Abstract

Words are credited with extraordinary powers—from making Christ present in the Eucharist to constituting scientific objects. Comparative anthropology has largely examined these powers through participants’ theories of symbols, producing an impression of ontological diversity. This article argues that such approaches begin one step too late. Within a process ontology of living coordination, symbolization is one of five mediations and operates through three universal modes: mesocosmic coordination in the present, reference to absence, and re-presentation of prior occurrences. Differences between Catholic sacrament, Hindu invocation, and social construction reflect divergent commitments about what returns, not different operations. The result is a single analytic framework that distinguishes what symbols do from what people claim they do.

I. Introduction: The Strange Powers of Words

Words are credited with remarkable powers. In the Catholic Eucharist, words spoken by an ordained priest over bread and wine, within an authorized rite, are understood to make Christ sacramentally present. In many Hindu ritual traditions, mantras and invocations take part in the manifestation of a deity, who is invited into a consecrated image and later bidden farewell. In physics, mathematical and linguistic formulations allowed black holes to be identified decades before any instrument registered one. In some theories of social construction, discursive and social practices are credited with constituting the very objects of knowledge. And in the anthropology of ritual, from Malinowski's studies of Trobriand garden magic to Tambiah's essay on the magical power of words, the question of what words can do has been a founding preoccupation.

These cases raise a deceptively simple question: what exactly happens when symbols are used? The answers differ dramatically. Some concern remembering, some referring, some making present, some coordinating people, and some the creation or transformation of reality. Anthropology has studied these differences largely by examining how particular people or traditions understand the powers of symbols, and the resulting literature on language ideologies, semiotic ideologies, performativity and ontological difference is sophisticated and wide-ranging. What it has not established is whether the apparently different powers correspond to different kinds of symbolic operation.

This article proposes that comparative anthropology has often begun its investigation one step too late. Instead of first identifying the operations through which living beings coordinate symbolically, it has concentrated on culturally articulated explanations of those operations. The result can be a misleading impression of ontological diversity, in which differences in people's theories of symbols are read as differences in the nature of symbolization itself.

The alternative proposed here rests on a shift in starting point so small that it is easily mistaken for agreement. Linguistic anthropology has long insisted that language is a form of social action rather than a set of labels, and Living Value Theory agrees. The difference lies in what that action is taken to be part of. For linguistic anthropology, language belongs together with culture and social relations in a single circle of mutual constitution. For LVT, symbolization is one of five mediations of living coordination, and it arrives late, within coordinations already organized without it. This small shift produces a radically different theory of symbolization: one in which symbols cease to be the medium of everything, in which their operations can be distinguished by how they relate the present to what is absent, and in which people's theories of symbols become objects of comparison rather than evidence of separate worlds.

LVT proposes that symbolization operates through three modes: mesocosmic coordination, reference to absence, and re-presentation. Their universality has not been demonstrated, and the claim is offered as a hypothesis, stated precisely enough that counterexamples, boundary cases and possible further modes can be identified. If it holds, a Catholic Eucharist, a Hindu invocation of divine presence, a scientific description of a black hole and an anthropologist's account of ritual language could be analysed within one process ontology, without assuming that their participants share an understanding of reality.

The argument therefore distinguishes three things that have repeatedly been run together: the operations of multisymbolization, meaning what symbols do in living coordination; participants' ontological commitments about multisymbolization, meaning what they understand symbols to be doing or capable of doing; and anthropological interpretations of those commitments, meaning what researchers conclude about reality from participants' explanations. Each deserves investigation, and none can stand in for another. The first question is whether different peoples perform different symbolic operations or understand the same operations through different ontological commitments, and only after that whether they inhabit different symbolic worlds.

II. A Small Shift in Starting Point

Laura Ahearn's Living Language, the most widely used introduction to linguistic anthropology, states the field's founding principle plainly: language is inherently social, and "the use of language is itself a form of social action" (Ahearn 2021: 8). Following Duranti, it treats language as "a form of action that both presupposes and at the same time brings about ways of being in the world", and it holds that social contexts and linguistic practices "mutually constitute each other" (2021: 8). Language, culture and social relations are "so thoroughly intertwined that they must be studied in connection with one another" (2021: 34). Against the popular view of language as an empty vehicle for pre-existing meanings, the textbook sets Jakobson's model of multifunctionality, in which the referential function is only one of six present in every speech event (2021: 22–23).

LVT shares much of this. It too rejects the picture of language as a set of labels, and it too holds that symbols are consequential in what people do. The two approaches can look so close that the difference seems a matter of emphasis. The difference lies in what the action of symbols is taken to belong to. In the textbook's framing, language is one of three terms, language, culture and social relations, that constitute one another, and nothing lies outside the circle they form. Ahearn reports that some linguistic anthropologists, following Hill and Mannheim, go further and hold that "everything is linguistically mediated", treating this as an axiom to be judged by the productive questions it opens rather than as a testable claim (2021: 96). LVT begins instead with living coordination: the continuous activity through which living beings maintain and transform their relations with their own bodies, with other beings, with found conditions and, in the human case, with made things and symbols. Symbolization is one mediation within that activity, irreducible to the others and dependent on them.

Three consequences follow from this shift, and together they amount to a different theory of symbolization. The first concerns scope. If symbolization is one mediation among five, it cannot be the medium of everything. Most living coordination on earth proceeds without symbols, and much human coordination does as well: a body digesting, a hand finding a railing in the dark, a sleeping infant settling against a parent. From this standpoint the axiom that everything is linguistically mediated becomes a claim with precise empirical content, and a false one. Symbols reorganize embodiment, being-with and dwelling profoundly, often down to perception itself, but they arise within coordinations that did not need them and remain answerable to those coordinations.

