Wir können uns nicht in sie finden
"If a lion could speak, we could not understand him" (Wenn ein Löwe sprechen könnte, wir könnten ihn nicht verstehen). Wittgenstein writes this sentence in the second part of the Philosophical Investigations, the section long printed as Part II, xi and now published as Philosophy of Psychology: A Fragment. It appears within a sprawling, fragmented reflection on understanding other forms of life, shortly after a passage in which he observes that one human being can be a complete enigma to another. We discover this, he writes, when we come into a strange country with entirely strange traditions, and even when we have mastered the country's language. We may understand what people say, and even what they say to themselves, and still not understand them: Wir können uns nicht in sie finden. Anscombe rendered the phrase as "we cannot find our feet with them"; literally, we cannot find ourselves into them.
In two short fragments Wittgenstein leaves one of philosophy's most consequential distinctions, between two modes of understanding. The first concerns another's symbolic expression: their words, their grammar, the regularities of their signs. The second concerns what might be called getting the other: grasping what their activity is bound up with, what their words are doing within the life in which they are spoken, what would be lost if that life came apart. Living Value Theory can state the distinction as one between symbolic decipherment and mesocosmic understanding. Mesocosmic understanding is whatever access one being can achieve, across multisensorial embodiment, being-with, multiversal dwelling, multimateriality, and multisymbolization, to the coordination through which another being keeps its world livable. Stakes arise within that entire mediational field. They form no hidden layer underneath symbols, waiting to be reached once the symbols have been cracked; symbols are one mediation among the five through which stakes are lived, articulated, and sometimes obscured.
Xenolinguistics, the study of how humans might communicate with extraterrestrial intelligence, has organised itself almost entirely around the first mode. Its central problems are detection, structure, and decoding: how to recognise an artificial signal, how to identify its grammar, how to map its symbols onto ours. The title of this article makes a claim that must be stated precisely. If success means detecting signals or describing their formal organisation, xenolinguistics may well succeed. If success means mesocosmic understanding, then decoding cannot by itself produce it. The argument does not hold that understanding across radically different forms of life is metaphysically impossible. Two beings who begin very far apart could, through sustained and repeated interrecursive contact, gradually establish shared coordination where none existed before. What the argument rules out is the idea that the formal decoding of a detached signal is sufficient.
The reason can be stated in advance. A detached symbolic trace radically underdetermines the multimediated, recursively organised mesocosm that made it consequential in the first place. This is no exotic condition created by extraterrestrials, since every ordinary human utterance already has exactly this structure. Xenolinguistics merely removes the shared multimediated background that normally prevents us from noticing it.
I. The fantasy and its ground
There is something quietly comic about the Voyager Golden Record. Launched in 1977, it carries greetings in fifty-five languages, ninety minutes of music ranging from Bach to Chuck Berry, a montage of earthly sounds that includes wind, surf, a heartbeat, laughter, and a kiss, and 115 encoded images: mathematical and physical definitions, the solar system, diagrams of DNA and human anatomy, a sequence from conception to birth, a photograph of a birth, a nursing mother, people eating and drinking, a woman in a supermarket (Sagan et al. 1978). It was assembled by a committee chaired by Carl Sagan and sent into interstellar space as humanity's calling card to any intelligence that might one day intercept it. The committee made deliberate exclusions. War was left out so that the record would not seem aggressive; politics and religion were avoided; a nude photograph of a man and a pregnant woman was vetoed by NASA and replaced with silhouettes.
Far from neglecting biology, the record's makers took evident care to include eating, birth, and nursing. Yet what they could include were pictures of eating, and what could not travel was hunger. A photograph of a woman eating a grape is a symbolic deposit whose intelligibility, for any human viewer, rests on everything the viewer already brings to it: the experience of appetite, of taste and satiety, the sociality of meals, the supermarket as a material arrangement through which food becomes available by exchange. For a human viewer the image suffices because that ground is shared. For a recipient without it, the image is a pattern of light and dark in which the stakes of eating are nowhere to be found. Pain, illness, ageing, and death are absent altogether, and war was removed by design. The record is a curated self-portrait of the species, composed largely by and for those whose working lives are organised around symbolic articulation, and it represents humanity as such a group would wish it to be seen.
This is a structural diagnosis rather than a criticism of the individuals involved. The Voyager record is what happens when thoughtful people attempt to represent their species to an imagined alien audience without a theory of how symbols come to work at all. Absent such a theory, they select what can be detached (images, sounds, numbers, music) and treat detachability as though it were transmissibility. Everything that made each item sufficient for a human recipient remains behind on Earth, in bodies, relationships, places, objects, and histories, unremarked because it is so pervasive that it has ceased to appear as anything in particular.
The Voyager record is xenolinguistics in miniature. It encapsulates the field's core assumption: that communication between intelligent species is primarily a problem of transmitting symbolic content across the gap between different coding systems. Get the code right, build the right bridge between symbol systems, and understanding will follow. This article argues that the assumption is structurally inverted. Symbolic content is the thinnest part of what makes communication work, and the part that survives detachment is precisely the part that says least about what was at stake.
