I. A Recurring Frustration
Anyone who has spent time in anthropology seminars will recognise a particular genre of concluding sentence. After an hour of theoretical exposition, the lecturer arrives at what is presented as the decisive insight: culture, power, identity, whatever the object under discussion, is not a fixed entity. It is constantly changing, fluid, hybrid, in flux. Students nod. The point feels correct. And yet something is wrong. The wrongness is hard to name, because the statement is not false. It is, if anything, obviously true. So why does it feel so unsatisfying? Why does it have to be repeated, semester after semester, generation after generation, as if it were still news?
The answer, from the perspective of Living Value Theory (LVT), is that the statement is not a solution. It is a symptom. It tries to repair a problem without naming the problem. And the problem is not that the concept of culture was being held in a static form. The problem is that it was being treated as an entity at all, and adding the adjective "dynamic" to an entity-form leaves the entity-form entirely intact. What you get is not a corrected concept. You get a moving thing where before you had a stationary thing. The underlying error is reproduced, just with more motion added.
This essay argues that the social sciences have been suffering from a structural confusion that has never been explicitly named or resolved: the systematic failure to distinguish between concepts used to track processes and concepts used to stabilise entities. This distinction, once seen, is almost impossible to unsee. It dissolves entire literatures of debate, retroactively explains why certain concepts have worked while others have not, and offers a precise criterion for what good conceptual work in the social sciences should look like. The argument is developed here through Living Value Theory, a framework that began as a theory of value and health in South Asian contexts but has expanded into a general theory of mesocosmic coordination, that is, a theory of how life actually holds together across embodiment, being-with others, dwelling, material engagement, and the use of symbols.
II. The Ontological Underdetermination of Concepts
The first and most fundamental insight is one that has no established name in social theory: concepts are ontologically underdetermined with respect to their mode of use. A word does not carry within it any signal of whether it is being deployed to track a coordination process or to stabilise an entity. Nothing in the surface form of the term "culture," "power," "identity," or "the market" tells you whether it is functioning as a way of noticing ongoing processes or as a way of naming a bounded thing that can be pointed to, measured, counted, or owned. That decision lives entirely in the mode of use, and it can shift silently, within a single text, within a single paragraph, without any marker that would alert a reader or even the writer themselves.
This is not a minor observation about language. It is a structural feature of symbolic mediation with very large consequences. Because symbols can be stabilised, detached from their original contexts, and circulated widely, they naturally accumulate the appearance of referring to stable objects. The same term that was introduced as a heuristic for tracking coordination, as a way of noticing patterns in motion, can, through repetition and circulation, begin to behave as if it named a thing. And once it behaves that way, it starts to generate the kinds of questions that things generate: does it really exist? where are its boundaries? is it the same thing across different places? These are not bad questions in general. They are questions that are perfectly appropriate when applied to tables, or stars, or chemical compounds. They are almost always the wrong questions when applied to phenomena in inter-recursive social domains.
The underdetermination of concepts is therefore not merely an epistemic puzzle. It is the source of most of the conceptual turbulence that the social sciences have been experiencing since they were founded. Once you see that a concept can be used either as an entity or as a process, and that the concept itself does not tell you which is happening, the history of anthropological theory becomes legible in an entirely new way.
III. Inter-Recursive Domains and the Impossibility of Entity Ontology
The deeper claim, from which everything else follows, is this: in inter-recursive domains, entity ontology is structurally incompatible with the nature of the phenomena being studied. This is not a matter of preference or theoretical style. It is an ontological constraint.
An inter-recursive domain is one in which the entities involved respond to each other, to how they are described, to the expectations placed on them, and to the categories through which they are organised. Social life is the paradigm case. Human beings in coordination are not passive objects that can be measured from the outside without being changed by the measurement. They adjust, resist, internalise, strategise, and reorganise in response to how they are categorised. The moment a category enters social life, "the unemployed," "the mentally ill," "the consumer", it begins to reorganise the very coordination it claims to describe. It becomes performative. It is no longer a representation of something fixed. It is an intervention into something that was already in motion.
