I. Introduction: The Lost Manuscript
Among the papers held at the Bielefeld University archive are obscure references to a book manuscript that, according to at least two disciples of Niklas Luhmann, was completed in the final months of his life and then vanished. The manuscript has never been formally catalogued. Its existence is known only through the testimony of people who heard Luhmann describe it, and whose accounts have the quality of things preserved in oral tradition rather than institutional record. The title, as his editor at Suhrkamp Verlag confirmed, was Der Verkehr der Gesellschaft: on the traffic of society.
The argument of the manuscript, as these accounts reconstruct it, was that the traffic system is a functionally differentiated system whose autopoietic reproduction consists in the continuous production and reproduction of vehicular communication through its own operational elements. Traffic lights are not, in this account, merely mechanical devices for coordinating vehicular flow. They are the operative symbols of the traffic system's binary code, which distinguishes movement from non-movement and thereby enables the system's self-referential reproduction. Roads do not merely provide physical surface for automobiles. They are the structural couplings that channel environmental perturbations from the physical and biological environments of the traffic system into forms the system can process through its own code. Drivers are not, in the relevant sense, social actors. They are psychic systems in the environment of the traffic system, capable of irritating the system through structural coupling but unable to participate directly in the system's autopoietic operations.
Congestion, in this framework, is the traffic system reproducing traffic through the specific modality of blockage: a reflexive intensification of vehicular communication that demonstrates the system's capacity to generate self-referential complexity within its own operational domain. The horn is a communication medium through which the traffic system's binary distinction is reaffirmed under conditions of operational strain. Lane markings are the institutional crystallisation of the traffic system's self-description: a moment in which the system's expectation structures achieve material anchoring in the physical environment through which it operates. Roundabouts represent a sophisticated evolution in the traffic system's management of contingency, distributing the reduction of movement-complexity across multiple simultaneous operational sequences without requiring the binary closure of traffic-light communication.
The fictional manuscript continues for some time in this vein. And then, as the oral tradition has it, something happened. Luhmann, carrying the full type-written manuscript to the post office, encountered another pedestrian on a narrow pavement. Both stepped left. Both stepped right. Both stepped left again. The small recursive oscillation, that brief moment of awkward mutual obstruction familiar to everyone who has walked in a city, repeated itself two or three times before the other man laughed, apologized, and stepped decisively aside. Luhmann, however, froze. He seemed shocked. The manuscript never reached Suhrkamp Verlag, and vanished without a trace.
The story is certainly invented. The lesson it encodes is not. The moment of pavement oscillation is the moment in which the manuscript becomes visibly absurd: not because the language is wrong but because the ontology it projects onto the world is wrong. The pavement problem is not a problem that any traffic system solves. It is solved, when it is solved, by two recursive beings who stop recursively mirroring each other's anticipations and allow one of them to make a definitive move that the other accepts. What stabilizes the encounter is a small act of recursive surrender, not the operation of a functionally differentiated communication system. And yet the language of the fictional manuscript, applied to the problem, generates perfectly grammatical sentences. This is the joke. And the joke, as will become clear, already contains the theory.
This article argues that systems theory mistakes stabilized effects for autonomous ontological entities. Social institutions are not autopoietic systems whose purpose is the reproduction of their own operational closure. They are multimediated stabilization architectures designed to compress interrecursive unpredictability among recursive beings sufficiently for large-scale coordination to become metastably regular. The distinction is not terminological. It is an ontological inversion with consequences for everything: what institutions are, how they function, why they fail, whose labor they depend on, and what the appropriate analytical response to their constitutive features must be.
II. The Corridor Dance: Interrecursivity in Miniature
Begin with two strangers approaching each other along a narrow corridor. Each is attempting something that deserves the name courtesy: a sideways movement to allow the other to pass without contact. Each anticipates that the other is making a similar attempt. The first moves right. The second, having anticipated this, also moves right, to leave space for the first. Now both have moved to the same side. The first, perceiving this, moves back left. The second, perceiving this, also moves left. The oscillation continues for two or three cycles before one of them laughs, apologizes, and steps decisively aside.
