerns the work a distinction does. The significance of an L4 binary lies not in whether it corresponds to an underlying binary in the mesocosm, but in what it enables or obstructs within living coordination. What problems does it make manageable? What decisions does it support? What possibilities does it open or close, what does it make difficult to discern, and for whom does it work? Usefulness in this sense is neither an abstract property of a classification nor a ledger of benefits and costs; it has to be investigated in relation to particular living processes and particular vital stakes, and its consequences may be uneven, conflicting, changing and hard to articulate. A person/nonperson distinction may serve one institutional setting well and obstruct another; the individual/dividual distinction may bring certain transactions into focus for an anthropologist while obscuring other consequential coordination. Neither consequence establishes the ontological reality of the binary.
Law shows the process with unusual clarity, provided it is treated as a development of symbolic stabilisation rather than as a separate theory of persons. Its sequence can be stated in three steps: an L4 classification; its multimaterial and multisymbolic stabilisation; and consequences for living coordination. Modern institutional law depends on a dense coupling of symbols and made things: courts, registries, documents, files, precedents, archives, digital records, signatures and authorised offices sustain distinctions far beyond the situations in which they were made. These deposits are nonrecursive and persist beyond the activity that produced them, and the recursion they make possible lies in their uptake, in reading, applying, maintaining and contesting them (Ecks 2026c). Such dense infrastructures are characteristic of modern bureaucratic law and are no universal prerequisite; customary and oral legal orders stabilise status through memory, witnesses, genealogy and repeated performance.
A birth registration shows the sequence at work. In Scotland a birth must be registered within twenty-one days; an entry in a register, held by a national records office and copied into databases, creates a status that will be retrieved when the child enrols at school, applies for a passport, marries or inherits. The status persists while the infant sleeps and feeds in complete ignorance of it, and its consequences reach into embodiment, through who may consent to its medical treatment, into dwelling, through where it may live, and into material provision, through what its household can claim. Correcting the entry is possible, and it too depends on stabilised forms: an application, a prescribed procedure, evidence, an authorised office. L5 reconsideration in a modern legal system is made possible by L4 deposits and regulated by them.
The same sequence can be traced for the other cases on the lecture slide, and in each the decisive question is what becomes consequential through the classification rather than whether the entity really is a person. When New Zealand's Te Awa Tupua Act 2017 declared the Whanganui River a legal person, what changed was representation and governance: an office of two people, one nominated by iwi with interests in the river and one by the Crown, now acts and speaks on its behalf, within a statute that settled a long-standing Māori claim. The vital stakes that classification addresses belong to the river's flows, the beings that depend on them, and the communities for whom the river is an ancestor; whether it serves them is a question for the coordination it now organises. When a company is treated as a legal person, as in Salomon v A Salomon & Co Ltd (1897), what becomes possible is contract, ownership, litigation and limited liability, and the consequences are distributed unevenly, since the risk the company's members escape falls on its creditors. Debates over foetal personhood involve several vital stakes at once, those of the pregnant woman, of the developing foetus, of families and of clinicians, and they are poorly served by attempts to settle them through a philosophical definition of the person; what matters is the concrete institutional consequences of different classifications for each of those stakes. And when New York's Court of Appeals refused habeas corpus to Happy, an elephant held at the Bronx Zoo (Nonhuman Rights Project v Breheny 2022), the institution's classification of her status must be distinguished from the coordination it concerned: an elephant's movement, space and companionship, which the classification of her legal standing neither describes nor exhausts. In none of these cases do the law's conceptual distinctions reveal an intrinsic boundary within the mesocosm. They matter because of the work they do.
Stabilisation also stabilises positions. Imagine sitting in a room while a committee discusses whether you are a person. Before its members begin, they occupy the position of legitimate questioners and you that of the questioned, and your answers count as evidence on terms they control. Even an affirmative verdict leaves this unchanged: the committee was entitled to make your personhood questionable, and you were not entitled to make its authority questionable in the same way. L4 stabilises the roles entitled to apply a category as well as the category itself, and access to L5 reopening may require standing and representation allocated by the institution being challenged, so that the burden of reopening falls on the classified (Ecks 2026d). Here too the question is whose vital stakes are at issue. The committee's members risk little in deliberating; the person in the chair may risk everything. The positions of a lecturer, an ethnographer and a registrar should not be collapsed into one kind of authority, but each occupies a place from which another being's living can be made to answer a question it did not pose.