II. Transmission and interrecursivity
The dominant picture of communication in cognitive science, artificial intelligence, and popular imagination treats language as a code through which already-formed parties transmit information to one another. A speaker encodes a message, a channel carries it, a listener decodes it, and communication succeeds when the decoded message matches the encoded one. This is the linguistic form of what Living Value Theory calls transactive dualism: the assumption that social life is made up of exchanges between parties who exist, complete, before the exchange begins.
Critics of this picture often attribute it to "formal linguistics" in general, as though Saussure, Chomsky, information theory, and transformer architectures were all variations on one mistake. They are doing quite different things. Saussure analysed the sign as a relation within a system of differences; Chomsky modelled the generative competence of an idealised speaker; transformer models learn statistical dependencies across vast corpora of text. The precise source of the transmission picture is Claude Shannon's mathematical theory of communication (Shannon 1948). Shannon modelled an information source, a transmitter, a channel subject to noise, a receiver, and a destination, and he stated explicitly that the semantic aspects of communication were irrelevant to the engineering problem he was solving. The model was superb. It gave telecommunications and computing a rigorous account of how signals can be moved reliably through noisy channels, and it did so by treating sender and receiver as endpoints whose own activity did not enter into the message.
The problem lies in what happened to the model afterwards: a superb model of largely nonrecursive machine transmission became inflated into an ontology of living communication. Warren Weaver's introduction to the 1949 book edition already suggested that the mathematical theory might illuminate questions of meaning and of effect on conduct as well as questions of transmission (Weaver 1949), and from there the sender-channel-receiver architecture migrated into linguistics, psychology, and social theory as an implicit account of what communication is. Xenolinguistics inherits this ontology directly. If communication is transmission, an alien language is simply a harder code, requiring more ingenuity, more computing power, and more time.
What the transmission ontology cannot accommodate is interrecursivity. In actual conversation, the supposed receiver enters recursively into the production of what the supposed sender goes on to say. I monitor whether you understand me; you monitor me monitoring you; you interrupt, and I reformulate; you hesitate, and I slow down; you laugh too early, and I realise that you have seen where I am going. Human communication consists of ongoing reciprocal modification rather than the passage of a finished informational package between fixed endpoints. Interrecursivity names this relation between living counterparts in which each one's responses alter the other's subsequent activity, so that the course of the exchange is jointly generated rather than determined by either party alone. Mathematical models can represent patterns in such exchanges, and some do so with great success, but the representation of an interrecursive process should not be confused with the living process it represents.
The same inflation governs a concept central to information-theoretic linguistics: redundancy. Researchers ask how much redundancy a language contains and whether repeated grammatical marking or predictable words make speech easier to learn or process. Redundancy, however, can be measured only after an analyst has stipulated what counts as the message and what counts as its successful transmission. Relative to propositional content, please, please be careful says no more than be careful. Relative to the living relation between a frightened speaker and someone leaving on a dangerous journey, the doubled please is the most important part of the utterance. Repetition, pausing, reformulation, and apparently superfluous markers are often doing interrecursive work, keeping two counterparts in step with each other. More fundamentally, even the most redundant utterance remains radically impoverished relative to the multimediated event in which it occurs. The basic condition of human language is constitutive symbolic underspecification, and redundancy is at most a local ripple on its surface.
For xenolinguistics the consequence is immediate. A one-way signal received across interstellar distance is the limiting case of communication without interrecursivity. We cannot look puzzled at it, and it cannot reformulate. We cannot interrupt, and it cannot slow down. Whatever made the signal sufficient for its makers, the recursive adjustment of counterparts to one another is exactly what has been stripped away before it reaches us.
III. The radical underspecification of language
Consider the instruction Put it over there. As a linguistic string it is spectacularly impoverished. It could refer to anything; there could be anywhere; no method of moving the thing is specified, no reason is given, and no deadline appears. Yet two people standing in a kitchen, one holding a hot pan and the other clearing space on a counter, understand it instantly. Multisensorial embodiment supplies gaze, gesture, orientation, and the bodily recognition that the pan is heavy and the holder under strain. Being-with supplies a history of mutual knowledge and anticipation. Multiversal dwelling supplies the unmade conditions of the scene: gravity pulling on the pan, the light by which both can see, the air through which a quiet word carries across a small room. Multimateriality supplies the counter, the stove, and a layout that already proposes where things go. Multisymbolization supplies, beyond the utterance itself, the sedimented expectations of household order. The words mark a single point of adjustment within a field that is already richly coordinated.