The phenomenology of this encounter rewards careful attention. Each participant is attempting to coordinate with the other. Each is doing so by modeling the other's anticipated movement. Each is adjusting their own movement in light of this model. And each adjustment changes the conditions under which the other's next adjustment will be made. The encounter is informationally trivial: two bodies, one narrow space, the goal of mutual passage. A weather simulation of equivalent spatial complexity would involve millions of interacting variables. This encounter involves two. The instability is not generated by information surplus. It is generated by something structurally different: the mutual recursive orientation of two beings who are each modeling the other's modeling of them, in loops that have no natural stopping point because each response constitutes a new input for the other's model.
This is interrecursivity, and it is constitutively distinct from two other sources of unpredictability that the intellectual tradition has persistently conflated with it. The first is the unpredictability of non-recursive systems: a weather system is unpredictable because it amplifies tiny initial condition differences exponentially, but it does not respond to being modeled. Predict the weather however accurately you like; the weather will not adjust its behavior in response to your prediction. The second is computational complexity: the intractability of problems with many interacting variables where brute-force calculation is insufficient. Both of these are real phenomena. Neither is what generates the corridor oscillation. The corridor oscillation is generated by two beings who are recursively modeling each other's modeling, where each model update is itself visible as a new input to the other, in loops that generate instability without requiring informational surplus or chaotic sensitivity. The pavement problem is, computationally, trivial. Its difficulty is entirely and exclusively a function of interrecursivity.
Now consider how the problem is solved at scale. Two strangers negotiating corridor passage, with enough social competence, can resolve the oscillation through a combination of micro-signals: a raised hand, a slight tilt of the head, a more decisive movement to one side that reads as commitment rather than tentative anticipation. These are interrecursive solutions: they work by changing the recursive dynamics of the encounter rather than by eliminating them. But two strangers can only do this within a specific range of social conditions. Thousands of strangers simultaneously navigating a busy train station cannot solve the oscillation problem through interrecursive negotiation in real time. The scale exceeds the capacity of live recursive exchange to stabilize coordination.
The institutional solution is the suppression of recursive negotiation. Keep left. Keep right. Lane markings. Arrows. Traffic lights. These mechanisms work precisely because they eliminate the recursive problem rather than solving it. They convert what would otherwise be a constitutively open interrecursive negotiation into something closer to a non-recursive coordination problem: given that everyone moves on the same side, the question of which side to move to is no longer a recursive problem but a categorical one. The institution works by reducing live recursive negotiation to zero in the relevant domain. This is what successful institutions always do. They do not optimize recursive coordination. They externalize it into stabilized mediational infrastructure, from which the recursive burden has been architecturally removed.
III. What Systems Theory Saw Correctly
Luhmann's systems theory was, for a generation, the most intellectually powerful framework available for analyzing modern social institutions, and giving it the credit it deserves is essential before specifying where and why it went wrong. Luhmann saw something genuinely important: that modern institutions exhibit a peculiar kind of self-maintaining stability that seems to persist beyond the intentions, actions, or even the continued existence of any individual participant. Law continues to produce legal decisions long after the judge who trained any given generation of lawyers has retired. The economy continues to process payments through mechanisms whose logic no individual fully grasps. Science continues to distinguish true from false through procedures whose epistemic legitimacy has been contested since their inception. Institutions, in this sense, look like they are running themselves.
Luhmann also saw something equally important: that this self-maintenance is achieved through communication that recursively refers to prior communication. Legal decisions cite precedents. Scientific publications respond to prior publications. Economic contracts are interpreted through prior economic arrangements. The self-referential character of institutional communication is not accidental. It is the mechanism through which institutions generate continuity across time and across changing personnel. The judge who retires is replaced by another judge whose decisions will be made against the background of the decisions made by their predecessor, not because they personally knew their predecessor but because they were trained within a communication system that preserved and transmitted those prior decisions as the operative context for new ones.
The concept of double contingency is Luhmann's most important contribution and the one that comes closest to identifying the real problem. Ego cannot know how Alter will act without knowing how Alter will respond to Ego's action, and Alter's action depends on Alter's anticipation of Ego's response, which depends on Ego's anticipation of Alter's anticipation, in potentially infinite recursive loops. This is, in Living Value Theory's vocabulary, the recognition of interrecursivity. Luhmann saw the phenomenon. He understood that it was the constitutive problem facing social coordination. He also saw that face-to-face interaction as a self-referential circle is inherently unstable and can dissolve at any moment, which is why he insisted that interaction needs a system that treats the self-referential circle as an element of its own self-reproduction. This is closer to the LVT position than critics often acknowledge: Luhmann knew that interrecursive encounters cannot generate institutional stability by themselves.
