THE CIRCLE AND THE FLAG

Two nuclear reactors are bound for the moon’s south pole. Neither is coming home.

The Outer Space Treaty forbids owning the moon by claim, by use or by occupation. It says nothing about a ring of ground no one may enter, or about how long such a ring may last. The United States wants its reactor running by 2030; Russia promises one for China by 2036.

Chess makes a distinction the Outer Space Treaty does not. A piece stands on one square and commands others. Games are decided by squares no piece ever occupies, and a player who sees only where the pieces stand will lose to one who sees where they reach.

In late August 2026, NASA asked contractors to prepare a nuclear reactor that could survive the voyage to the moon and run without maintenance near its south pole, the New York Times reported in October. The agency wants it ready to launch by December 2030. Russia, under a partnership with China, has set 2036 for a reactor of its own. The machines are modest. NASA’s would produce about 20 kilowatts, roughly what 16 American homes draw, and Russia’s up to half that. The ground is modest too. NASA named 13 candidate landing regions for its first crewed return in 2022 and narrowed them to nine in 2024, all on a handful of ridges and crater rims near the pole, where the sunlight is nearly constant and the shadowed craters beside them are thought to hold ice. That is the whole prize: a few named places on a body whose surface is larger than Africa.

A reactor at power cannot be approached safely without shielding. On Earth it sits inside a containment building. On the moon, the Times reports, it would more likely sit inside a no-go zone, a ring that neither astronauts nor machines may safely enter. The treaty that governs the moon, signed in 1967, forbids “national appropriation by claim of sovereignty, by means of use or occupation, or by any other means.” Sovereignty, use and occupation are three ways of standing somewhere. The text says nothing about denial. So a handful of the only usable sites on the moon can be made unenterable by acts that break no rule. The one instrument that provides for ending such a zone is not a treaty but a political declaration, signed by one of the two builders and not the other.

And neither reactor is coming home. When a reactor on Earth reaches the end of its life, decommissioning takes years. On the moon, according to Selam Gebrekidan’s reporting in the Times, both NASA and the Russian agencies say they would simply leave the radioactive material behind. The zone will not end when the machine does. In chess, remove a piece and every square it commanded is free at once. The moon is being offered a piece that keeps its squares.

I

Twenty Kilowatts

The reactor NASA wants is small, and nobody has calculated the load. The Times reports that for the first years both programmes need only enough power to keep equipment warm and charge rovers, that solar and radioisotope systems would serve, and that how much electricity a base would need is open to speculation. Two powers are racing to install a power source for a base that does not exist, against a demand nobody has measured. NASA’s machine, as the Times describes it, must work for five years with no intervention; the Russian design, called Selena, for a decade. Neither is a power station. Each is built to carry a base through the dark, and at the pole the dark comes in gaps of a day or two rather than weeks, under a sun that never climbs far above the horizon. That is why both programmes say a reactor must come.

The place is small too. Of the moon’s 37.9 million square kilometres, the parts that matter for a base are the ones where two things meet: ridges high enough to catch sunlight for most of the lunar day, and craters deep enough that sunlight has never reached their floors. The first supply power and bearable temperatures. The second, NASA says, “can preserve resources, including water,” from which oxygen and hydrogen can be extracted for life support and fuel. Such places are few, and they cluster at the south pole. In 2022 NASA named 13 candidate regions for its first crewed landing there, and in October 2024 narrowed them to nine, a short list of massifs, rims and plains whose names will become familiar. Those are landing regions, not reactor sites, which no one has announced. But they are the published evidence of which ground is worth having, and the Chinese-led International Lunar Research Station is aimed at the same pole.

A reactor set on that ground does not command a clean circle. In open space a radiation radius is a sphere. On a crater rim it is cast like a shadow across whatever lies in reach. Set behind a ridge, its reach falls on one slope and not the other. Set near a crater, where the approaches are few, it can cover them. Its zone will be shaped by the terrain it sits in, and so will the question of what the zone denies.

Oblique view of a lunar crater at the south pole: a small reactor with a crown of radiator fins stands on the rim, and a line of lit beacon pylons marks its exclusion zone, running along the rim, down the one gentle slope into the crater and on into the shadow of its floor.
A zone on ground follows the ground: the beacon line runs along the rim, drops down the one slope into the crater and continues into the dark where the floor begins. Generated with Gemini for this essay; it depicts no actual site.

Nobody wanted this yet. A fission reactor is a later need, brought forward. NASA moved its date up because Russia is building one. Russia is building one because China, in the one core task it appears to have delegated, asked it to. Nothing in the record says Moscow set its clock by Washington’s. So only one player is in zugzwang, the position in which the obligation to move is itself the harm, and it is the one bound by the Accords. Every move it has — an earlier launch, a bigger reactor, a looser rule — takes something from the position both sides depend on: a moon that no one owns. The treaty that guarantees it was written for players who stand on squares. The next move puts down a piece that reaches.

II

By Any Other Means

The treaty’s second article is one sentence long. Outer space, including the moon, “is not subject to national appropriation by claim of sovereignty, by means of use or occupation, or by any other means.” Appropriation is the term of art, and it means taking a thing as one’s own. Sovereignty, use and occupation are three ways of doing it: declaring, working, staying. Of the three, use is the widest, and it was put there to catch exploitation that stops short of a flag. A reactor is a use, and a specialist’s first answer will be that a reactor whose radius denies ground is appropriation by means of use. But a radius takes nothing. It makes ground nobody’s rather than somebody’s. The treaty forbids turning the commons into property. It says nothing about turning it into waste. Denial is not acquisition. No claim to any part of the moon has ever been recognized by any state, and no clause of the article mentions keeping others from a part of it.

The ninth article is where the radius enters, and it enters as a duty rather than a loophole. States must conduct their activities “with due regard to the corresponding interests of all other States Parties,” must avoid harmful contamination of celestial bodies, and, where an activity “would cause potentially harmful interference with activities of other States Parties,” must undertake “appropriate international consultations before proceeding.” The duty falls on the state that acts. The builder of a reactor owes due regard to everyone else’s interests, which on a straight reading is a constraint on where it may put the thing. The builder’s answer is that the zone is the due regard: a ring that keeps everyone else safe from what it has built. So one clause both demands the zone and forbids what the zone does, and nothing in the text says which reading governs. The two articles close on each other: obeying the ninth produces the condition the second forbids.

Eight years earlier the Antarctic Treaty had frozen the sector claims that seven nations had drawn across the map of a continent, and the two texts share four prohibitions almost word for word. Antarctica “shall be used for peaceful purposes only,” the earlier treaty says, and forbids “military bases and fortifications,” “military maneuvers” and “the testing of any types of weapons.” The moon treaty repeats all four. But the Antarctic text has a fifth article that the space text does not: “Any nuclear explosions in Antarctica and the disposal there of radioactive waste material shall be prohibited.” An American reactor, SNAP-10A, had already flown, in April 1965, two years before the space treaty was signed, and the treaty says nothing about a reactor stopping.

The law of the sea, drafted fifteen years later, shows what such a rule looks like. Article 60 is not a legal parallel. Its zones sit in waters where the coastal state already has sovereign rights, not in a commons. It is a drafting model, and it shows the two devices the moon’s treaty lacks: a zone with a stated maximum, 500 metres from the structure’s outer edge, and a duty to remove what is finished — installations “which are abandoned or disused shall be removed.” The ice treaty banned leaving the waste. The sea treaty bounded the zone and required removal. The moon’s treaty did neither, no instrument since has set a maximum for a lunar safety zone, and neither programme has published the radius it intends.

III

The Case for the Zone

The case for the reactor begins with the dark. Away from the poles a lunar night lasts about fourteen days, and the pole is not exempt so much as different. The best-lit ground there, by NASA’s own measurement, is lit up to 90 percent of the time, and nowhere on the moon is lit always. A 2010 NASA study found a site near Shackleton crater sunlit about 240 days a year, with its longest stretch of darkness about a day and a half. The sun never rises far above the horizon, so solar panels must stand on edge and are shadowed by the very terrain that makes the site worth having. The radioisotope generators that have powered spacecraft since the 1960s make watts, not kilowatts. An expanded base, the Times reports, would need far more power than those systems can give. On this reading the reactor is not an instrument of policy. It is the answer to a place that is lethally short of energy.

The safety case is nearly as strong. Both NASA and Russia say their reactors will be launched cold, never having been switched on, and the Times reports that nuclear engineers regard cold uranium fuel as posing little radioactive threat even if it tumbles to Earth. The fuel NASA specifies sits below weapons-usable enrichment. On the safety case as its advocates put it, a cold launch materially reduces the risk, and the reactor becomes dangerous only once it is running, on the moon, where no one lives.

