Abstract
Social theory routinely compares heterogeneous phenomena under common concepts, and this article examines what happens when the possibility of such comparison is taken as evidence of its relevance. Drawing on Nelson Goodman’s critique of similarity, it argues that any two phenomena resemble each other in indefinitely many respects, so that a concept’s applicability establishes little about its explanatory significance. The article defines conceptual overcommensuration as the process through which the possibility of comparing phenomena under a common concept comes to be treated as sufficient grounds for the relevance, explanatory priority or general significance of that comparison. The problem is most acute with unmarked concepts such as “exchange,” “transaction” or “consumption.” These appear to describe what people “obviously” do and so escape the scrutiny that explicitly theoretical concepts invite. Through cases ranging from a birthday party and a free piano to Bengali idioms of eating and an expensive watch, the article identifies twelve fallacies through which possible comparison is promoted into presumed explanation. It shows how institutional classification can produce the relevance it appears to discover. Living Value Theory grounds relevance in mesocosmic coordination, and on that basis the article proposes a ten-step method of recursive relevance and discernment. The method asks which comparisons matter for which questions, and why.
I. The problem is not comparison but relevance
Someone eats dinner. Someone listens to a favourite pop song. Both can be described as consumption. Whether that description is useful depends entirely on what one is trying to find out. A household expenditure survey has excellent reasons to treat food and music together, since both draw on the same budget and compete for the same income, and an economist who wants to know how households respond to a rise in prices needs exactly that comparison. A study of nourishment, digestion and metabolism, on the other hand, has little reason to place the song beside the dinner, and a study of why someone returns to one song through the years of a life has little reason to place the dinner beside the song. The same comparison is illuminating for one question and uninformative for another. The classification does not determine its own relevance.
Most theoretical concepts announce themselves as choices. A sociologist who describes a hospital in terms of biopower signals a theoretical tradition, and readers can ask what that tradition assumes. Exchange, transaction and consumption rarely arrive with such a warning label. They look like ordinary descriptions of what people obviously do, and so a research field can grow up around one of them before anyone has examined whether the comparisons it makes possible are relevant to the questions being asked.
This article investigates how possible comparisons become presumed explanations: how the fact that heterogeneous activities can be brought under a common concept comes to be treated as evidence that the concept identifies what matters about them, why this transformation is especially hard to see when the concept is unmarked, and how a set of twelve recurrent fallacies sustains it. A birthday party shows the stakes. A cake is bought, presents are given, a room is decorated, candles are lit and a child eats a slice of cake. An economist can identify expenditure, an economic anthropologist reciprocity, an anthropologist of consumption the symbolic life of goods. All of these comparisons may be legitimate, and none of them is equally relevant to every question one might ask about the party, least of all to the question of why it mattered to the people in the room. We can almost always describe an activity as exchange or as consumption. What needs to be shown is why doing so should help us understand what matters about it.
II. Goodman, comparison and mesocosmic relevance
Applying a common concept brings phenomena into a relation of comparison. To call two activities consumption is to identify a respect in which they may be treated together, and it does not establish that they share one fundamental organization. Every concept foregrounds some properties and leaves others in the background. The question is always which properties are selected, and why.
Nelson Goodman's "Seven Strictures on Similarity" (1972) supplies the philosophical ground for that question. "Similarity, ever ready to solve philosophical problems and overcome obstacles," Goodman wrote, "is a pretender, an impostor, a quack." His strictures deny similarity the work that philosophers had asked it to do. Similarity does not provide the grounds for counting two occurrences as performances of the same work or repetitions of the same behaviour. It does not explain metaphor. It does not account for our inductive practice. And it cannot be equated with, or measured by, the possession of common characteristics, since "any two things have exactly as many properties in common as any other two" once properties are counted without restriction. Three insights follow for the argument here. The first is that any two phenomena resemble each other in indefinitely many respects, so that the discovery of a similarity has, by itself, very limited explanatory significance. The second is that judgements of similarity "often require not merely selection of relevant properties but a weighting of their relative importance", and that importance "is a highly volatile matter, varying with every shift of context and interest" (Goodman 1972). Grouping things together therefore depends on the purposes, practices and questions that make some characteristics matter. The third concerns metaphor. Against the view that metaphor records a pre-existing likeness, Goodman suggested that "metaphorical use may serve to explain the similarity better than, or at least as well as, the similarity explains the metaphor" (Goodman 1972). Applying a term can generate the similarity it then appears to record. To call listening to music consumption helps to establish the very comparison that it seems merely to describe. Goodman's fifth stricture adds a connection to practice: predictions cannot rest on the bald principle that the future will be like the past, since among countless lines of similarity the question is always along which of them our predictions run. Established inductive practices may set standards of similarity more than similarity sets standards for practice.