What he concluded from these observations is that institutions must themselves be recursive systems: that the response to interrecursive unpredictability among human beings must be the emergence of communication systems with their own recursive self-reproduction, operational closure, and autopoietic dynamics. He saw the stabilizing effect and attributed it to a new level of recursive being: the institution as organism. This is where the theory goes wrong. Not in what it observed but in the ontological conclusion it drew from the observation.
IV. The Biological Category Mistake: Autopoiesis Extended Without Warrant
Humberto Maturana and Francisco Varela developed the concept of autopoiesis in the early 1970s as an account of what makes living systems living systems. An autopoietic system is one that continuously produces and reproduces the components that constitute it. The cell produces its own membrane, its own enzymes, its own organelles, in a self-referential material process that maintains the cell as a distinct living system against the entropic tendency of its environment. Autopoiesis in this sense is a metabolic achievement: it requires specific chemical processes, specific thermodynamic conditions, specific material organization. The cell that ceases to metabolize ceases to be. The boundary that it maintains against its environment is a material boundary, maintained by ongoing material processes that require energy inputs from outside the system.
A precise account of the LVT critique requires acknowledging something that unsympathetic readings of Luhmann often miss. He was not naive about the difference between biological and communicative autopoiesis. He explicitly distinguished the two and was aware that social systems lack material metabolic closure. Maturana debated Luhmann on exactly this point, and Luhmann's response was not confusion but principled argument: his claim was that the formal structure of self-referential reproduction is analogous across different substrates, such that the concept of autopoiesis can be meaningfully extended to communication systems even though the material conditions of biological and communicative reproduction differ fundamentally. Whether this extension is warranted is a genuine theoretical question on which Luhmann took a considered position, not a confusion he failed to notice.
The LVT critique is therefore not that Luhmann confused social systems with organisms. It is that his extension of autopoiesis across substrates is unjustified, and unjustified in a specific way. The properties that make biological autopoiesis explanatorily powerful are precisely the material metabolic properties that social systems lack. A concept stripped of the features that give it its analytical force cannot be extended to new domains on the grounds that the formal structure is preserved. Formal structural analogy without material substrate equivalence does not support the ontological claims Luhmann builds on the extended concept. He retained the vocabulary of self-reproduction, operational closure, and organismic identity while removing the material conditions that make biological autopoiesis what it actually is. The result is a framework that attributes to institutions a kind of autonomous being that only the material metabolic closure of living organisms can actually sustain.
Institutions do not metabolize. Remove the electricity supply and the legal system does not metabolize its way to a new operational equilibrium. It stops. Biological organisms have unified embodiment: they have a body that can be located in space, that can be exhausted, injured, or killed. Social institutions have no such body. The legal system is not located in any particular space, cannot be exhausted or injured in any ordinary sense, and cannot be killed in the way an organism can be killed, though it can collapse. Most importantly, biological organisms possess genuine self-recursive interiority: they are metabolically responsive to their own states, registering depletion and mounting repair processes in ways that constitute, in more complex organisms, the embodied feeling of need, discomfort, and urgency. Social institutions have no such interiority. The legal system does not feel the pressure of its caseload. The economy does not feel the tension between inflation and unemployment. Recursive beings feel all of these things, and it is their feeling, their embodied self-recursive response to institutional conditions, that constitutes the living substance on which institutional functioning depends.
The decisive formulation follows from this analysis. Institutions are not recursive beings. They are architectures constructed around recursive beings. The university is not an organism that communicates. It is a multimediated stabilization architecture within which sleep-deprived academics teach, maintenance workers repair aging infrastructure, students study under conditions of financial anxiety, and administrators improvise responses to situations that institutional procedures did not anticipate. Remove these recursive beings, and there is no university. Remove the academic from their cell and there is still a cell. The difference is the difference between a being and an architecture, and it is the difference that systems theory's extension of autopoiesis systematically obscures.
V. What Social Institutions Actually Are
A social institution is a multimediated stabilization system that compresses interrecursive possibility-space sufficiently for aggregate coordination to acquire quasi-non-recursive regularity. This definition is precise and requires unpacking through each of its terms.