Language is therefore remarkable because it contains so little, rather than because it contains redundant information. One could spend a day trying to describe a single second of multisensorial bodily experience (the pressure of the floor, the temperature of the air, a faint tension in the jaw, the smell of the room, the half-noticed rhythm of breathing) and still fail to reproduce it symbolically. Every utterance is drastically short of what the situation contains. It works because it never works alone. Most of the coordination was never in the linguistic signal at all. Embodiment, being-with, dwelling, and multimateriality reduce the symbolic burden because speaker and hearer already inhabit a richly coordinated world, and symbolisation selectively articulates within that world rather than carrying it.
When symbolisation is separated from its situation, it must expand to compensate. Legal documents grow elaborate because they must function among parties who may share neither situation, history, nor goodwill. Written instructions must specify what joint activity would leave tacit. Text messages develop emoji and conventions of punctuation because gaze, gesture, and prosody are absent. Designed interstellar languages are the extreme case of this compensatory verbosity. Hans Freudenthal's Lincos attempted to build a language for cosmic intercourse from first principles, beginning with numbers and logical relations and working step by step toward behaviour and conversation (Freudenthal 1960). The ambition was to specify everything the recipient would need. Yet every step presupposed that the recipient would take the first steps as instructions, would treat pulses as countable units, and would recognise a sequence as a lesson. The attempt to specify everything shows how much any specification already assumes.
A further dimension of underspecification concerns what kind of operation a symbol is performing at all. Living Value Theory distinguishes three symbolic modes. In coordination mode, a symbol participates directly in the situation it names, as a shouted warning does. In reference-to-absence mode, it makes something not presently available matter to current coordination, as a report or a memory does. In re-presentation mode, it claims to make present again what is absent, lost, imagined, or never directly encountered, as ritual, art, and theory do. The symbolic surface does not reliably announce which mode is operative. The word fire can empty a building, report yesterday's blaze, or open a poem about purification, and in each case listeners discern the mode through situation, prosody, history, and the responsiveness of the speaker rather than through anything in the word.
This alone dramatically strengthens the case against decoding. An alien signal might be formally analysable in exhaustive detail while leaving us unable to determine whether it is participating in some current action, referring to something absent, or re-presenting something imagined, remembered, or ritualised. We might identify its grammar perfectly and still not know whether we were reading an alarm, a chronicle, or a liturgy. Among humans the mode is almost never marked because it almost never needs to be; the shared situation settles it. A detached trace arrives without its situation, and so without the very resource on which the discernment of mode depends.
IV. What language makes possible
Evolutionary accounts of language often speak as if language emerged in order to solve a problem, as if the capacity were a solution waiting for its need. Living Value Theory avoids this teleological vocabulary. It is more exact to speak of recurrent pressures, of affordances that some forms of life came to exploit, and of consequences that followed once certain capacities were in place. The question is what language makes possible, and under what coordination conditions that possibility became consequential.
The five mediations fall into two groups. Multisensorial embodiment, being-with, and multiversal dwelling are the zoetic mediations. Together they constitute a complete living mesocosm: an organism coordinating with its own body, with other living beings, and with the found conditions of its environment is fully alive and can sustain extraordinarily complex forms of coordination, including those of predators, migrants, parents, and offspring across great ranges of space and time. Multimateriality and multisymbolization are the poietic mediations. They enable recursive life to deposit prior coordination into relatively durable, nonrecursive ground: material ground in the case of a path, a nest, a hearth, or a tool, and symbolic ground in the case of a name, a count, a rule, or a record.
Rudimentary poietic capacities are not uniquely human. Birds build nests, termites raise mounds that regulate their own climate, crows and chimpanzees shape and reuse tools, and many animals produce calls that reliably distinguish kinds of threat. The distinctively human threshold therefore lies in the sustained, open-ended recursive coupling of the two poietic mediations rather than in simply having tools and symbols: material deposits that are named, counted, taught, and regulated, and symbolic deposits carried by material media and projected onto material arrangements, each opening new possibilities for the other across generations.
The coupling must be distinguished from identity. A tool is first a multimaterial deposit. Its edge, weight, and grip hold a form that makes certain actions easier and others possible, and this affordance persists whether or not anyone has named it. A tool may also be symbolically classified, named, taught, represented, and institutionally regulated, but its affordance does not make it a symbol. Conversely, symbolic ground cannot be reduced to the material substrate that carries it. A road can persist materially long after every symbolic system that once explained it has vanished, and an equation can survive as symbolic ground through countless changes of paper, chalk, and screen without being any of them. Multimateriality and multisymbolization are coupled but irreducible, and their extraordinary human power comes from their recursive coupling rather than from their being two faces of a single capacity.
A second architecture cuts across the first. Embodiment and being-with are characteristically recursive counterpart relations. In embodiment the counterpart is selfrecursive: the living body whose own prior activity alters its subsequent activity. In being-with the counterpart can be interrecursive: another living being whose responses respond to one's own. Dwelling, multimateriality, and multisymbolization are relations to nonrecursive ground: found in the case of dwelling, materially made in the case of multimateriality, and symbolically made in the case of multisymbolization. A mountain, a road, and a written law do not respond to our responses, however consequential our engagement with them may be.