Goodman's argument is a philosophical criticism of the explanatory pretensions of similarity, and inflating it into a general theory of relevance would be a mistake; later work has shown both its force and its limits, since some notions of similarity survive his strictures in restricted forms (Decock and Douven 2011). Its value here lies in what it forbids: the inference from the availability of a resemblance to the importance of a comparison.
That prohibition clarifies a decisive reversal between two ways of working. In the first, a researcher encounters a problem, investigates what matters about it, and determines which comparisons help to explain it. In the second, the researcher begins with a common concept, finds indefinitely many instances of it, and interprets their common classification as evidence of explanatory relevance. The second procedure may well produce correct comparisons, and its failure lies in never justifying their relevance.
Goodman shows why relevance cannot be inferred from similarity. He says much less about where relevance comes from, beyond the shifting interests of those who compare. Here Living Value Theory makes its distinctive contribution. Within a mesocosm, the world at the scale at which a living being actually lives it, certain properties and differences matter because living beings must coordinate with conditions, counterparts, histories, expectations and possibilities that are consequential for their continuing lives. Relevance in this sense, neither an intrinsic property of things nor a preference of the observer, arises in coordination, and much of it is never consciously selected or symbolically articulated. A change in a familiar person's voice can matter, as a felt misalignment at L2, long before anyone can say what has changed, and long before any researcher decides it is worth studying. Mesocosmic relevance thus gives a deeper account of what makes a comparison matter than the investigator's intentions alone.
Three relations of relevance need to be distinguished: relevance within the living coordination, relevance to an analytical question, and relevance created through institutional classification. They may coincide, and often they do not. Consumption is relevant to an expenditure survey and comparatively peripheral to someone's attachment to a song. A hospital can make patient-satisfaction scores administratively decisive while the scores reveal little about the experience of illness. And a conceptual classification can disclose something that participants had never noticed, so that the relevant distinctions cannot simply be read off participants' own articulations. The aim is to establish how a particular comparison relates to the processes and stakes under investigation, and swapping the researcher's concepts for the participants' would not achieve it.
Relevance must not itself become another unexamined universal. Relevant to whom, to which coordination, to which question, at what moment, with what consequences? Different inquiries can legitimately set different priorities, and the priorities within a coordination change as it develops. What Living Value Theory offers is a way of making the grounds of relevance investigable, so that relevance is treated neither as subjective taste nor as a property stamped on phenomena. Two cautions follow. A comparison can be relevant because it identifies material or relational conditions that participants have not articulated or that the investigator did not anticipate, so mesocosmic relevance is more than what someone happens to find interesting. And the discovery of a surprising similarity remains a legitimate theoretical achievement. Goodman does not show that comparisons are arbitrary or worthless, and a powerful theory can reveal a consequential similarity between things previously thought unrelated; much of Living Value Theory's own promise lies in such unexpected comparisons. What the theory must then establish is why the similarity matters.
The concept at the centre of this article can now be defined. Conceptual overcommensuration is the process through which the possibility of comparing heterogeneous phenomena under a common concept comes to be treated as sufficient grounds for the relevance, explanatory priority or general significance of that comparison, without adequate examination of what matters within the coordinations being compared. Once established, such classifications can organize inquiry, privilege particular similarities, obscure consequential differences, and eventually reshape the activities through which they acquire apparent confirmation. The definition does not require exchange or consumption to have one objectively correct domain of application. An activity may be productively described as consumption for one question and unproductively for another, and the problem lies in the promotion of conceptual possibility into presumed relevance, never in conceptual permissiveness as such.
III. Unmarked concepts and invisible theoretical decisions
Biopower points to an identifiable theory of modern power, associated with Michel Foucault, and invites questions about what that theory assumes (Foucault 1976). Exchange appears to be a plain fact. Identifying an activity as exchange may nonetheless involve equally consequential assumptions about transfers, reciprocity, separable parties and the organization of relationships. Both kinds of word are equally theoretical, and the distinction between marked and unmarked concepts concerns whether the theoretical operation involved in a word's use becomes visible.
W. B. Gallie described essentially contested concepts, such as democracy, art and social justice, whose proper use "inevitably involves endless disputes about their proper uses on the part of their users" (Gallie 1956). Exchange, transaction and consumption sit at the opposite pole. Where Gallie's concepts invite visible disputes over their interpretation, these function as though their applicability required no special justification at all. One might provocatively call them essentially uncontested concepts. They can certainly be contested, and economic anthropology has contested exchange for a century, but their theoretical character habitually escapes attention, so that disputes concern the kinds and meanings of exchange while its relevance as a frame passes unexamined. Unmarked concept is therefore the more precise designation.