Interrecursive possibility-space is the field of possible futures generated by the constitutive openness of recursive beings recursively orienting toward each other. Without institutional scaffolding, this field is too large and too volatile for the densities of cooperation that complex social life requires. Any sufficiently large group of recursive beings, all simultaneously modeling each other's likely actions, generates interrecursive dynamics too unstable and too computationally demanding for unscaffolded coordination to sustain. Institutions compress this field: they reduce the range of effective interrecursive responses by establishing shared expectations, material constraints, symbolic frameworks, spatial channels, and temporal rhythms that make many recursive possibilities practically unavailable, and thereby make the remaining possibilities tractable.
The compression operates simultaneously across all five mediations of the mesocosm. Embodiment stabilizes institutions through disciplined bodies: queues require bodies trained to inhibit the immediate impulse toward the resource and to wait in spatial arrangement. Sleep disciplines align with institutional time: the factory shift, the school day, the office hour. Affect regulation suppresses the spontaneous recursive responses that embodied beings would otherwise display under institutional pressure: the restraint of anger in bureaucratic encounters, the maintenance of professional composure under provocation, the inhibition of fear responses in formal legal proceedings. These are not trivial achievements. They represent the sedimentation of institutional requirements into the embodied dispositions of the recursive beings who carry them, through years of training, repetition, and social reinforcement. Norbert Elias documented this sedimentation with extraordinary historical precision. What he called the civilizing process is, in Living Value Theory's vocabulary, the recursive suppression that institutional embodiment requires.
Dwelling stabilizes institutions through the spatial channeling of activity: roads, corridors, pavements, lanes, rooms, buildings, cities, borders. These spatial arrangements do not merely accommodate institutional activity. They are constitutive of it. The road is not a physical substrate on which the traffic system operates. The road is an institutional mediation: a material arrangement that channels human movement into patterns sufficiently regular for coordinated driving to be possible at all. Without roads, the traffic institution does not merely face difficult conditions. It ceases to exist as a coordinating architecture. The spatial organization of institutional activity is not background infrastructure. It is one of the five dimensions through which institutional compression is simultaneously achieved.
Multimateriality stabilizes institutions through the material instruments that carry institutional content across time and space: documents, clocks, servers, barriers, contracts, money, credentials, machines, weapons, measuring instruments. Each of these performs a specific compression function. Clocks convert the constitutively open negotiation of when to do what into a non-recursive fact that requires no interpersonal coordination to establish. Contracts convert the constitutively open interrecursive negotiation of future obligation into a quasi-non-recursive fixed point from which deviation requires institutional justification rather than simple assertion of preference. Traffic lights convert the interrecursive negotiation of intersection priority into a binary mechanical sequence that removes the recursive problem from the intersection entirely. The material instruments of institutional life are not the tools of self-reproducing communication systems. They are the compression devices through which recursive possibility-space is narrowed.
Multisymbolism stabilizes institutions through the symbolic frameworks that organize the interpretation and articulation of institutional activity: laws, rules, signs, credentials, categories, codes, standards, algorithms. These are what Luhmann's analysis most accurately describes. The legal code of legal and illegal, the economic code of payment and non-payment, the scientific code of true and false: these are L4 stabilization mechanisms that compress the interrecursive openness of activity within their domains into forms sufficiently regular for institutional processing. Luhmann was right that these symbolic frameworks are self-referential: legal decisions cite prior legal decisions; scientific publications respond to prior publications. He was wrong that this self-referentiality makes the symbolic frameworks themselves recursive beings. The self-referentiality is a property of the symbolic dimension of institutional compression. It is not evidence that the institution as a whole is an autopoietic organism.
Being-with stabilizes institutions through the relational frameworks within which institutional activity is conducted: roles, trust, reputation, hierarchy, expectation, tacit coordination. The trust that allows strangers to transact is not generated by the economy's autopoietic self-reproduction. It is generated by the accumulated experience of recursive beings who have repeatedly engaged in institutional encounters that proceeded as expected, and who have learned to extend anticipatory trust to new institutional encounters on the basis of this experience. When trust collapses, the institution does not draw on internal reserves to restore it. It waits for the recursive beings whose accumulated experience constitutes it to rebuild it through the slow accumulation of successful encounters.