This second architecture allows a precise statement of what a detached alien transmission is. It is a nonrecursive symbolic deposit produced by a recursive process that is no longer present to us. We confront the trace without access to the recursive activity that made it. In this respect an alien signal resembles every archive, inscription, and letter from the dead. The difference is that when humans take up a human deposit they reactivate it from within a shared mesocosm, bringing to it bodies, relationships, dwellings, and material worlds sufficiently like those of its makers. With an alien deposit that reactivation has nothing comparable to draw upon.
What the poietic mediations make possible, above all, is coordination that outlasts co-presence. A group of hunters moving through terrain together can coordinate through posture, gaze, and the modulation of shared attention, but such coordination cannot by itself reach forward to next season, backward to what worked last time, or laterally to members who are not present. Symbols allow recursively relevant distinctions to travel, persist, and reorganise later coordination. It is plausible that early symbolic articulation clustered around food, danger, offspring, and the organisation of the group, and some evolutionary accounts make this case, but the evidence is thin and the argument here does not depend on it. The structural claim is sufficient: symbols become consequential because they carry distinctions beyond the occasion in which those distinctions first mattered.
Nor do symbols gain their traction only from persistent material things, since they also coordinate bodily states (it hurts here), relationships (she is my sister), absences (he has gone to the city), imagined futures (after the rains), promises, legal statuses, memories, and purely intersymbolic operations such as proofs and calendars. What gives symbolisation its traction is its embedding in the whole five-mediational mesocosm. Remove that embedding and the symbol remains, but the traction goes with the world.
V. Vital stakes
Bronisław Malinowski placed two demands at the foundation of modern ethnography. One was to grasp the native's point of view, to render the world as others articulate it. The other, stated in the same breath, was to study what concerns people most intimately, "the hold which life has on him" (Malinowski 1922). Anthropology developed the first demand with enormous sophistication and progressively neglected the second. Stakes, in the sense this second demand points toward, are what coordination depends on: what would cause a world to come apart if it failed, what is being held together in the ongoing and largely unarticulated work of keeping a life livable. Stakes differ from what people say about their lives, though what they say may express, obscure, or reorganise them.
Living Value Theory now names the ground of stakes more exactly as vital stakes: the viability of L1 coordination across all five mediations. L1 is the level at which coordination proceeds without becoming separately salient, the smooth functioning of a body, a relationship, a dwelling, a material world, and a symbolic world that one is not currently attending to. What is vital is whatever that functioning depends on. Metabolic stakes remain fundamental, and evolutionarily basal. Every living being must take in energy, maintain its boundaries, and avoid destruction; hunger, injury, temperature, and fatigue constrain everything else a living being can do. But metabolic stakes are not synonymous with vital stakes. What is at stake in a friendship, a betrayed promise, a desecrated place, or a lost archive cannot be translated back into calories or reproductive success. Because the human mesocosm has become dependent on poietic deposits, the destruction of material and symbolic ground can itself be vital: the loss of a home, a language, or a legal status can make a life unlivable while every metabolic need is met. Metabolic stakes form a basal constraint nested within vital stakes rather than their definition.
Recursivity must be defined with corresponding care. It is the capacity for the consequences of prior mediation to alter subsequent mediation. The registering of misalignment and the attempt to restore coordination are crucial cases of recursivity, perhaps the cases through which it is most easily recognised, but they are cases rather than the definition. A body that grows stronger through use, a relationship that deepens through repeated encounter, and a skill refined through practice are all recursive without any misalignment needing repair.
Wittgenstein's lion can now be reconsidered. His claim that a talking lion could not be understood, because understanding requires a shared form of life, moves in the right direction. It would be a mistake, however, to conclude that a speaking lion's words would be empty. The lion has an extremely full mesocosm: an olfactory world of scent-marked ground, a being-with organised by the pride and its coalitions, a metabolic existence built around the hunt and the long rest that follows it. The difficulty lies in the limited mesocosmic fit between our articulations and its living coordination. Nor is the overlap zero, for humans and lions share gravity, fatigue, injury, hunger, threat, sleep, locomotion, heat, and attachment to their young. The words that might pass most easily between us would be those anchored in such shared zoetic constraints. I am hungry might be the most intelligible sentence a lion could utter, and even then it would underdetermine almost everything about what hunger meant within the pride's coordination: who eats first, what the hunt costs, which prey is worth the risk. This is mine would be far more obscure, because ownership among lions is organised through scent, patrol, and combat rather than through the material and symbolic deposits on which human claims to property depend. The problem is uncertain and partial overlap rather than an absence of common ground.
VI. What decoding can and cannot do
The field of xenolinguistics has been organised around the expectation that alien communication, if it exists, will take a form that can in principle be decoded: a signal with structure, a structure with grammar, a grammar with semantics. The search for extraterrestrial intelligence looks for signals that differ from natural background in ways that suggest artificial organisation. Lincos was designed to communicate through mathematical and logical primitives assumed to be universal. Generative linguists, Chomsky among them, have speculated that the basic combinatorial operation of human syntax might be shared by any intelligent beings, providing a structural common ground. These are serious undertakings, and decoding is capable of real achievements. It can detect artificiality, identify combinatorial structure, and map correspondences between the elements of one system and another.