This is fundamentally different from the situation in domains where things are effectively non-recursive. A table does not reorganise itself in response to how it is described. The sun does not change its trajectory because astronomers have named it. In those domains, entity-ontology is not just acceptable but appropriate. Point-mode stabilisation, the act of treating something as a fixed referent that can be named and measured, works precisely because the thing does not respond to its own naming. This is what allows physics and chemistry to function with the conceptual precision they achieve. The world they study stays put while being described.
Social life does not stay put. The moment social theory installs a stable referent, a culture, a society, a class, a subject, and treats it as if it were as stable as a table or a star, it has done something that cannot be undone by adding qualifiers. It has imposed a non-responsive ontology onto a domain that is irreducibly responsive. And the consequences of that imposition propagate forward, generating endless secondary debates that are, at root, attempts to manage the consequences of an initial error.
This means that LVT is proposing something stronger than a methodological preference for process thinking. It is proposing that in inter-recursive domains, entity-use is always a closure device. It may be a necessary closure device, sometimes coordination requires acting as if things were more stable than they are, and there is no disgrace in that. But it must never be mistaken for ontology. When it is mistaken for ontology, the distortions become systematic, and the debates they generate are irresolvable, because they are asking the wrong questions.
IV. The Entity/Process Diagnostic
From these two claims, that concepts are ontologically underdetermined, and that entity-forms are structurally incompatible with inter-recursive domains, a simple but powerful diagnostic follows. When encountering any concept in social theory, ask: is this being used to track a process of coordination, or to install a stable referent? If it is tracking coordination, ask further: what specific coordination is being made visible here that would otherwise be missed? If it is installing a referent, ask: what closure is being performed, what is being hidden by that closure, and what pseudo-problems will follow from it?
This diagnostic has the unusual property of dissolving debates rather than advancing them. Take the case of culture. When applied here, the diagnostic immediately reveals that the concept is doing double duty. Used processually, it can be a productive way of tracking how practices, expectations, symbols, and materials circulate and recombine across social contexts. Used as an entity, it immediately generates the familiar set of pseudo-problems: where does one culture end and another begin? do individuals "have" cultures? is it the same culture if some people don't share all of it? These questions have consumed enormous intellectual energy and produced almost no resolution. From an LVT perspective, this is expected, because the questions arise not from genuine complexity in the world but from the wrong mode of use of the concept. The entity-form generates its own problems.
What is striking about the anthropological literature on culture is not that it failed to notice this. Many anthropologists have been exquisitely sensitive to the dangers of reification, bounded-group thinking, and essentialism. What the literature consistently failed to do is state the underlying rule. It noticed the symptoms without naming the disease. Christoph Brumann's widely cited defence of the culture concept, for instance, offers a sophisticated and largely processual description of what culture actually does, it is social sharing, uneven distribution, constant recombination, overlapping circles of participation. This is almost exactly what an LVT-compliant account would look like. But then the text reinstalls "culture" as a quasi-entity that people can "have," "use," "capture," or "distribute." The entity-form returns, softened and qualified, but structurally unchanged. The corrective "culture is dynamic and unevenly distributed" is not a different concept. It is the same concept wearing more modest clothing.
The diagnostic does not call for abandoning culture or any other particular concept. It calls for a continuous attention to mode of use. A concept that was introduced processually can become entity-like as it circulates. A concept that has hardened into an entity-form can be reclaimed for process work by deliberately shifting how it is held. The concept is not the problem. The unreflective solidification of its use is the problem.
V. Process-Entity Drift and the Cycle of Theoretical Crisis
Once the entity/process distinction is in place, the history of social theory becomes readable as a pattern of recurring oscillation. Call this process-entity drift: the mechanism by which concepts that begin as process-tracking insights gradually solidify into entity-forms, generate pseudo-problems, attract critique, and are eventually abandoned in favour of new concepts that begin the cycle again.
The trajectory is remarkably consistent. A theorist or ethnographer encounters a phenomenon that existing vocabulary cannot adequately capture. They introduce a concept that is, at this stage, a way of noticing coordination, a gesture toward something that is happening that was not previously visible. The concept is productive. It travels. It enables comparisons, generates research programmes, becomes teachable. In order to travel well, it has to be stabilised. Stabilisation means simplification, compression, and detachment from the original context in which the coordination was observed. The concept becomes a shorthand. Shorthands behave like entities. Once it behaves like an entity, it starts generating the wrong questions, and eventually a new generation of scholars declares it flawed, ideologically suspect, or theoretically obsolete.