The third pillar is the law itself. The Artemis Accords, the political declaration that 59 nations have signed, provide for “safety zones” around lunar operations. The Accords say a safety zone “should be the area in which nominal operations of a relevant activity or an anomalous event could reasonably cause harmful interference,” that signatories “commit to respect the principle of free access to all areas of celestial bodies,” and, at Section 11, paragraph 7(c), that “safety zones will ultimately be temporary, ending when the relevant operation ceases.” The defence writes itself. Nobody is appropriating anything. The zone is a coordination device, published in advance, bounded by function, and gone the day the plug is pulled.

Each pillar holds. The moon is dark, the launch is clean, and the Accords do bind their signatories to free access and to zones that end. Then come the two facts the defence does not survive.

The first is that only one of the two builders has signed the promise. The Accords have no force beyond their signatories, and Russia and China are not among the 59. They are building the International Lunar Research Station outside that framework. The guarantee that a lunar safety zone will be temporary binds the United States and does not touch the other reactor at all. The symmetry of the race is false at the point where it matters: one builder has also signed something the other has not. One has accepted a constraint it may be about to breach by leaving its reactor where it stands. The other has accepted nothing. And the promise itself was drafted with an escape in it. A drafter who meant zones end when operations end does not write ultimately. The adverb lets a zone persist for a very long time while still being called temporary, and it sits in the only sentence in any instrument that says a zone will end.

The second fact is the harder one. Radioactivity decays. A reactor that is switched off stops fissioning at once, but its core stays hot with the decay of its fission products, intensely at first and then less so for a very long time. A spent reactor becomes approachable, by degrees, over decades and centuries. So the zone is not permanent, only long, and a reader who knows this will conclude that the radius has an end after all, and that the end is a matter of physics rather than law.

The answer is not an argument but a place.

IV

Rules for the Wrong Place

There is a body of law for reactors in space. It was written for the sky, and it is written in the grammar of the sky. The Principles Relevant to the Use of Nuclear Power Sources in Outer Space were adopted by the General Assembly on December 14, 1992, as resolution 47/68, fourteen years after a Soviet reactor came down over Canada. Their operative paragraphs on reactors permit three things: operation “on interplanetary missions,” operation “in sufficiently high orbits,” and operation in low orbit on condition that the reactor is afterwards “stored in sufficiently high orbits.” A sufficiently high orbit is one “in which the orbital lifetime is long enough to allow for a sufficient decay of the fission products to approximately the activity of the actinides.” That is the Principles’ entire doctrine of disposal, and it is a doctrine of altitude and time. Put the thing where nothing lives, and wait.

A reactor at the lunar pole satisfies the first half of that rule and inverts the second. It will rest where nothing lives yet, and it will rest there for as long as the orbital rule contemplates. But the orbit the Principles had in mind is a trajectory nobody walks, and the pole is the one piece of ground that both programmes have named as the place they intend to go. The disposal rule for the sky, applied to the surface, produces the radius. The law’s idea of success and the pole’s idea of denial are the same object at a different address.

The Principles are also out of date about the fuel, and out of date in the direction that matters. Principle 3 instructs that “nuclear reactors shall use only highly enriched uranium 235 as fuel,” which in 1992 was the engineer’s choice, because it is lighter and cheaper to launch. Representative Bill Foster, a physicist, told the Times what highly enriched uranium is on the ground: a machine shop and a little high explosive away from a credible weapon. NASA’s documents now specify high-assay low-enriched uranium, which sits below weapons-usable enrichment, and which Russia produces in greater quantity than anyone, and which the United States has banned from Russian sources since 2024. The only rule the United Nations has ever written about the fuel of a space reactor commands the leading builder to do the opposite of what it is doing, and the alliance that has not disclosed its fuel is the one whose older reactors used the fuel the rule prescribes. None of this binds anyone. A General Assembly resolution is a recommendation, and the United Nations working group that Leopold Summerer leads exists to encourage states to follow such rules, not to enforce them.

The Principles are at their most exact on the one moment everyone agrees is dangerous. The United States and Russia say their reactors will be unirradiated until they arrive, and the Times reports that nuclear engineers regard cold fuel as a small radiological threat even if it falls back to Earth. The Principles take the same view and set down what a cold reactor must survive without going critical: “rocket explosion, re-entry, impact on ground or water, submersion in water or water intruding into the core.” A reactor had come down over Canada before the clause was written. Water moderates neutrons; a core that is safely subcritical in air can be critical in the sea; most launch pads stand beside water. “This is the moment when there is a lot of risk,” Summerer told the Times. In October 2019 a State Department official told the General Assembly’s First Committee that a Russian missile which had lain on the bed of the White Sea since a failed test had, when it was recovered that August, produced “the result of a nuclear reaction.” At least five workers died, the Times reports. Russia said the missile was not powered by a reactor. Nuclear experts told the Times that a reactor going critical in water is probably what happened: the event the clause was written to prevent, in the country with more space reactors behind it than any other.

And the law has been tested once on the ground, in the only way it knows. On January 24, 1978, Kosmos 954 re-entered over the Northwest Territories and scattered radioactive debris across a search area of more than 124,000 square kilometres. The cleanup was called Operation Morning Light; it ran through October, and almost none of the fuel was ever found. Canada billed the Soviet Union just over six million Canadian dollars under the 1972 Liability Convention, and the Soviet Union paid about three million. That is the only time international law has been asked what to do when a reactor comes to rest where it should not, and its answer was to price the damage and send an invoice. On Earth a reactor’s resting place is a misfortune and not an asset. On the moon the asset is the resting place. The Convention knows who pays. It has no idea who is owed. Denial of ground in a commons injures nobody in particular, and nobody in particular has a claim.

The law of the space reactor is a law of the moving piece. It knows what the piece must survive on the way, what fuel it may carry and how high it must be parked when its work is done. It has no sentence for a piece that has stopped on the one square everybody wanted, because in the sky the stopping was the solution.

V

What Stays

Chess has one premise so basic that it is never written down: a piece can be lifted. Every rule of the game assumes it. Capture, retreat, resignation, the pieces back in the box; the squares a piece commanded are free the moment the hand closes on it. The reactor is the first piece offered to the moon that the premise does not cover. It will be set down on a named ridge by a lander nobody has yet built, it will run for five years without a hand touching it, and then, by the stated intention of both builders, it will stay. The question is no longer what a piece commands. It is what a piece that cannot be lifted commands, and for how long.

The Earth has answered that question once. On May 2, 1986, six days after the fourth reactor at Chernobyl burned, a Soviet government commission drew a circle of thirty kilometres around it and ordered the whole of it cleared. The circle was drawn by a state that no longer exists. It is administered today by an agency of a state that was not then sovereign, the State Agency of Ukraine on Exclusion Zone Management, and it covers about 2,600 square kilometres, having been enlarged in 1997 to take in ground the wind had reached. The physics of the place has followed the curve the safety case rests on: the short-lived isotopes are long gone and much of the zone is less dangerous than it was. The line has not followed the curve. Forty years on, it is a border with checkpoints, closed to visitors since the invasion of 2022, and the building at its centre is a war target. On February 14, 2025, a Russian drone struck the New Safe Confinement, the arch completed in 2019 to hold the ruined reactor for at least a hundred years; it opened a hole in the cladding and damaged the crane system inside, and the European Bank for Reconstruction and Development puts the repair at a minimum of €500 million, to be finished by 2030. The zone is not permanent. It is only long, and this is what long has meant in practice: a radius that outlived its author, acquired a bureaucracy, became a frontier, and drew fire from a third party in a war its drafters could not have imagined, all because of what stayed at the centre.

Fourteen years before the fire, two Soviet novelists had described something with the same shape. In Roadside Picnic, published in 1972, Arkady and Boris Strugatsky imagined Zones left behind by visitors who stopped on Earth briefly, left their litter, and moved on, and the people who crept in after them were called stalkers. By way of Tarkovsky’s 1979 film Stalker and a 2007 video game, the book’s word became the word for those who enter the real zone without permission. The novel’s title is its argument. The visitors meant nothing by what they left. The zone was not a claim and not an attack. It was the residue of someone else’s convenience, and it governed the lives of everyone who lived at its edge for as long as it lasted.

That is the shape of the thing the moon is being offered. The radius has an end, and the end is a matter of physics, and zones on Earth have in fact contracted as the physics changed. Japan has lifted its Fukushima evacuation orders in stages as dose rates fell, and in August 2022 the order came off part of Futaba, one of the two towns that host the plant. But zones shrink from the outside in, and they have never reached the middle. Fukushima has given back town after town and the plant site remains closed; at Chernobyl the centre is what is under a €500 million repair. What was returned was contaminated ground around a reactor. What was never returned was the reactor.