The second consequence concerns what is classified. A theory that begins with language classifies either kinds of sign, as Peirce's icon, index and symbol do, or functions of the speech event, as Jakobson's six functions do. A theory that begins with living coordination asks a different question of any symbolic act: how does it relate the coordination under way to what is present, to what is absent, or to what is being made present again? The answer yields operations rather than sign types or functions, and it brings into view an operation, re-presentation, that neither sign typologies nor functional models contain.

The third consequence concerns people's theories of symbols. If language, culture and social relations constitute one another without remainder, a community's theory of language is part of what its language is, and there is no independent ground from which to ask whether the theory is right about what its words do. Gal and Irvine accordingly describe language ideologies as "neither true nor false" but as "positioned and partial visions of the world" (Gal and Irvine 2019: 2). If symbolization is a mediation of living coordination, a theory of language is an articulation, at L3 or above, about operations that occur whether or not anyone articulates them. It can be compared with those operations, found more or less fitting, and set beside other theories of the same operations.

The order on which LVT relies is an order of dependency, and it must be distinguished from the order of experience or of individual development. Human infants are surrounded by language from birth and in some respects before it, and within their first year their perception of speech sounds narrows towards the contrasts their language uses (Werker and Tees 1984). No human being encounters a world untouched by symbols, and no human perception can be assumed to be symbolically neutral. The claim that symbolization is late concerns evolution and derivation: symbolization depends on the zoetic mediations, which occur without it, while they do not depend on it. That human life is saturated with symbols from the start shows how deeply a late mediation can reach, and it leaves the order of dependency intact.

III. Multisymbolization within a Process Ontology

In human mesocosms, LVT distinguishes five mediations that co-mediate continuously: multisensorial embodiment, multispecies being-with, multiversal dwelling, multimateriality and multisymbolization. These are analytical distinctions within ongoing processes, and none of them is a separate substance, institution or causal force. They are irreducible to one another, so that no amount of symbolization can do the work of a body and no body can do the work of a found terrain, yet in living coordination they always operate together. Multisymbolization is accordingly one of the ways in which living coordination happens, and never an autonomous domain floating above it.

The five mediations are not evolutionarily equal. Embodiment, being-with and dwelling are zoetic, belonging to living process itself, and together they constitute a complete living mesocosm without any further mediation: a plant maintaining itself through drought, a herd moving with the seasons. Multimateriality and multisymbolization are poietic, the coupled capacity to deposit coordination into forms that persist beyond the activity that produced them. Multisymbolization in LVT's technical sense, the cumulative deposition and recombination of symbols across many symbolic forms, is an evolutionarily late achievement, provisionally understood as specifically human. Animal communication, learned signs and the wider field of nonhuman semiosis pose serious questions for that exclusivity claim, and the article need not settle all of them. It is enough to establish that human symbolic operations arise within more fundamental living processes and remain answerable to them.

LVT also distinguishes five levels of recursivity, which describe what is happening within coordination at a given moment. L1 is seamless coordination, the smooth proceeding of life without explicit attention. L2 is felt misalignment, a snag that registers before anyone can say what it is. L3 is articulation, in which the snag becomes a question or a statement. L4 is stabilization, in which an articulation is fixed as a category, rule, decision or doctrine. L5 is meta-reflection on the categories and conditions of coordination themselves. The levels form no ladder of worth. L5 is no more accurate than L1, and a theoretical explanation can misrecognize a practice that its participants coordinate perfectly well.

One observation follows directly, and much of this article depends on it. A theory of language, whether religious, indigenous, philosophical or anthropological, is itself a recursively elaborated symbolic achievement, produced at L3, L4 and L5. It cannot be treated as a transparent revelation of the ontology of language merely because someone has articulated it. A ritual specialist's account of why a mantra works and a linguist's account of why a performative succeeds stand in the same relation to the practices they explain: both are symbolic operations on symbolic operations.

Universality, in this framework, has a precise meaning. The five mediations define the architecture of human living coordination; the three zoetic mediations provide the broader evolutionary baseline shared with all life; the five recursivity levels describe the forms recursive coordination can take. The three modes would similarly identify the universal repertoire of operations wherever multisymbolization occurs. The claim is that any human multisymbolic coordination should in principle be analysable through these three operations, singly or in combination, a claim that requires neither that every organism symbolize nor that all three modes be active in every symbolic event.

IV. The Three Modes

The first mode, mesocosmic coordination, is symbolization that participates directly in organizing, sustaining, modifying or repairing a living coordination in the present. Its defining feature is immediate participation in an ongoing coordination, as distinct from introducing a counterpart that remains absent or reinstating an earlier coordination. Greeting someone, giving directions, shouting a warning, negotiating who will do what, reassuring a frightened child: in each case the symbols are part of what participants are doing together now, rather than descriptions of something elsewhere. "Come here" alters the coordination between speaker and listener, provided the conditions for its uptake hold: that the listener hears, understands, is free to come and is inclined to. A greeting succeeds through bodies within sight or earshot, through a relationship it presupposes or opens, in a place where meeting is possible, often with the help of made things such as a door or a telephone, and through a symbolic form both parties recognize. Its efficacy is usually evident at L1, without anyone needing a theory of it.

One boundary must be drawn firmly. Austin, having begun by separating performative utterances from statements, ended by asking, "When we issue any utterance whatsoever, are we not 'doing something'?" (1962: 91), and linguistic anthropology has largely answered that question, in Ahearn's words, with "a resounding 'Yes!'" (2021: 190). In one sense the answer is right, since every symbolic event takes place within some coordination. If that alone qualified an event for the first mode, however, the mode would become trivially universal and would distinguish nothing. What defines the first mode is the role of the symbol in presently coordinating activity, rather than the fact that every utterance is an act. A history lecture is delivered now and coordinates a room of students, and in that respect it involves the first mode; what it says about the Peloponnesian War involves the second.

