The Grammar of Collapse: Why We Mistake Language for Physics (And AI for Salvation)

If you spend enough time trying to explain the thermodynamic limits of the global economy to highly educated people, you will eventually hit a cognitive brick wall.

You will point to the depleting net exergy of the US shale patch, the collapse of global aquifers, and the physical mechanism of the Eurodollar vacuum. In response, a standard Neoliberal economist will look you dead in the eye and say something to the effect of: “The Second Law of Thermodynamics is formidable, but it has yet to pass a national budget.” (Thanks to my commenter, ‘The Gadfly Doctrine’.)

This is not a joke. This is an actual sentiment, variations of which appear in my comment sections with alarming regularity. When confronted with the absolute, physical boundaries of a dissipative structure, the mainstream mind retreats into a bizarre, anthropocentric hallucination: it assumes that human legislation outranks the laws of physics.

To understand why this delusion is so deeply ingrained, we cannot just look at economics. We have to look at the evolutionary neurology of the human brain.

The User Interface of ‘Universal Grammar’

In the mid-20th century, Noam Chomsky proposed that humans possess a ‘Universal Grammar’—a hardwired neurological structure dedicated to linguistic cognition. From an evolutionary standpoint, this was our species’ superpower. The ability to construct complex syntax, form abstract narratives, and communicate rationalisations allowed early humans to coordinate and build massive social structures.

It is a compressive symbolic layer sitting atop a biophysical substrate.

Compression discards detail. It simplifies complexity into archetypes, categories, and narratives. That simplification is adaptive for social coordination — but it is not selected for high-fidelity modelling of nonlinear thermodynamic systems.

Language evolved to stabilise groups, not to calculate planetary energy descent.

Language is, fundamentally, just a User Interface (UI). The fatal flaw of the Strong Enlightenment tradition—the philosophical bedrock of both Neoliberalism and modern state-planning—is that it became entirely hypnotised by its own UI. Because our brains are so incredibly adept at forming narratives, we began to believe that the symbols we created actually governed the physics of the world.

We invented words like ‘Fiat’, ‘Debt’, ‘Yield’, and ‘Quantitative Easing’. Because these concepts make perfect grammatical sense within our neurological language centres, we assume they must be physically real. We mistake the map for the territory.

Neoliberal economics is not a physical science. It is a branch of linguistics. It is the study of how humans use specialised vocabulary to rationalise the temporary dissipation of highly concentrated fossil fuels. When economists talk about ‘compressing demand’ or ‘repricing the supply chain to achieve a soft landing’, they are just using grammar to politely describe the brutal, physical rationing of energy. They genuinely believe that if they can construct a grammatically correct sentence, the physical universe is obligated to obey it.

The Consistency Trap: How to Spot a Linguistic Thinker

To understand why this hallucination persists, we must recognise a fundamental divide in dominant human cognition: the difference between Linguistic Thinkers and Relational Thinkers.

Human cognition is modular and context-sensitive. Different processing architectures dominate in different environments.

In materially constrained environments — geology, farming, field engineering — relational constraint tracking dominates. Feedback is tight. Error is punished immediately.

In institutional and policy environments, linguistic coherence dominates. Narrative stability and propositional consistency are rewarded.

Most of humanity operates relationally, most of the time. They respond to immediate physical environments, social cues, embodied experience, and shifting contexts. For the Relational Thinker, language is not the medium of thought; it is merely the post-hoc report. The narrative is generated after the physical or social reality has been processed.

Those trained in propositional reasoning are conditioned to treat cross-domain coherence as a marker of intelligence. Incoherence feels like failure.

But thermodynamic systems are nonlinear, phase-shifting, and often domain-incoherent. Geological reality does not owe ideological symmetry to macroeconomic models.

The problem is not contradiction.

It is compartmentalisation.

One can observe a petroleum geologist such as Art Berman speak with relational clarity about depletion curves — and then shift seamlessly into policy language that abstracts away those same constraints. This is not hypocrisy. It is cognitive mode-switching under institutional context.

Linguistic training overgeneralises the demand for coherence across domains that operate under fundamentally different constraint regimes.

