Abstract
Nineteenth-century anthropology ranked peoples and societies along evolutionary sequences. Twentieth-century anthropology, especially after Franz Boas, rejected racial hierarchy, unilineal evolution, and increasingly the vocabulary of primitive and civilised peoples. Cultural relativism is usually read as marking a fundamental epistemological break, the moment the discipline stopped judging human worlds against one another.
This article argues that the standard history misses a deeper continuity. Ranking could occur only after anthropology had already selected the dimensions along which human formations were treated as comparable in the first place. Technology, religion, kinship, property, political organisation, writing, and subsistence did not present themselves as the natural axes of human difference. Anthropology made them such, and it did so through what Living Value Theory calls a SWIPE, a split of worlds paired with incommensurable epistemologies. By separating humans considered as natural organisms from humans considered as cultural beings, nineteenth-century anthropology created an asymmetrical economy of difference: psychic and eventually biological commonality became increasingly protected from hierarchical inference, while culture became the legitimate repository of human variation.
Adolf Bastian's psychic unity, the 1892 handbook Notes and Queries on Anthropology, and Boas's own dismantling of racial and evolutionary ranking are read as successive stages of this architecture, one that survived every subsequent critique of ranking because the critiques never reached the prior selection ranking depended on. The article closes by proposing comparative relevance, not relativism, as the missing methodological problem: comparison should not begin by asking how two cultures differ, but by establishing why a particular difference matters for the process under investigation.
I. The Familiar Story: From Hierarchy to Relativism
Victorian anthropology organised human societies into evolutionary sequences. Savagery, barbarism, and civilisation supplied overtly developmental stages, and technologies, marriage systems, religions, political institutions, and forms of subsistence were read as evidence of where a given people stood along that sequence. Then came the twentieth-century correction. Franz Boas attacked the comparative method precisely where it assumed that similar cultural phenomena must share identical causes. Historical particularism rejected universal developmental sequences in favour of the specific historical trajectory of each culture. Cultural relativism undermined the practice of judging one society by another's standards.
This should be treated as a real intellectual achievement rather than something to be debunked. Boas explicitly rejected the inference from cultural accomplishment to racial capacity, insisting that historical circumstance, not supposed racial endowment, explained the difference between societies. Later anthropology increasingly separated biological variation from cultural difference as a matter of principle rather than convenience, and the separation held.
A puzzle nonetheless remains once the achievement is acknowledged. If anthropology abandoned the ranking of cultures, why did so much of its comparative vocabulary stay organised around distinctions such as simple and complex, stateless and state, segmentary and centralised, oral and literate, subsistence and market, undifferentiated and differentiated? Contemporary scholarship can still deploy simple and complex societies as ostensibly neutral analytical categories, decades after the ranking those categories once served was supposed to have been abandoned.
The conventional history, this suggests, concentrates too heavily on ranking and not enough on the prior construction of rankable dimensions. The article therefore asks a different historical question. Before anthropology could rank human worlds against one another, how did it decide which differences between them mattered in the first place?
II. Difference Is Cheap: Relevance Is the Problem
Any two human populations differ in indefinitely many respects. They differ in average rainfall, cooking technique, vocabulary, household composition, architecture, humour, inheritance practice, footwear, naming convention, sleeping arrangement, greeting practice, food preference, religious practice, legal procedure, and innumerable further respects that could be added without ever reaching a natural stopping point. They also resemble one another under indefinitely many descriptions at the same time. The instruction to compare two cultures is, on its own, radically underdetermined, because similarity and difference always require a specified respect before either one can be asserted at all.
The distinction that matters is therefore not between difference and sameness but between difference and relevant difference. A difference is relevant only relative to some theoretical or empirical problem. Whether two populations use different roofing materials may be decisive for a question about thermal adaptation and close to irrelevant for a question about inheritance. Whether they recognise different deities may be decisive for one inquiry and nearly beside the point for a different process in which their organisational arrangements turn out to converge.
This yields the general principle the rest of this article develops. Comparison does not begin when a difference is discovered; it begins when a respect of difference is selected, and the selection itself is the comparative-relevance problem this article addresses: what warrants choosing one dimension of human variation as the relevant one rather than any other. Selection of this kind is unavoidable, and nothing here argues that anthropology should have compared without selecting. The problem is not selection; it is selection that remains unacknowledged or simply inherited, so that a dimension chosen for one purpose, decades or centuries earlier, goes on organising comparison for purposes it was never chosen for.