V. Whose Question? Whose Stakes?
Any two societies can be compared along innumerable dimensions, and neither society supplies the dimension. Twentieth-century anthropology did much to dismantle the ranking of societies and was far less consistent in examining the selection of the dimensions along which they were compared; one can refuse to rank and still inherit an unexamined grid of what is worth comparing (Ecks 2026b). Personhood entered that grid early. Marcel Mauss's history of the category of the person, from the role and mask of the personnage through the Roman legal persona and the Christian metaphysical person to the modern self (Mauss 1985 [1938]), is indispensable precisely because it shows how differently the word has been used. It cannot show that all these uses refer to one object, yet the anthropology that followed tended to assume one, and to compare societies by the concepts of it they held.
Personhood classifications need not be salient in everyday living. Much coordination proceeds without anyone asking whether a being is a person or what a person consists of. An ethnographer can nevertheless put precisely those questions in an interview, and a respondent may produce a meaningful and sincere answer from symbolic resources already available. What has then been established? A respondent formulated an answer within a particular encounter, perhaps drawing on distinctions already significant in everyday symbolic practice. It does not follow that a classification of personhood routinely organises the respondent's living, nor that the answer represents a collectively stabilised L4 category. LVT names the risk precisely: symbolic overreach occurs "when an imported question dictates in advance what the coordination must contain, or is read back into L1 or L2 as though it had been there all along" (Ecks 2026c: 21).
The research encounter does more than retrieve answers. It can introduce a problem, foreground a distinction, or alter what participants attend to, and the stakes it creates may be real. Three things therefore have to be distinguished as separate empirical matters: what was consequential in participants' living before the research question was asked; what became consequential through the ethnographic encounter; and what became consequential through subsequent anthropological analysis and publication. When the anthropologist extracts an answer, translates it, generalises it and publishes it as a cultural theory of personhood, a situated L3 articulation can become a portable comparative category, cited and taught by people who were never present. The monograph is a deposit in LVT's sense, and its uptake can give an elicited answer the standing of a doctrine about how a people understands itself. None of this implies that elicited responses are insincere, meaningless or illegitimate. The problem lies in the inference from an elicited articulation to previously existing salience, and where prior salience has not been established, an ethnography should say so.
The question of vital stakes should not be confused with anthropological reflexivity or positionality. Examining why an ethnographer asks whether a stone is a person may clarify the researcher's assumptions, training, institutional position and interests, and such reflection can be valuable. But it cannot establish whether the distinction mattered to those being studied. That requires inquiry into their living coordination. A person/nonperson distinction may become symbolically articulated during an interview without having previously organised the respondent's concerns; the research question may itself introduce new stakes; and the ethnographic account may then stabilise the resulting articulation into a portable comparative category. The task is therefore not simply to make the researcher reflexively visible, but to distinguish stakes already consequential in participants' living from stakes generated by the inquiry. This is no claim that researchers can stand outside the mesocosm or avoid becoming consequential to those they study; it is precisely because research encounters enter ongoing living coordination that their effects must be distinguished from what preceded them.
The ethnographer's own stakes belong in the same analysis. A question about personhood may matter a great deal within a research project: it may build an argument, challenge a disciplinary assumption, produce a comparative category, or make a finding communicable to other anthropologists. These are real stakes, located in the living coordination of a discipline, its careers, journals and debates. They do not establish that the same question was already consequential to participants. Three possibilities must remain distinguishable: the researcher's question may be central to anthropology and marginal to those studied; it may intersect with a pressing local problem; or it may generate a problem that was previously absent and becomes consequential through the encounter. A strong anthropological reason to ask a question is not evidence that participants had a corresponding reason to answer it before they were asked.
