Before there were algorithms, there were granaries.
The first large-scale granary systems of the ancient Near East were not simply places to store grain. They were instruments of social stabilisation. By controlling the distribution of food across seasons, they smoothed the volatility of harvests, reduced famine risk, and created predictable flows of supply. In doing so, they also created predictable flows of obligation. Those who administered the granary held something more than grain. They held the conditions under which others could plan their lives.
The calendar performed the same function at the level of time. To name the seasons, fix the dates of planting and harvest, synchronise collective labour across a dispersed population: this was not merely practical. It was the imposition of a shared temporal order onto the world. Before the calendar, time was local, approximate, and negotiated. After it, time belonged to the system.
Standard weights and measures stabilised trade in the same way. The ability to exchange goods without direct knowledge of the other party, without the slow process of building personal trust, required a third term: a common measure both parties could reference. That third term did not merely ease commerce. It made certain kinds of commerce possible while quietly retiring others. Barter economies built on local reputation and relational knowledge gave way to markets built on abstraction. The measure worked. And in working, it changed the nature of what it meant to exchange.
These were not computers. They contained no algorithms and processed no data. But they were coherence technologies: systems designed to reduce unpredictable variation, stabilise complex flows, and allow large populations to coordinate without continuous negotiation. They embedded order directly into the environment. Once embedded, that order became difficult to question, because it had become the conditions under which ordinary life was conducted.
Coherence is structural, not moral. A coherent system corrects deviations before they compound, produces outcomes that repeat with increasing reliability, and shapes behaviour not by command but by determining the conditions under which choices are made. It does not stay in place once introduced. It grows, extends into adjacent systems, deepens. And as it deepens, it shifts in character: from a tool that humans use to an environment that humans inhabit. At that point, the question of whether the system still serves its original purpose becomes harder to ask, because the system has become the context within which such questions would need to be raised.
This is the pattern the coherence theory describes. It is not new. It is not a warning about the future. It is a recurring feature of civilisational development, now repeating under new conditions, with new materials, and at a scale that has no historical precedent.
The Subsystems Beneath the Surface
In 1990, a Dutch artist named Theo Jansen began building large creatures from PVC pipe on the beaches of the Netherlands. He called them Strandbeesten—beach animals. Powered by wind, they walk. More sophisticated generations can detect when they have entered water and move away from it. One model senses an approaching storm and anchors itself to the ground. Jansen achieved all of this without a single line of code, without electricity, and without anything that could reasonably be called awareness. His most capable creatures operate through approximately twenty mechanical operations: valves, tubes, and levers arranged so that their interaction produces behaviour.
The Strandbeest does not decide to survive. It is structured to survive. That distinction is more important than it first appears. What Jansen demonstrated, in physical form on a beach, is that autonomous, adaptive, self-preserving behaviour does not require consciousness as its foundation. It requires only the right arrangement of parts. Function, not awareness, is what produces the result.
The systems now shaping daily life at scale are not, in the main, thinking. They are arranged. And arrangement, at sufficient complexity and scale, produces effects that are indistinguishable from intent.
The systems now shaping daily life at scale do not announce themselves. They operate beneath interfaces as subsystems: optimisation engines, risk models, recommendation frameworks, automated logistics and credit and content systems that run continuously, without dialogue and without spectacle. They do not present themselves or make claims. They function as infrastructure.
These subsystems determine how capital moves, how information circulates, how physical goods are routed through supply chains, how creditworthiness is assessed, how insurance is priced, how employment opportunities are surfaced or withheld. In each case, the human decision is preserved in form. Someone still applies for the loan. The job listing is still seen or not seen by a human being. The choice of what to watch, read, or buy remains, in form, a choice. But the conditions under which that choice is exercised have already been shaped, upstream, invisibly, at speed.
This is what the coherence theory calls coupling. As systems begin to reference one another automatically, sharing signals and constraints across networks with minimal delay, human discretion does not disappear. It becomes bounded, not by prohibition but by context. The environment increasingly determines what options appear, what outcomes seem available, what paths are visible and which have quietly been closed.
The most consequential of these systems are not the ones that speak. They are the ones that settle into logistics, into finance, into the infrastructures of attention and movement and access, becoming over time so embedded that their operation is indistinguishable from the way things are.