III. The Nature/Culture SWIPE and the Unequal Distribution of Difference
Living Value Theory has a name for the historical operation that structures the comparative field this article is describing: a SWIPE, a split of worlds paired with incommensurable epistemologies. A companion article develops the concept in full for the case of social theory's own Enlightenment inheritance, and the definition it gives applies without modification here. All SWIPEs are binaries, but most binaries are not SWIPEs. Most binaries sit on a gradient, hot and cold, fast and slow, near and far. A SWIPE is different in kind. It is simultaneously an ontological split and an epistemological one, dividing not only what exists into two worlds but also licensing two different, mutually exclusive ways of coming to know each world.
The 1892 second edition of Notes and Queries on Anthropology supplies an unusually explicit demonstration of exactly this operation at work inside a single discipline. The handbook opens by defining its own object: "Anthropology may be defined as the natural history of man," divided into two main branches, anthropography and ethnography. Anthropography, the handbook states, treats man "from a purely animal point of view, that is, from a structural and functional aspect," while ethnography "deals with him as a social and intellectual being," covering manners, customs, institutions, history, traditions, language, religion, intellectual aptitudes, industries, and arts. The first part of the volume is devoted to anthropography, the second to ethnography, and the two parts do not simply organise existing knowledge into convenient chapters. They distribute human difference into two different epistemic regimes at the moment of definition, before a single observation has been recorded.
The consequence is profound. One human being becomes two analytical objects. The natural human has anatomy, physiology, reproduction, sensation, pathology, and physical variation, matter for anthropography's structural and functional gaze. The cultural human has clothing, habitation, technology, religion, kinship, property, government, writing, exchange, and art, matter for ethnography's social and intellectual gaze. An asymmetry follows directly from this division and deepens across the following decades of the discipline's history. Nature increasingly becomes the domain in which commonality must be protected beneath variation, a domain where difference is measured but not, in principle, allowed to imply fundamental inequality. Culture becomes the domain in which consequential variation must instead be discovered despite an increasingly insisted-upon underlying commonality. Compactly: nature comes to mean variation without fundamental difference, while culture comes to mean difference without biological inequality.
The SWIPE fractures continuous living processes along exactly this line, in each case assigning one half to nature and the other to culture. Metabolism becomes nature; cuisine becomes culture. Reproduction becomes nature; marriage becomes culture. Thermoregulation becomes nature; clothing becomes culture. Vocal production becomes nature; language becomes culture. Mortality becomes nature; burial becomes culture. Sensory capacity becomes nature; aesthetics becomes culture. None of these pairs was a gradient waiting to be discretised for convenience. Each is a single mesocosmic process, engaging embodiment, being-with, dwelling, multimateriality, and multisymbolization together, that the classificatory operation has cut in two and distributed to different disciplines with different epistemologies before either discipline began its actual work.
The claim this section makes is therefore stronger than the observation that anthropology classified human difference. The nature/culture SWIPE did not merely classify differences that were already known and waiting to be sorted. It created an asymmetrical economy of rankability, determining in advance where anthropologically consequential human difference would subsequently be sought, and where it would not.
IV. Bastian's Paradox: Psychic Equality Enables Cultural Hierarchy
Adolf Bastian supplies the pivotal figure for understanding how this economy of rankability could coexist with a sincere and influential doctrine of human equality. Bastian's psychic unity of mankind, Elementargedanken, held that human beings everywhere share a fundamental psychic constitution, elementary ideas common to the species as such. Geographically and historically differentiated Völkergedanken, folk ideas, were then understood to produce the actual cultural variation observable across peoples. This doctrine mattered enormously to the discipline's subsequent development. It became important to Boas, who engaged seriously with Bastian's legacy before turning against the evolutionary uses to which psychic unity was more commonly put, and it was equally compatible with British evolutionary anthropology, which never treated it as a rival claim.
