The Cognitive and Scientific Life

A civilization thinks with the minds of its people, and minds can be strip-mined as surely as forests. This chapter is about the two axes that treat a population's capacity to attend, to reason, and to discover as assets a state must sense - and may deplete or engineer only at its peril.

The quietest extraction in history

It is 11:40 on a weekday night, and a man of twenty-six is watching his ninth consecutive short video. He is not enjoying himself, exactly; if you asked him what he wants to be doing, he would say reading, or sleeping, or almost anything else, and he would mean it. No single video harmed him. Each one, taken alone, was mildly diverting - a small, freely chosen transaction between a willing viewer and a platform that asked nothing but his attention. Yet across three years of such evenings something real has been taken from him: the capacity for sustained attention he had at nineteen, the patience for a long book, a measure of his sleep, and some harder-to-name portion of his ability to direct his own mind. Nothing was stolen. Everything was consented to. And a sensor that reads only money recorded the whole process as growth.

Book I established that financial success is not civilizational success - that a platform can add trillions in market value while degrading the very capacities its users think with. The cognitive axis is where that insight becomes an instrument. It treats a population's attention, its reasoning capacity, and its mental health as what they are: civilizational assets, the substrate on which every other axis depends. An economy is only as good as the judgment of the people running it; a democracy is only as good as the minds doing the voting; and both can be quietly liquidated while every conventional indicator rises.

Attention as a depletable commons

The doctrine's analysis of this asset borrows its physics from Donella Meadows. Attention is a renewable stock, but renewable in the demanding sense: it regenerates - through sleep, through rest, through undemanded time - and it can therefore support extraction indefinitely, but only at a rate no faster than its regeneration. Harvested beyond that rate, a renewable stock does not gracefully decline; it falls below a critical threshold and behaves, for practical purposes, like a nonrenewable one. A fishery collapses. A soil exhausts. An attentional commons - the shared capacity of a population to concentrate, deliberate, and think in stretches longer than a swipe - can do the same. And Meadows adds the warning that matters most in an age of machine persuasion: for a renewable stock, a gain in extraction efficiency is high leverage in the wrong direction. Better sonar destroyed fisheries that centuries of worse sonar had spared. An AI-driven leap in the efficiency of harvesting human attention is, on this analysis, not an innovation to celebrate but a trigger to tighten constraint - the moment the trawler arrives in waters that had only ever known rowboats.

Polanyi supplies the complementary premise. Attention was not produced for sale. Like land and labour, it exists prior to any market and for other purposes; to treat it as an ordinary commodity, priced and cleared and consumed, is the same fiction that once commodified the soil, and it licenses the same slow devastation. Book I drew the consequence on the property side - data, compute, and attention as a commons whose rents belong to everyone. This chapter draws it on the human side: the polity senses the stock itself. The cognitive axis reads, in aggregate, whether a population's attention is regenerating or being mined; whether its reasoning capacity - numeracy, argument, the ability to follow a chain of inference - is holding; whether its mental health, which Aristotle already understood as a public capability rather than a private luxury, is a rising or a falling curve. As everywhere in the doctrine, the sensor is aggregate and statistical. It reads flows and distributions across cohorts; it does not read your mind, your feed, or your diagnosis. There is no register of individual attention spans, for the same reason there is no register of souls.

The harm no single act commits

What makes cognitive harm politically difficult is its shape. The doctrine's anti-value taxonomy classifies it precisely: systemic, not edge; long-lagged, not immediate. No individual video addicts anyone - inspect any single session and it reads as neutral or mildly positive, a willing exchange. That is the trap. The harm exists only in the pattern across thousands of flows and many months: measured erosion of a cohort's attention, dependency where there was choice, and that most telling of early signals, the widening gap between what people say they value and where their hours actually go. The vote-vs-spend gap is the sensor's first tell that a demand field has been engineered against the demanders.

Because the harm lives in the pattern, so does the accounting. When citizens designate attention erosion and engineered dependency as anti-value - and this designation is theirs to make through the weighting vote, not a ministry's to assume - the measured systemic effect is attributed back across the platforms whose flows produced it, in proportion to their contribution, with a heavier share for any actor who concealed what its optimization was doing. Because the lag is long, enforcement is retrospective: revaluation and clawback, the machinery Book I built for slow harms, reaching back to profits that were booked as value before the pattern surfaced. And the correction, when it comes, takes the form the whole doctrine mandates. The responsible organ reshapes the demand field - prices the harm, changes the incentive gradient under which engagement is engineered - and does not ban the videos, license the platforms' content, or ration anyone's evenings. The state corrects the conditions under which the man of twenty-six scrolls. What he does at 11:40 remains, unnegotiably, his own business.

The scientific axis: what the market underprices

If the cognitive axis guards the substrate of thought, the scientific axis values its highest product. Discovery is the paradigm case of value the market structurally underprices. A validated result spills into every sector that touches it, compounding across decades and borders, while the price mechanism returns to its discoverer a sliver of that value or nothing at all - which is why the market, left alone, buys systematically less discovery than a civilization is worth. Book I's principle that the earned is left alone while the unearned is captured has a bright counterpart here: where the market underpays a positive spillover as systematically as it underprices a negative one, the commons that captures unearned rent also funds the science the price system cannot see.