A. Irving Hallowell's Ojibwa ethnography is the best test of these distinctions, because it contains so much more than an interview answer. Hallowell noticed that the Ojibwe word for stone belongs grammatically to the animate gender, and he asked an old man whether all the stones around them were alive. "No! But some are," came the reply (Hallowell 1960). The principal question to put to this material concerns what was at stake in encounters with particular stones, what practical and symbolic differences mattered, and under what circumstances those distinctions became consequential, rather than whether the Ojibwa understood stones to be persons. Hallowell's evidence is layered. There is the grammatical classification, a stabilised feature of the language that settles nothing by itself about experience. There is the elicited answer, prompted by the anthropologist's curiosity about that grammar, which shows that grammar and experience diverged: animacy in the language, responsiveness only in some cases in life. There are narratives of encounters in which particular stones had been seen to move, and there is observed conduct toward such stones and toward Thunder Birds and winds. And there is Hallowell's own generalising construction, "other-than-human persons", which gathers all of this under a single heading. The ethnography can establish that some stones, in some circumstances, were engaged as counterparts that answer back, and that such encounters could matter. It cannot establish, from the elicited answer alone, that a general person/nonperson binary organised everyday Ojibwa living, and much of Hallowell's broader argument does not depend on that answer. What formulating the material as a question of personhood adds is the anthropologist's comparative category. The case also shows that consequential differences among beings require no underlying binary in the mesocosm: some stones mattered differently from others, and the difference lay in coordination, not in a two-valued division of existence.
The same care applies to ethnonyms conventionally translated as "people". The Inuktitut inuk means a person and its plural Inuit the people; the Yup'ik self-designation is commonly glossed as "real people", from yuk, person, with an intensifying suffix. Such terms suggest that being human, belonging to a community, being a social counterpart and possessing a general category of personhood need not be the same classification. They establish nothing more without local evidence: a gloss such as "real people" does not imply that others are regarded as less than human, and how the terms are used, and what is at stake in their use, has to be established case by case rather than read off the translation.
VI. How Anthropology Constructed Individuals and Dividuals
The most influential comparison in the anthropology of personhood sets a Western individual against a non-Western dividual. Its two sides are usually presented as two findings of the same kind. Examined through recursivity levels and vital stakes, they turn out to be built differently, and the difference follows three inferential movements, none of them automatically invalid, each of which has often been left unjustified.
The first movement runs from an L4 conception of the bounded individual to an assumed description of Western everyday living. Geertz gave the conception its most quoted form: "The Western conception of the person as a bounded, unique, more or less integrated motivational and cognitive universe, a dynamic centre of awareness, emotion, judgment, and action organised into a distinctive whole and set contrastively both against other such wholes and against its social and natural background, is, however incorrigible it may seem to us, a rather peculiar idea within the context of the world's cultures" (Geertz 1974: 31). What evidence underlies the picture? It could be an account of how people in Europe and North America coordinate their lives, of how they articulate themselves, of a philosophical conception such as Locke's "forensic" person (Locke 1975 [1694]: II.xxvii.26), of a legal model of rights-bearing agency, or of an economic ideal of autonomous choice. These are different sources at different levels. Geertz's essay offers sustained ethnography of Java, Bali and Morocco and none of the West; the Western conception enters as a known quantity, in a vocabulary drawn from philosophy and psychology. The bounded individual is primarily an L4 stabilisation within Western intellectual and institutional traditions: abstracted from many coordinations, deposited in treatises, statutes and textbooks, and available anywhere. Its availability does not show that Western people live as bounded, self-contained individuals. A student is one kind of person in a lecture hall and another at a party at midnight, is fed, housed and comforted through dense dependencies, and would find it odd to be told that she experiences herself as an integrated universe set against her background. Melford Spiro pressed a version of this objection long ago (Spiro 1993).
The Balinese side of Geertz's own comparison shows that the asymmetry takes more than one form. In "Person, Time, and Conduct in Bali", the person is reconstructed from orders of symbolic designation: personal names, birth-order names, kinship terms, teknonyms, status titles, public titles and the calendars that organise ceremonial time (Geertz 1973c). These are documented, publicly maintained L4 classifications. Geertz's Bali and Geertz's West are therefore both represented by stabilised symbols, but of different kinds: an observed system of public designation on one side, a philosophical portrait of the self on the other. What the Balinese do with their names at L1 and L2, in the coordination of households, work, illness and sleep, is as absent from the contrast as what Westerners do with theirs, and so is any account of what is vitally at stake in either.