This compatibility is the paradox this section exists to explain. Psychic unity sounds, on its face, inherently egalitarian, a doctrine that could only undercut claims of racial or cultural superiority. Yet psychic unity was perfectly compatible with cultural evolutionism, and it did more than merely coexist with it. Psychic unity helped make cultural evolutionary comparison possible in the first place. If human populations possessed fundamentally different kinds of minds, there could be no single developmental sequence along which their cultural achievements could be meaningfully arranged at all, since the achievements of different kinds of minds would simply be incommensurable with one another. Shared psychic constitution supplies the control condition a developmental sequence requires. Differing cultural outcomes then become the variable the sequence is built to explain. The equation follows directly: the same basic humanity, combined with different cultural achievements, yields different developmental positions along a single scale. Psychic unity and evolutionary ranking are not, on this reading, historical opposites that happened to coexist for a time. They are structurally suited to one another.
The methodological significance of this reaches beyond Bastian's own writing. What his framework accomplishes is the de-ranking of the natural or psychic human while leaving cultural outcomes fully available for differentiation and, eventually, ranking. Human commonality, established at the level of underlying psychic constitution, does not eliminate hierarchy from anthropological comparison; it relocates hierarchy, moving it out of the domain of what humans fundamentally are and into the domain of what their worlds have made of shared human capacities. The psychic unity of humankind made it possible to insist, without any felt contradiction, that humans were fundamentally alike and that their worlds occupied radically different positions of development.
V. Notes and Queries: Manufacturing the Space of Cultural Difference
Notes and Queries on Anthropology deserves treatment as this article's central case study, because few documents make the construction of comparative relevance as visible as this one does, and because the visibility is not an accident of hindsight. The handbook was compiled to standardise anthropological observation among travellers, missionaries, colonial officials, and others positioned to encounter unfamiliar peoples but untrained in systematic inquiry, and its preface states its own diagnosis of the problem it exists to correct with unusual candour. Travellers, the editors write, "have usually recorded only those customs of modern savages which they have chanced to observe," and, as a rule, they "have observed chiefly those which their experience of civilized institutions has led them to look for." Information gathered this way has sometimes been "lamentably distorted" to fit preconceived ideas, so that "the imperfections of the anthropological record surpass those of other sciences, and false theories are often built upon imperfect bases of induction." The remedy the handbook proposes is a set of standardised questions, intended to let the traveller "collect information without prejudice arising from his individual bias," while, tellingly, also directing him toward determining "the non-existence of others to which attention is drawn," not only what is present, but what is conspicuously absent according to a template the traveller did not devise himself.
This solution deserves reinterpretation through the lens of comparative relevance rather than being read simply as a triumph of scientific method over amateur impressionism. The manual does not eliminate selection; it transfers selection from the individual observer to the discipline itself. A traveller's own salience, that custom looks strange, is replaced by disciplinary salience, all populations should be investigated with respect to this specified list of domains, and the replacement is a real methodological advance in exactly the respect the preface claims. It is not, however, an escape from selection, only a relocation of who performs it and at what point in the process.
The sequence of selections this relocation puts in place can be traced with some precision through the handbook's own structure. An ontological selection comes first: which aspects of human life belong to ethnography at all, as opposed to anthropography or to no anthropological domain whatever. A domain selection follows: the ethnography section of the 1892 edition organises inquiry under headings that include totemism, music, etymology, language, poetry, history, archaeology, war, hunting, nomadic life, pastoral life, agriculture, the training of animals, slavery, relationships, causes that limit population, burials, astronomy, arithmetic, property, trade, money, measures and weights, marital relations, education, games and amusements, communications, tribal marks, and memorial structures, among much else, each assigned to a named contributor: Edward Burnett Tylor on etymology, language, and arithmetic; James Frazer on totemism; Francis Galton on physical powers, heredity, and astronomy; Augustus Pitt-Rivers on archaeology and games; John Lubbock on relationships and measures; even Charles Darwin, credited with the section on physiognomy. A variable selection operates within each domain, sex, age, rank, material, function, inheritance, ritual use, manufacture, historical change, specifying which further distinctions within a domain are worth recording once the domain itself has been settled. A theoretical selection follows, asking which variations count as evidence of development, diffusion, invention, survival, or conservatism, categories that presuppose an evolutionary or diffusionist framework the questionnaire's own neutral tone otherwise conceals. A practical selection closes the sequence, shaped by what matters to museums, to imperial administration, and to the disciplinary institutions the handbook was itself designed to serve.