The second mode, reference to absence, is symbolization that makes something not presently encountered available for symbolic coordination, without thereby reinstating that thing or the living coordination in which it was once encountered. It names the extraordinary human capacity to coordinate with what is not here: people elsewhere, events that have ended, anticipated futures, hypothetical situations, imagined possibilities, abstract relations and fictional beings. Absence need not be temporal, since a colleague in the next room is absent from the present encounter without belonging to the past. Absence also leaves open whether the referent exists, so that the operation is the same whether one refers to a distant friend or an invented character. And reference leaves its referent unencountered. Talking about Aristotle does not bring Aristotle into being-with, talking about rainfall does not make it rain, and describing an imaginary city builds nothing.

Reference to absence is nonetheless anything but inconsequential. It makes possible planning, coordination across distance, historical continuity, anticipation, fiction and every complex form of collective organization. Its achievement lies in allowing living coordination to proceed in relation to something absent while the absent thing stays where it is. This requires a distinction between a symbolically available counterpart and an actually present one. A symbolically available counterpart can be discussed, planned for, feared or loved; an actually present one can be touched, answered, eaten or avoided. Confusing the two is a recurring source of philosophical and anthropological difficulty.

The third mode, re-presentation, is symbolization through which a prior articulation, performance, relationship, event or configuration of living coordination is made operative in a present coordination again. A historical account refers to a past event, while a ritual reenactment, a theatrical performance or a renewed recitation may seek to make something of that earlier occurrence present and effective again. The central distinction lies between describing a previous occurrence and participating in its renewed actualization. Precision about what is re-presented matters from the outset. A performance of Henry V makes Shakespeare's composition operative again in a new coordination of actors and audience, while the historical king remains absent and no one in the theatre supposes otherwise. A recording of a voice may be a present encounter with a material trace, a reference to an absent speaker, or an occasion on which the speaker's words become operative again, as when a recorded lecture still teaches or the recorded voice of someone who has died still consoles. The same physical event can support each possibility, and they must not be conflated.

Re-presentation raises questions that reference to absence does not. What exactly is made present again: a symbolic articulation, a bodily experience, a relationship, an event, an identity, or a divine presence as participants understand it? What establishes continuity between the earlier occurrence and the present one? Who is authorized to carry out the re-presentation? Does success require the original words, materials, actors or circumstances, or only some of them? And what happens when participants disagree about whether re-presentation has occurred at all?

The third mode faces a danger symmetrical to the one that threatened the first. Silverstein showed that indexical signs can be creative as well as presupposing, bringing into being the contexts they appear to point to (Silverstein 1976). Butler, following Derrida, argued that performatives work by citing earlier performances, so that every utterance repeats prior ones (Butler 1993; Derrida 1988). If every utterance makes earlier articulations operative again, every utterance is a re-presentation, and the third mode becomes as trivially universal as the first threatened to be. The answer lies in distinguishing re-presentation from sedimentation. All symbolic activity draws on sedimented forms, words, genres, grammatical patterns, ways of greeting, that earlier coordination has deposited and that later coordination takes up without returning to any particular earlier occasion. That is how symbols work at all. Re-presentation is narrower. In it, a particular prior occurrence, this composition, this event, this relationship, is made operative again, and its renewed presence is constitutive of the act and attributed to it by those who take part. A speaker who uses a conventional greeting takes up a sediment; mourners singing the hymn a dead friend chose re-present something of a particular earlier coordination. Creative indexicality, likewise, belongs to the first mode: the formality brought into being by a formal register is a reorganization of the coordination under way, and nothing prior returns in it.

A funeral shows how ordinarily the three modes combine. The officiant's instructions, to stand, to sit, to follow the coffin, coordinate the mourners in the present. The eulogy refers to a life that has ended and to a person now absent. And when the mourners sing the hymn the dead person chose, or recite words they once recited together, something of an earlier coordination may be understood to become operative again among them, without anyone supposing that the dead person has returned. Three operations unfold within an hour, often within a single sentence, and the mourners need no theory of any of them.

The modes must therefore be distinguished from existing classifications. Peirce's icon, index and symbol classify relations between signs and their objects: resemblance, physical connection, convention. Jakobson's six functions classify orientations of a speech event towards its constitutive factors: speaker, addressee, context, contact, code and message. The conative and phatic functions resemble the first mode, and the referential function overlaps with the second, but the functions are aspects present in every speech event, while the modes are operations defined by how a symbolic act relates present coordination to presence and absence. No Peircean sign type and no Jakobsonian function corresponds to re-presentation. A single word, image or gesture can participate in any of the three modes: the words "this is my body" refer to an absent event when read in a history of the Last Supper and, for Catholics, take part in a re-presentation when spoken at the consecration. The unit of analysis must therefore be the symbolic operation, rather than a whole language, text, ceremony, culture or tradition.

What holds the three modes together is a single analytical question: how does a symbolic operation relate present living coordination to something presently encountered, to something absent, or to something being made operative again? The success of the hypothesis depends on keeping to this common criterion. It cannot be shown merely by assigning every imaginable case to whichever of the three categories seems nearest.

V. Re-presentation across the Five Mediations

Umberto Eco's The Name of the Rose offers a case of re-presentation unusually free of theological commitments. The novel turns on the survival, in a fourteenth-century abbey library, of the lost second book of Aristotle's Poetics, the treatise on comedy. Aristotle is long dead and his living coordinations cannot be restored, yet the manuscript retains traces of his symbolic articulations, and these can enter the coordinations of readers seventeen centuries later. The blind librarian Jorge of Burgos poisons the pages and in the end burns the library because he fears exactly this possibility: a book that taught the dignity of laughter would, he believes, free people from fear, and with fear from the order that fear sustains. His terror is the terror of re-presentation, that an earlier symbolic argument might become newly operative and transform what later readers think and do. What would be re-presented is an articulation, and the case shows that an articulation can be made operative again while nothing else from its original coordination returns with it.