Unmarked concepts are protected by a characteristic oscillation between a narrow, uncontroversial use and a broad, theoretically demanding one. People obviously consume food, and consumption therefore seems an obvious category. The concept is then applied to identity, cultural experience, attention and relationships. When the expanded use is questioned, the narrow meaning supplies reassurance: surely nobody denies that people consume things? The self-evidence of the narrow usage shelters the explanatory ambitions of the broad one. Exchange works in the same way. Nobody denies that people trade at markets, and the obviousness of the market stall lends its authority to descriptions of love, conversation and ritual as exchange.
Unmarked concepts also travel freely between registers, and each crossing adds to their apparent self-evidence. The consumer of household budget surveys, the consumer whose rights are protected by legislation, the consumer addressed by advertising and the consumer analysed in social theory share a word, and the borrowed authority of each use supports the others. When a policy document calls patients, passengers or students consumers, it draws on the obviousness of the shopper; when a social theorist then finds consumers everywhere, the policy usage appears to confirm the theory. Nobody along the chain has to argue that these figures were usefully compared, since the word itself seems to have done the comparing.
Disciplinary organization then naturalizes the broad use. Once there is an anthropology of consumption, researchers investigate consumption practices, consumption cultures, consumption identities and consumption inequalities, and the category's legitimacy is embedded in journals, handbooks, courses and job titles. Increasingly sophisticated theories can be produced without anyone asking whether the activities grouped under consumption have been compared on relevant grounds, because the field itself appears to answer the question.
Unmarkedness leaves the relevance of any particular comparison an open question while making an unjustified assumption of relevance easier to overlook, and it multiplies every mechanism of overcommensuration discussed below. Marked concepts can be overcommensurated too. Biopower has been stretched to cover almost any relation between institutions and living bodies, and the same promotion of applicability into relevance can occur with it. The difference is that its markedness provides a standing invitation to challenge: readers who meet biopower know they are being offered a theory and ask whether it fits. Readers who meet consumption believe they are being offered a description, and the question of fit rarely arises. That is why conceptual overcommensuration can persist among scholars who are otherwise highly reflexive. They examine the concepts that look like theory, and leave untouched the ones that look like description.
IV. What earlier critiques have established
The problem has distinguished predecessors. Karl Marx showed that commensurability is a historical accomplishment: returning to Aristotle, who had seen that exchange requires equality and equality requires commensurability but could not say what made a bed and a house commensurable, Marx located the answer in societies organized by commodity production (Marx 1867, ch. 1). Karl Polanyi named the economistic fallacy, the projection of market organization onto human livelihood in general, and distinguished the formal meaning of economy, economizing among scarce means, from its substantive meaning, the instituted process of making a living (Polanyi 1957, 1977). Giovanni Sartori distinguished legitimate increases in a concept's generality from conceptual stretching, which enlarges a concept's extension while blurring its defining properties (Sartori 1970). Ludwig Wittgenstein found, in place of a common essence of games, "a complicated network of similarities overlapping and criss-crossing" (Wittgenstein 1953, §66), and Rodney Needham carried the point into anthropology, arguing that its comparative terms are polythetic, that "there is no such thing as kinship" and that marriage is an "odd-job word" (Needham 1971, 1975). Wendy Espeland and Mitchell Stevens analysed commensuration, the comparison of different entities through a common metric, as a social process that simplifies, redirects attention and suppresses difference (Espeland and Stevens 1998). David Graeber asked why buying refrigerator magnets, putting on eyeliner and cooking dinner should all be called consumption (Graeber 2011).
Each critique leaves a part of the problem open. Polanyi's substantive economy still presupposes that making a living forms a separable domain to which economic concepts properly belong. Sartori's ladder of abstraction treats the difficulty as one of logical extension and intension, without asking why some stretched concepts escape notice while others are challenged. Needham dissolves the unity of comparative terms without saying which of their many overlapping resemblances matters for a particular inquiry. Espeland and Stevens analyse practical commensuration through metrics, and the conceptual commensuration that happens when a scholar applies an unmarked word lies largely outside their frame. These approaches identify problems of conceptual scope, classification, commensurability and theoretical universalization. Conceptual overcommensuration concerns something more specific: the unwarranted promotion of possible comparison into presumed explanatory relevance, especially where the ordinary familiarity of a concept conceals that promotion. The contribution claimed here lies less in the discovery that concepts can be overextended, which these predecessors established long ago, than in an account of the systematic relations between comparability, relevance, unmarkedness and recursive theoretical authority, and of the living coordinations in which relevance is grounded.