The key distinction that the five-mediation analysis requires is the distinction between mechanical stability and metastable recursive stabilization. Mechanical stability belongs to genuinely non-recursive systems: the bridge that holds, the chemical compound that maintains its structure, the physical law that governs a repeatable process. Metastable recursive stabilization belongs to institutional architectures that have compressed interrecursive openness sufficiently for aggregate regularity to emerge. The bridge will hold whether or not anyone cares about it. The institution will function only as long as the recursive beings who constitute it continue to coordinate within its topology. Remove the recursive beings and the metastable stability instantly resolves into nothing.
VI. Traffic as the Paradigmatic Institution
Traffic is the paradigmatic institution precisely because it strips away the mystification that surrounds more prestigious institutional forms. Nobody imagines that roads exist for their own sake. Nobody proposes that the traffic system has interests that transcend the interests of the beings who use it. Nobody suggests that congestion is the traffic system reproducing itself in a higher mode of self-referential complexity. The starkness of the functional situation, recursive beings physically obstructing each other while attempting to move in the same finite space, makes the compression function of traffic institutions visible in a way that is less immediately obvious in law or science or the economy.
Consider lane discipline. The painted lines that divide a road into lanes are not communications within a self-reproducing traffic system. They are compression devices. They convert the constitutively open interrecursive negotiation of road positioning, who will move where in relation to whom, into a non-recursive fact. If everyone maintains their lane, the recursive problem of anticipating what other drivers will do is reduced to the manageable problem of anticipating whether other drivers will maintain the convention everyone is following. The convention does not eliminate uncertainty. It narrows the field of relevant uncertainty to a manageable range.
Stop signs are more radical: they eliminate the recursive problem of intersection priority entirely by substituting a binary mechanical rule. All drivers approaching the intersection stop. Then they proceed in the order of arrival. The interrecursive negotiation of who goes first, which in the absence of rules would require continuous mutual modeling of other drivers' intentions and speed and determination, is replaced by a simple sequential rule that requires no interrecursive modeling to apply. The institution achieves this by being, in its operational logic, anti-recursive: it suppresses the recursive problem rather than solving it.
The most illuminating feature of traffic as a paradigmatic institution is what happens when institutional stabilization succeeds completely. The best traffic signs stop functioning as explicit symbols. The stop sign, properly internalized by a skilled driver, does not operate as a sign that the driver consciously interprets and then decides to obey. It operates as a trigger for an embodied response sequence that has been repeated so many times it has descended from L3 symbolic articulation through L4 habitual rule-following to L1 smooth embodied coordination. The sign disappears into the mesocosm. The driver does not think: this is a stop sign, which means I should stop, therefore I will apply the brake. The driver stops. The success of institutional symbols is measured by how thoroughly they erase themselves in the act of functioning.
This self-erasure is the deepest feature of institutional compression, and it is what Luhmann's framework is constitutively unable to describe. The communication system that Luhmann analyzes is always L3 and L4: it always involves explicit coding, explicit binary distinction, explicit self-referential communication. But the traffic institution's deepest achievement is the elimination of explicit communication from the coordination problem it addresses. The driver who needs to consciously decode the stop sign is a driver who is not yet fully institutionalized. The fully institutionalized driver has moved the relevant behavior to L1, where it no longer needs symbolic articulation to operate.
The contrast between European compressed traffic systems and highly interrecursive systems like those of Kolkata reveals the degree to which different civilizations distribute recursive burdens across different mediations. In a highly compressed traffic system, the institutional architecture carries the coordination burden: lane discipline, signals, right-of-way rules, standardized road design, and traffic light timing together generate sufficient compression that individual drivers need to carry only a modest interrecursive burden. In a highly interrecursive traffic system, the institutional architecture carries far less of the coordination burden, and individual drivers carry far more. The result is not chaos but a different mode of coordination: continuous embodied negotiation through micro-signals, horn patterns, speed modulation, gap assessment, and the fine-grained mutual attunement of drivers who have developed extraordinary interrecursive competence within a less compressed institutional field.
Neither system is simply better or worse. They represent different distributions of the recursive stabilization burden across mediations. The compressed European traffic system externalizes coordination into infrastructure, symbolic rules, and material organization. The interrecursive Kolkata traffic system externalizes coordination into the embodied and relational competence of drivers who have learned to navigate constitutively open interrecursive dynamics in real time. Everyone honks all the time, so that others know that they are there. Both achieve coordination at scale. Both are institutional achievements. What differs is the mediational distribution of the compression work, and therefore the form of competence that each system demands from its participants. Kolkata traffic is so extremely interrecursive that I would never dare to drive a car there myself. In fact most private car owners employ a driver because they don’t dare to drive either. I wouldn’t last five minutes without an accident.