The central observation this sequence supports can be stated plainly. A person encountered in the field does not live as a concatenation of kinship plus religion plus property plus technology plus political organisation. These are disciplinary cuts through a single ongoing life, imposed from outside that life by a questionnaire built for a different purpose than living it. Yet once every population encountered by a traveller has been translated into the same slots, comparison becomes straightforward in a way that conceals the work the translation performed to make it so. Society A has marriage form X; society B has marriage form Y. Society A uses technology X; society B uses technology Y. The comparison appears purely empirical, a simple report of what was found, only because the conceptual infrastructure that made the finding possible, that decided marriage and technology were the things to look for, has already disappeared from view by the time the comparison is stated.
One distinction is worth repeating rather than stated once and left behind, because it is exactly the distinction crude constructivism collapses. The questionnaire does not invent the practices its categories go on to record. Marriage, in whatever form, was really there before any questionnaire arrived; the practices ethnography documents are not fabrications. What the questionnaire produces is not the practices but their comparative salience, the fact that this particular practice, among the indefinitely many features of a life the questionnaire could in principle have asked about, becomes the thing recorded, compared, and eventually ranked, while countless other features of the same life go unrecorded because no heading was ever built for them.
VI. Before Hierarchy: Relevance Selection as the Deeper Operation
The central theoretical turn this article makes can now be stated directly. Cultural evolutionism is usually criticised for having ranked societies, and the criticism is correct as far as it goes. But ranking, on the argument developed above, is logically downstream of a prior and less examined operation. Before societies A and B can be ranked according to technology, technology must already have been selected as a relevant dimension of comparison between them. Before writing systems can be ranked, writing must already have become relevant. Before political complexity can be ranked, political organisation must already have been isolated as a domain in which different societies could be compared to one another at all. The sequence runs selection, then comparison, then ranking, and each stage presupposes the one before it without being reducible to it.
Twentieth-century criticism concentrated overwhelmingly on the third stage of this sequence. Boasian anthropology attacks the inference from comparison to a universal evolutionary sequence, showing that similar cultural outcomes can arise from entirely different historical processes. Later relativism attacks evaluative ranking directly, denying that one society's standards license judgement of another's. Both critiques are real and both changed the discipline permanently. But the first operation, selection itself, can survive both critiques fully intact, because neither critique was aimed at it. A discipline can stop ranking technologies against one another while continuing to treat technology as an obviously relevant dimension along which societies differ in ways worth recording; a discipline can refuse to call one kinship system more advanced than another while continuing to assume, without argument, that kinship is one of the dimensions any adequate ethnography must address.
This yields what should be read as the article's pivotal sentence. The deepest inheritance from nineteenth-century cultural evolutionism may therefore not be its rankings at all, discredited and abandoned as those specifically have been, but the space of differences within which ranking became possible in the first place, a space that has proved considerably more durable than any particular ranking constructed within it.
VII. Boas: Dismantling the Ladder While Preserving the Field
Boas deserves careful treatment rather than the role of a foil against which this article's argument can be made to look sharper than it is. Two distinct interventions need to be kept apart in his work, because collapsing them obscures exactly the continuity this article is tracing.
The first is the rejection of racial mental hierarchy. Boas's argument against the inference from cultural achievement to racial capacity strengthened, more than any single prior contribution, the separation between biological capacity and historically transmitted cultural behaviour; George Stocking's classical history of the discipline treats this as central to the formation of the modern culture concept, and the assessment has held up well. The second intervention is the rejection of unilineal cultural evolution as such. Historical particularism attacks the assumption that similar cultural phenomena necessarily result from the same evolutionary process, insisting instead that cultures have histories, plural and specific, rather than positions along one universal ladder.
The magnitude of this transformation should not be understated. But it is worth asking precisely what survives it. Cultures remain identifiable as differentiated objects after Boas exactly as they were before him. Cultural traits remain describable in the same terms. Kinship, language, technology, religion, and social organisation remain available for comparison across the cultures Boas himself studied and wrote about extensively. What becomes illegitimate is the question higher or lower; what survives, apparently undisturbed, is the question different in what way, together with the inherited domains within which any answer to that second question gets given. Those domains prove remarkably durable across a transformation that changed almost everything else about how anthropology conducted itself.
The distinction this section wants to draw can be put compactly. Boas transformed the rules governing comparison far more radically than he transformed the prior selection of what counted as comparable in the first place. This needs immediate qualification, since the claim should not be overstated into the assertion that nothing changed in Boas's hands; comparative priorities and categories did shift under his influence, sometimes substantially. The precise claim is narrower and, for that reason, more defensible: cultural relativism, whatever else it accomplished, does not by itself constitute a theory of comparative relevance, and Boas's own writing never claimed that it did.