What decoding cannot do by itself is reach the recursive process from which a signal came. Its product is a formal description of a symbolic surface, a set of intersymbolic relations, which may be complete on its own terms while leaving untouched what the signal was doing, which mode it was operating in, and what was at stake for those who produced it.
A recent line of laboratory research shows how easily this gap is overlooked even on Earth. In studies of artificial-language learning, participants are told that they are learning an alien language. They see pictures of unfamiliar creatures and objects, learn novel words and grammatical markers, and are tested on whether they can extend a grammatical category, such as a noun class, to entities they have not seen before. The design novelises the symbols while holding the world substantially constant. The creatures still have faces and limbs; the objects still look like artefacts or inert things. Participants arrive already able to distinguish beings that move and respond from things that lie where they are placed, and the alien grammar merely attaches new markers to these pre-existing mesocosmic attributions. Whatever such studies show about learning, they are studies of terrestrial ontology relabelled. A truly alien experiment would require a world whose relevant categorial distinctions were themselves unavailable in advance: entities that resisted the distinctions between animate and inanimate, agent and object, individual and collective, inside and outside. Participants would have to discover what distinctions the world afforded before they could learn which word applied. That such an experiment is so hard to design is itself instructive. Decoding an actual alien signal would present the reverse of the laboratory situation: an unfamiliar world with no labels attached, rather than familiar distinctions dressed in unfamiliar words.
Research on animal communication makes the same point with a more living subject. Some animal signalling can indeed be mapped, decoded, and associated with context. Karl von Frisch showed that the honeybee's waggle dance conveys the direction and distance of a food source (von Frisch 1967), and alarm calls in several primate and bird species reliably distinguish kinds of predator. These are real decodings. Yet decoding the vector encoded in a dance does not exhaust the coordination in which the dance participates: the scent the forager carries, the condition of the colony, the recruitment of some foragers and not others, the hive as a material and climatic world. Whatever can be decoded from formal regularity does not exhaust the living coordination in which the signals participate.
Sperm whales provide the most instructive current case. Sperm whales communicate through sequences of clicks called codas, and recent analysis of recorded exchanges has shown that codas have far more structure than earlier classifications recognised. Beyond rhythm and tempo, they display context-sensitive features the researchers call rubato and ornamentation, which are controlled by the whales and imitated across whales during exchanges, and these features combine to produce a repertoire nearly an order of magnitude larger than previously described (Sharma et al. 2024). The authors themselves note that the communicative function of many codas remains an open question. This work does not refute the argument of this article; it strengthens the need for it. Increasingly sophisticated structural decoding reveals more and more organisation while leaving open which distinctions matter within sperm-whale life. The finding that whales modulate and imitate features of one another's codas in the course of an exchange is especially telling, because it shows interrecursive adjustment at work, the very dimension that decoding recorded sequences can describe but not enter.
Two errors must therefore be avoided. The first treats the absence of translation as proof that translation is structurally impossible. Nothing in the argument licenses that inference, and research that continues to reveal structure in cetacean communication should be expected to reveal more. The second treats the progress of structural decoding as progress toward understanding in the same measure. The two achievements can advance at very different rates, and success in one does not entail success in the other.
VII. The Malinowski loop at cosmic scale
The trajectory of anthropology after Malinowski illustrates what happens when the representational demand crowds out the demand concerning stakes. Clifford Geertz narrowed significance from consequence to meaning, redefining the study of culture as the interpretation of texts (Geertz 1973). The Writing Culture moment deepened the representational turn by making the ethnographer's own textual authority the object of critique (Clifford and Marcus 1986). At its furthest extension, the principal stake of anthropological writing became the anthropologist's relation to the impossibility of representing others without violence. This followed from treating symbolic articulation as the primary object of inquiry rather than from any individual failure. Multisymbolization has no intrinsic signal of its own detachment: a system of symbols can go on elaborating, complexifying, and examining its own operations indefinitely, achieving ever greater intersymbolic fit, without any internal indication that its contact with the coordination it was meant to illuminate has thinned.
Xenolinguistics repeats this trajectory at cosmic scale. It begins with an important question: is there other intelligent life, and could we communicate with it? It then narrows the question to its representational dimension: how would we decode such communication, which symbolic structures might be universal, which mathematical or logical primitives might provide common ground? The coordinative question, what we would actually be coordinating and what stakes could make such communication consequential, is either not asked or dispatched with brief optimism that intelligent beings will surely have interests like ours. The first demand has received enormous technical elaboration. The second has been set aside, because it is harder and less amenable to the methods the field employs.