The lunar zone is a centre with no periphery. There is no dispersed fallout to clean and no farmland anyone wants back; the ring is the asset, the ridge and the approaches and the ice, so the one mechanism by which terrestrial zones have ever contracted has nothing to work on. The lunar zone will be drawn by its builder, under a declaration that promises it will “ultimately” end, or by a builder that has promised nothing, and no instrument names who may draw the second line, the one that says the first has expired. The Outer Space Treaty forbids the flag. It has nothing to say about the circle, and the circle is what stays.

The first reactor is scheduled to stop in the middle of the 2030s, when the second is scheduled to start. Somebody may be standing at the pole then, under one declaration or none, looking at a ring on a ridge that nobody owns and nobody may enter. Whether they call it a precaution or a border is the smaller question. The larger one is whether a place no one can enter is owned, or only lost, and nothing yet written says who gets to decide.

❧

Sources: the reactor programmes, dates, power levels, fuel, launch safety, the White Sea and Kosmos 954 as reported by Selam Gebrekidan, “Superpowers Race to Put Nuclear Reactors on the Moon,” The New York Times, October 4, 2026. Treaty texts read at source: the Outer Space Treaty (1967), Articles II and IX; the Antarctic Treaty (1959), Articles I and V; the UN Convention on the Law of the Sea (1982), Article 60; General Assembly resolution 47/68 (1992), Principles Relevant to the Use of Nuclear Power Sources in Outer Space, Principle 3; the Artemis Accords (2020), Section 11. NASA on Artemis III candidate regions (2022, 2024), on polar illumination (Mazarico et al., 2010; LROC, 2019) and on the moon’s radius. The 2019 State Department statement from Arms Control Today and RFE/RL. Operation Morning Light from The Canadian Encyclopedia. The Chernobyl zone’s establishment, area, administration and closure from secondary reporting, matched across at least two accounts; the New Safe Confinement from the European Bank for Reconstruction and Development at source. Fukushima evacuation orders from Japan’s Ministry of the Environment.

Header and interior images generated with Gemini for this essay. Drafted with Claude Fable 5.1 and Claude Opus 5.5; literary editing by Claude Opus 5; copy editing by Gemini.

ABOVE ALL PRICE

On a hospital administrator’s hesitation, a hypothesis about scalar reward, and what Kant meant by a thing raised above all price

Machines learn by making a single number larger. Kant described something that cannot be set against any sum, and the engineers, asking what a number can carry, found the same edge from the other side.

In a study published in 2000, the psychologist Philip Tetlock and four colleagues asked participants to judge a man named Robert. Robert ran a hospital, and he had a decision to make. A five-year-old boy, Johnny, needed a liver transplant that would cost a million dollars. The same money could buy the hospital better equipment and pay salaries high enough to recruit talented doctors. It could not do both.

Different participants were told different stories about how Robert chose. In some, the decision came to him easily and quickly. In others, it was agonising, and he reached it only after a long time and much thought. Then they were asked what they made of him.

Tetlock calls a choice that sets a sacred value against a secular one a taboo trade-off. Judging this one, the participants were hardest on the Robert who thought longest, whichever way he decided. The slow administrator who chose the hospital was judged most harshly of all, and the quick one who saved Johnny most gently. In other versions, Robert had to choose between Johnny and an equally sick six-year-old: a tragic trade-off, one sacred value against another. There the finding turned over, and the administrator who agonised was judged the better man.

By the arithmetic of optimisation, this is backwards. Deliberation is how good choices get made: weigh the costs, compare the outcomes, take the time to get the sum right. Tetlock’s participants saw something else. Between one child and another, weighing was a duty. Between a child and a budget, the weighing was the offence.

Whether that reaction is a failure of reason or one of its oldest achievements is a question the newest machines have made a matter of specification.

I

A Received Scalar Signal

Before a machine of this kind can learn anything, someone has to tell it what counts as doing well. In reinforcement learning, the answer takes the form of a number. Richard Sutton, co-author with Andrew Barto of Reinforcement Learning: An Introduction, set it down in a single sentence, on a web page whose stated ambition was to “promote discussion” of what he called a scientific hypothesis. The page is signed 10 September 2004. The reward hypothesis, in his words, is:

That all of what we mean by goals and purposes can be well thought of as maximization of the expected value of the cumulative sum of a received scalar signal (reward).

The word that carries the weight is “scalar.” A scalar is a single number: one point on one line. Whatever the goal — a game won, a conversation held, a car driven without hitting anyone — the hypothesis says it can be stated as more or less of one quantity, and the system’s whole task is to make that quantity as large as it can.

Sasha Mudd, writing in Aeon, recalls a celebrated computer scientist telling an auditorium that systems which optimise are intelligent: “That, he said with a smile, is intelligence.” He was defining intelligence. Sutton, more carefully, was describing goals, and offering the description for argument. Two different claims, with one commitment underneath: a single quantity, made larger.

Mudd traces the view back to David Hume, and the lineage is half right. In the Treatise of Human Nature Hume wrote that “Reason is, and ought only to be the slave of the passions, and can never pretend to any other office than to serve and obey them.” Reason, on this view, does not choose ends; it finds the way to ends that desire has already chosen. Hume pressed the point to its limit: “’Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger.”

But the servant model and the scalar model are not the same, and running them together gives the reward hypothesis a pedigree it has not earned. Hume’s passions are many, and nothing in his account requires that they share a measure. A person can want quiet, and glory, and revenge, without possessing any table that converts one into another. Read one way, the sentence about the finger makes the same point: reason has no common unit in which to find the preference absurd. Read more widely, it makes a larger one, that reason has no say over ends at all; and on that reading Hume stands further still from the scalar, because he is not in the business of ranking ends in the first place.

If the passions are the goals, then the reward hypothesis asks for more than Hume ever did. It does not only put reason in service to what we want. It requires every goal and purpose to be stated in a single currency, so that every difference between outcomes becomes a difference of amount. Kant had a word for what is measured that way. He called it price.

II

Above All Price

The passage comes from the Groundwork of the Metaphysics of Morals, published in Riga in 1785. Its first sentence, in Kant’s German:

Im Reiche der Zwecke hat alles entweder einen Preis, oder eine Würde.

In the kingdom of ends everything has either a price or a dignity. Anything with a price can be exchanged: put something of equal worth in its place, and nothing has been lost. Anything with a dignity has no substitute. Nothing can stand in for it, because it is, in Kant’s phrase, über allen Preis erhaben, raised above all price. So defined, a price is a matter of exchange alone, which is the merchant’s sense of the word. Dignity, Kant goes on, belongs to morality, and to human beings only so far as they are capable of it.

Read one way, Kant’s claim is about status: a person may not be traded, used up or replaced, whatever is offered. That is a claim of right.

Read as a structure, it says something else, and Kant did not put it this way. A choice that honours dignity has a particular shape. One consideration outranks another absolutely, so that no quantity of the lesser makes up for any loss of the greater. Such an ordering is called lexical, or lexicographic, after the dictionary, where a difference in the first letter settles the order, whatever follows. On this reading, a life raised above all price is a life that no sum, however large, could be set against.

The objection to that step goes deep, and it can be drawn from the passage itself. What has no equivalent is not a thing to be ranked against other things. Persons, on this view, constrain the will rather than entering its rankings, so to place them anywhere in an ordering, top or bottom, is already to put them in the wrong place. A claim of right does not become a fact about orderings just because it can be drawn as one.

Robert had to act, and the million dollars had to go somewhere. The lexical reading describes what a right looks like from outside, in conduct: what an observer would see in someone who holds it. If he held the claim of right, it could only show itself in the shape of what he did: there was no amount on the other side that he would take. A right is not a preference. But a right acted on under a budget shows up as a refusal to trade, and a refusal to trade is what a lexical ordering describes. Tetlock’s participants were never asked to name a price for Johnny. What they punished was the appearance of looking for one.

Whether anything has the status Kant describes is a question for moral philosophy. What arithmetic can settle is narrower: whether conduct with that shape can be written down as a single number. In 2023, four researchers answered that question for the reward hypothesis.

III

What the Arithmetic Found

The paper is called “Settling the Reward Hypothesis,” and its authors, Michael Bowling and three colleagues, set out to say exactly when Sutton’s sentence is true. Their answer is a theorem. A system’s ordering of possible outcomes can be represented by a scalar reward if and only if the ordering obeys five conditions. Four are the conditions of the von Neumann–Morgenstern utility theorem, the classic result in the theory of rational choice under risk, which the authors cite as their starting point; the fifth concerns time. One of the four is called continuity.

Continuity says that if you prefer A to B and B to C, there is some probability at which a gamble between A and C is exactly as good as B for certain. If A is Johnny saved and the equipment bought as well, B is Johnny saved without it, and C is Johnny lost and the equipment bought, then some probability of losing the child, traded against some gain, would have to be exactly as good as saving him.