VII. Why Durkheim Misunderstood Suicide Statistics
In Suicide, Emile Durkheim observed a phenomenon that has never ceased to fascinate students of social life: aggregate suicide rates remain remarkably stable over time within a given society, vary characteristically across different societies, and correlate with institutional variables like religious affiliation, marital status, and economic integration, despite the fact that any individual suicide is a deeply personal, existentially constituted, and radically unpredictable event. How can the aggregate be so stable when the constituents are so volatile?
Durkheim's answer was to treat the aggregate stability as evidence of a social force operating at the aggregate level: society exercises coercive force on individuals through the mechanisms of moral integration and regulation, and variations in this force generate predictable variations in suicide rates. The social fact of the suicide rate is the expression of a collective moral state that constrains individual behavior at the aggregate level. This was an important insight, and the empirical correlations Durkheim documented remain impressive more than a century later. But the mechanism he proposed for producing the stability was wrong in a way that matters for everything that follows.
The error becomes visible through a comparison with genuinely stochastic phenomena. When a fair coin is tossed a million times, approximately half the outcomes will be heads. This aggregate regularity is produced by the non-recursive character of the individual events: each coin toss is entirely independent of every other coin toss. The coin has no memory, no anticipation, no social embedding, no recursive responsiveness to its own history or to the behavior of other coins. The statistical regularity of the aggregate is literally and entirely generated by the properties of the underlying individual events, which are genuinely stochastic, genuinely independent, and genuinely non-recursive.
A suicide is none of these things. An individual suicide is recursively constituted all the way down. The person considering suicide is modeling their social world: the responses of others to their distress, the perceived availability of support, the anticipated consequences of their action, the social meanings attached to self-destruction in their cultural context, the institutional classifications that would be applied to their death. They are responding to a field of recursive relations that includes family dynamics, institutional encounters with psychiatric or welfare systems, economic pressures, religious frameworks of meaning, and the accumulated experience of all the interrecursive encounters that have constituted their particular form of social embedding. The individual suicide is an interrecursive event of extraordinary depth.
And yet aggregate suicide rates are stable. The stability is not produced by the same mechanism that produces the stability of coin-toss aggregates. It is produced by institutional compression. The kinship structures, religious frameworks, labor systems, psychiatric classifications, media representations, and bureaucratic procedures that constitute the institutional field of any given society compress the interrecursive possibility-space within which individual lives are constituted. This compression does not determine individual outcomes. It constrains the topology of possible outcomes. The stability of the aggregate reflects the stability of the compression architecture, not the statistical independence of the underlying events. This is the missing mechanism that Durkheim approached but could not name.
VIII. Against Complexity Reduction
Luhmann's central claim about what institutions do is that they reduce complexity. The world presents an overwhelming surplus of possibilities, and social systems reduce this surplus by establishing stable patterns of selection that allow coordination to proceed without eliminating contingency but managing its local consequences. Luhmann's concept of complexity is technically specific: it is a combination of selectivity, contingency, and self-constitution of a system through its own selections. It is not simply informational overload but a precise characterization of the ratio between possible and realized states in a system's operation.
Living Value Theory's counter-claim is that institutions compress interrecursive volatility, which is a different problem requiring a different mechanism and generating different consequences. Luhmann's complexity, however precisely defined within his framework, is ultimately a property of the world as experienced from the standpoint of a symbolic system attempting to select from a surplus of possibilities. It is the perspective of a communication system observing its environment and finding more there than it can process. This is a real experience of L4 symbolic systems. It is not the ontological ground of social life. Most social coordination does not occur at L4. It occurs at L1, the level of seamless practical engagement where the most valuable coordinations are the least visible because they have no need to become objects of attention or selection.
Interrecursivity is specific in a way that Luhmann's complexity concept is not. A weather system may be enormously complex by any measure while remaining ontologically non-recursive: it does not respond to being modeled, does not anticipate the anticipations of others, does not adjust its behavior in light of predictions made about it. Interrecursivity, by contrast, is the mutual recursive orientation of beings who model each other's modeling, generating instability that has nothing to do with informational surplus. The corridor dance is informationally trivial and interrecursively destabilizing. A suicidal crisis is not complex in the same way a weather system is complex. It is interrecursively constituted, embedded in relational fields of mutual modeling, and compressed by institutional architectures in ways that bear no resemblance to the information-processing complexity that Luhmann's framework addresses. Computational intractability and recursive openness are different ontological conditions, and the failure to distinguish them is the failure that runs through systems theory from its foundations.