This happened with culture. It happened with kinship. It happened with society, structure, symbol, and identity. Each of these terms had an original processual insight somewhere in its genesis. Each became dominant through stabilisation. Each eventually attracted a major critique that was, in effect, a response to the entity-form the concept had settled into. And in most cases, the resolution involved replacing the concept with a new one, hybridity, practice, assemblage, affect, that itself was processual in its initial form and quickly began to repeat the cycle.
The crucial point is that this is not evidence of theoretical failure in the ordinary sense. It is evidence of a structural feature of symbolic mediation. Symbols can be stabilised and circulated in ways that other forms of coordination cannot. That is precisely what makes them so powerful. But the same property that makes them travel well is what makes them drift from process to entity. The concept-generating power of symbolization and its drift toward reification are two sides of the same capacity.
LVT does not propose to escape this cycle by finding the right concepts. It proposes to escape it by changing what a concept is allowed to be, by holding, as a methodological commitment, that in inter-recursive domains, concepts are valid only insofar as they track coordination without closing it.
VI. What Coordination Concepts Actually Look Like
One of the most useful pieces of evidence for this argument is retrospective. Looking back at the concepts that have proven most durable and most empirically productive in the anthropology of health and medicine, terms like polyiatrogenesis, pharmaceutical citizenship, and biocommensuration, a striking pattern emerges. These are not concepts that try to name entities or bound populations. They are concepts that track coordination patterns: how different kinds of intervention, classification, expectation, and material process intersect and sometimes catastrophically misalign within the lives of people trying to remain well.
Polyiatrogenesis, for instance, does not name a thing. It names a process in which multiple therapeutic interventions, each individually defensible, compound in ways that produce harm across mediations, embodied, material, institutional, symbolic. The concept is not tied to any particular substance, disease category, or health system. It is tied to a pattern of coordination failure that can recur across very different settings. This is why it travels well, but it travels differently from entity concepts. It travels as a pattern, not as a label. You do not sort populations into those who have polyiatrogenesis and those who do not. You ask: is this pattern of multi-level intervention failure present in this situation? The concept remains a question about coordination, not a claim about kinds of things.
This retrospective observation is generalisable as a criterion. A good concept in the social sciences is one that makes a specific coordination pattern visible that would otherwise be missed or misdescribed. Its value lies in its capacity to orient attention toward something in motion. It should not compress heterogeneous coordination into a single stable label, and its generality, when it has generality, should come not from abstracting away from specificity but from recognising recurrent process configurations.
When applied as a criterion to the existing conceptual repertoire of social theory, the results are both illuminating and slightly uncomfortable. A large proportion of canonical concepts in anthropology and sociology turn out to be primarily entity-stabilising rather than coordination-tracking. They classify, they count, they delimit. The most productive concepts, by contrast, tend to track thresholds, misalignments, transitions, and breakdowns. They are precisely the concepts that are hardest to fit into a standard glossary because they resist being defined as if they named a class of things.
VII. Why Multi-Symbolization Dominates the Conceptual Field
LVT understands social life as occurring within what it calls the mesocosm: the site in which embodiment, being-with others, dwelling in place, material engagement, and symbolic mediation are always already co-present and irreducibly entangled. None of these five mediations is more real than the others. None can be reduced to any of the others. Coordination is always multi-mediational, and its characteristic forms, its stabilities, its breakdowns, its innovations and pathologies, emerge from the interplay among them.
Given this framework, one would expect a roughly even distribution of conceptual resources across the five mediations. In fact, the distribution is dramatically uneven. The domain of multi-symbolization, the use of language, categories, narratives, classifications, numbers, and other symbolic forms, is by far the most conceptually dense. The domain of embodiment is the least. Dwelling and being-with occupy intermediate positions. Multi-materiality, which concerns the way objects, technologies, substances, and physical infrastructures mediate coordination, has been conspicuously underdeveloped in the LVT framework thus far, despite being clearly a domain of high consequence.