This suggests a hypothesis: the difficulty of re-presentation depends partly on what is supposed to be re-presented, and each mediation sets different conditions. An earlier symbolic articulation can survive through writing, recitation and recording, and it is the easiest to make operative again, because symbols are made to be taken up repeatedly. An earlier multimaterial configuration can be reconstructed, as a building is restored or an instrument rebuilt, although materials and their relations can never be assumed identical. An embodied event can be reenacted, but the original bodily experience is not thereby recovered; the battle reenactor does not bleed. A prior relationship of being-with cannot be restored by reproducing its descriptions or its setting, because the counterparts have changed and now answer differently. An earlier configuration of dwelling may be revisited, but the found conditions and the history of the living being who returns have both moved on.

Historically informed performance of early music makes the point vivid. Musicians play Bach on gut strings and at period pitch in an attempt to re-present a multimaterial and sonic configuration, and the attempt has been fiercely debated precisely because no one can re-present the listeners, whose ears have been trained by two further centuries of music. The instruments can be rebuilt and the score is intact, while the being-with of the original performance, a congregation in Leipzig on a particular Sunday, is beyond recovery. This might be called the mediational difficulty of re-presentation. It forms no fixed ranking in which one mediation is always easier to re-present than another, since each effort meets its own conditions of continuity and discontinuity, but a practice that seeks to re-present a relationship or a presence takes on a far heavier burden than one that seeks to re-present an argument.

A further distinction is indispensable. Every act of re-presentation involves all the mediations active in the coordination in which it occurs, while the mediation being re-presented need not be the mediation through which re-presentation is attempted. Aristotle's argument would be re-presented through parchment and through the embodiment of readers bending over it in a cold scriptorium, but what is re-presented is multisymbolic. The Eucharist, as Catholics understand it, works through words, bread, wine, vessels, an ordained priest and a gathered assembly, and what it is understood to make present is Christ himself and the sacrifice of the cross. Re-presenting Aristotle's argument and re-presenting Christ's bodily presence may both involve texts, institutions, materials and ritualized action, but their claimed achievements are ontologically very different. The difference lies in what is claimed to return, and that is where religious traditions part company.

VI. Religious Difference and Prior Ontological Commitments

The Catholic Eucharist poses the problem of re-presentation in its most demanding form. According to Catholic doctrine, Christ's sacrifice on the cross was a historically singular event, offered "once for all". The Eucharist does more than commemorate it by referring to its absence. The Catechism of the Catholic Church states that the Eucharist is a sacrifice "because it re-presents (makes present) the sacrifice of the cross" (§1366), insists that the sacrifice of the Mass and that of the cross are "one single sacrifice", differing only in the manner of offering (§1367), and holds that Christ's presence under the appearances of bread and wine "endures as long as the Eucharistic species subsist" (§1377). The theological problem is how something historically singular can be made present again without being repeated as a new historical event. The question of authorization sharpens it. Catholic theology locates the efficacy of the sacrament in Christ, "the principal agent of the Eucharist" (§1348), with the priest acting in his person within an authorized sacramental order, so that the words of institution, printed in every missal, consecrate nothing when read in a book.

Rather than evidence that Catholicism possesses a different fundamental mode of symbolization, the complexity of Eucharistic theology arises from the particular ontological commitments that its account of re-presentation must preserve and reconcile: the uniqueness of the incarnation, the singularity of the crucifixion, the distinction between bread and body worked out in the doctrine of transubstantiation, and the distinction between priestly activity and divine agency. A tradition that held the crucifixion to be repeatable, or the priest to be the source of the sacrament's power, would face a far simpler problem. The elaborate distinctions are the price of holding all these commitments at once, and they are commitments about the referent, the agent and the material, rather than about what symbols as such can do.

Hindu traditions offer a revealing contrast, provided they are not treated as a single ontology. Each autumn in Bengal, clay images of Durga are made by specialist artisans and consecrated in a rite of prāṇa pratiṣṭhā, the establishing of life-breath, through which the goddess is invoked into the image. She is worshipped over several days and then bidden farewell before the image is carried to the river and immersed. Bengali devotional song frames the festival as Uma's annual return to her parents' home from Shiva's mountain, with songs of welcome, āgamanī, and songs of parting, vijayā. Something is understood to be made present, and the language of return is explicit, yet the theological burden differs sharply from that of the Eucharist. The goddess is understood as enduring, and her presence in the image as a renewed manifestation of a reality that never ceased, rather than the making-present of a singular event that happened once and is closed. The question is how an enduring reality becomes manifest, accessible and efficacious here and now. The return is also cyclical and expected, scheduled in the calendar like the season it accompanies.

The sacred syllable Om sharpens the comparison. The Māṇḍūkya Upaniṣad opens by declaring that Om is this whole world, that past, present and future are Om, and that whatever transcends the three times is also Om. In traditions shaped by such texts, and by the Mīmāṃsā doctrine that the Veda is eternal and authorless, the renewed utterance of sacred sound raises no problem of reproducing a singular event from a distant past, since what is uttered was never absent in time. Practices involving comparable operations can therefore carry radically different theological burdens. Hinduism possesses no homogeneous ontology of continuity, however, and Christianity no uniform commitment to separation: Hindu traditions recount singular divine descents, avatāras born at particular moments of particular ages, while Christian mystical traditions speak of a God continuously present. And Om may become one of the most important tests of the three-mode hypothesis, since an utterance that makes manifest something understood as eternally present may be re-presentation, mesocosmic coordination, or an operation that challenges the distinction.