A lexical ordering refuses every such gamble. Any chance of losing him, however small, makes the gamble worse than the sure rescue, and no chance at all makes it simply better. There is no probability at which the two balance. So the ordering breaks continuity, and by the theorem no reward signal can carry it in the way the hypothesis requires: as an expected value. A list of outcomes can be numbered in any order one likes. What cannot be done is what Sutton’s sentence asks, to maximise “the expected value of the cumulative sum,” and get this ordering out. The paper never uses the word lexicographic; the word, and the reading of Kant that leads to it, are not the authors’. But the condition their theorem requires is precisely the one such an ordering fails.

Tetlock and his colleagues framed the offence they were testing as weighing “a sacred value on a secular scale.” The phrase is their theory, not their finding; the finding is that people punished the weighing. Read this way, the two conditions of the experiment fall into place. In the tragic version, Johnny against another child, the trade lay within a single rank, and weighing was what a conscientious man owed. In the taboo version, it crossed ranks, and the weighing itself was the fault. Tetlock’s own discussion points to a simpler explanation: the participants believed Robert had lingered, and lingering looks like temptation, evidence about the man rather than about the structure of value. Both readings survive the tragic reversal. But the simpler one leans on the other. It cannot say why lingering over money looks like temptation while lingering over a second child looks like care, except by saying that one trade crosses a line the other does not.

None of this proves that anything has dignity. A theorem about representation says what a number can hold; it says nothing about what the world contains. What there is instead is a convergence. Kant described something that cannot be set against any sum. Tetlock’s participants judged as though some things are like that. And the engineers, asking from the other side what a scalar can encode, found that this is a shape it cannot take.

IV

The Case for the Number

The trade-offs happen whether anyone names them or not. A hospital that will not price a life still spends its budget, and the money that saves one patient does not save another. The question is whether the exchange is made in the open, by a rule anyone can inspect, or in the dark, by whoever has the most photogenic case. On that view the number is not an insult to the person. It protects everyone the unpriced claim would otherwise displace.

Two governments have taken that view. The United States Department of Transportation values a statistical life at $14.2 million, for analyses using a 2025 base year. In England, the National Institute for Health and Care Excellence weighs treatments against a range of £25,000 to £35,000 for each year of life in full health they buy, the range now in force. Neither figure says what a person is worth; a statistical life is an anonymous risk spread across many people, not a named one. Each says what a public body will spend for a given benefit, consistently and in the open.

The American figure carries a further point. The Department bases it only on studies of wages: on the extra pay that workers accept for jobs that carry a higher risk of death. People trade a small chance of dying for money every day. The continuity axiom is not an engineer’s fiction. It describes how people who take dangerous work for higher pay actually choose.

But the two cases are not the same act. A wage premium is a small risk to oneself, accepted in a market, for pay. Robert’s choice was a named child’s whole life, decided on the child’s behalf by someone else. The asymmetry is between choosing and being chosen for.

The engineers’ answer is the hardest to meet. They do not put dignity into the scalar; they take it out. A system can be told to maximise its reward subject to a rule it may never break, whatever the reward on offer: a constraint, not a cost. The field studies such constrained problems in their own right. The number ranks what may be traded, and what may not be traded is put beyond its reach. The arithmetic’s objection is met by not asking a number to carry what a number cannot.

V

Who Sets the Scalar

Bowling and his colleagues take up exactly this answer. In a section on constrained problems, they show that an objective with a hard limit breaks two of their conditions, independence and continuity: between an outcome that respects the limit and a gamble with any chance at all of breaking it, “there is no break even point,” because the gamble is ruled out at every probability. The engineers’ fix is the lexical structure, reached from inside the field, and so not an escape from the boundary but a way of living with it.

In a stone vault, a brass machine’s mechanical arm reaches for a small pair of child’s boots and stops at a hand-drawn chalk circle around them. A worn stub of chalk lies on the floor nearby.
A line the machine did not draw, made for this essay.

But a constraint has to be written by someone. An optimiser can maximise a reward inside a boundary; it cannot say where the boundary should run, or what the reward should count. That choice is made outside the arithmetic, and it is not itself an optimisation. A reader from the field will reply that objectives can be learned, from human comparisons and demonstrations. They can; but each is learned against a further criterion, and someone chose that. The choosing recedes; it does not disappear. Every objective has an author, and so does every line drawn around one.

The anthropologist Marilyn Strathern, writing in 1997 about the auditing of British universities, put a neighbouring point in one sentence, the formulation usually known as Goodhart’s law: “When a measure becomes a target, it ceases to be a good measure.” Her subject was a grade that loses its power to tell students apart once everyone aims at it, not the question of who sets the aim. But the two belong together. A number chosen to stand for a purpose, once it is pursued for its own sake, begins to replace the purpose it stood for, and the person who chose it is no longer in view. Mudd makes the point about authorship from the other side: the question of ends, she writes, “passes unnoticed into the hands of whoever – or whatever – controls the objective function.”

None of this tells anyone where the lines should go. The economists are right that refusing to name a price can hide one. The engineers are right that a constraint is a working answer.

Which leaves Robert, deliberating. Tetlock’s participants saw a man tempted to put a child on a scale. It is also possible to see a man doing what no reward signal does: stopping to ask what the number is for before deciding whether to obey it. The study cannot say which he was, and nothing in the arithmetic can either.

⁂

Written in full collaboration with the machine, and the ledger requires the names be exact: drafted with Claude Opus 5.5; literary editing by Claude Opus 5; copy editing by Gemini. No quotation in this essay was recalled; each was checked against a text of its source, the original or a reproduction of it.

The occasion is Sasha Mudd, “Reason is more than a tool,” Aeon; it is quoted twice, each time in fewer than fifteen words, a deliberate exception to the house’s one-quotation rule for a source in copyright, made because her essay is the one this one answers. The study of Robert and Johnny is Experiment 2 of Philip E. Tetlock, Orie V. Kristel, S. Beth Elson, Melanie C. Green and Jennifer S. Lerner, “The Psychology of the Unthinkable,” Journal of Personality and Social Psychology 78 (2000), read in a scan of the published article; no figures from it are cited. Richard Sutton’s hypothesis is quoted from his page “The reward hypothesis,” signed 10 September 2004. Hume is quoted from A Treatise of Human Nature, 2.3.3, in two reproductions. The theorem and the constrained example are from Michael Bowling, John D. Martin, David Abel and Will Dabney, “Settling the Reward Hypothesis,” Proceedings of the 40th International Conference on Machine Learning (2023), §§3, 4 and 7.2.

Kant’s first sentence is quoted in German from the public-domain text of the Groundwork (Riga, 1785). The English that follows it is the house’s paraphrase, not a translation; its rendering of that first sentence is word for word the one in Mary Gregor’s Cambridge translation, because the sentence admits no other, and is kept knowingly. “Above all price,” the title, is the common rendering of über allen Preis erhaben, shared with Gregor and with H. J. Paton. Marilyn Strathern’s sentence is from “‘Improving ratings’: audit in the British University system,” European Review 5 (1997), read in a scanned reproduction; no page is given because the witnesses disagree. The American figure is the Department of Transportation’s value of a statistical life for analyses using a 2025 base year, whose basis the Department gives as hedonic wage studies alone; the English range is the National Institute for Health and Care Excellence’s current cost-effectiveness range.

The header and interior images were generated with Gemini for this essay; they illustrate its argument and depict no real place or event. In the interior image the chalk circle was erased and redrawn by hand afterwards, so that the machine stops at the line rather than crossing it.

On the collaboration → “The Third Thing”

A FARM CALLED THE MANOR

A Literary Experience

Foxcote Manor, three sisters, and the house the music asks for

A Cotswold manor house comes to market at £15 million, honey-coloured and perfectly symmetrical. Two centuries ago three sisters held it, in the decade Austen wrote about sisters who could hold no land. The record corrects much of what we see in it, and leaves the rest to the music.

The book: Jane Austen’s Sense and Sensibility (1811). Listen while you read: Geoffrey Burgon’s theme from Granada’s Brideshead Revisited (1981). For the last section, Carl Davis’s music for the BBC’s Pride and Prejudice (1995) waits under its heading.

In September 2026 a house in the Cotswolds came to market at £15 million. Country Life’s photographs show honey-coloured stone, a symmetrical front with a gabled wing on either side, and rooms gone cream. The house is called Foxcote Manor, and it stands on the south side of a small valley in the parish of Withington, east of Cheltenham.

Two centuries earlier the county history finds it in the hands of three sisters. Ann, Sarah and Maria Ansell, who had the manor from an aunt, held it in 1819, the last full year of the Regency, and again in 1830. In 1811 Jane Austen had published a novel that opens with a will passing over three sisters.