This asymmetry is not a failure of theoretical balance. It is a faithful map of a real difference in how coordination becomes available for conceptualisation. Multi-symbolization is uniquely concept-generative because it is the only mediation that systematically externalises and stabilises its own outputs. When coordination is achieved through symbolic means, the symbols persist. They can be stored, repeated, compared, transported across contexts, and recursively reworked. They produce artefacts that can themselves become objects of coordination. This makes symbolically mediated coordination unusually visible, unusually manipulable, and therefore unusually available for conceptual work.
Embodiment, by contrast, does most of its coordinating work below the threshold of symbolic availability. The vast majority of what the body is doing at any moment, metabolic regulation, postural adjustment, temperature management, immunological activity, is deeply consequential for how coordination can proceed, but it does not externalise itself in a form that invites conceptual multiplication. It is not that embodiment is less important than multi-symbolization. It is that its processes resist the kind of stabilisation that produces concepts. The concepts that do emerge from embodiment, pain, fatigue, disruption, somatic foregrounding, metabolic drag, are not entities. They are threshold-events: moments when the body's ongoing coordination becomes conspicuous within the wider field of mesocosmic coordination, usually because something has gone wrong or is under stress.
The principle that governs this distribution is what LVT calls mesocosmic salience. A process becomes mesocosmically salient when it enters shared coordination in a way that is visible, consequential, and trackable across contexts. Salience, in this sense, is not the same as importance. A process can be deeply important without being salient, the liver performing its ongoing biochemical work is an example, and can be highly salient without being particularly important in the long run. Salience is a function of whether a process can be noticed, shared, and followed across actors and situations. Multi-symbolization manufactures salience: it is, in a sense, a machine for converting processes into shareable, revisable form. That is why concepts proliferate there.
VIII. The Stratification of Salience
A crucial refinement prevents mesocosmic salience from collapsing into a merely symbolic concept. Salience is itself stratified across levels of recursivity. The assumption that salience means "what is noticed and talked about" already encodes a bias toward higher-level symbolic articulation. But a process can be mesocosmically salient at levels that are entirely below symbolic articulation, and this lower-level salience is often the more decisive kind.
LVT distinguishes at minimum four levels at which salience operates. At the first level, what might be called latent salience, a process is fully operative in coordination, structuring everything that follows, but entirely backgrounded. It does not stand out. It does not call attention to itself. It is salient precisely because it is what makes everything else possible, and it only becomes visible when it ceases to function. Much of what embodiment does operates at this level. So does the taken-for-granted infrastructure of dwelling: the reliability of floors, walls, and streets, the predictability of ambient temperature, the stability of the spatial arrangement within which social life unfolds. These things coordinate at L1 without appearing in coordination.
At the second level, felt salience, something begins to register as misalignment, tension, or disruption. It is present to a recursive being as a disturbance, but it has not yet been articulated. Pain before it is named. Unease before it is theorised. The sense that something is wrong before there is any concept of what is wrong. This is where a great deal of what matters most in social life actually lives: in the space between the fully backgrounded and the articulable. Social theory has great difficulty with this zone, because it is, by definition, resistant to the symbolic processing that produces theoretical discourse.
At the third level, articulated salience, something can be named, discussed, and compared. This is where concepts operate. At the fourth level, stabilised salience, something has been compressed into a category or decision-machine that can be applied rapidly and without reflection.
The implications for conceptual work are significant. LVT concepts, like all concepts, operate at the third and fourth levels. They can only ever capture what has crossed the threshold into articulability. But the phenomena they most need to track often reside at the first and second levels. This creates a permanent structural tension: the most fundamental processes are the least conceptually tractable, while the most conceptually available processes, those that have already been symbolically stabilised, are not necessarily the most fundamental.
This means that a good LVT concept must do something paradoxical: it must be articulable enough to function as a concept while pointing back toward something that resists full articulation. The concept of polyiatrogenesis does this. It can be explained, defined, and applied. But what it points to, the way multiple mediational pressures compound within a life, is something that is felt long before it is named, and is often erased by the very classifications that medical and institutional systems use to manage it.
IX. The Power of Multi-Mediational Concepts
If the richest conceptual territory lies at points where mediations couple and misalign, then the strongest concepts will be those that are irreducibly multi-mediational. This is not a methodological preference but a consequence of how the mesocosm actually works. Phenomena that can be explained within a single mediation are relatively rare. What makes social life interesting, and what makes it capable of generating genuine surprises, is precisely the entanglement of mediations that do not reduce to each other.