V. Transaction: the relevance of transfers
A commercial transaction can be identified as a bounded operation: a payment is made, goods change hands, a contract is settled, ownership or claims are transferred. The classification is extraordinarily useful, and nothing here diminishes it. A transaction is an episode within a wider process, and its boundaries, institutionally defined, documented and recognized, need not coincide with the boundaries of the living coordination within which it occurs. The receipt records a moment; the household goes on. When an analyst adopts the transaction as the unit of inquiry, the analyst also inherits the boundaries that contracts, receipts and accounting systems draw. Those boundaries were made for purposes of enforcement, taxation and record-keeping, and they are highly relevant to those purposes. Whether they also mark the edges of what matters to the people involved has to be investigated, since a purchase that closes neatly in the accounts may open a long period of worry, use, repair, pride or regret that no ledger records.
Consider a neighbour giving away a piano because it takes up too much space, and notice how many comparisons the case supports. The piano's change of ownership matters for legal questions: who is liable if it is damaged on the stairs, and who may sell it later. The recipient's provision of removal labour matters if the arrangement is examined as an exchange, since the piano is in a sense paid for with a van and a Saturday morning. The piano's weight and dimensions matter for understanding how it can be moved at all. The owner's need for space matters for understanding why the piano is leaving the house. None of these descriptions is false. What differs is their relevance to particular questions, and the case is badly served by any approach that decides in advance which comparison is the real one.
Transactive dualism appears, in this light, as an error of relevance. It is less the belief that living consists only of transactions than the treatment of transactions as the privileged elements through which all wider coordinations should be understood. The relevant similarity is selected in advance: whatever else is going on, the analyst will look for the transfer. The existence of transfers, which can be found in almost every situation, then becomes evidence that their analysis deserves explanatory priority. A parent comforting a frightened child shows where this leads. A sufficiently broad conception of transaction can describe reassurance, attention and emotional benefit as things transacted, and at that point the category loses the features that made it useful for identifying actual transactions. Rather than a packet handed over, the parent's calm is a way of breathing, holding and speaking that the child's body takes up, and the child's quieting changes the parent's breathing in turn.
The useful question for the transaction is therefore what its identification explains about the wider coordination, and what would remain unexplained if it were treated as the principal unit. For the piano, the transaction explains the legal change of owner and almost nothing about why a room had become uninhabitable or how two strangers got an instrument down a narrow staircase. For a mortgage, it may explain a great deal about a household's next thirty years.
VI. Exchange: the possibility of finding it everywhere
Exchange is extraordinarily adaptable. It can describe transfers, reciprocal obligations, cooperation, communication and relationships through ever more abstract ideas of giving and receiving. Such descriptions need not be wrong. Their weakness is that they select similarities from an almost unlimited range of possible comparisons, and Goodman's point applies with full force: that a conversation and a sale can both be seen as giving and receiving tells us very little until we know why that respect of comparison matters. Goodman's remark on metaphor applies as well. The everyday phrase "an exchange of words" seems to report a likeness between talk and trade, and repeated use helps to create the likeness it reports, so that by the time a theorist describes conversation as exchange the comparison feels like an observation. What the metaphor brings forward is turn-taking and response, and what it leaves behind is everything about a conversation that has no counterpart in a transfer: tone, hesitation, the shared attention to something that neither party gives the other.
The useful question is the relevance of exchange to an investigation, more than its bare presence. At a market stall, exchange explains price formation, changes of ownership and the reciprocal expectations of buyer and seller. Within a family, exchange may be highly relevant to disputes about money, inheritance, gifts or obligations, and an analysis that ignored it would miss real grievances. During an evening spent comforting a frightened child, exchange offers a possible interpretation, and the investigator must show why it illuminates the ongoing coordination before relying on it. Some situations are principally organized by exchange, some contain exchange in a subordinate role, and some can be described as exchange only by metaphor; an adequate theory must be able to tell which is which, and the telling is a matter of relevance to the question as much as of presence in the situation.
The distinction between transaction and interrecursive being-with remains crucial. A transaction identifies a bounded arrangement involving a transfer or a change of claims; exchange emphasizes a relation between contributions given in view of one another; interrecursive being-with is coordination in which living beings continuously respond to one another's responses. A relationship can involve all three at once. A couple may share bills under an explicit agreement, keep a rough tally of favours, and live through years of mutual adjustment that neither could describe as contributions. What matters is that the researcher recognizes the different operations and determines their relevance to the question. Interrecursive responsiveness may be what sustains a relationship even when transfers and reciprocity are also present, and an exchange framework that sees only the second and third will explain the bills and the favours and miss the relationship.