The second movement runs from observations of non-Western living coordination to a purportedly alternative, generalised L4 conception of the person. McKim Marriott introduced the term "dividual" for South Asian persons understood as divisible beings who continually absorb heterogeneous material influences and give out particles of their own coded substances, through food, touch, service, gifts, marriage and residence (Marriott 1976). Marilyn Strathern developed the idea for Melanesia: "Far from being regarded as unique entities, Melanesian persons are as dividually as they are individually conceived. They contain a generalised sociality within. Indeed, persons are frequently constructed as the plural and composite site of the relationships that produced them" (Strathern 1988: 13). These accounts draw on rich material: feeding, exchange, incorporation, ritual, reproduction, kinship, gifts, names and obligations. That material mixes observed coordination, participants' articulations, locally stabilised conventions and the anthropologist's constructions. Strathern's synthesis draws on her fieldwork in the Mount Hagen area of the Papua New Guinea Highlands and on many other Melanesian ethnographies, much of it concerning ceremonial exchange, initiation, marriage payments and mortuary distributions, the most public and most elaborated forms of coordination, recorded in part because they were conspicuous. Everyday coordination in a Highland household, how sleep is arranged, how a sick child is tended, how a garden is weeded on an ordinary morning, enters chiefly where it can be read as exchange. The dividual that results is an L4 anthropological construction elaborated from ethnographic practices and articulations.
The third movement runs from an elicited L3 articulation to the claim that a particular L4 classification already organises a community's living. It is the movement examined in the previous section, and in the individual/dividual literature it operates wherever an answer to an anthropologist's question about what persons are made of becomes evidence for a local theory of the person.
Fairness to Strathern requires a further distinction. Marriott and Strathern were not reproducing participants' statements; they were making sophisticated theoretical arguments, and Strathern in particular used comparison deliberately to interrogate Western analytical assumptions. Her book treats society, the individual and the commodity as the analyst's own cultural equipment and builds its Melanesian counterpart as a device for exposing that equipment. Within her argument, the asymmetry between the two sides is therefore no oversight, and it would be wrong to say that she mistook an invented contrast for a directly observed native ontology. The difficulty lies in the inferential warrant the contrast has been given in the wider literature, where it has circulated as a finding about two ways of being. Edward LiPuma pressed a related objection, arguing that individual and dividual aspects of persons are present in all societies and that their apparent distribution between West and non-West owes much to the comparison itself (LiPuma 1998).
The asymmetry can now be stated plainly. One side of the comparison is frequently an L4 philosophical or institutional abstraction; the other is a theoretical interpretation of observed living coordination, supplemented by elicited articulations and presented as an alternative L4 theory. The comparison thereby risks mistaking a difference between recursivity levels for a difference between cultural conceptions of persons. A Western individual reconstructed from feeding, care, gift-giving and kinship in an English town might look far less bounded; a Melanesian individual reconstructed from formal rules of land tenure and the conventions of public oratory might look far less dividual.
The criterion of vital stakes adds a second question to that of evidential level. Does the comparison establish that individual/dividual classifications are consequential for participants in the ways attributed to them? And what does the binary enable anthropologists to understand, and what does it obscure? Its ontological status is clear: it is an L4 stabilisation, and there is nothing in the mesocosm for it to correspond to. Its usefulness is an empirical and analytical question. It has made visible the transfers of substance and the circulation of parts of persons that bounded individualism had hidden, and in that respect its usefulness to anthropology has been considerable. It also directs attention toward exchange and away from everything that does not appear as exchange, and it organises whole regions of the world around a contrast with a Western self that was never ethnographically described. Whether it helps to understand what is vitally at stake in a Highland household or a Bengali kitchen has to be shown in each case, and it cannot be shown by the elegance of the contrast. The binary need not be discarded because it is a binary; it has to earn its place by what it allows anthropologists to see about the living they study.
Behind this comparison lies a longer history. Inverting a historically specific model of bounded, transactional agency, so that transactions constitute persons rather than persons entering transactions, does not escape the framework that made persons and relations the central explanatory problem (Ecks 2026a). That genealogical argument is a separate one, and the asymmetry identified here stands for a reader who is not persuaded by it.