Consider again the case of polyiatrogenesis. What it tracks is the convergence of symbolic classification (the diagnostic categories that determine which treatments are appropriate), multimaterial intervention (the drugs, devices, and procedures that are applied to the body), institutional coordination (the structures of care that sequence and combine interventions), and embodied response (the body's actual capacity to absorb, metabolise, and recover from cumulative intervention). None of these mediations produces the phenomenon alone. The harm is emergent: it arises from the way they combine, and it cannot be identified, let alone addressed, without a concept that holds all of them together.
Pharmaceutical citizenship works similarly. It does not name a group of people or a kind of drug or a type of institution. It names the configuration of coordination through which a person's political standing, biological status, access to care, and capacity for self-presentation become entangled in ways that are mediated by pharmaceutical objects. The concept makes a specific form of mediation visible, one that could not be seen from within the perspective of pharmacology, or political theory, or medical sociology alone.
The general principle is that good social concepts should be expected to straddle mediations, because what they are tracking is coordination, and coordination is always multi-mediational. A concept that sits entirely within one mediation is probably either too narrow to be interesting or too abstract to be informative. The concepts that do the most work are those that name configurations of entanglement, specific ways in which embodied, material, relational, spatial, and symbolic processes intersect to produce outcomes that are surprising, harmful, generative, or all three at once.
X. Against Hierarchical Ordering
A persistent temptation in building any theoretical framework is to organise its concepts hierarchically, to identify some as foundational, others as derived, and to rank them by centrality or importance. LVT resists this temptation, and the resistance is not incidental but principled. Once concepts are understood as tracking different kinds of mesocosmic processes rather than naming different levels of a single reality, the grounds for hierarchical ordering dissolve.
Consider two concepts that might seem to be obvious candidates for different positions in a hierarchy: ontological fit and transactive dualism. Ontological fit is a diagnostic operator. It asks, of any conceptualisation: how well does this track the coordination it claims to describe? It applies everywhere. It is relevant whenever any concept is in use, in any domain. Transactive dualism, by contrast, is a historically specific process-pattern. It names the particular way in which modern market economies have installed a symbolic overlay on the mesocosm, one in which transactions are understood as exchanges between self-contained entities carrying equivalent values, and in which this symbolic structure has become so deep and so pervasive that it reorganises the very coordination it claims to describe. Transactive dualism has enormous scope and enormous consequences. But it is a specific historical configuration, not a universal diagnostic tool.
Which is more important? The question cannot be answered, because the two concepts are not competing for the same role. They are performing different kinds of work. Asking which is more important is like asking whether a scalpel or a blood pressure monitor is more important in medicine. The answer depends entirely on what needs to be done in a specific situation.
What this means for the construction of a theoretical lexicon is that hierarchy must be abandoned in favour of what might be called orthogonality: concepts differ not by how fundamental they are but by what kind of work they do. Some concepts track recurring coordination patterns that are widespread across contexts, perturbation, closure, recursive misalignment. Others name historically specific and massively consequential formations, transactive dualism, the Enlightenment binary purifications of nature and culture, sacred and secular. Still others function as diagnostic operators, ontological fit, mesocosmic salience, the entity/process distinction itself. And others mark thresholds and breakdowns that are fragile and context-dependent, somatic foregrounding, mediational mismatch, point-mode stabilisation.
These are not the same kind of concept, and they cannot be ranked. Removing any of them would leave a specific kind of blind spot. Their value lies not in their position in a hierarchy but in their precision and in the particular coordination pattern each one makes visible.
XI. The Deeper Genealogy of Entity-Thinking
The argument made so far might seem to imply that the entity-form is simply a mistake, an error that better theory could have avoided. But this would itself be too simple. Entity-thinking in the social sciences has a genealogy, and that genealogy explains both its power and its persistence.
The drive to render social life legible through stable entities is not merely a conceptual habit. It is connected to a deep and historically specific anxiety about verification. Modern institutions, law, science, medicine, bureaucracy, and markets, are built around the need to demonstrate, to prove, and to account for. They require that claims be made in a form that can be checked, compared, and contested. This requires entities: things that can be pointed to, counted, bounded, and treated as the same across different contexts and times.