The gift shows the logic of competing comparisons. A gift can be studied as a transfer of ownership, a reciprocal obligation, a demonstration of affection, an attempt at recognition, a material inconvenience to the recipient, or a response to someone's need. All of these comparisons are available, and the researcher has to explain why one or several of them are relevant to the case at hand. Marcel Mauss remains an indispensable predecessor on this view (Mauss 1925). His demonstration that obligation and enduring relationship pervade what looked like free giving was a major discovery about many gifts, and it neither makes gift exchange a universal explanation nor turns it into an invalid category. An unwanted present passed on to a charity shop, a coat given to a neighbour who is cold, and a necklace circulating in a ceremonial ring are all gifts, and the comparisons that illuminate them differ.
Universal theories of exchange therefore fail less because exchange cannot be found in every case than because finding it is too weak a justification for making it the centre of explanation.
VII. Consumption: when a useful comparison becomes a field
Consumption shows most clearly how relevance can disappear beneath a familiar classification. Return to the dinner and the song. For research on household expenditure, food and music are meaningfully compared as spending categories. For research on embodiment, digestion and nourishment, the consumption of food is centrally relevant, and music has no place. For research on a song's recurring importance in someone's life, calling listening consumption is much less illuminating, since nothing is used up and the song is, if anything, more present after the hundredth hearing than after the first. The category's relevance shifts with the investigation.
Conspicuous consumption sharpens the point. Thorstein Veblen's account of reputability, pecuniary emulation and the display of wealth remains important (Veblen 1899). Consumption and ostentation can overlap, and distinguishing them requires no denial that an expensive watch was bought or that it is worn, only the recognition that expenditure and recognition-seeking can have different explanatory significance. For a study of household debt, the expenditure may matter most. For a study of prestige, the practices of display and recognition may matter most. For a study of inheritance, continuity of ownership and family history may matter most, as when the watch passes to a grandson. The same object supports several relevant comparisons, and none of them determines the others. A further consequence is the concealment of similarity. An inherited watch worn to be noticed and a charitable donation made to be noticed fall under different economic headings, ownership and giving, though both may involve closely related processes of status display; a hospital nurse turning a patient at night and a daughter turning her bedridden mother are classified as paid work and unpaid care, though the lifting, the timing and the broken sleep may be the same. A classification that establishes the wrong commonality can hide the right one, so that the whole field of comparison is distorted by misleading proximities and misleading distances.
The ethnographic cases point the same way. Bengali idioms of eating bring taking a bribe, being beaten, being deceived, embezzling, losing one's livelihood and taking medicine into the vocabulary of ingestion. For a study of the semantics of eating, the comparison is the object; for a study of corruption or of violence, the shared verb is a curiosity. The refusal of school meals cooked by a Dalit cook can be described as non-consumption, a comparison relevant to a nutritional survey of the midday-meal scheme, while for understanding what happened the relevant comparisons are with other practices of caste exclusion and commensality. Taking an antidepressant involves ingestion and metabolic effects that matter for pharmacology, and consumption is far from irrelevant to medication. For understanding why someone keeps taking a drug, or stops, the social classification of the illness, expectations of the medicine and the relationship with the prescriber may be decisive. The question is always what the concept captures relative to the explanatory task.
Here the problem of unmarkedness returns with most force. The anthropology of consumption has become so institutionally recognizable that consumption is treated as an empirical object, something found in the world, and its status as a comparative classification, a respect in which an analyst has chosen to treat things together, drops out of view. Researchers debate the motivations, meanings and consequences of consumption without asking why the activities they study were assembled into that category. Once the field exists, its boundaries expand by analogy: places are consumed by tourists, images by viewers, news by readers, experiences by everyone. Each extension is easy to make, because each activity can be compared to buying and using up in some respect, and each one enlarges the field without anyone having to show that the comparison illuminates the activity better than the alternatives. Symbolic meaning often supplies the apparent unity, as when Mary Douglas and Baron Isherwood treated goods as an information system and Pierre Bourdieu showed how taste classifies the classifier (Douglas and Isherwood 1979; Bourdieu 1979). Daniel Miller's finding that everyday shopping in north London was largely an expression of love for particular family members shows what happens when the evidence pulls against the frame: if the stakes of shopping are care, then consumption may not be the most useful comparison for understanding it (Miller 1998).
The productive question is therefore what treating these activities as varieties of consumption allows us to discover, and which similarities and differences the classification makes harder to recognize. That question replaces a less useful one, whether a given activity really is consumption, and the replacement is more than a change of terminology. It turns a dispute about membership into an investigation of relevance.
VIII. The twelve fallacies: mechanisms of misallocated relevance
Twelve recurrent fallacies explain how possible comparison is promoted into presumed relevance. They are distinguishable mechanisms, not twelve names for one mistake, and conceptual overcommensuration is the overall transformation to which they contribute, not a thirteenth fallacy beside them. Unmarkedness and definitional oscillation belong to a different order, as the conditions under which the mechanisms operate without scrutiny. Not every case involves all twelve, and they need not occur in a fixed order.