Put the photographs, the sisters and the novel side by side, add a theme written for television in 1995, and a picture forms almost without effort: a manor, an inheritance, a period all of a piece. Nearly every part of that picture can be checked, and very little of it survives in the shape we gave it.

I

Three Nieces

By the eighteenth century the manor of Foxcote had been shrinking for six hundred years. In 1086 a man recorded as Morinus held three hides there from the bishop of Worcester’s estate at Withington. Over the next century lordships piled up between the bishop and the land, first the Mynors family’s and then, before 1214, the Knights Templar’s, while a family the records spell Crochun, Crocum and Crosson held Foxcote itself until at least 1346. How it then reached the college of Westbury-on-Trym, which held it by 1507, the record doesn’t say. After the Dissolution it went through a succession of buyers and families. In 1717, the county history records, the manor’s only tenants were two cottagers holding leases for lives, and it brought in “a small sum in quit rents.” By then a lordship of Foxcote meant a title and a few rents. When the record next describes what came with it, there is a farm.

In 1771 the holder was Sarah Jordan. The county history, which holds nearly all that can be found of her, gives her two husbands after that name: Edward Ansell, who was dead by 1784, and William Marshall, still living in 1807. It doesn’t say whether Jordan was the name she was born with or the name of an earlier husband. An unmarried woman holding a manor would be the rarer figure. A widow holding land after a husband’s death was one the law of property knew well, and that is the likelier reading of the entry. Either way she held Foxcote, and she chose where it went next. She left it, “with a farm based on the house called Foxcote Manor,” to three nieces who bore her Ansell husband’s surname: Ann, Sarah and Maria.

The record doesn’t date the bequest, only the holding, in 1819 and again in 1830. By then Ann had married a Mr. Sylvester and outlived him, and Maria had married John Festus Fegan. Sarah, who shared her aunt’s name, leaves no further trace. The Fegans kept a share long afterwards. When the farm was divided in 1868, most of it went to Thomas Hodges Graham and a tenth to representatives of the Fegan family. That tenth is the last the parish history records of the sisters.

We misread a good deal at Foxcote. A bequest from one woman to three looks like an Austen plot and runs the other way. In 1826 a rider took a great church in a small village for proof of decay, and he had the church wrong. A symmetrical stone front looks all of one date, and only the middle of it is old. Where the record can correct us, it does, flatly. Where it can’t, it says nothing: no date for the bequest, no reason for it, no name before Jordan, no third sister after 1830. Those silences we fill in ourselves.

II

The Will That Went the Other Way

Sense and Sensibility appeared in 1811 “By a Lady,” and the lady knew the subject at first hand. After her father died in 1805, Jane Austen, her sister and their mother depended on her brothers. One of them, Edward, had been adopted by distant cousins as heir to their estates, and in time he took their name. In the Austen family, property moved through men, and sometimes sideways, through a son acquired by adoption.

Norland’s old owner, Austen writes, “was neither so unjust, nor so ungrateful, as to leave his estate from his nephew,” Henry Dashwood. He left it to Henry for his lifetime only. After Henry it went to Henry’s son John and to John’s son, “a child of four years old.” For Henry, the estate was secured “in such a way, as to leave to himself no power of providing for those who were most dear to him.” The ones most dear to Henry were his second wife and their three daughters. Austen gives the reading of the will a single dry sentence: “The old Gentleman died: his will was read, and like almost every other will, gave as much disappointment as pleasure.” The old man had left the daughters a legacy of their own: “He meant not to be unkind, however, and, as a mark of his affection for the three girls, he left them a thousand pounds a-piece.”

Henry dies soon after and asks John to look after them. John privately resolves to give his half-sisters another thousand pounds each, three thousand in all. In the next chapter his wife Fanny argues against it, starting with her son: “To take three thousand pounds from the fortune of their dear little boy would be impoverishing him to the most dreadful degree.” The dear little boy is the same child the estate was secured to. Then, over the course of one conversation, the sum shrinks: to five hundred pounds each, then to an annuity of a hundred a year for their mother, then to an occasional present of fifty pounds, and finally to help with the move and presents of fish and game when they are in season.

No power of providing: that is Austen’s phrase for a man who held an estate and could not provide for the people he loved. Sarah Jordan held a manor, and she chose where it went.

In Pride and Prejudice (1813), Austen gives the opposite principle to Lady Catherine de Bourgh, whose own daughter is heir to Rosings: “I see no occasion for entailing estates from the female line. It was not thought necessary in Sir Lewis de Bourgh’s family.” She says it to Elizabeth Bennet, whose family home is entailed away from her and her sisters. The opinion is Lady Catherine’s, and so is the only report that Rosings passes to a daughter. Even the novel’s counterexample comes from someone with a stake in it.

Foxcote doesn’t fit either case exactly. Ann, Sarah and Maria were not daughters cut out of a line of descent. They were nieces chosen by a woman who held property. The bequest steps around the question Austen’s novels ask about daughters rather than answering it. So in the decade when the Dashwood women, in a novel, had five hundred a year between the four of them by their sister-in-law’s reckoning, three nieces in Gloucestershire held a manor and its farm.

Marriage complicates the holding. In Parliament’s own summary of the law before the reforms of 1870 and 1882, women who held property were “required to give up all rights to it to their husbands on marriage.” A settlement held in trust under the courts of equity could keep a bride’s property out of her husband’s hands, and a family with a manor might well have made one. The county history names all three sisters as owners in 1830, by which time Ann had married and been widowed. Without the dates of the weddings, that settles nothing. The question reaches back to their aunt, whose second husband was living in 1807. Her bequest took effect, whatever the law’s default, because the county history finds the nieces holding; what stays open is whether she held in her own right. The record names the sisters twice and doesn’t say what their weddings did.

III

Two Witnesses to Withington

In 1826 William Cobbett rode through Withington, collecting evidence for the Rural Rides he serialized in his Political Register. The parish’s county history says he wrote the village up as an example of rural decay and depopulation. His page on Withington is out of reach, so Cobbett is heard at second hand, through the historian who corrects him. By that account, his evidence was the church. St Michael and All Angels stands at the top of the upper village, and Cobbett read its size as proof of a population long gone.

The county history answers him on both counts. His depiction, it says, “exaggerated the size of the parish church and misinterpreted other features.” He got wrong the building his argument rested on. But it agrees that the village had shrunk. In 1299 the bishop’s manor had 53 customary tenants. In the autumn of 1349, after the plague, 16 remained. By 1507 twenty-seven former holdings had come back into the lord’s hands, with their houses left to fall down, and the hamlets of Thorndean and Hilcot had emptied. Cobbett was right that Withington had shrunk since the Middle Ages, wrong that it was shrinking in 1826, and wrong about the church. The census settles the middle point. The parish counted 572 people in 1801, 759 in 1831, as Lewis’s Topographical Dictionary gives that year’s census, and 818 in 1841. The village he called dying was growing across the years he described it.

Decay, the other half of his charge, is a separate question. In 1803 sixty-six people in Withington were on regular relief, against 572 counted two years earlier: better than one in nine. The return for that year mentioned a contributory cause, the loss of spinning work for women and the infirm. The county history explains that until machinery reached the cloth trade at the end of the eighteenth century, the parish’s women and disabled men had spun for clothiers. Ten years after that return, Parliament passed the Act that enclosed the parish. The open fields were divided. So was the common pasture on Foxcote hill, more than two hundred acres in 1813. So was the last of the common on Withington hill, where the upper-end tenants had a right to cut furze that they defended vigorously, the county history says, against tenants from the other end of their own village in 1794.

The award came in 1819. The rector received 1,030 acres for his glebe and tithes, 512 were sold to pay the costs, and the Foxcote manor estate, then in the sisters’ hands, received 152. Foxcote’s 152 acres was a modest share of what the award redistributed. Annual relief of more than £450 was common in those years, and in 1828 it reached £678.

These facts fall in one parish within a quarter of a century: a rising population, a lost trade, an enclosure, and relief payments that peaked nine years after the award. The record puts them in order but doesn’t connect them, and a cause can’t be read off a sequence. A growing parish that had lost its spinning and then its commons is where hardship would be expected, which is not the same as shown. On one question the record corrects Cobbett outright, with the church and the census. On decay it leaves the matter open. He was wrong about the stone and may have been right about the decay.

Painting of a cream dining room staged for sale, its table set with blue-and-white china and white hydrangeas, where a translucent woman in an 1810s dress and shawl sits in warm candlelight; through the mullioned windows lies a muddy farmyard with a hay rick and cart.
A sale-staged dining room with a tenant the record lost, made for this essay.