The Protestant Reformation introduced into Western culture an unusually intense preoccupation with the problem of trust and its verification. In theological terms, this took the form of a demand that claims about salvation be grounded in something demonstrable rather than simply delegated to institutional authority. But the consequences spread far beyond theology. The verification drive, the need to render claims checkable through some form of external evidence, became a pervasive feature of modern institutional life. And verification, as a structural demand, favours entity-ontology. You cannot verify a process in the same way you can verify a thing.
This helps explain why entity-concepts are not simply errors. They are responses to real institutional pressures. Disciplines that wanted to exist, to be funded, to train students, and to accumulate knowledge in a form that could be passed on, had very strong reasons to stabilise their objects of study into entity-like forms. The alternative, sustaining the full complexity of coordination in motion, across all its mediations, at all its levels, is both intellectually demanding and institutionally impractical. The entity-form is a compression that enables circulation. The price of that compression is the systematic distortion it introduces.
The history of anthropology becomes, from this perspective, less a story of intellectual failure and more a story of a discipline caught between two irreconcilable demands: the demand to study phenomena that are irreducibly processual and inter-recursive, and the institutional demand to produce stable, transmissible knowledge in entity-like form. The endless debates about culture, structure, and power are, in significant part, the fallout from this tension. They cannot be resolved within the existing framework because they arise from the framework's own constitutive contradiction.
XII. What an LVT Concept Is Allowed to Be
All of this leads to a reformulation of what conceptual work in the social sciences should look like. The decisive move of LVT is not the introduction of new concepts as such, though new concepts do emerge from it. The decisive move is a change in what a concept is allowed to be.
A concept, in the LVT framework, is not a label for an entity. It is a way of tracking mesocosmic coordination, a way of making a specific process visible that would otherwise remain invisible or be misdescribed. This means that the value of a concept lies not in its precision as a definition but in its sensitivity as an instrument. The question to ask of any LVT concept is not "what does this refer to?" but "what coordination does this allow me to see?"
This shift has several consequences.
First, it means that concepts will not be evenly distributed across domains. They will cluster where coordination becomes mesocosmically salient, where processes cross into shared visibility in ways that invite conceptual work. This explains the density of LVT concepts around multi-symbolization and their relative sparseness around embodiment: not because symbolization is more important, but because it manufactures salience in a way that the other mediations do not.
Second, it means that the strongest concepts will be multi-mediational. They will arise at points where different kinds of coordination intersect and misalign, because it is precisely there that existing single-mediation vocabularies fail and a new concept earns its place.
Third, it means that concepts cannot be ranked by importance. They perform different kinds of work, and the work each performs is irreducible to the work performed by others. A theory that is built this way will look less like a pyramid and more like a field, a set of instruments, each calibrated to a specific kind of coordination, none of which can be reduced to or derived from the others.
Fourth, and most importantly, it means that the continuous vigilance required by this framework can never be discharged. Even the best concepts can drift. Even process-tracking concepts can harden into entity-forms as they circulate. The entity/process diagnostic is not something applied once and then set aside. It is a standing orientation toward one's own conceptual work, a reminder that the question is never settled, that every concept must be held in use, not just in definition.
XIII. A Note on Anthropology's Best Work
There is a generous version of this argument that is worth stating. Anthropology's best ethnographic work has always been, implicitly, process work. Ethnography at its most powerful does not name entities. It shows coordination in action, the way people navigate competing demands, manage misalignments, repair breakdowns, and innovate within constraints. The finest ethnographies are full of processes. They show rather than classify. They follow rather than bound.
The problem is that this process sensitivity at the level of empirical work has never been matched by a corresponding clarity at the level of theory. The gap between what ethnography shows and what theory says has generated enormous productive tension, but it has also generated enormous confusion. When ethnographers produce rich descriptions of coordination and theorists respond by asking which entity their work belongs to, which culture, which structure, which discourse, the disconnect can be disorienting for everyone involved.
LVT provides a vocabulary that is adequate to what ethnography was already doing. The insight that all good anthropological concepts are process concepts is not a criticism of anthropology. It is, in the deepest sense, an act of theoretical fidelity to its own best practice.