The search for extraterrestrial intelligence has now continued for more than six decades without a confirmed detection. Standard interpretations appeal to the rarity of technological civilisations, to their distance in space and time, or to the possibility that they transmit in ways we have not anticipated. It would be tempting to add that the search has been looking for the wrong thing, but the null result cannot support that inference. Many explanations are compatible with silence, and nothing in the absence of a detection tells us which of them is true. What Living Value Theory can say is more limited and more secure. The search's detection criteria embody assumptions about what a recognisable poietic trace would look like: narrowband radio emission, laser pulses, the waste heat of large-scale engineering. These criteria are modelled on human material and symbolic deposits, and they define in advance what would count as evidence of another poietic life. That definition is itself an L4 stabilisation, reasonable as a starting point, whose assumptions should be kept visible rather than mistaken for a neutral description of what intelligence must produce. Beings whose poietic mediations were coupled differently from ours, or scarcely developed at all, might coordinate richly across their own forms of absence without leaving any trace our instruments are built to notice. Whether they do is an open question that the silence cannot answer.
VIII. Cosmic Esperanto and three kinds of fit
The dream of a universal language for interstellar communication, whether Lincos, mathematical primitives, or the logical structures of a universal grammar, rests on the hope that some symbolic system can operate independently of the mesocosm that gives symbols their traction. Mathematics is the favoured candidate precisely because it appears the most detached, the most formally self-sufficient, and the least dependent on any particular embodied or environmental ground.
Living Value Theory distinguishes three kinds of fit that talk of universal language tends to merge. Intersymbolic fit concerns coherence among symbolisations themselves: the consistency of a proof, the systematicity of a grammar, the correspondence between two notations. Cosmic fit concerns the relation between a nonrecursive symbolic system, paradigmatically mathematics, and the aspect of the cosmos to which it is applied. Mesocosmic fit concerns the relation between a symbolic articulation and the lived coordination of beings across the five mediations. The three are independent achievements, and success in one does not guarantee success in another.
Mathematics exhibits extraordinarily high intersymbolic fit. Some mathematics also has remarkable cosmic fit, in that physical predictions formulated in it hold regardless of human preferences. It is entirely possible that beings elsewhere, investigating the same cosmos, would converge on some of the same invariances. The cover of the Voyager record uses the transition of the hydrogen atom as its unit of time and length, on the reasonable assumption that any technological civilisation would know it. Suppose the assumption is correct. Convergence on the same physical invariance would establish a point of shared cosmic fit, and perhaps a point of intersymbolic correspondence between two notations. It would tell us nothing about what that invariance mattered for in the recipients' lived world. Shared cosmic fit establishes no mesocosmic fit. Two beings could agree exactly on the hydrogen line and have no idea what the other ate, feared, or cared for.
Highly formal systems also need to be placed correctly among the recursivity levels. L3 is situated symbolic articulation: this pain, this danger, this question, articulated within a particular interaction. L4 is portable stabilisation: abstraction, classification, and institutionalised symbolic ground that can travel beyond its occasion. L5 recursively repositions such grounds, asking why a question is askable or a classification valid. A formally coherent exchange of mathematical or logical structures between species would therefore be an achievement of intersymbolic fit operating at L4 and L5 rather than situated articulation at L3: internally rigorous and potentially portable across vast distances, whose contact with the living coordination of either party would remain an open question.
Translation must be reframed accordingly. It has often been imagined as the mapping of one code onto another, with the alternative being some deeper sharing of concerns. The three kinds of fit show that translation comprises several achievements that can come apart. Translation can achieve high intersymbolic fit between two sign systems without achieving mesocosmic fit: a flawless correspondence between two formal languages, or a grammatically perfect rendering of a text whose referents neither translator has encountered. Conversely, extremely crude symbolic exchange can achieve high mesocosmic fit if it successfully coordinates something consequential. A pointing gesture, an offered hand, or a shared rhythm between a rescuer and a trapped animal can coordinate life and death with almost no symbolic content at all. The universal translator of science fiction is incoherent because it collapses these achievements into one, assuming that perfect intersymbolic correspondence would deliver mesocosmic understanding automatically.
No symbol is freestanding in the sense the universal-language project requires. Every symbolic deposit is nonrecursive, and every uptake of it is recursive: a living being must reactivate it from within some mesocosm. The universality that mathematics appears to possess among humans is real within the enormous shared background of human embodiment, dwelling, material practice, and teaching. Whether it extends further is a question about cosmic fit that may well be answered in its favour. Whether it would carry any mesocosmic significance across the gap is a different question, and the formal properties of mathematics cannot answer it.
IX. Misunderstanding and recursive opening
If communication across a real mediational gap depends on more than shared code, misunderstanding must be revalued. Xenolinguistics treats misunderstanding as a problem to be solved: a gap in decoding competence, a mismatch between symbol systems, a technical failure awaiting better methods. It is better understood as the honest starting point of any encounter across a real mediational difference. When beings with different vital stakes, different embodied sensory architectures, different dwelling conditions, different material worlds, and different symbolic systems encounter one another, misunderstanding is evidence that there is a real difference to be understood. A translation that seemed to bypass all of this would amount to the projection of one mediational structure onto another, the imposition of one form of life's categories onto another's stakes. Anthropology at its worst has done this to the humans it studies, and a triumphant xenolinguistics would be tempted to do it to any aliens it encountered.