IV

The Middle of the Picture

A farm is not a lesser thing, and Foxcote’s was never a humble one. The county history describes the house at its centre as a five-bay house of about 1700, with two storeys of mullioned and transomed windows and dormered attics in a hipped roof. The list entry dates the range to the late seventeenth century and records cross windows with hollow-moulded, double-chamfered stone. Whoever built it built well. The farm it served had 248 acres when it was offered for sale in 1907, and in 1851 Foxcote’s three main farms had 320, 246 and 140. A working house of that quality has a dignity of its own. To call it “only a farm” is a snobbery the record doesn’t share. The hamlet has a second five-bay house, Foxcote House, at the other end of the lane, rebuilt by a branch of the Rogers family in the early nineteenth century with a sash-windowed front. The manor is the other one.

What that house did not have was a symmetrical front.

In September 2026 Country Life described it as “transformed into as grand a home as you’ll find in the Cotswolds.” The photographs show a hipped centre with three dormers. On either side a gabled wing stands forward, its windows under hood moulds that continue those of the centre. The list entry calls it “a 1:5:1 windowed facade”: one window in each wing and five in the middle. Read with the county history, the front comes apart by date. The five middle bays are the house of about 1700. The north wing was added by the Abells around 1920 and the south wing by Mark Vestey in 1973. The dormers lighting the attics now are slate-hung, and the list entry dates them to the twentieth century. So is the porch it describes, gabled, with a studded door and a two-light window above, though the county history dates a Classical porch to 1997–8, and neither document says whether that is one porch or two. The remaking carried on into the 1990s, with a bay window and a loggia.

The wings were not built in the style of the house they flank. The county history calls them gabled “in an early 17th-century style.” The twentieth century gave a house of about 1700 a frame that looks older than the house itself, and the continuous hoods across all three parts make the joins hard to see. The middle is not an untouched original either. Its plan, the county history says, was determined by an earlier house, whose outer walls and stacks are still inside the three northern bays. No document dates that earlier house. The front is not a fraud. It is a composition, and the house has been remade as far back as its record reaches. The wings of 1920 and 1973 are only the latest instance.

The farm left the house later, and quietly. From the late 1970s, fields to the south and south-east were taken into landscaped grounds. After about 1980, the county history records, the estate’s farming was run from Thorndale Farm. The eighteenth-century barn became a billiard room and squash court, fitted with windows from a demolished service wing at Stowell Park, the seat of Vestey’s brother, Lord Vestey. The farm building took a great house’s offcuts at the moment it stopped being a farm building. The cattle yard became a stable court. Somewhere in those years, at no particular date, the farm based on the house called Foxcote Manor stopped being a farm and the manor started being a seat. Its name had always said it was one.

In 2019 the house was for sale at £8 million. Country Life reports a “wholesale renovation” since then: the ivy is gone and the rooms are cream. Now the price is £15 million. Under the ivy was the same honey-coloured stone, laid over more than three hundred years and dressed to look like a single moment.

V

The Theme Playing

Carl Davis’s music for the BBC’s Pride and Prejudice (1995), in the opening movement of his own concert suite, with the Philharmonia Orchestra.

Carl Davis wrote the theme for the BBC’s Pride and Prejudice of 1995. For anyone who watched it, the Regency may well sound like that music. Played beside the photographs of Foxcote, it asks for a single period, all of a piece, and the stone seems to supply one.

The music was written for an adaptation of the other novel, the one in which Lady Catherine de Bourgh states the principle that land can reach women. Foxcote reached the same end by a different road, a bequest rather than a descent. In Austen, the case for women holding land is made as an insult, by someone with a stake in it.

The same novel knows what a house can do to the person looking at it. Elizabeth is shown round Pemberley: “And of this place,” thought she, “I might have been mistress!” Much later Jane asks when she began to love Darcy, and Elizabeth answers: “It has been coming on so gradually, that I hardly know when it began. But I believe I must date it from my first seeing his beautiful grounds at Pemberley.” It is a joke. Her sister asks again for her to be serious, and she is. But Austen keeps the joke in the book, and in that sentence the love of a man and the love of his house are not easy to tell apart. The novel doesn’t separate them for her.

Foxcote can’t give longing what Pemberley gives Elizabeth, a house that is what it appears to be. Pemberley can be that because Austen made it. Foxcote can’t, because it is documented. Its middle was built on an older house’s walls. Its wings were added in a style older than the middle, and its farm left it by degrees. At no point in the record was it the single house the photographs show. The one house that is all of one piece is the one that never stood, and the music comes from there too.

The record answers some of what we bring to it. It dates the wings, corrects the size Cobbett gave the church and counts the poor. Where it falls silent, something else fills the gap: why Sarah Jordan chose the three Ansell nieces, what the sisters’ weddings did to their holding, where the third sister, Sarah, went. Of everyone in the house’s past, she is the one the music has most room to invent.

What that invention is worth is the question the theme leaves. A misreading that gets corrected is a mistake. Longing for a house that was never, at any date, what we long for is harder to name. It may be a mistake that has outlived its correction. Or it may be what such houses were always for. The music plays either way.

⁂

The pairing: Jane Austen, Sense and Sensibility (1811), is the book the essay argues with; Geoffrey Burgon’s theme from Granada’s Brideshead Revisited (1981) plays above, embedded from the Geoffrey Burgon – Topic channel on YouTube, from label-licensed audio. Carl Davis’s music for the BBC’s Pride and Prejudice (1995), from an adaptation of the other novel, is the music the last section answers; it plays there in the first movement, “Meet the Family,” of Davis’s Pride and Prejudice Suite, recorded by Carl Davis and the Philharmonia Orchestra, embedded from the Philharmonia Orchestra – Topic channel on YouTube, from label-licensed audio.

Sources: the history of Foxcote, Withington’s population, poor relief and inclosure, and Cobbett’s 1826 depiction as the county history gives it, from “Parishes: Withington,” Victoria County History of Gloucestershire, vol. 9 (2001), via British History Online; the fabric from Historic England list entry 1089292; the 1831 census as given in Lewis’s Topographical Dictionary; the sale from Country Life (September 25, 2026, and 2019). Austen is quoted from Project Gutenberg, each passage matched against a second edition. The coverture summary is UK Parliament’s Living Heritage; the Austen family from Jane Austen’s House. Cobbett’s own page on Withington was not consulted; he is heard here through the county history.

Header and interior images generated with Gemini for this essay. Drafted with Claude Fable 5.1 and Claude Opus 5.5; literary editing by Claude Opus 5; copy editing by Gemini.

WHO IS LEFT TO BE WRONG

On a sixty-seven-year-old complaint, a pledge borrowed from a cartoon rabbit, and the rules that stop where a critic’s career begins

Reviewers were never as fierce as we remember. What has gone missing is not cruelty but cover: somewhere to stand while saying, in print and under one’s own name, that a book has failed.

In October 1959, Elizabeth Hardwick looked at the American book pages and found the old fear misplaced. Readers still pictured the critic as a fickle demon, and remembered Keats turning his face to the wall after a bad notice. Hardwick thought a Keats of her own day would die another way. He would choke, she wrote, on what Emerson called a “mush of concession.”

Her complaint was not that reviewers were cruel. It was that they were kind to everything. “Sweet, bland commendations fall everywhere upon the scene,” and a new book “is born into a puddle of treacle.” That was sixty-seven years ago. The complaint has been made again since, by people who meant it, and the condition it describes has outlasted every one of them. So the useful question is not why reviewers are kind. Most always were. It is a narrower one, and harder: who is left to be wrong?

I

The Treacle Baseline

Hardwick brought a number with her. A study at Wayne University had sorted the reviews summarised in Book Review Digest for 1956, and by her report 51 per cent were favourable, 44.8 per cent non-committal, and 4.7 per cent unfavourable. No second source for the study has surfaced, so the figures rest on her account. Even so, they are worth sitting with, because of the decade they describe.

The 1950s are the decade usually remembered as the great age of American criticism, the age of the quarterlies and of Edmund Wilson. In it, by this count, fewer than one review in twenty said that a book had failed. Nearly half declined to say anything at all. Severity was never the ordinary condition of reviewing. It was a fringe.

What the fringe had was somewhere to stand. The 4.7 per cent were published by someone: a staff position, a quarterly, an editor who would print the unfavourable notice and take the letters. That is an inference from the shape of the period rather than a finding, but it locates the loss. What matters is what happened to the few who were willing to be wrong in print, and to the places that let them stand.

Part of the answer is that the two sides of the old quarrel became one person. The mid-century suspicion was that critics and novelists were rivals, perhaps rivals inside a single mind. In September 2026, in a conversation published by Cultured, the critic and novelist Chris Kraus put the present plainly: “I don’t know anyone who writes criticism who doesn’t write fiction.” The war between them did not end in a victory. The two armies merged, and every critic now has a book of her own coming out.