A small minority of the population—the academics, the policymakers, the economists, the lawyers, and the pundit class—are Linguistic Thinkers. They construct internal models of the world using propositional logic and vocabulary. For them, the UI is the world.

This creates what we might call the ‘Consistency Trap’. We are trained to view ideological or intellectual consistency as truth. But in a biophysical world defined by chaotic thermodynamics and nonlinear depletion, rigid consistency is actually a marker of cognitive bias. It proves the individual is captive to the UI.

Linguistic Thinkers will fiercely defend the coherence of their economic models across time and domains because their linguistic brains demand it, entirely ignoring the physical reality breaking down outside their window. They suffer from an intellectual compartmental fracture: a geologist might understand the relational, physical limits of oil extraction in the field, but the moment he sits down to write a policy paper, he switches into linguistic mode and generates a beautifully coherent, grammatically flawless narrative about infinite ‘green growth’ that has absolutely no anchor in physical reality.

The Neurodivergent Observer

Why is it so difficult for the ruling class to see this fracture? Because for the neurotypical majority, the linguistic interface feels like ‘reality’. Hallucination is invisible.

It often requires a neurodivergent mind to perceive the UI for what it actually is.

For many neurodivergent individuals, language processing is not an automatic, invisible transparency; it is a distinct, systematic, and often effortful layer of computation. When you experience the construction of linguistic output as a mechanical process, you are inherently less likely to mistake fluency for truth. You can see the pixels on the screen.

When a Neoliberal economist smoothly explains how a 10% tariff will magically optimise the global supply chain without acknowledging the thermodynamic drag, the neurodivergent observer does not marvel at the logical coherence of the sentence. They recognise it as a highly optimised post-rationalisation. Fluency is no guarantee of truth.

Physics does not care about consensus. It does not care about narrative coherence. For a mind that has learned to distrust the automatic social and linguistic wiring of the crowd, the cold, non-negotiable laws of thermodynamics are a relief. They are the one thing that isn’t just another social performance.

Physics Is Not Grammar

Mathematics and physics are not the same category as social-symbolic language.

Physics does not tell us how reality ‘works’ in a mythic sense. It constrains what is possible.

The Second Law of Thermodynamics does not narrate civilisation. It bounds it. It eliminates impossibilities. It prunes the phase space.

Physics is abstraction under experimental duress. Social-symbolic systems, by contrast, are abstraction under social reinforcement. The difference is feedback discipline.

When abstraction is disciplined by material constraint, error collapses quickly.

When abstraction is disciplined by consensus, error can accumulate for decades — especially under conditions of energy surplus.

Even professional scientists, trained to operate under constraint, are human. They can sometimes treat their discipline mythically, relying on consensus, prestige, or narrative coherence rather than rigorous testing. The laws themselves do not change, but the way humans perceive, communicate, and operationalise them is subject to the same cognitive dynamics as anyone else.

In short: physics constrains reality, but humans—including scientists—often translate those constraints into language, stories, or policy frameworks. Hallucination can appear anywhere linguistic fluency is rewarded over relational feedback. Neurodivergent observers are often more sensitive to these translation layers, noticing when symbolic coherence drifts away from physical possibility—even when the underlying science is sound.

Many contemporary policy frameworks are SGCs—Structurally Guaranteed Compromises (see my paper “The Path to the Singularity”). They are internally coherent, grammatically elegant, and rhetorically persuasive. But they are physically impossible under current energetic and material constraints. Net Zero by 2050, perpetual green growth, full electrification without grid transformation, and global-scale carbon capture are all examples: flawless linguistic constructs, doomed by thermodynamics.

Surplus, Hallucination, and Collapse

Energy surplus masks compression error. When net exergy is abundant, symbolic simplifications can float above material reality. Slack absorbs misallocation. Inefficiency is subsidised. Under contraction, slack disappears. Maintenance power (P_maint​) rises. Constraint signals sharpen.

When surplus declines sufficiently, the system does not debate correction.

It collapses.

Collapse is not primarily ideological or moral. It is mathematical. It occurs when maintenance power exceeds available surplus and structural complexity can no longer be sustained. Much like—in a structural, not literal sense—the collapse of a wave function in QM, the 'collapse' of a complex system is the moment where a cloud of probabilistic narratives is forced by physical interaction to resolve into a single, non-negotiable state. The maths is the logic of the constraint; the physics is the result. We are currently living in a linguistic superposition that thermodynamics is about to collapse.