Four mechanisms produce an apparent commonality. In historical universalization, a comparison established within one historical arrangement is treated as relevant to others without renewed justification, so that the consumer becomes a universal human figure and market exchange a model for cooperation everywhere. In transactionalization, transactions become the preferred units of comparison even where other dimensions of living coordination matter more. In categorical conflation, heterogeneous activities are grouped under one classification, and the different respects in which they might usefully be compared disappear behind a shared word, as when buying, eating and listening all become consumption. In false decomposition, analytical partitions such as production and consumption are treated as the relevant divisions within continuous living processes, as though the life of a coat really divided into its making and its using up. The general error in all four is the unjustified selection of the respects in which phenomena are to be compared, and the classifications themselves may be perfectly accurate. Each mechanism answers the question of which respect of comparison to use before anyone has asked which respect is relevant, and the answer is supplied by habit, history or the available vocabulary.
Four mechanisms confer explanatory relevance, and they form the heart of the problem. In the move from occurrence to centrality, an operation occurs and therefore becomes the presumed focus of explanation: the birthday includes a purchased cake, so the birthday becomes a consumption event. This can happen even when the category is correctly applied. In the move from applicability to explanation, a concept can describe an activity and the description is therefore treated as explanatory: an intimate relationship can be described as an exchange of benefits, and the description is taken to explain its persistence. In unlimited extension, ever more phenomena become comparable through a concept, creating an appearance of growing explanatory reach with no corresponding gain in discriminating power, until giving and withholding, buying and abstaining, all fall under the same heading. In symbolic substitution, an available symbolic classification determines what appears worth investigating, displacing attention from other consequential aspects of coordination, so that status categories replace the investigation of enjoyment and prices replace the investigation of what people stand to lose.
Two mechanisms allocate explanatory priority. In mediational supremacy, one dimension of coordination, material production, symbolic distinction, economic incentive or interpersonal obligation, is presumed universally relevant or determining without its significance being established in the case at hand. In inferred motivation, a familiar analytical classification supplies a motive, expensive ownership becoming status competition, shopping becoming identity construction, cooperation becoming reciprocal advantage, and the investigation of what is actually at stake is bypassed. The distinctive problem in both is that theoretical relevance is assigned before empirical relevance has been examined. Consider a woman who buys an expensive coat. A theory committed to symbolic distinction will find status display, one committed to economic incentive will find a calculation of durability against price, and one committed to obligation will find the expectations of the workplace she dresses for. Any of these may be right. Each arrives at the case already knowing which mediation will turn out to matter, and the possibility that the coat matters because she is cold on a long walk to work, or because it resembles one her mother wore, is excluded before the investigation begins. The fallacy lies in the advance allocation, whichever answer it happens to favour.
Two mechanisms maintain the allocation. The birthday shows how these four work together. The cake was bought, so the party contains a purchase; the purchase is classified as consumption; consumption is described as identity work; and the identity work is then said to be what the party was about. Every step makes a comparison that is available, and no step establishes that the comparison illuminates what made the afternoon matter to a six-year-old and her grandmother.
In critical inversion, a theory contests the evaluation of an activity while keeping the same classification and its assumed relevance, condemning as exploitation or commodification what another approach celebrated as exchange or consumer freedom; the criticism may be justified, and it leaves the relevance of the frame untouched. A critique of the commodification of care, for instance, can denounce the reduction of care to an exchange of services while continuing to analyse care through the same categories of provision, payment and receipt, so that the critical position borrows its frame from the position it opposes. In recursive confirmation, classifications reorganize inquiry and practical activity, producing new evidence that appears to justify the classifications themselves.
The mechanisms reinforce one another. The first group supplies a common frame, the second promotes the frame into an explanation, the third decides which part of the explanation matters most, and the fourth protects the result from challenge. Overcommensuration arises when the resources of conceptual comparison are mistaken for criteria of relevance, and it persists when established classifications determine which comparisons are subsequently recognized as important.
IX. Money and the seduction of common measures
Money is the clearest practical case of the difference between comparability and relevance. It makes rice, rent, medicine, labour, artworks and bus fares comparable in one dimension, and that comparison is relevant to budgets, purchases, accounts and economic models. Two things can have identical prices, and the equality can be decisive for a household deciding what it can afford. The same equality may say very little about the different significance of the two things for anyone's continued living: a month's rent and a course of treatment may cost the same, and nothing about that tells us what either does in a life. Georg Simmel described both sides. Money becomes "the common denominator of all values", and in doing so "hollows out the core of things, their individuality, their specific value, and their incomparability" (Simmel 1903). Practical commensuration produces real and consequential comparisons, and it does not establish their explanatory priority for every investigation.