Genesis poses a different challenge. "And God said, Let there be light: and there was light" (Genesis 1:3). At first sight this appears to show a mode the three do not cover, originary creation through speech. It is essential, however, to distinguish the divine speech represented in the narrative from the human symbolic operations through which the narrative is told, read, interpreted and transmitted. The narrative describes God's creative action and offers no independently observed instance of human multisymbolization creating light. Theological arguments about the Logos and the divine ideas add enormously consequential commitments about divine intelligibility and creation, and they do not by themselves demonstrate a fourth human symbolic operation. The case should nonetheless be kept as a potential challenge, since human beings also claim, in more modest ways, to bring things into being by speaking.

The most important comparative proposition of the article can now be stated. Apparently radical differences in the powers attributed to symbols may arise from different ontological commitments about their referents, rather than from differences in the fundamental operations of multisymbolization. Religious traditions differ in their commitments about divine agency, temporality, material identity, historical singularity, spiritual presence and the conditions under which an act is effective, and those commitments shape which symbolic practices become intelligible, problematic, obligatory, forbidden or contested. The commitments are prior to particular symbolic operations in the sense that participants inherit them, and in no absolute sense prior to symbolization itself. They are recursively produced, transmitted, modified and sometimes rejected through living coordination, largely at L4, and remain open to L2 snags and L3 questioning. The comparative task is to distinguish the recurrent symbolic operation from the historically particular account through which its efficacy is understood, which offers a firmer basis for comparison than either assuming all religious claims to be equivalent or treating each as evidence of an incommensurable ontology.

VII. The Anthropology of Other People's Theories of Symbols

Linguistic anthropology moved, over the last decades of the twentieth century, from studying linguistic structures and practices towards studying speakers' understandings of what language is, how it works and what it can do. Michael Silverstein's account of linguistic ideology opened the field; Kathryn Woolard and Bambi Schieffelin surveyed its reach; Asif Agha traced how registers become recognizable through reflexive processes of enregisterment. Webb Keane's concept of semiotic ideology broadened the inquiry to people's assumptions about what signs are, what functions they serve and what consequences they can produce. Introductory teaching now presents language ideologies as one of the discipline's four key concepts, alongside multifunctionality, practice and indexicality (Ahearn 2021: 21–33).

The achievement of this literature is substantial. It shows that symbolic practices are governed by more than linguistic rules, and that participants' assumptions about signs help organize their actual coordination: a Protestant suspicion of formulaic prayer changes how people pray, and a purist ideology of a national language changes what is taught in schools. Keane, moreover, distinguishes general semiotic processes from historically particular semiotic ideologies, and his work is an important interlocutor rather than an example of the confusion this article criticizes. The difficulty lies elsewhere, in a step the field's framing makes easy. If language ideologies are "neither true nor false", and if, as Ahearn adds, scholars' own views about language are "also, of course, language ideologies" (2021: 24), then every account of what symbols do, including the analyst's, becomes one positioned vision among others. Reflexivity is preserved, but evaluation is abandoned.

The objection can be turned against LVT. Its distinction between what symbols do and what people say they do is itself an articulation at L5, produced within the symbolic class it analyses, and it enjoys no privileged access to operations. LVT accepts the first half of this. It does not accept the conclusion, because its articulations remain answerable in a way that the ideology framework declines to be. Operations can be identified through observable differences in coordination, before any theory of them is adopted: what changes when the same words are written rather than spoken in an authorized rite; who can perform an act and who cannot; what follows when it fails; which bodies, things, places and times it requires. The Catholic missal shows the method in miniature, since the same words do nothing on the page and, on Catholic teaching, consecrate when enacted by a priest over bread and wine. Any theory of symbolization, LVT's included, has to accommodate such differences, and a theory that cannot is less fitting than one that can. That is a standard of mesocosmic fit, and it applies to the analyst as much as to the analysed.

Consider what an ethnography of somebody else's theory of language involves. An anthropologist observes a ritual. Participants attribute a particular power to the words used, and afterwards an experienced ritual specialist explains why those words are effective. At one level there is the ritual coordination itself, involving embodiment, being-with, dwelling, multimateriality and multisymbolization: people sitting, chanting and handling objects, in a place, at a time of day and year, with words. At a second level there are the participants' articulated explanations. At a third there is the anthropologist's description and interpretation of those explanations. The recursivity levels distinguish these. At L1, participants may carry out the ritual fluently without explicit reflection. At L2, something may snag: a mispronounced syllable, an unexpected event, a disagreement about what should happen next. At L3, participants articulate an explanation, often prompted by the snag or by the anthropologist's question. At L4, an account is stabilized as correct or authoritative by a lineage of specialists, a text or an institution. At L5, participants may question the nature of ritual language itself, as Mīmāṃsā philosophers did for the Vedic injunction and scholastic theologians for the sacramental formula. The anthropologist then produces further articulations, which may travel back into the field when a monograph is read by the community it describes.

An anthropological account of another person's L5 theory of symbolization is therefore itself an L3, L4 or L5 operation of multisymbolization. There is nothing wrong with this, since it is what scholarship involves. The difficulty begins when the documented theory is treated as evidence that the underlying ontology of symbolization must be different. Three levels of ontological claim have to be kept apart: the ontology of living coordination, what processes and mediations are actually occurring; participants' ontological commitments, what they understand to exist and what powers they attribute to symbols, persons, materials or divine beings; and the anthropologist's ontological interpretation, what the researcher infers about reality from those commitments. An ethnography can establish that people attribute agency to particular ritual words and can investigate, often with great precision, how that attribution organizes relations and practices. It cannot establish from the attribution alone that the words possess the agency attributed to them, nor that the participants inhabit a different process ontology.