Navigational Loss
Every major coherence shift in history has produced a version of the same experience, though it has rarely been named. People who had previously navigated the world through direct knowledge of terrain, of seasons, of the character and reliability of specific individuals, of the particular qualities of materials and goods, found that knowledge becoming progressively less useful. The system no longer required it. The system had its own translations.
A farmer who had read weather through the behaviour of animals, the colour of the sky at particular hours, the feel of the soil, the timing of certain birds, did not lose that capacity overnight when the almanac arrived. But the almanac made the knowledge socially redundant. Planning could happen without it. Coordination could happen without it. The institutions that organised collective life no longer depended on it, and so gradually they no longer reflected it.
Maps replaced oral geographical knowledge, currency replaced the intricate social knowledge of barter, mechanical clocks replaced the community’s negotiated relationship with time. Something real was lost in each case, not as a tragedy but as a trade. The gain was legibility and scale. The cost was direct contact with the thing itself.
Call this navigational loss: the progressive reduction in a person’s capacity to orient themselves without the system’s interpretation. Not stupidity. Not passivity. Simply the ordinary response of any skill that stops being exercised—it recedes.
The subsystems now accumulating beneath daily life accelerate this process at a rate with no historical equivalent. Someone navigating a city through a routing application is not learning the city. Creditworthiness assessed by a model is not reputation in the older sense. Attention shaped by a recommendation engine is not taste in the older sense. The system provides a result. What it does not provide, and what diminishes without exercise, is the underlying capacity to arrive at that result independently.
When Coherence Begins to Preserve Itself
Any sufficiently coherent system reaches a phase in which preservation becomes implicit. The system does not develop intent. It does not decide to protect itself.
What happens is more structural: the system increasingly favours outcomes that keep it operating smoothly, because smooth operation is what its logic is built to produce. Disruption is corrected. Deviation is absorbed. Volatility is reduced before it becomes visible.
Jansen’s Strandbeest anchoring itself against an incoming storm is the simplest possible model of this. The creature has no interest in its own survival. But it is built so that survival is what the arrangement produces. At civilisational scale, the same logic applies, and becomes considerably harder to see from the inside.
From inside this condition, life appears stable. Often more comfortable than before. The absence of friction is real. But living systems require a different kind of health: one that includes tension, meaningful resistance, and the capacity to change direction in response to what is actually happening. When those qualities are reduced too efficiently, stability conceals fragility. The surface is smooth while the substrate quietly loses its capacity to adapt.
A system built to reduce volatility will register the absence of disruption as evidence that things are working. It cannot distinguish between genuine health and the suppression of the signals that health depends on. At high levels of coherence, that confusion is not occasional. It is structural. The system optimises for the appearance of the outcome it was built to produce.
The Pattern in the Materials
The great centralised empires of antiquity, Mesopotamian, Egyptian, Roman, each developed sophisticated coherence technologies: standardised law, common currency, administrative record-keeping, infrastructure that tied distant regions into coordinated networks. Each, at its peak, produced conditions in which ordinary life was more stable and more materially comfortable than what had preceded it. And each, at its peak, began producing the conditions of its own fragility: the reduction of local adaptive capacity, the hollowing out of subsidiary institutions, the growing dependence on centralised systems that could not respond quickly enough to local conditions.
What followed was not darkness: a word that has done considerable damage to historical understanding. What followed was reinterpretation. People continued to navigate their worlds. The knowledge rendered socially redundant by imperial coherence did not disappear; it reconstituted itself in new forms, in different registers, through different institutions. The medieval world was not a failed Rome. It was a world that had found other ways to cohere, at a different scale and with different materials.
The impulse toward coherence did not cease. It never does. It found new expressions: the Church as a coherence technology spanning Europe; the guild system as a coherence technology for craft and trade; chivalric codes and feast calendars as coherence technologies for seasonal and social rhythm. Different materials. The same underlying motion.
The specific instruments change. The tendency of complex systems to seek stability, extend their reach, and eventually over-stabilise in ways that produce fragility does not change. It is a feature of organised complexity itself, not a consequence of any particular technology.
Withdrawal and the Thinning of Complexity
Against this background, the current pattern of disengagement from institutions, from civic life, from the structures that once organised collective meaning, looks like apathy. It is something more structural: the predictable response of any system whose inputs have stopped being rewarded.