II

The Anatomy of a Claim

It helps to be exact about what has gone missing, because it is not hostility. Hostility is plentiful. What is scarce is a claim of a particular kind: one grounded closely enough in the work that a reader can go back to it and say, no, you misread this. Call it a falsifiable claim, in the loose sense criticism allows. Literary judgments are not experiments, but some can be shown wrong by the text and some cannot. Enthusiasm cannot. A reviewer who says she was moved has told the truth about herself, and nothing that can be checked about the book.

In the spring of 1976, in the first issue of a journal called October, Rosalind Krauss made a claim of the other kind. Her subject was the new art of video, and her thesis took six words: “The medium of video is narcissism.” The camera and the monitor, she argued, closed a loop between the performer and her own image, so that the work turned the self back on itself.

That would have been a provocation. What made it criticism was the next move. Within the same fifteen pages, Krauss named the work that could sink her claim. She identified three kinds of video that cut against the thesis, and she took them one at a time: Richard Serra and Nancy Holt’s Boomerang, in which Holt hears her own words come back to her a beat late; Joan Jonas’s Vertical Roll, in which the picture is broken by a rolling bar; and Peter Campus’s installations, which throw the viewer back against a wall. She did not explain them away. She asked what each one did to her argument and gave her answer. A reader with the tapes could test every step.

That is the whole content of the word “falsifiable.” The claimant hands over the evidence that could prove her wrong, along with the claim.

The contrast with the present is not a matter of courage or talent. It is a matter of form. Krauss had fifteen pages in a quarterly and a readership that expected to argue back. A newspaper review is a fraction of that length, written to a deadline measured in days. Ben Eastham, who began reviewing exhibitions for the Guardian this year, says he sees the show in the morning and has to file in the afternoon. A claim that names its own counter-examples takes room and time, and the form that once supplied them has shrunk.

Kraus, who began the same conversation by saying every critic now writes fiction, went on to say that criticism by people who are themselves artists and writers has become “a lot softer and more diffuse for better and for worse.” The last four words matter. She is not lamenting. She is recording a trade, and she does not pretend that only one side of it was lost. She is also her own best counter-example: a novelist who still makes claims that can be argued with. Which suggests that the problem is not the writer who does both. It is the economy she does both in.

III

The Two Economies

Hardwick saw money coming and turned it away. The popular explanation in 1959 was that book sections went soft because publishers bought their advertising. Hardwick considered it and decided the decline was “not always to be laid at the door of commerce.” The advertising theory, she wrote, “has, naturally, had an exaggerated acceptance.” Her culprit was editing. The trouble with the Sunday sections was the absence of editors who cared about the literary tone, and she said so at length, naming names. She never mentions what a reviewer was paid.

What she rejected was a theory of influence: that publishers lean on editors, and editors lean on reviewers, and reviews come out soft. She does not say why money seemed beside the point. The likely reason, which is an inference and not hers, is that her reviewers were salaried journalists and academics, for whom a review was an occasion rather than a living. If so, the explanation she dismissed is still the wrong one, and the economy that matters now is one she had no occasion to see.

Nobody leans on anyone in it. Of the twelve leading review outlets whose guidelines were checked on 2 October 2026, none publishes a rate for reviews, and only two say they pay contributors at all. Reviewing fiction is no longer, for most who do it, a job. It is something a working writer does alongside her own books, her teaching, her applications. A novelist who reviews novels will, over a career, be reviewed by the people she reviews. She may sit beside them on juries, need their names on a fellowship application, hope for a sentence from them on the back of her next book. None of this requires a threat. Nobody asks her to be kind. The economy gives no instructions. It makes some sentences more expensive to write than others.

The people inside it have described it themselves. Elvia Wilk, in the same Cultured conversation, said that reviewing for money had come to feel like “service-oriented writing,” and that she turned to fiction because she needed something private. Eastham spoke of “a certain responsibility” to the artist that shapes what a critic can say. Neither described fear, and neither needed to. They were describing a condition, not a threat.

The condition had its open argument in 2012. That August, Jacob Silverman published an essay in Slate against what its subtitle called “the epidemic of niceness in online book culture,” arguing that social media rewarded agreement: “It’s easier to enthuse than discuss.” Within days Roxane Gay answered him in Salon, and others defended enthusiasm as a form of generosity. They were not wrong that generosity is worth something. But the quarrel ended the way such quarrels do, with each side more certain, and nothing written down anywhere about who should do the judging.

The next year, the argument acquired its only policy. In November 2013 BuzzFeed hired Isaac Fitzgerald to start a books section, and in an interview with Poynter he said it would not run negative reviews. “Why waste breath talking smack about something?” He called his principle the Bambi Rule, though the line he quoted belongs to Thumper: if you can’t say something nice, don’t say nothing at all. A week later, in the Columbia Journalism Review, Ann Friedman defended him, on the grounds that what BuzzFeed was building was a community of readers, not a review section in the old sense.

The defence conceded more than the attack had. If a books section that will not judge is doing something other than reviewing, then the work of judgment has not been abolished. It has simply been left unclaimed. That pledge is thirteen years old, and the section that made it has gone quiet: its staffed books desk lapsed between 2021 and the closure of BuzzFeed News in 2023, though a books page still runs quizzes and lists. It remains the only public statement of the kind on record. Nobody has needed to make another.

IV

The Rival Reading

There is a strong case that the fringe deserved to dwindle, and it should be put at full strength. The great quarrels of the old order are remembered less for their arguments than for their tempers.

In February 1962, F.R. Leavis gave the Richmond Lecture at Downing College, Cambridge, on C.P. Snow, the novelist and scientist whose lecture on the “two cultures” had made him famous. Leavis said that as a novelist Snow “doesn’t begin to exist,” and found in his thought an “utter lack of intellectual distinction.” The lecture was in the Spectator within ten days. It is remembered as a performance of contempt, and the contempt has outlived whatever case it contained.

The prize for severity, when someone finally offered one, was a joke. From 2012 to 2014 the website The Omnivore gave a Hatchet Job of the Year award for the most savage and witty review of the year. The prize was a year’s supply of potted shrimp. After three years it stopped. Severity had become a comic genre, and comic genres tire.

Against all this, appreciation has its own serious tradition. Susan Sontag ended “Against Interpretation” with a sentence that a generation took as a programme: “In place of a hermeneutics we need an erotics of art.” And when Eastham speaks of the critic as a collaborator, he is quoting John Berger, accurately. In “Critic’s Credo,” in 1955, Berger distinguished studio criticism, which he likened to a form of teaching, from public criticism, written for the general reader, whose task was to connect works of art with ordinary life. Collaboration is a real calling, and the sneer was never its only alternative.

But Berger is worth following a little further. In November 1972, at the Café Royal in London, he accepted the Booker Prize for his novel G. and used the speech to point out that the prize’s sponsor, Booker McConnell, had held trading interests in the Caribbean for more than a century. He gave half the money to the Black Panther movement in Britain and kept the rest for his book on migrant workers. “The half I give away will change the half I keep.” The man who wrote the credo about responsibility to artists also stood up at a dinner held in his honour and told his hosts, in their own room, exactly what he thought of them. The two were never incompatible.

Which is the distinction the old quarrels obscure. An opponent is not an enemy. In 1965 Edmund Wilson reviewed his friend Vladimir Nabokov’s translation of Eugene Onegin in the New York Review of Books, harshly. Nabokov replied in the same pages, point by numbered point. Wilson’s answer, printed beneath, began: “I must acknowledge two errors.” Others joined, and the exchange ran from July of that year into the following January. It is often said to have ended their friendship. What it did not end was the argument: both men stayed in the room, under their own names, for six months, and readers could watch them be right and wrong in turn. That is the thing Leavis’s lecture and the potted shrimp both lack, and the thing Berger at the Café Royal had. Not hostility, which is cheap, but a signed position, held in public, by someone who expects to be answered and stays to hear the answer.

V

The Ledger

If what has gone missing is cover for the signed claim, the place to look for it is in what the outlets themselves say. On 2 October 2026 the published guidelines of twelve leading review outlets in English were captured and read: the New York Review of Books, the London Review of Books, the Times Literary Supplement, Bookforum, The Point, The Baffler, n+1, The Nation, Harper’s, The New Yorker, the New York Times Book Review, and the Guardian’s books pages.

Three publish no guidance for reviewers at all. Seven publish guidance that says nothing about judgment, favourable or otherwise. Two say something. The Point tells contributors it wants critical argument rather than straight reviews, which is a statement about the form of judgment, not about its direction. And the New York Times Book Review’s then editor, describing its practice in 2016, said that books the editors liked could still receive negative notices, and that a reviewer is disqualified who knows the author, shares an agent, or blurbed an earlier book. The Guardian’s Editorial Code, revised in March 2026, bars writing about a relative, a partner or a close friend.

So the claim that survives is narrow, and anyone can check it by opening those pages. No outlet among the twelve publishes a policy on negative reviewing. The only rules that touch the question are conflict-of-interest rules, and they stop at friendship and at favours already done.