Economic concepts such as exchange, transaction and consumption share money's peculiar combination of familiarity and reach. They are used every day, they seem to describe obvious activities, and they open almost unlimited possibilities of comparison. That combination makes them especially prone to the promotion of possible comparison into presumed relevance. Whether pervasive monetary commensuration has historically encouraged broader habits of conceptual commensuration is a plausible hypothesis, and it remains a hypothesis until historical evidence is brought to bear on it.
X. Recursive confirmation and the production of relevance
A classification can be poorly aligned with a living coordination when it is introduced and become practically consequential once institutions reorganize their activities around it. Classifying students as consumers may at first foreground one limited aspect of university life, their rights against poor provision. If universities then reorganize teaching, administration, evaluation and funding around consumer-oriented measures, the classification acquires a relevance it did not have before: satisfaction scores now affect league tables, recruitment and budgets, and students learn to approach their studies through expectations of delivery. That newly produced relevance does not show that the original classification captured what mattered about higher education. A classification can become relevant because people organize their activities through it without becoming an adequate explanation of the wider coordination, and its practical effectiveness is a different matter from its explanatory adequacy, as studies of how economic models shape the markets they describe have shown (MacKenzie 2006).
The recursivity levels locate the process. At L1 a coordination proceeds without articulation, at L2 a misalignment is felt, at L3 it is articulated, at L4 classifications stabilize interpretations and decisions, and at L5 those classifications can themselves be examined. Inherited L4 classifications can organize sophisticated L5 theories, and they can become increasingly consequential through institutional adoption while those theories proceed. Researchers at L5 also work within institutions that reward the use of established categories: grant calls ask for studies of consumption, journals publish within recognized fields, and a dissertation framed around an unmarked concept needs no defence of its frame. The theorist's reflexivity is thereby channelled toward refining the classification and away from examining its relevance. The ability to articulate and criticize a classification differs from the ability to discern its relevance, and a theory can be intensely reflexive about exchange while remaining convinced that everything important must be exchange.
XI. A method of recursive relevance and discernment
The positive contribution is a sequence of questions that can be applied to exchange, transaction, consumption or any other general concept, and its aim is discernment of relevance more than the policing of categories.
The first step is to make the concept visible. Whenever an ordinary category appears to describe an activity directly, its use is recognized as an analytical choice: the claim that an activity is exchange becomes the proposition that it is being investigated as exchange. This denies nothing; it makes the choice examinable. The second step is to identify the specific comparison proposed. Which properties, operations or relations are being selected? If dinner and listening to music are both called consumption, precisely what similarity is claimed, and does it go beyond a shared word? The third step is to establish relevance to the particular question. What does the comparison help to explain, what becomes newly intelligible through it, and would a different comparison expose something more consequential?
The fourth step is to reconstruct mesocosmic relevance, without limiting relevance to the investigator's interests. How does relevance arise within the coordination itself, through embodiment, being-with, dwelling, multimateriality and multisymbolization? What do participants sustain, anticipate, depend on or struggle with, and which differences become consequential through felt misalignment or disruption? The fifth step is to distinguish the recursive counterparts, establishing where nonrecursive, selfrecursive and interrecursive dynamics are consequential. A comparison between a commercial transaction and an intimate relationship may be revealing precisely because their counterpart dynamics differ, even though both can be described as exchange. The sixth step is to test alternative comparisons. Instead of asking only whether exchange or consumption applies, the same event is investigated through several plausible comparisons, and the investigator establishes which disclose the relevant processes and which remain peripheral.
The seventh step is to examine what a comparison excludes. Which important similarity might be hidden because activities have been assigned to different categories, and which misleading similarity is created by placing them in the same one? The eighth step is to use null cases appropriately. A concept's null case remains useful for testing its boundaries, especially where a researcher claims that a specific operation actually occurs. Null cases are not, however, the final criterion of theoretical success. An activity may unambiguously involve exchange while exchange has limited relevance to what matters about it, and a deliberately metaphorical comparison may be highly illuminating where no literal exchange occurs. The decisive test is relevance, and categorical membership is only part of it. The ninth step is to investigate recursive changes in relevance. Has the classification altered the activity? Have particular comparisons become more consequential because institutions, participants or researchers now organize their activities through them? Prior relevance must be separated from relevance produced by classification.
The tenth step is to apply the same test to Living Value Theory. The concept of living coordination does not escape overcommensuration merely because it names a more general process. Its value has to be demonstrated through its capacity to identify relevant differences in actual coordinations, and it must not replace the indiscriminate attribution of exchange with an equally indiscriminate attribution of living coordination. The theory has to show why its mediations, counterpart distinctions and recursivity levels allow more adequate investigation of what matters.