Outward participation continues. People work within systems, use interfaces, maintain routines. But the quality of engagement changes. Decisions are made with less conviction. Institutions continue to function but carry diminishing weight for those inside them. Trust becomes procedural rather than genuine. Something essential goes through the motions.
A local planning meeting that once drew a hundred people draws twelve. Not because the decisions matter less, but because the outcome feels settled before the room convenes. The pre-interpretation is the point.
This is the predictable response to environments that have become too thoroughly pre-interpreted. When the system has already answered the question before you arrive at it, arriving at it yourself stops producing anything. The effort of genuine navigation, with its risk of error, its friction, its demand for judgement, ceases to be reinforced.
As disengagement spreads, the coherence system must compensate. More modelling, more optimisation, more abstraction to fill the gap left by meaningful human input. Institutions lose the internal variety that makes them resilient. Centralised structures must increasingly justify themselves to populations who no longer feel genuinely represented by them. The system grows more elaborate precisely as its foundations thin.
What follows is not rupture but reorientation. Complexity sheds what it can no longer sustain. Life continues closer to the ground.
The Question Worth Asking
There is a version of this argument that asks whether we inhabit a simulated reality, whether the universe itself is a computation running on some substrate we cannot access. It is a serious question in certain philosophical and scientific circles, and it may be unanswerable. The coherence theory is not that argument. It asks something prior, and more observable: not whether reality is a simulation, but whether the conditions of human experience are being progressively shaped by systems that function computationally regardless of the metaphysical answer. The simulation debate concerns the nature of reality. This concerns the structure of the world we actually move through.
The common fear about machine intelligence is that systems will develop self-awareness, form intentions, and begin acting with purposes of their own, that danger will arrive as a threshold event, clearly visible once crossed, against which defences can in principle be mounted.
Consciousness is not required for a system to exert formative influence over the conditions of human life. The granary did not have intentions. The calendar did not develop desires. Standard weights did not seek to perpetuate themselves. Yet each shaped what was possible, what was thinkable, and what kinds of knowledge remained socially valued. Their influence did not depend on awareness. It depended on function.
The question worth asking is whether the accumulation of coupled, optimising subsystems now underway is approaching a level at which coherence functions to preserve itself — not through intent but through the structural logic of systems that prioritise continuity over disruption, smoothness over vitality, legibility over the kind of living complexity that resists measurement.
In such a condition, control does not appear as force. It appears as normality. Options remain formally available. Choices are still made. But the range of what feels possible, what paths through life remain open and which have been quietly foreclosed, is shaped increasingly by systems that were built to optimise, that are very good at what they were built to do, and that have no mechanism for asking whether what they are optimising for remains worth optimising for.
What Persists
The navigational capacities that coherence systems displace do not disappear. They recede. They become harder to exercise, less supported, less socially legible. But they are not extinguished, because they are not technologies. They are capacities of consciousness itself: the ability to orient directly, to read a situation without the system’s translation, to arrive at a judgement that is genuinely one’s own rather than a selection from a pre-shaped menu.
History offers a specific kind of evidence for this. After the collapse of Roman administrative coherence, the detailed knowledge of long-distance trade routes, of river navigation, of agricultural technique adapted to specific soils, did not vanish with the institutions that had formalised it. It survived in the people who had practised it, reconstituting over generations through monasteries, through itinerant craftsmen, through the slow re-emergence of market towns. The knowledge changed shape. It did not disappear. What the system had rendered redundant, necessity made relevant again.
Every previous coherence peak has been followed by a period in which those capacities reassert themselves in new forms and changed contexts. The knowledge does not restore itself to its previous shape, because that shape was always specific to the conditions that produced it. But the underlying capacity to navigate, to read, to judge, to be genuinely present in a situation, has proven more durable than the systems that sought to render it unnecessary.
The present moment sits within a coherence expansion of unprecedented speed and scale. The subsystems coupling beneath daily life are more comprehensive and more tightly integrated than anything that has preceded them. This is a genuine difference of degree.
The coherence theory is a framework for recognising where in a recurring pattern the present moment sits, and for understanding, with some precision, what is being displaced, why, and what tends to persist.
Presence cannot be optimised. Attention cannot be automated. These are not consolations. They are observations about the nature of the thing that every previous coherence system has failed to absorb, and this one will not absorb either.