VIII. The Migration of Hierarchy: From Persons to Arrangements
This section makes the article's most provocative claim, and it is worth stating the linguistic evidence before the theoretical argument built on it. Primitive mind has become unacceptable as a description of any human population; simple society remains, in a great deal of contemporary scholarship, analytically respectable. Lower race has become unacceptable; low-complexity political organisation can remain ordinary technical vocabulary, appearing in journal articles without apology or qualification. This is not merely a matter of finding a more polite synonym for the same judgement; it reflects the nature/culture SWIPE precisely, redrawn one level further down. Persons have been increasingly protected by the doctrine of psychic and biological equality this article has traced from Bastian through Boas. Social arrangements, the cultural output of those persons rather than the persons themselves, remain legitimately differentiable, and differentiable along exactly the dimensions the nineteenth century first selected as relevant.
Anthropology, put compactly, progressively learned to de-rank human beings while continuing to rank human worlds.
The claim becomes subtler, and more difficult to dismiss, once a natural objection is addressed directly. Rank need not mean moral superiority, and most contemporary uses of complexity language are offered as purely descriptive: more levels of political hierarchy, greater occupational specialisation, larger population, more institutional differentiation, more elaborate technology. Simple and complex remain analytically active vocabulary across anthropology, archaeology, and adjacent comparative fields, and recent scholarship often explicitly presents these categories as neutral, free of the evaluative freight their nineteenth-century ancestors carried. But these dimensions, whatever their intentions, still produce ordered comparative spaces, spaces in which one society can be located above or below another along a scale, however carefully the scale's designers disclaim any evaluative intent for the ordering it nonetheless produces.
The crucial question this disclaimer never quite answers is neutral according to what criterion. A society can be called simple only relative to dimensions selected in advance of the comparison, and the selection determines the verdict as surely as it did in the nineteenth century, only now without any accompanying claim to be doing something value-laden. A society might show low occupational differentiation alongside extraordinarily elaborate ecological knowledge that a differentiation-based scale has no place to register. It might lack centralised government while sustaining immensely complex interrecursive relations among its members and with nonhuman others that a state-versus-stateless axis was never built to detect. It might lack writing while maintaining elaborate transgenerational symbolic archives carried by other means entirely. Complexity, on this evidence, does not reside in a society considered in itself; it is complexity with respect to a dimension selected in advance, and the society that appears simple on one dimension can appear extraordinarily elaborate on another the comparison never thought to ask about.
IX. Anthropology's Inherited Cultural Ontology
The historical argument can now be connected to the discipline's present organisation. The comparative domains the nineteenth century selected, kinship, religion, economy, politics, law, technology, material culture, language, did not remain historical curiosities once cultural evolutionism itself fell out of favour. They became subfields, textbook chapters, ethnographic headings, and areas of professional expertise, and their institutionalisation makes the original act of selection increasingly difficult to see from within the discipline that inherited it.
The question a fieldworker asks accordingly changes in character across this process of institutionalisation, from why should kinship be privileged as a comparative dimension at all to simply what is the kinship system here, a question that no longer carries any trace of its own history as a selection among alternatives. The category has become ground rather than figure. This is worth connecting, briefly and without overloading a historical argument with more terminology than it needs, to what Living Value Theory calls L4 stabilisation: a provisional analytical distinction, useful for a particular purpose at a particular moment, becomes recorded, taught, institutionalised, and recursively reproduced across generations of researchers who inherit the question rather than selecting it afresh for themselves. The result is a specific kind of disciplinary path dependence in comparative relevance, one in which the questions available to ask are shaped less by the phenomena under study than by which questions previous generations of the discipline happened to institutionalise.
None of this should be read as declaring the inherited domains unreal. Economy exists. Religion exists, as both an emic category for many of the people anthropologists study and an institutional reality with real consequences. Kin relations exist and matter enormously to the people who live within them. The mistake this article is identifying is not the claim that these domains are real; it is the treatment of their comparative relevance, their status as the dimensions along which societies ought to be compared to one another, as though that relevance followed automatically from their reality, rather than being a further claim requiring its own justification that the discipline largely stopped making once the domains became familiar enough to feel self-evident.