VII. What Must Anthropological Comparison Establish?
Comparative anthropology faces two distinct requirements, and neither replaces the other. The first is relevance: why is personhood a consequential dimension for the living under study? The second is comparability: are the phenomena or claims being compared of equivalent evidential and analytical status? A comparison can fail by choosing a dimension that is irrelevant to the lives in question, by comparing unlike phenomena, or by doing both. An anthropologist can construct a technically symmetrical comparison, with equally deep evidence at matching levels on both sides, that is nonetheless insignificant for understanding what matters in either. Relevance comes first: only once a dimension has been shown to have consequential relevance to the living coordination under investigation is it worth determining what evidence supports the comparison and whether the phenomena are commensurable.
Comparability then imposes its own discipline. Four kinds of evidence must be kept apart for Western and non-Western cases alike: observed coordination, what has actually been documented; participant articulation, what people have said, named, narrated or classified, and in what circumstances; institutional stabilisation, what classifications have acquired portable, repeatable, consequential standing; and anthropological construction, what the researcher has inferred, generalised, translated or selected. A statement at one level is not sufficient evidence for another without argument. Western first-person life should not be presumed bounded because a Western philosophical concept emphasises boundedness; a non-Western conception should not be presumed relational because observed exchanges are consequential; an elicited formulation should not be presumed to be an enduring local category; a published concept should not be presumed to capture how participants understand themselves; and an LVT redescription should not be presumed to show what matters most to participants without appropriate evidence.
A comparison may also be legitimate as an explicitly scholarly construction that reproduces no salient participant distinction. Comparing the ways different legal orders allocate standing, or using an analytical contrast to expose an anthropologist's own assumptions, as Strathern did, can be valuable. In that case the author must identify and justify the analytical stakes of the comparison, rather than attributing them silently to participants. The failure lies in presenting the stakes of a discipline as the stakes of the people it studies.
The disciplinary organisation of personhood returns here as a warning. If an inquiry begins by asking what philosophy, law, psychology and anthropology each say about persons, it has already allowed historically stabilised divisions of knowledge to decide what the phenomenon contains. The divisions are useful for tracing how particular concepts of the person were formed and circulated. They are no guide to how personhood becomes consequential in living, where a legal status, a philosophical idea, an interview question and a meal may all bear on the same afternoon.
The student who asked, after the lecture, whether anthropology can have theory at all if it must not impose its categories, raised the obvious objection to these requirements. The alternative to premature classification is not theorylessness. Every inquiry begins from a position, and there is no concept-free description of living. The relevant distinction lies between theory used as a practice of attention, articulating what would otherwise go unnoticed, and theory used to determine in advance what the material must mean. My own fieldwork in Kolkata was planned as a study of ideas of pollution. Interviewees kept returning to the stomach: to digestion, constipation, appetite and the daily work of keeping a belly in order. The research began with a question, and the question changed because something more consequential for the people concerned kept surfacing. LVT's method, mesocosmography, makes such redirection a principle: it begins by asking what is vital in a case and lets the case declare its own coordinative organisation before the architecture is brought to bear (Ecks 2026c).
VIII. Conclusion: What Is at Stake in Making Personhood the Question?
The anthropology of personhood is not invalid because personhood is a symbolic classification. Many of the discipline's most meaningful objects are symbolic classifications, and studying how they are made, used, contested and enforced is one of its central tasks. The problem arises when an L4 stabilisation is allowed to determine in advance what must matter in the mesocosm: when it is treated as a description of the fundamental organisation of everyday living, or compared with ethnographic observations from a different level and the resulting difference declared a cultural fact.
The question with which this article began, what a person is, therefore gives way to another: what is at stake in making personhood the question? It can be asked at every level the article has distinguished. In the living coordination of those studied, it asks whether any vital stake depends on a classification of persons, and how we know. In the ethnographic encounter, it asks whether the question found a concern that was already there or created one. In the work of stabilised classifications, it asks what a given distinction enables and obstructs, and for whom. And in anthropological comparison, it asks what the comparison does for the discipline, and whether those stakes have been mistaken for the stakes of the people compared. Symbolic articulations can become consequential without encompassing all that is consequential, and a classification can matter enormously at one moment of a life while most of that life proceeds without it.
Two achievements should be distinguished. One is the study of personhood classifications where participants themselves use and maintain them, as legal systems, religious traditions and many ritual orders plainly do; there the question of what the classification enables and obstructs, and for whom, is open to direct investigation. The other is the study of how living proceeds without assuming that personhood is the category through which it must be understood. The first remains a legitimate and important project. The second is necessary if the relevance of the first is to be established rather than assumed.