Fiction has often explored this terrain more candidly than the science of xenolinguistics. In Ted Chiang's "Story of Your Life", adapted by Denis Villeneuve as Arrival, the heptapods' written language is not a transcription of speech, and learning it reorganises the linguist's relation to time (Chiang 1998). What first appears as a decoding problem becomes a problem of dwelling differently in temporality, and the linguist's understanding advances only through sustained, reciprocal sessions in which each side responds to the other's responses. In China Miéville's Embassytown, the Ariekei can speak only of what is, and communication with them requires a peculiar form of shared presence embodied in paired human ambassadors, so that the obstacle concerns the architecture of being-with as much as vocabulary (Miéville 2011). In Stanisław Lem's Solaris, the planetary ocean responds to the scientists by materialising figures from their memories, an intervention in the embodied and relational ground of their lives that never becomes communication in any symbolic sense, and whose purpose, if it has one, remains undecidable (Lem 1961).
Each of these fictions points toward what might be shared across radically different forms of life. Shared code, shared grammar, and shared mathematics all prove insufficient. What may be shared is more elementary: recursive susceptibility to misalignment. The recursivity levels allow this to be stated precisely. At L1, coordination proceeds in operative closure, without anything becoming separately salient. At L2, something ceases to disappear into smooth functioning; a disturbance, a pull, or an unease becomes felt, a questionability that does not yet contain a question. At L3, the disturbance is articulated, and explicit askability becomes possible: a question can be posed within an interaction. L4 stabilises and institutionalises articulations, determining which questions are legitimate and which answers acceptable. L5 recursively repositions such grounds, asking why a question is askable at all. Opening need not be linguistic. An animal startled by an unfamiliar sound, a plant redirecting its growth, and a person who senses that something in a room has changed are all undergoing recursive opening without formulating anything.
Alien life, if it is recursive, may share the more elementary of these conditions: the capacity for something to go wrong, to press against current coordination, to become salient at L2. This must be carefully distinguished from the capacity for linguistically formulable questions. The former may be common to all recursive life; the latter is a late and specific achievement whose forms we cannot assume to be shared. Shared susceptibility to misalignment provides a common structural position without providing common content. Two beings could both be unsettled, each in its own way, and the recognition of that shared unsettledness might produce the mutual acknowledgment of profound strangeness rather than conversation. Such acknowledgment is already something. It is also where any further understanding would have to begin.
X. What could become shareable?
Even if contact occurred and some symbolic exchange were achieved, the question of what we would actually say to each other would remain open. It might be tempting to answer that communication becomes substantive only when a common problem presses on both parties, so that beings without shared urgent problems would have nothing to talk about. That answer is too narrow. Human beings communicate at length when nothing presses at all. They tell stories, gossip, play, joke, sing, and discuss Homer precisely when no urgent problem demands resolution, often in the periods of relative security that metabolic buffering makes possible. Much of the richest symbolic life of any community arises from such non-pressing activity. Shared stakes make communication intelligible without needing to be currently pressing. We can argue about justice because we both know what it is to be harmed; we can discuss beauty because we share embodied sensory architectures; we can grieve together because we both face mortality. The overlap among humans is so pervasive that it is invisible, and it makes even idle talk mutually intelligible.
With beings whose vital stakes diverge radically from ours, this pervasive background could not be assumed. But it would be a mistake to treat the matter statically, as though two beings either already share a field of concern or have no foundation for communication at all. Contact itself can generate new shared stakes. If I do something and you respond, and I respond to your response, regularities begin to emerge; each of us starts to discern the other, to anticipate, to adjust. Where there was no interrecursive field before, one can form. The long history of humans and dogs shows how a new interrecursive field can be built across species over thousands of years, until each partner reads the other's gaze and gesture with a fluency neither could have had alone. On a smaller scale, every researcher who has spent years habituating to a group of primates, or following the same whale families season after season, has participated in the slow formation of such a field.
Far from being a minor qualification, this tells us what kind of xenolinguistics could actually work. A xenolinguistics based on a one-way radio signal is profoundly limited, because the signal is a nonrecursive deposit, received long after the recursive process that produced it, with no possibility of response entering into its production. A xenolinguistics based on sustained reciprocal encounter would be far richer, because it would allow an interrecursive field to form in which shared stakes could emerge rather than needing to be presupposed. The interesting possibilities lie there, in encounter rather than in decipherment.