The scientific axis therefore books discovery as first-class created value - and, in one of the doctrine's most deliberate corrections to the legacy incentive landscape, it books replication the same way. The independent replication of a load-bearing result counts as value comparable to the original, because the polity's exposure is to the reliability of the claims its decisions rest on, and the act that independently confirms or destroys such a claim is exactly as valuable as the act that first asserted it. A failed replication that removes a false result from circulation is not failure; refutation is production. The novelty-only incentive that hollowed out twentieth-century science - a million untested assertions chasing priority - is dead by construction. So is its bureaucratic shadow: citation counts, publication tallies, and venue prestige may circulate as diagnostics, but nothing in the polity may fund, appoint, or charge on a bibliometric proxy alone. Fraud and fabrication, meanwhile, are treated as what they are - concealment, carrying the same clawback as any concealed anti-value - while honest error, contested findings, and good-faith retraction are the system working, and are never punished. A science that punishes honest error gets dishonest science; the designers knew it.

Missions: steered by corridors, not by plans

How does a polity that has renounced central planning pursue aims as large as a science program, a demographic renewal, an ecological restoration? Through the instrument the doctrine calls a mission: a chartered, registered parcel of strategic demand. A mission pays for verified outcomes - milestone contracts, prizes, advance purchase commitments - and pulls production toward an under-served axis by being a large, honest customer. It never commands. It cannot direct a producer's output, set a price, compel participation, or requisition capacity; it buys under ordinary market rules, capped so that no mission ever becomes the monopsonist that dictates a sector by weight of purse. The market keeps clearing; the state simply shows up on the demand side of it, on behalf of value the demand side otherwise fails to express. Nor may a funder buy conclusions: a grant purchases effort and capacity against a question, and conditioning payment on a result's direction - or paying for a result's suppression - is concealment by the funder, not science policy.

Steering, in this design, is done by corridors rather than plans. The polity does not decide which theory should win or which laboratory deserves to exist; it holds each axis within a band - floors it will not let capability fall beneath, ceilings that check runaway concentration, including any single funder's grip on a field - and lets a plural portfolio of researchers, contrarians, and replicators contend inside the corridor. Every mission is born from a measured imbalance, carries falsifiable success criteria and kill criteria from its first day, and dies on schedule or by its own stated tests - never quietly living forever as a budget line, which is how mission apparatus becomes a planning state through the back door. Floors outrank missions absolutely: no moonshot may draw down the dividend or the real-capability floor, and no projected future glory licenses a present breach. And a failed mission pays its epistemic debt in full - its results, negative ones included, are booked as knowledge, because a civilization that hides its failures buys them twice.

What minds are for

Behind both axes stands the question the rupture forces. Machine cognition now performs a growing share of what used to be the distinctive work of human minds - calculating, drafting, diagnosing, optimizing, and increasingly discovering. If machines think for their living, what are human minds for? The industrial answer - minds are for employment - dies with the wage. Two other answers are on offer, and the doctrine rejects them both: minds as consumers of engineered content, which is the strip-mine with better manners; and minds as objects of state cultivation toward an approved excellence, which is the garden with worse ones.

The doctrine's own answer is structural, and it is already built into the constitution described in these chapters. Machines may produce; only citizens confer worth. The weighting vote, the designation of anti-value, the judgment of what a good life and a good society are - these cannot be delegated to the machinery, not because the machinery is stupid but because the question is not one of computation. Value, in this system, flows from the subjective judgments of persons, made objective-for-governance only by their collective vote. Every advance in machine cognition therefore raises rather than lowers the demand on human cognition: the fewer things citizens must think about to survive, the more everything depends on how well they think about what matters. Aristotle held that the user of a thing, not its maker, is its proper judge. In a civilization whose maker is increasingly the machine, the human vocation migrates to exactly the capacities the two axes of this chapter protect: attention, judgment, and the unforced curiosity that turns leisure into discovery. Freed minds are for valuing, for judging, and for wanting to know - the three activities no machine can perform on our behalf without ending the experiment.

Between neglect and engineering

A state equipped with a cognitive axis faces two symmetrical temptations, and the doctrine's defences must be stated against both. The first is neglect: a state that reads its dashboards while the attentional commons collapses, treating cognition as a private matter until the deliberative capacity its own legitimacy requires has been mined out from under it. Against this stands everything above - the axis itself, the anti-value designation, the retrospective reach of the clawback.

The second temptation is darker precisely because it is high-minded: a state that, having measured minds, sets out to improve them - optimizing attention like a crop, prescribing cognitive hygiene, engineering its citizens toward some certified mental excellence. Here the doctrine's oldest walls hold. The state never claims to know the Good, and never polices thoughts. Sight is not reach: the ability to measure an axis, however finely, licenses correction of conditions and nothing more - and cognition is the axis where that rule matters most, because a mind is the one territory where even a benevolent occupation is a conquest. So the state prices the strip-mine and funds the library; it guarantees the regenerative conditions - the floor of time, the unharvested hours, the plural sciences - and lets minds do with them what minds will. Some of what they do will be idle, eccentric, or wasted, and the unsteered zone exists partly so that no one ever has to prove otherwise. It corrects the frame; it does not direct your life - and nowhere is that sentence more literal than here, where the frame is everything around a mind, and the life is the thinking itself.

In the doctrine

The aggregate sensing behind the cognitive axis is bounded in Measurement and Correct, Not Direct; the commons analysis of attention draws on the lineage pages for Meadows and Polanyi, and the slow-harm accounting on Book I's Markets Kept Honest.