IX. Civilization and the Redistribution of Recursive Burden
Norbert Elias documented, with formidable historical scholarship, the progressive transformation of European social life between the medieval period and early modernity. The use of shared eating utensils gave way to individual forks and knives. Bodily functions that had been performed openly in social spaces were progressively relegated to private spaces and then to shameful concealment. Violence that had been a normal feature of public social life was progressively monopolized by the state and removed from everyday interpersonal relations. Emotional volatility that had been openly expressed was replaced by affect restraint, deferred gratification, and the inhibition of spontaneous recursive response.
Elias called this the civilizing process and located its primary mechanism in the progressive lengthening of interdependency chains: as social life became more complex and more interdependent, individuals required greater capacity to anticipate the effects of their actions on others at greater temporal and spatial distance, which required greater affective restraint, greater embodied discipline, and greater capacity for delayed recursive response. The civilizing process was, for Elias, fundamentally a process of embodied self-discipline, the internalization of external constraints into internal psychic and somatic structures.
This account is accurate in its description but incomplete in its mechanism. Living Value Theory's correction is that the civilizing process is better understood as the progressive redistribution of recursive stabilization burdens across the five mediations. Medieval European social life carried much of its recursive stabilization burden through embodied and relational mediations: personal honor, physical violence, direct interpersonal negotiation, and the immediate recursive encounters of face-to-face community life. Modern European social life progressively transferred this burden to multimaterial and multisymbolic mediations: infrastructure, bureaucracy, law, technology, architecture, and the vast institutional apparatus of the modern state.
This transfer explains features of modernity that Elias's account cannot adequately accommodate. The most heavily institutionalized modern societies exhibit not only greater affect restraint in formal institutional contexts but also considerably greater informality and spontaneity in informal ones. The Swedish office worker who maintains absolute procedural correctness in bureaucratic encounters and then performs elaborate uninhibited celebrations at the midsummer party is not alternating between civilized and uncivilized modes. They are distributing their recursive burden appropriately across contexts that make different mediational demands. The institutional infrastructure carries the recursive stabilization burden in the work context, liberating embodied expressiveness in the social context.
The civilizational trajectory can therefore be described as the historical expansion of multimediated recursive compression. As interrecursive density increased through population growth, urbanization, technological development, and the integration of previously separate social worlds, societies required increasingly elaborate stabilization architectures to prevent coordination collapse. States emerged. Bureaucracies expanded. Roads extended. Writing systems preserved institutional memory. Money stabilized exchange. Law stabilized obligation. Clocks synchronized activity. Algorithms increasingly stabilize behavioral pathways that explicit institutional rules cannot reach. Each of these developments represents the transfer of recursive stabilization burden from embodied and relational mediations, where it was being carried at increasing cost, to multimaterial and multisymbolic mediations that could carry it more efficiently at scale.
The limit of this transfer is also visible: the more thoroughly recursive stabilization is externalized into multimaterial and multisymbolic infrastructure, the more vulnerable the resulting institutional architecture becomes to infrastructure failure. Medieval European social life was resilient against infrastructure collapse because it carried most of its recursive burden through embodied and relational mediations that did not depend on electricity, supply chains, or server uptime. Modern European social life is extraordinarily vulnerable to infrastructure collapse because it has transferred most of its recursive burden to infrastructural mediations whose failure cascades instantly through the institutional architecture that depends on them.
X. Conclusion: The Traffic System Does Not Traffic
Return to the lost manuscript. The joke is now fully intelligible. Der Verkehr der Gesellschaft fails because it misidentifies what the traffic system is. The traffic system is not an autopoietic communication system that reproduces itself through the medium of vehicular communication. It is a multimediated compression architecture that constrains the interrecursive volatility of recursive beings attempting to navigate shared space, sufficiently for large-scale coordinated movement to become metastably possible.