This pruning of the possible entropic phase space—constrained by path dependency and systemic inertia—is exactly what I model with SETE.¹

At that moment, symbolic coherence loses authority. The substrate reasserts itself.

Hallucination thrives under surplus, invisible to most. Collapse enforces realignment.

Ecological Myth and Succession

Many cultures embed ecological constraint in mythology. This is most clearly retold in contemporary Western mythology as the Eden parable, but throughout the world there have been many cultures who managed to attain relatively sustainable practices.

It is unprovable whether all such cosmologies emerged after collapse phases. It is conceivable that foresight alone could produce restraint. But historically, collapse is a powerful feedback amplifier. It shortens time horizons. It eliminates slack. It makes constraint visible.

Ecological cosmologies may function as succession adaptations — stabilisation algorithms encoded in narrative form. The social transition from r-strategy to K-strategy.

Modern industrial mythology, by contrast, is an expansion-phase cosmology. It emerged under unprecedented surplus.

It assumes continuation because surplus has always been present.

The Ultimate Syntax: The AI Delusion

The irony of this cognitive blind spot is that we are currently repeating the exact same mistake, at a much grander scale, with Artificial Intelligence.

Silicon Valley and the global financial elite are currently betting trillions of dollars that Large Language Models (LLMs) will invent our way out of the Resource Entropy Singularity. They believe AI will ‘solve’ the energy crisis, optimise the grid, and decouple economic growth from biophysical constraints.

But what actually is an LLM?

It is the purest, most concentrated manifestation of Chomskyan Universal Grammar ever created. It is a highly advanced statistical engine designed to perfectly predict and replicate the structure of human language. It has mastered syntax, logic, and narrative.

Because we are neurologically hardwired to interpret linguistic fluency as ‘understanding’ and ‘consciousness’, we look at an LLM and see a god. We assume that because the machine can speak flawlessly, it must possess the power to alter reality. AI amplifies hallucination.

But AI is pure ‘software’. An LLM cannot code a new law of thermodynamics. It cannot generate a single joule of physical exergy. It can only rearrange the symbolic, linguistic data that humans have already generated—and as we have established, 99% of that data is infected with Neoliberal blindness. AI is just a mirror reflecting our own UI back at us.

The Thermodynamic Reality of the Algorithm

Far from saving us from the exergy cliff, AI is dramatically accelerating our sprint toward the edge.

The techno-optimists view AI as a source of infinite, ethereal growth. But biophysically, AI is the most aggressive, exergy-hungry dissipative structure we have ever built. The physical hardware required to run this linguistic software is staggering. Training these models and querying them requires gigawatt-scale data centres, millions of gallons of fresh water for thermal cooling, and massive amounts of entropically dispersed rare earth metals that must be mined using diesel fuel.

AI is increasing the system’s Maintenance Power at the exact historical moment that our global net surplus exergy is entering terminal decline.

We are effectively building a supercomputer on the deck of the Titanic to calculate the most grammatically elegant seating arrangement, while completely ignoring the iceberg of net energy depletion tearing open the hull.

The Limits of Language

A legal budget cannot negotiate with a depleted oil well. An algorithm cannot prompt a stripped copper mine into yielding higher-grade ore.

You can use Universal Grammar to write a brilliant, logical, internally consistent story about how an economy can grow forever without fuel. You can pass laws declaring it so. You can build an AI that writes a million beautiful essays agreeing with you.

But gravity does not speak English. And the laws of thermodynamics have never read your code.

We built an economy that mistook money for energy, and now we are building an AI that mistakes language for reality. When the Resource Entropy Singularity finally forces the reduction of our global complexity, the Relational Thinkers will adapt to the physical pain, while the Linguistic Thinkers will stare at their flawlessly consistent models in disbelief. It will be the ultimate proof that the linguistic layer can never override the physical one.

The vocabulary loses its magic the moment the power goes out: “Colourless green ideas sleep furiously.”

1. For a full breakdown of the SETE model, see the formal paper [here], or the introductory essay series [here]).