It is sometimes said that the probability of achieving shared understanding with aliens is much smaller than the probability of decoding their signals. No basis exists for quantifying either probability, and the comparison obscures the more important point. The two problems are categorically different. Decoding concerns intersymbolic structure; understanding concerns mesocosmic fit and, where it is possible at all, the formation of shared interrecursive coordination. Success at one does not entail success at the other. An alien receiving the Voyager record and decoding every symbol on it would know more about human symbolic deposits than any human institution currently knows about cetacean communication. It would still have received the trace without the process, and the surface without the world that made that surface sufficient.
XI. Toward an honest xenolinguistics
What would a xenolinguistics look like that took this argument seriously? It would treat its task as mesocosmography under extreme conditions: the attempt to describe another being's mesocosm when almost none of the usual shared background can be assumed. Its method would move through a sequence of questions rather than beginning with a codebook.
It would ask first what kind of counterpart it was encountering: whether the source of a signal was recursive at all, whether it was responding to anything, and whether it might respond to us. It would ask which mediations were observable, whether bodies, relations among individuals, the found conditions of a world, made things, or symbolic deposits, and which could only be inferred. It would ask what appeared to be recursively relevant to the being in question, which differences it responded to and which it ignored, and which distinctions seemed to become salient in its coordination. It would ask what symbolic mode, if any, was operating in a given signal, and it would treat the answer as uncertain by default. It would ask what kind of fit had actually been established at each stage, whether intersymbolic, cosmic, or mesocosmic, and it would refuse to let success in one pass for success in another. Above all it would ask what changed when the other responded to us, because only there would an interrecursive field begin to form.
Such a xenolinguistics would take animal communication seriously as a set of coordination systems whose relation to vital stakes must be understood on its own terms, rather than as a set of codes awaiting better decryption. It would welcome every advance in structural analysis, such as the recent work on sperm-whale codas, while insisting that structure revealed is not yet significance understood. It would begin with the beings already around us, whose bodies, dwellings, and histories overlap with ours far more than any extraterrestrial's plausibly could, and it would learn from how hard understanding remains even there.
It would treat misunderstanding as its primary datum, the evidence that real mediational difference exists and must be respected rather than engineered away. And it would be honest about its speculative character. The stakes of another form of life, the hold that life has on alien beings, to adapt Malinowski's phrase, may not be available for representation without distortion, and an honest xenolinguistics would hold that possibility at its centre rather than assuming that successful decoding would dissolve it.
The Voyager record is still travelling, carrying humanity's most careful symbolic self-portrait into the dark. If it is ever received and decoded, its finders will learn that somewhere in the galaxy a species makes music and mathematics, and even that it eats, gives birth, and nurses its young. Yet even if they decoded every word and every image, almost everything that made those words and images communicatively sufficient for humans was never in them. It was in bodies, relationships, places, objects, histories, timings, and recursively responsive others. That is why the symbolic surface cannot reconstruct the living world from which it came, and why the problem was never peculiar to aliens: every human utterance has the same structure, and only the shared world we bring to it hides the fact. The finders will hold our words without the world that made them enough. They will have no idea that we get hungry.
Selected References
Chiang, Ted. 1998. "Story of Your Life." In Starlight 2, edited by Patrick Nielsen Hayden. New York: Tor.
Clifford, James, and George E. Marcus, eds. 1986. Writing Culture: The Poetics and Politics of Ethnography. Berkeley: University of California Press.
Freudenthal, Hans. 1960. Lincos: Design of a Language for Cosmic Intercourse. Amsterdam: North-Holland.
Geertz, Clifford. 1973. The Interpretation of Cultures. New York: Basic Books.
Lem, Stanisław. 1961. Solaris. Warsaw: Wydawnictwo Ministerstwa Obrony Narodowej.
Malinowski, Bronisław. 1922. Argonauts of the Western Pacific. London: Routledge.
Miéville, China. 2011. Embassytown. London: Macmillan.
Sagan, Carl, F. D. Drake, Ann Druyan, Timothy Ferris, Jon Lomberg, and Linda Salzman Sagan. 1978. Murmurs of Earth: The Voyager Interstellar Record. New York: Random House.
Shannon, Claude E. 1948. "A Mathematical Theory of Communication." Bell System Technical Journal 27: 379–423, 623–656.
Sharma, Pratyusha, Shane Gero, Roger Payne, David F. Gruber, Daniela Rus, Antonio Torralba, and Jacob Andreas. 2024. "Contextual and Combinatorial Structure in Sperm Whale Vocalisations." Nature Communications 15: 3617.
von Frisch, Karl. 1967. The Dance Language and Orientation of Bees. Cambridge, MA: Belknap Press of Harvard University Press.
Weaver, Warren. 1949. "Recent Contributions to the Mathematical Theory of Communication." In Claude E. Shannon and Warren Weaver, The Mathematical Theory of Communication. Urbana: University of Illinois Press.
Wittgenstein, Ludwig. 2009. Philosophical Investigations. Revised 4th edition. Translated by G. E. M. Anscombe, P. M. S. Hacker, and Joachim Schulte. Oxford: Wiley-Blackwell.