The argument extends to approaches that infer ontological difference from grammar or classification. Whorf argued that Hopi grammar embodied a metaphysics without the flowing, spatialized time of European languages, and Malotki's later documentation of Hopi temporal expressions showed how much of that inference rested on the analyst's reading of grammar rather than on Hopi coordination. Levinson's work on spatial frames of reference presents a stronger case. Speakers of Guugu Yimithirr and Tzeltal, whose languages describe location in cardinal rather than relative terms, rebuild an arrangement of objects after rotation according to those fixed coordinates, and keep track of cardinal directions with remarkable accuracy (Levinson 2003). Here differences in linguistic classification do correlate with measurable differences in nonlinguistic coordination, and specifically coordination with dwelling. LVT has every reason to accept this, since a mediation that reorganizes the others can be expected to do so, sometimes down to L1. What the evidence does not show is a different mode of symbolization or a different architecture of coordination. Guugu Yimithirr speakers coordinate with the same found terrain, sun and wind through the same five mediations; their language has made one aspect of dwelling, cardinal orientation, continuously relevant, much as navigation makes it relevant to a sailor. The line falls between the reorganization of coordination, which symbolic classification can certainly produce, and a difference in the architecture of coordination, which it cannot.

Anthropology's own categories are not exempt. "Culture", "society", "ontology", "language", "representation" and "construction" are symbolic stabilizations through which anthropologists try to understand living processes, and they can be useful while concealing what they claim to explain. An L5 commitment to reflexivity is only half met by recognizing that other people's ontologies differ from one's own. It also requires investigating how the anthropologist's categories are constructed, where their applicability ends, and whether they obscure the coordinations under study. What results is a clearer specification of what the ethnography of language ideologies can and cannot establish, rather than its abandonment.

VIII. Social Construction and the Limits of Symbolic Efficacy

Academic theory has its own theories of magical words. An anthropologist studying rainmaking may reasonably ask how ritual words could alter atmospheric processes. Yet some expansive formulations of social construction attribute reality-constituting powers to discourse without specifying how the constitution occurs. What exactly does it mean to say that a black hole is socially constructed? If the claim concerns scientific categories, mathematical representations, the instruments of the Event Horizon Telescope that produced the first image of one in 2019, or the processes through which astronomers came to accept that black holes exist, identifiable historical processes can be studied, and the study is valuable. If the claim concerns the physical existence of black holes, it requires a very different explanation, and none has been offered. The gravitational waves detected in 2015 from two merging black holes more than a billion light years away had been travelling since long before anyone on Earth could construct anything.

Ian Hacking's insistence on asking "the social construction of what?" anticipates this argument. Hacking distinguished the construction of ideas from the construction of the objects those ideas classify, and he distinguished interactive kinds, such as diagnostic categories that the people classified learn about and respond to, from indifferent kinds such as quarks, which are unaffected by being named. In LVT's terms the distinction tracks the type of recursive counterpart: classifications applied to interrecursive counterparts loop back upon them, while classifications applied to nonrecursive counterparts do not. A person diagnosed with depression may come to understand their past, their sleep and their relationships through the diagnosis and change in ways the category then registers. A black hole does nothing of the kind.

The decisive question is never whether symbols have effects, since they plainly do. Promises, contracts, theories, diagnoses, declarations and rituals transform living coordination on a vast scale, and some declarations help to constitute institutional facts in Searle's sense: an authorized pronouncement creates a marriage, a verdict a convict, a count of votes a government. Those effects depend, however, on the conditions under which the symbolic operation takes place, and a declaration cannot by itself produce every state of affairs its words describe. A registrar can pronounce a couple married, while no pronouncement can make it rain or make a bridge bear a load it was not built for. LVT locates symbolic efficacy in the co-mediation of all five mediations, rather than attributing an unlimited world-creating power to symbols themselves. A methodological principle follows, applying equally to religious explanations and academic theories: claims about symbolic efficacy must specify what is transformed, through which mediations, under what conditions, and on what evidence. The existence of a symbolic theory of reality does not establish the reality of the symbolic powers that the theory attributes to itself.

IX. Why the Question Has Rarely Been Asked

If the shift in starting point is small, it is fair to ask why it has rarely been made. Each of the three modes has been noticed, often brilliantly. Coordination was seen by Malinowski, who described language in use as "a mode of action" and named "phatic communion" (Malinowski 1923), by Austin, and by Maturana and Varela, who described language as the coordination of coordinations of behaviour (Maturana and Varela 1987). Reference to absence has been the central concern of the philosophy of language since Frege, and Hockett listed displacement among the design features of human language (Hockett 1960). Re-presentation has been the concern of theology, hermeneutics and ritual studies, and Gadamer's account of the festival, which exists only in being celebrated and is contemporaneous with each celebration, comes close to it (Gadamer 2004). Bühler's distinction between pointing to what is perceptually present and pointing within the imagination to what is absent, his Deixis am Phantasma, comes closer than most to separating the first two modes within one theory (Bühler 1990). What has been missing is the combination: the three treated as modes of a single mediation, defined by a single criterion, within a theory in which symbolization depends on four other mediations.

Three reasons for the absence suggest themselves. The first is disciplinary. Each mode has been the property of a different discipline, and each discipline has taken its own mode to be what symbols basically do. Reference-first approaches turn coordination into "pragmatics" and re-presentation into "belief"; coordination-first approaches turn every utterance into an act, and so lose the distinction between acts. No single field contains all three, and so none has had reason to define them against one another.