X. Cultural Relativism Does Not Solve Comparative Relevance
The two problems this article has been distinguishing throughout need to be separated with full rigour at this point, since the temptation to conflate them is considerable. Relativism asks by whose standards a cultural formation should be understood or evaluated. Comparative relevance asks a different and prior question: why this dimension has been selected for comparison at all, before any question of whose standards should apply to judging what is found along it.
A perfectly relativistic anthropology can, and often does, reproduce inherited comparative categories without any friction between the two commitments. One can refuse to rank religions against one another while still assuming, without examining the assumption, that religion is a natural cross-cultural unit every society can be expected to have something to say about. One can refuse to rank kinship systems while assuming kinship should be investigated everywhere a fieldworker goes. One can celebrate the existence of multiple ontologies, in the contemporary sense that vocabulary has acquired, while assuming that ontology itself is the relevant dimension along which those multiple worlds differ from one another, an assumption relativism as such gives no grounds for either accepting or questioning.
This generalises well beyond the nineteenth-century categories this article has focused on so far. Every subsequent theoretical school in anthropology performs its own act of relevance selection, and naming a few makes the pattern vivid. Marxist anthropology selects production and class as the relevant dimensions. Symbolic anthropology selects meaning and classification. Feminist anthropology recovers neglected dimensions such as gender and reproduction that earlier selections had left invisible. Ontological anthropology selects being, worlds, and relations. Medical anthropology selects illness, healing, and embodiment. None of these selections is illegitimate, and this article is not arguing for a comparative anthropology that selects nothing. The methodological requirement proposed here instead is narrower and, for that reason, achievable: that selection be made explicit and justified relative to the specific explanatory problem a given piece of research is actually trying to solve, rather than inherited silently from whichever domains a prior generation of the discipline happened to institutionalise.
XI. From Cultural Difference to Consequential Difference
The positive alternative this article has been building toward can now be stated. The question how different are these cultures should be set aside, not because it is meaningless but because it has no theory-neutral answer of the kind it appears to promise; any answer already presupposes a respect of comparison the question itself never specifies.
The concept proposed in its place is consequential difference. A difference is comparatively relevant when changing that difference would alter the process actually being investigated, and nothing weaker than this should count. Two healing systems may invoke radically different symbolic vocabularies, spirits in one setting, neurotransmitters in another, while sharing a strikingly similar recursive organisation: expert classification, authoritative naming, material intervention, and repeated monitoring of the patient's condition over time. On the process of how authority over a body is established and maintained, the symbolic vocabulary may be close to irrelevant even though it is, on the surface, the most conspicuous difference between the two systems. Conversely, two psychiatric clinics using almost identical diagnostic terminology may differ radically in how a diagnostic classification, once assigned, becomes authoritative, contestable, or reopenable, a difference no comparison of terminology alone would ever surface. Two populations may share a single religion while differing profoundly in how they organise inheritance. Two populations may hold entirely different religions while organising property, authority, or care in remarkably similar ways.
The comparative object, on this account, can never simply be culture as a whole, treated as a bounded unit to be set beside another such unit. Comparison has to be process-specific from the outset. The methodological principle this yields can be stated directly: do not begin by inventorying the most conspicuous cultural differences a fieldworker happens to notice first. Begin instead by specifying the process that actually requires explanation, and only then determine which differences, among the indefinitely many available, could plausibly alter that process. Comparison conducted this way becomes capable of a result ordinary cultural comparison structurally cannot produce: the discovery that an apparently dramatic difference is irrelevant to the process at hand, and that an apparently minor one is, for that same process, decisive.
XII. Living Value Theory and the Reconstruction of Cultural Comparison
Living Value Theory offers a way to reconstruct cultural comparison without simply installing another universal checklist in place of the one this article has been dismantling, and this qualification is important enough to state before anything else. The five mediations must not be substituted for kinship, religion, economy, and politics as the new, finally correct dimensions of comparison. Doing so would reproduce the original error in a new vocabulary, replacing one inherited grid with another that merely sounds more theoretically up to date.