The free piano shows the method at work without any need to declare exchange the wrong classification. Made visible, the concepts on offer are exchange, gift, transfer, removal, household reorganization and material remediation. Each selects something real: the van and the Saturday as a return, the absence of payment, the change of owner, the carrying, the reclaimed room, the instrument's new life in another house. For the question of why the piano left, the owner's need for space and the reconfiguration of her household are most relevant, and exchange contributes little. For the question of how the recipient's family came to have a piano, the gift and the labour of moving it matter more, and the recipient's stake, perhaps a child's lessons, enters the analysis. For a dispute about damage on the stairs, the transfer of ownership becomes decisive. Mesocosmically, the weight of the piano and the geometry of two houses dominate the coordination, embodiment enters through the strain of lifting, and being-with through the brief cooperation of strangers. Instead of one correct label or a list of interchangeable ones, the method yields an account of which comparison matters for which question, and why.
XII. Conclusion: What makes a comparison matter?
Dinner and a favourite song can both be described as consumption. Sometimes the comparison is useful, as when a household is deciding what it can afford, and sometimes it obscures far more than it reveals, as when we want to know why a song has stayed with someone for twenty years. The possibility of placing phenomena under a common classification does not establish the relevance of that classification. The twelve fallacies explain how this possibility is repeatedly promoted into explanatory authority, unmarked concepts explain why the promotion so often goes unnoticed, and institutional practices can reproduce and amplify its consequences until the classification seems confirmed by the world it has helped to make. Social theory has always compared things that are different, and must; its central problem lies in mistaking an unlimited capacity to generate comparisons for an understanding of which comparisons matter.
References
Bourdieu, Pierre. 1979. La distinction: Critique sociale du jugement. Paris: Éditions de Minuit.
Decock, Lieven, and Igor Douven. 2011. "Similarity After Goodman." Review of Philosophy and Psychology 2 (1): 61–75.
Douglas, Mary, and Baron Isherwood. 1979. The World of Goods: Towards an Anthropology of Consumption. New York: Basic Books.
Espeland, Wendy Nelson, and Mitchell L. Stevens. 1998. "Commensuration as a Social Process." Annual Review of Sociology 24: 313–343.
Foucault, Michel. 1976. Histoire de la sexualité. Vol. 1, La volonté de savoir. Paris: Gallimard.
Gallie, W. B. 1956. "Essentially Contested Concepts." Proceedings of the Aristotelian Society 56: 167–198.
Goodman, Nelson. 1972. "Seven Strictures on Similarity." In Problems and Projects, 437–446. Indianapolis: Bobbs-Merrill.
Graeber, David. 2011. "Consumption." Current Anthropology 52 (4): 489–511.
MacKenzie, Donald. 2006. An Engine, Not a Camera: How Financial Models Shape Markets. Cambridge, MA: MIT Press.
Marx, Karl. 1867. Das Kapital: Kritik der politischen Ökonomie. Vol. 1. Hamburg: Otto Meissner.
Mauss, Marcel. 1925. "Essai sur le don: Forme et raison de l'échange dans les sociétés archaïques." L'Année sociologique, nouvelle série, 1: 30–186.
Miller, Daniel. 1998. A Theory of Shopping. Cambridge: Polity Press.
Needham, Rodney. 1971. "Remarks on the Analysis of Kinship and Marriage." In Rethinking Kinship and Marriage, edited by Rodney Needham, 1–34. London: Tavistock.
Needham, Rodney. 1975. "Polythetic Classification: Convergence and Consequences." Man, new series, 10 (3): 349–369.
Polanyi, Karl. 1957. "The Economy as Instituted Process." In Trade and Market in the Early Empires, edited by Karl Polanyi, Conrad M. Arensberg and Harry W. Pearson, 243–270. Glencoe, IL: Free Press.
Polanyi, Karl. 1977. The Livelihood of Man. Edited by Harry W. Pearson. New York: Academic Press.
Sartori, Giovanni. 1970. "Concept Misformation in Comparative Politics." American Political Science Review 64 (4): 1033–1053.
Simmel, Georg. 1903. "Die Großstädte und das Geistesleben." In Die Großstadt: Vorträge und Aufsätze zur Städteausstellung, edited by Theodor Petermann, 185–206. Dresden: von Zahn und Jaensch.
Veblen, Thorstein. 1899. The Theory of the Leisure Class: An Economic Study of Institutions. New York: Macmillan.
Wittgenstein, Ludwig. 1953. Philosophical Investigations. Oxford: Blackwell.