The second reason is the starting point. Asking how symbols appear within living coordination requires beginning from living coordination, and no discipline is organized around it. Semiotics begins with the sign, linguistics with the language system, anthropology with culture or the human. A theory that begins with the sign produces classifications of signs; a theory that begins with the speech event produces functions; neither produces operations defined by their relation to presence and absence. Classifications of signs are also far easier to apply to what scholars actually handle, which is texts and transcripts.

The third reason may run deepest. Scholarship is the most institutionalized practice of reference to absence there is: written deposits about absent things, produced for absent readers. Bourdieu described the projection of the scholar's own relation to knowledge onto the practices studied as the scholastic fallacy (Bourdieu 2000), and the habit of treating reference as what symbols fundamentally do may be one of its forms. A further, historical hypothesis deserves investigation. Modern theories of the sign took shape in an intellectual world formed by the Reformation, in which "this is my body" had for many become "this signifies my body", and Keane has shown how deeply a Protestant semiotic ideology shaped the moral narrative of modernity (Keane 2007). If the arbitrary sign generalizes, in part, a Reformed theological position, then strong re-presentation could re-enter modern theory only as other people's belief, magic or ideology, something to be explained rather than an operation to be described. That would help to explain why re-presentation appears in linguistic anthropology almost exclusively under the heading of language ideologies.

X. Counterexamples, Falsification, and Comparative Research

Without an explicit search for counterexamples, the proposed universality of the three modes could become another L4 stabilization presenting itself as an ontological discovery. The first danger is conceptual. The modes may seem to be defined at different analytical levels: mesocosmic coordination by its immediate efficacy, reference to absence by the availability of its referent, re-presentation by renewed actualization. The criterion proposed here, the relation a symbolic operation establishes between present coordination and something encountered, absent or made operative again, must define each mode alone, and both the first and the third must be protected from trivial universality, the first by excluding mere presentness, the third by excluding mere sedimentation. The hypothesis succeeds only if its categories reveal differences that an undifferentiated account of symbolic activity would miss.

Five candidate counterexamples deserve serious attention. The first is originary symbolic creation. When someone invents a fictional character, defines a new mathematical object or draws a distinction no one has drawn before, are they referring to an absence, creating a new symbolic counterpart within present coordination, or doing something else? The referent need not have existed in any form before the act, and the case connects with Genesis: if human beings can call symbolic counterparts into being, the assumption that only divine speech is originary needs revisiting.

The second is the declaration and the promise. A promise changes a present relationship and projects an obligation into an absent future. It may be a combination of the first two modes, or the creation of an obligation may require a further operation that brings into existence a counterpart, the commitment, that did not exist before the words were spoken. If so, the promise and the fictional character may turn out to be two faces of a single unrecognized mode.

The third is manifestation without historical return. The invocation of an eternally present divinity, the utterance of Om and the annual return of Durga all challenge the temporal definition of re-presentation. If re-presentation requires prior absence, these practices may belong to a different operation; if it does not, the definition must say what "again" means.

The fourth is self-referential and non-referential symbolization. Music, mathematical reasoning, abstract diagrams and symbolic play may not fit a framework that tacitly assumes symbols to be primarily about identifiable referents. A fugue refers to nothing absent, re-presents no earlier event, and coordinates its listeners in ways that seem far richer than the first mode, as so far defined, allows.

The fifth concerns scope. Animal communication, symbolic capacities in other species and AI-generated symbolic activity all challenge the proposed domain of multisymbolization. AI systems in particular produce symbols at enormous scale without, on LVT's account, being living recursive counterparts, which raises the question whether symbolic outputs without a living producer operate in any mode until a living being takes them up.

One frequently cited case proves weaker than it first appears. Everett's account of Pirahã, a language of the Brazilian Amazon, holds that its speakers restrict talk to matters within the immediate experience of the speaker or of someone living who witnessed them (Everett 2005). If that were so, Pirahã would be a test of whether reference to absence is universal. The claims have been heavily contested (Nevins, Pesetsky and Rodrigues 2009), but even on Everett's own account Pirahã speakers refer to absent people and to past events witnessed by others. What varies is the reach of reference to absence, how far into the absent it is permitted to extend, rather than the presence of the mode. That is what one would expect if each mode, like each mediation, has a characteristic bandwidth that communities stabilize differently.

A comparative research protocol follows. Studies should first establish what participants are doing, through observable differences in coordination, before turning to their explanations. They should then identify the symbolic operations involved and the forms of presence, absence or renewed actualization at stake; distinguish participants' theories of efficacy from the observable consequences of the event; identify which mediations are involved, which counterpart is supposed to be transformed and which recursivity levels are operative; and ask finally whether anything remains unexplained by the three modes. A counterexample proper would identify a distinctive symbolic operation that cannot be explained through the three modes or their combinations without stretching their definitions until they lose explanatory value. The appropriate response would then be to revise the definitions, add a mode, or abandon the claim of exhaustiveness. The hypothesis becomes scientifically valuable when it permits cases that could show its inadequacy, rather than when it survives every possible reinterpretation.

XI. Conclusion: A Small Shift and a Different Theory

LVT and linguistic anthropology agree that language is a form of action. They part company over what the action belongs to. Placed in a circle of mutual constitution with culture and social relations, language becomes the medium of everything, its functions are aspects of every speech event, and people's theories of it are positioned visions that no one can assess. Placed within living coordination, as one of five mediations that arrived late and remains dependent on the others, symbolization becomes something with limits, its operations can be distinguished by how they relate the present to what is absent, and people's theories of it become articulations about those operations, open to comparison with what the operations observably do.