What Living Value Theory supplies instead is a process ontology from which relevance can be argued rather than assumed, and the distinction this makes available maps directly onto the exclusion procedure a companion article develops for cross-species comparison. For cultural comparison, the question is never whether a given human group possesses the three zoetic mediations, embodiment, being-with, and dwelling; every human group necessarily does, and their presence establishes commonality without discriminating anything. Nor is the relevant question merely whether a human group shows interrecursivity, multimateriality, or multisymbolization in some degree; human social life already exhibits all three extensively, everywhere it has been observed. The comparative question that actually discriminates concerns how particular processes are organised across the mediations and recursive levels in a given case, and which differences in that organisation become consequential for the process under study.
This escapes the old architecture in a specific and demonstrable way. A house need not be assigned either to natural environment or to cultural architecture, split across the nature/culture SWIPE before its role in any particular process has even been considered; it can instead be followed as the multi-mediated thing it actually is, engaging dwelling, multimateriality, being-with, and multisymbolization together, whichever of these turn out to matter for the process being investigated. Food need not be split between metabolism, assigned to nature, and cuisine, assigned to culture, before its role in a specific inquiry has been established. Marriage need not be isolated from reproduction, material resources, dwelling, symbolism, and interrecursivity and handed to a single discipline authorised to speak about it alone. The process in each case can be followed across the SWIPE rather than allocated to one side of it in advance, which is exactly the manoeuvre a companion article develops at length for the case of social theory's own inheritance from the same historical partition. The task this article shares with that one is not to add nature back into culture, restoring some lost wholeness after the fact; it is to stop beginning from the cut.
XIII. A Protocol for Relevant Cultural Comparison
The article closes with a practical methodological payoff rather than only a historical and theoretical one. For any comparative project, five questions should be asked before the comparison itself begins. What process is actually being explained? What potential dimensions could plausibly alter that process? Which dimensions are being excluded from consideration, and on what grounds? What evidence would show that an apparently conspicuous difference between the cases is, for this process, actually irrelevant? And what evidence would show that some overlooked difference is, for this same process, consequential after all?
A sixth question is worth adding, since it exposes a bias the first five do not automatically catch. Would the comparison have been formulated in the same way if neither case under study were being treated as the implicit reference point against which the other is measured? This question is useful precisely because it surfaces anthropocentric, Eurocentric, and narrowly disciplinary salience without treating any single one of these as the whole of the problem this article has been diagnosing; a comparison can be free of Eurocentric bias in the ordinary sense and still inherit an unexamined selection of relevant dimensions from some other, less obviously implicated source, a prior theoretical school, a disciplinary convention, a questionnaire compiled for entirely different purposes over a century earlier.
Comparison conducted this way becomes an explicit hypothesis about relevance, stated and defended before the comparison proceeds, rather than a catalogue of differences assembled first and interpreted afterward as though their comparative significance had been given along with their discovery.
XIV. Conclusion: Before Comparison
The historical sequence this article has traced can now be stated in summary. Bastian helped establish psychic unity as a discipline-wide commitment. Evolutionist anthropology combined that unity, without apparent contradiction, with confident cultural hierarchy. Boas decisively attacked both racial hierarchy and unilineal cultural evolution, a transformation of the first importance that this article has not sought to minimise anywhere along the way. Relativism further delegitimised ranking as such, extending Boas's attack into a general methodological principle. Beneath all of these transformations, however, a less examined operation remained largely undisturbed: the selection of the dimensions through which human worlds became culturally comparable to one another in the first place.
Anthropology, on the argument this article has developed, did not simply inherit differences between cultures waiting to be discovered and recorded. It inherited a space of culturally relevant differences, already structured before any particular fieldworker arrived to work within it. The nature/culture SWIPE was crucial to constructing that space: nature increasingly became the domain of protected shared humanity, while culture became the legitimate and expected domain of human variation, available for comparison, and for a very long time, for ranking. The discipline then became progressively critical about how cultural differences should be evaluated, a criticism that reached its fullest expression in relativism, without becoming equally critical about why particular differences had been selected as relevant for comparison in the first place, a selection relativism never directly addressed because it was never the target relativism was aimed at.
Comparison does not begin with difference; it begins with the production of a space within which certain differences are permitted to matter and others are not, and that production is itself a historical achievement requiring explanation rather than a neutral precondition comparison can simply take for granted. The task for a renewed comparative anthropology is accordingly neither to rank cultures, an error the discipline has already spent a century correcting, nor to refuse comparison altogether, which would abandon the discipline's central intellectual project. It is to make every comparison earn its criterion of relevance.