The economy does not exist to reproduce payment. It exists because recursive beings need to coordinate the exchange of goods and services across distances and timescales that exceed the range of direct embodied reciprocity, and because the interrecursive volatility of any attempt to do this without institutional stabilization would preclude the densities of exchange that complex social life requires. Law does not exist to reproduce legality. It exists because recursive beings in conflict require a mechanism that converts constitutively open interrecursive dispute into a procedurally governed sequence with recognized authority and accepted outcomes, sufficiently stabilized that coordination can continue while the dispute is being resolved. Science does not exist to reproduce true/false distinctions. It exists because recursive beings investigating a shared world need to coordinate their claims and methods across institutional time and space in ways that prevent recursive disagreement from preventing the accumulation of reliable knowledge.
In each case, the institution is a multimediated compression architecture that metabolizes interrecursive unpredictability into a form sufficiently regular for aggregate coordination to appear system-like. The appearance of autonomous self-reproduction is precisely the appearance produced by successful compression. When the traffic light changes smoothly, when legal decisions are made and accepted, when payments clear and economic activity proceeds, the recursive beings who constitute the institution become invisible in the smooth operation of the compressed result. It looks as if the institution is running itself. Luhmann observed this appearance, took it for ontological reality, and built a theoretical system on the mistake. The mistake was not naive: he made a principled decision to extend the concept of autopoiesis to communicative reproduction on the grounds of formal structural analogy. But the extension was not warranted, because the material metabolic properties that make biological autopoiesis self-sustaining are precisely what social institutions lack, and without them the appearance of self-reproduction is exactly that: an appearance sustained by the ongoing mediational labor of the recursive beings beneath it.
The pavement oscillation that allegedly ended the fictional manuscript was the moment in which the mistake became unavoidable. The traffic system had no resources for solving the pavement problem because the pavement problem was not generated by complexity requiring reduction. It was generated by interrecursive openness requiring compression. And the compression required was not the establishment of a new level of autopoietic communication. It was the small recursive surrender of one person laughing and stepping aside, allowing the other to pass. Recursive beings, not recursive systems, solved the problem. Recursive beings always do.
The machine never became a machine. It only became good enough at compressing recursion that it temporarily appeared to be one. This is the deepest correction that Living Value Theory offers to systems theory, and it is the correction from which everything else follows. Institutions are not recursive beings. They are multimediated compression architectures that partially suppress recursive openness among recursive beings sufficiently for aggregate coordination to appear system-like. The moment the recursive beings stop carrying the mediational burden, the system reveals itself to have been, all along, an architecture rather than an organism, a topology rather than a being, a compression rather than a life.
References
Durkheim, E. 1997 [1897]. Suicide: A Study in Sociology. Trans. Spaulding, J.A. and Simpson, G. New York: Free Press.
Ecks, S. 2022. Living Worth: Value and Values in Global Pharmaceutical Markets. Durham: Duke University Press.
Elias, N. 1994 [1939]. The Civilizing Process. Trans. Jephcott, E. Oxford: Blackwell.
Garfinkel, H. 1967. Studies in Ethnomethodology. Englewood Cliffs: Prentice-Hall.
Giddens, A. 1984. The Constitution of Society. Cambridge: Polity Press.
Goffman, E. 1963. Behavior in Public Places: Notes on the Social Organization of Gatherings. New York: Free Press.
Luhmann, N. 1984. Soziale Systeme: Grundriss einer allgemeinen Theorie. Frankfurt: Suhrkamp.
Luhmann, N. 1995 [1984]. Social Systems. Trans. Bednarz, J. and Baecker, D. Stanford: Stanford University Press.
Luhmann, N. 1997. Die Gesellschaft der Gesellschaft. Frankfurt: Suhrkamp.
Maturana, H. and Varela, F. 1980. Autopoiesis and Cognition: The Realization of the Living. Boston: Reidel.
Merleau-Ponty, M. 1962. Phenomenology of Perception. Trans. Smith, C. London: Routledge.
Parsons, T. 1951. The Social System. New York: Free Press.
Rogowski, R. (ed.). 2023. The Anthem Companion to Niklas Luhmann. London: Anthem Press.
Scott, J.C. 1998. Seeing Like a State. New Haven: Yale University Press.
Turner, V. 1969. The Ritual Process: Structure and Anti-Structure. Chicago: Aldine.
Varela, F., Thompson, E. and Rosch, E. 1991. The Embodied Mind. Cambridge, MA: MIT Press.