From that small shift a different theory of symbolization follows. Symbols coordinate what is happening now, make what is absent available without making it present, and, in some practices, are understood to make a prior occurrence operative again. What one tradition calls divine manifestation, another sacramental presence and another social construction may involve recognizably similar operations, understood through different commitments about reality. This takes nothing away from cultural difference and makes its investigation more precise, since the comparative task becomes explaining how different commitments concerning the existence, identity and powers of symbolic counterparts shape what people can do, expect, fear and regard as possible.

The anthropologist, equally, must recognize that describing another person's ontology of language is a further symbolic operation, rather than an immediate encounter with a separate fundamental ontology. An epistemic discipline follows. Symbolic operations should be assumed neither universally identical nor, on the strength of the diversity of theories about them, fundamentally different. They should be investigated directly, their consequences compared, and cases actively sought that the three modes cannot explain. That is the promise of a universal process ontology of multisymbolization, and it is also the standard by which it must be judged.

Selected References

Agha, Asif. 2007. Language and Social Relations. Cambridge: Cambridge University Press.

Ahearn, Laura M. 2021. Living Language: An Introduction to Linguistic Anthropology. Third edition. Hoboken, NJ: Wiley-Blackwell.

Austin, J. L. 1962. How to Do Things with Words. Oxford: Clarendon Press.

Bourdieu, Pierre. 2000. Pascalian Meditations. Translated by Richard Nice. Cambridge: Polity Press.

Bühler, Karl. 1990. Theory of Language: The Representational Function of Language. Translated by Donald Fraser Goodwin. Amsterdam: John Benjamins.

Butler, Judith. 1993. Bodies That Matter: On the Discursive Limits of "Sex". New York: Routledge.

Catechism of the Catholic Church. 1997. Second edition. Vatican City: Libreria Editrice Vaticana. Paragraphs 1348, 1366–1367, 1377.

Derrida, Jacques. 1988. Limited Inc. Evanston, IL: Northwestern University Press.

Duranti, Alessandro. 1997. Linguistic Anthropology. Cambridge: Cambridge University Press.

Eco, Umberto. 1983. The Name of the Rose. Translated by William Weaver. London: Secker and Warburg.

Everett, Daniel L. 2005. "Cultural Constraints on Grammar and Cognition in Pirahã." Current Anthropology 46 (4): 621–646.

Gadamer, Hans-Georg. 2004. Truth and Method. Second revised edition. Translated by Joel Weinsheimer and Donald G. Marshall. London: Continuum.

Gal, Susan, and Judith T. Irvine. 2019. Signs of Difference: Language and Ideology in Social Life. Cambridge: Cambridge University Press.

Hacking, Ian. 1999. The Social Construction of What? Cambridge, MA: Harvard University Press.

Hill, Jane H., and Bruce Mannheim. 1992. "Language and World View." Annual Review of Anthropology 21: 381–406.

Hockett, Charles F. 1960. "The Origin of Speech." Scientific American 203 (3): 88–96.

Jakobson, Roman. 1960. "Closing Statement: Linguistics and Poetics." In Style in Language, edited by Thomas A. Sebeok, 350–377. Cambridge, MA: MIT Press.

Keane, Webb. 2007. Christian Moderns: Freedom and Fetish in the Mission Encounter. Berkeley: University of California Press.

Keane, Webb. 2018. "On Semiotic Ideology." Signs and Society 6 (1): 64–87.

Levinson, Stephen C. 2003. Space in Language and Cognition: Explorations in Cognitive Diversity. Cambridge: Cambridge University Press.

Malinowski, Bronisław. 1923. "The Problem of Meaning in Primitive Languages." Supplement to C. K. Ogden and I. A. Richards, The Meaning of Meaning, 296–336. London: Kegan Paul, Trench, Trubner.

Malinowski, Bronisław. 1935. Coral Gardens and Their Magic. Volume II: The Language of Magic and Gardening. London: George Allen and Unwin.

Malotki, Ekkehart. 1983. Hopi Time: A Linguistic Analysis of the Temporal Concepts in the Hopi Language. Berlin: Mouton.

Maturana, Humberto R., and Francisco J. Varela. 1987. The Tree of Knowledge: The Biological Roots of Human Understanding. Boston: Shambhala.

Nevins, Andrew, David Pesetsky, and Cilene Rodrigues. 2009. "Pirahã Exceptionality: A Reassessment." Language 85 (2): 355–404.

Olivelle, Patrick, trans. 1998. The Early Upaniṣads: Annotated Text and Translation. New York: Oxford University Press.

Peirce, Charles Sanders. 1931–1958. Collected Papers of Charles Sanders Peirce. Edited by Charles Hartshorne, Paul Weiss and Arthur W. Burks. Cambridge, MA: Harvard University Press.

Searle, John R. 1995. The Construction of Social Reality. New York: Free Press.

Silverstein, Michael. 1976. "Shifters, Linguistic Categories, and Cultural Description." In Meaning in Anthropology, edited by Keith H. Basso and Henry A. Selby, 11–55. Albuquerque: University of New Mexico Press.

Silverstein, Michael. 1979. "Language Structure and Linguistic Ideology." In The Elements: A Parasession on Linguistic Units and Levels, edited by Paul R. Clyne, William F. Hanks and Carol L. Hofbauer, 193–247. Chicago: Chicago Linguistic Society.

Tambiah, S. J. 1968. "The Magical Power of Words." Man 3 (2): 175–208.

Werker, Janet F., and Richard C. Tees. 1984. "Cross-Language Speech Perception: Evidence for Perceptual Reorganization during the First Year of Life." Infant Behavior and Development 7 (1): 49–63.

Whorf, Benjamin Lee. 1956. Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf. Edited by John B. Carroll. Cambridge, MA: MIT Press.

Woolard, Kathryn A., and Bambi B. Schieffelin. 1994. "Language Ideology." Annual Review of Anthropology 23: 55–82.