The matter reaches beyond research. Introductory courses routinely teach that Westerners are individuals and that people elsewhere are relational, and students leave with a picture of themselves drawn from philosophy and law and a picture of others drawn from ethnography. Teaching the contrast as a finding stabilises it further, as publication stabilises an elicited answer, and each cohort inherits it as a fact about the world. An anthropology that takes its own methods seriously owes its students a different lesson: that every comparison has a construction and stakes of its own, and that what looks like a difference between peoples may be a difference between levels of evidence.
The lecture exercise shows the point in miniature. The difficulty in that room lay not in the uncertainty of the students' answers but in the fact that one symbolically stabilised distinction had been made the governing problem for nine radically different cases before anyone had asked what was at stake in any of them. Anthropology faces the same danger whenever it asks other people what a person is. Its safeguard is that ethnography must remain open to discovering that another problem, not personhood, was what mattered most in the lives it studies. The relevance of a question must be discovered through attention to living, not presumed from its availability within a symbolic or disciplinary framework.
References
Ecks, Stefan. 2026a. "Beyond the Compulsion of Relations: Transactive Dualism and the Anthropology of Personhood." Living Value Theory, livingvaluetheory.org.
Ecks, Stefan. 2026b. "Culture and Comparison." Living Value Theory, livingvaluetheory.org.
Ecks, Stefan. 2026c. "Living Value Theory: Core Principles (October 2026)." Living Value Theory, livingvaluetheory.org.
Ecks, Stefan. 2026d. "Who Am I? A Living Value Theory of Narrating the Self." Living Value Theory, livingvaluetheory.org.
Geertz, Clifford. 1973a. "Thick Description: Toward an Interpretive Theory of Culture." In The Interpretation of Cultures, 3–30. New York: Basic Books.
Geertz, Clifford. 1973b. "Deep Play: Notes on the Balinese Cockfight." In The Interpretation of Cultures, 412–453. New York: Basic Books.
Geertz, Clifford. 1973c [1966]. "Person, Time, and Conduct in Bali." In The Interpretation of Cultures, 360–411. New York: Basic Books.
Geertz, Clifford. 1974. "'From the Native's Point of View': On the Nature of Anthropological Understanding." Bulletin of the American Academy of Arts and Sciences 28 (1): 26–45.
Hallowell, A. Irving. 1960. "Ojibwa Ontology, Behavior, and World View." In Culture in History: Essays in Honor of Paul Radin, edited by Stanley Diamond, 19–52. New York: Columbia University Press.
LiPuma, Edward. 1998. "Modernity and Forms of Personhood in Melanesia." In Bodies and Persons: Comparative Perspectives from Africa and Melanesia, edited by Michael Lambek and Andrew Strathern, 53–79. Cambridge: Cambridge University Press.
Locke, John. 1975 [1694]. An Essay Concerning Human Understanding, edited by Peter H. Nidditch. Oxford: Clarendon Press.
Marriott, McKim. 1976. "Hindu Transactions: Diversity without Dualism." In Transaction and Meaning, edited by Bruce Kapferer, 109–142. Philadelphia: Institute for the Study of Human Issues.
Mauss, Marcel. 1985 [1938]. "A Category of the Human Mind: The Notion of Person; the Notion of Self." In The Category of the Person: Anthropology, Philosophy, History, edited by Michael Carrithers, Steven Collins and Steven Lukes, 1–25. Cambridge: Cambridge University Press.
Nonhuman Rights Project, Inc. v Breheny, 38 NY3d 555 (2022).
Salomon v A Salomon & Co Ltd [1897] AC 22 (HL).
Spiro, Melford E. 1993. "Is the Western Conception of the Self 'Peculiar' within the Context of the World Cultures?" Ethos 21 (2): 107–153.
Strathern, Marilyn. 1988. The Gender of the Gift: Problems with Women and Problems with Society in Melanesia. Berkeley: University of California Press.
Te Awa Tupua (Whanganui River Claims Settlement) Act 2017 (NZ).
Willerslev, Rane. 2007. Soul Hunters: Hunting, Animism, and Personhood among the Siberian Yukaghirs. Berkeley: University of California Press.