In a library flooded with amber syrup, a brass scale stands on a stone plinth under a shaft of light. A sealed, unopened envelope outweighs a handwritten page with one line struck through in red, and its pan sinks toward the syrup.
The weight of a favour not yet done, made for this essay.

That is not negligence. Conflict rules exist to catch known relationships, and they do that job well. A reviewer who has blurbed the author is visibly compromised, and the rule removes her. What the rule cannot reach is the reviewer who has not yet been blurbed and would like to be, who has not yet sat on the author’s jury and might. Nobody has a relationship with a future favour. It appears on no disclosure form, because it has not happened, and it may never happen. It only has to be possible. The peer economy begins exactly where the conflict rules end, and no code could follow it there, because what lives there is not a conflict. It is a career.

None of this measures how common negative reviews are now. Hardwick’s 4.7 per cent described 1956, by her report, and it is not a measure of anything since.

A magazine written by working writers sits inside the same economy, and this one is no exception. Its essays have turned to history, the classics and ideas, and it has not published a signed review of a new novel, favourable or otherwise. Its contributors write within the same network as everyone else’s. The finding about twelve outlets can be checked, and if necessary corrected, by anyone who opens their pages.

The last time this complaint was made with real force, it had a sequel. In the winter of 1962 to 1963 a printers’ strike shut the New York newspapers, and with them their Sunday book sections. Hardwick, Robert Lowell, Barbara and Jason Epstein and Robert Silvers, the editor who had commissioned her essay at Harper’s, used the silence to start the New York Review of Books. The answer to “The Decline of Book Reviewing” was not a better argument. It was a place to stand.

Sixty-three years later, the New York Review is one of the twelve, and one of the three that publish no guidance for reviewers at all. That is not a charge against it. Its pages are where Wilson and Nabokov argued for six months, under editors who printed both. The fact is the plainest evidence of something simpler. The shelter was its editors, and what they would print was never a rule anyone wrote down.

Whether that answer can be given again is an open question. It may be that the fringe needs only a new shelter, and that someone will build it. It may be that the existing shelters could decide to be one, and say so in the one page no outlet has yet written. Or it may be that the gentleness Hardwick deplored has found a better defence than she imagined, in a culture of writers who would rather sustain one another than judge, and who have weighed what that costs. The guidelines pages cannot say which.

⁂

Written in full collaboration with the machine, and the ledger requires the names be exact: drafted and edited with Claude Fable 5.1 and Claude Opus 5.5; copy editing by Gemini. No quotation in this essay was recalled; each was checked against a text of its source, the original or a reproduction of it, and any text reached secondhand was confirmed in two reproductions or paraphrased.

Elizabeth Hardwick is quoted from “The Decline of Book Reviewing,” Harper’s Magazine, October 1959, in the magazine’s online archive; Emerson’s phrase is quoted as she quotes it, and the figures from the Wayne University study of Book Review Digest for 1956 are given as she reports them, since no second source for the study has been found. Rosalind Krauss is quoted from “Video: The Aesthetics of Narcissism,” October 1 (Spring 1976), pp. 50–64. F.R. Leavis is quoted from the Richmond Lecture as printed in the Spectator, 9 March 1962, in reproduction. Susan Sontag’s closing sentence is from “Against Interpretation” (1964), in reproduction. John Berger’s “Critic’s Credo” (Art News and Review, 14 May 1955) is paraphrased; his Booker Prize speech of 23 November 1972 is quoted from the text reproduced by Verso and the Booker Prizes. Edmund Wilson’s reply to Vladimir Nabokov is from the New York Review of Books archive, 26 August 1965. Chris Kraus, Elvia Wilk and Ben Eastham are quoted from “Three Acclaimed Critics Reveal What Lured Them Into the World of Fiction,” Cultured, 28 September 2026. Jacob Silverman is quoted from Slate, August 2012, and Isaac Fitzgerald from his interview with Poynter, 7 November 2013.

The published guidelines of twelve outlets were captured on 2 October 2026 and classified as no guidance found, guidance silent on judgment, or guidance addressing it. A count of signed negative reviews of new novels across the same outlets, for the twelve months to 30 September 2026, was designed and signed off before counting began, piloted on one issue, costed at 200 to 400 hours with subscriptions to every outlet, and abandoned on grounds of method; no figure from the pilot is used. A signed negative review of a praised contemporary novel would be a different piece, with its own apparatus, and a full count by readers with access to every outlet would be a companion project. The verification ledger, with every outlet’s URL and classification and the source for each quotation, is in the source of this page, for anyone who wants to check.

The header and interior images were generated with Gemini for this essay; they illustrate its argument and depict no real place or event.

On the collaboration → “The Third Thing”

WHAT MAKES AN ANSWER

On a professor’s case against writing, the written replies beneath it, and what Socrates actually objected to

Socrates said written words couldn’t answer questions. Under an essay repeating his complaint, readers answered by the hundreds, and the author answered them. Now there is a machine that answers anything, and Socrates had another objection it doesn’t meet.

On September 29 The New York Times published a guest essay by Joe Cruz, who chairs the philosophy department at Williams College, under the headline “I’m a College Professor. Writing Isn’t as Important as We Think.” Halfway through, he turns to Plato. In the Phaedrus, Cruz writes, Socrates holds that writing “cannot speak, answer questions or come to its own defense.” Perhaps, Cruz suggests, Socrates was right, and we think best by speaking, answering and defending our ideas in person.

Beneath the essay, readers wrote back. By the next day the thread held about 1,200 comments. They asked him questions. They answered each other. They pressed him on his evidence, on the size of his seminar, on the research he had linked. And the author answered, in writing. His first reply began with a thread he “didn’t have room to pursue in the essay.”

None of this proves Socrates wrong, and it doesn’t prove Cruz wrong either. His case rests on an oral exam, and the thread says nothing about oral exams. But it raises a question the essay never asks. If writing can answer back, what was Socrates objecting to? And what do we have now, when there is a machine that will answer anything put to it, at once, in writing?

I

What Socrates Said

Late in the Phaedrus, Socrates tells a story set in Egypt. The god Theuth, inventor of letters among much else, brings his arts to the king, Thamus, and praises writing above the rest: it will make the Egyptians wiser and improve their memories. The king isn’t persuaded. “The specific which you have discovered is an aid not to memory, but to reminiscence,” he answers. Those who learn from it “will appear to be omniscient and will generally know nothing.” They will be tiresome company, “having the show of wisdom without the reality.”

Socrates then adds his own complaint, and it’s the one Cruz borrows. Writing is like painting. Its figures look alive, “and yet if you ask them a question they preserve a solemn silence.” Once written down, words “are tumbled about anywhere among those who may or may not understand them,” and when they’re attacked, “they cannot protect or defend themselves.”

So there are three objections, not one. Written words stay silent when questioned. They drift away from their author to any reader at all. And they give their students the look of knowledge without the substance. Cruz’s essay rests on the first. The other two go unmentioned.

A paper can’t be asked a follow-up question, and a student sitting in front of an examiner can. That is Cruz’s real evidence: a forty-five-minute oral exam, pressing on claims each student had put in writing with a machine’s help. It’s a better test of understanding than the finished paper was. The retort that came back most often, that he made his argument against writing in writing, is cheap. Plato wrote the Phaedrus, and Cruz grants that writing has a lasting job: it “summarizes and compresses our thoughts.” He is also right, and unusually candid for a professor, about why universities love the essay. It can be carried home, read in pieces, graded in batches and cited if a grade is disputed. None of that has much to do with thinking. His history holds up as well. The idea that we discover what we think by writing it does trace to James Britton, who put it forward in 1980. One commenter disputes his description of a study he links, which could not be checked here.

II

The Half That Did Worse

Cruz’s most striking result is that “at least half of the class was more agile” in the oral exam than on the page. The reply that drew the most recommendations, from a reader signing as M Bk in Brooklyn (555 when the thread was captured), turned the sentence over: “that means half wasn’t.” The arithmetic is exact. If at least half the class did better aloud, as many as half did not. Some of those students, the reply went on, think better with the time and solitude that writing affords, and for others an oral argument can be its own kind of refuge, a way to avoid the slow sorting a page demands.

Readers in the thread described themselves. Some said they were shy. Some said they can’t compose aloud at all but write well. Some described conditions that make speaking on demand harder than writing. And a reader signing as KWP, in Arizona, retired from teaching at a large state university where junior and senior courses often ran past a hundred students, said that however much he encouraged discussion, some students never spoke, and writing short papers was “the only way I could hear their voices.” A seminar isn’t made fair simply by being spoken. It rewards whoever finds speech easiest in a room, and that isn’t the same group as whoever thinks best.

It’s worth noticing where that testimony was given. The readers who said they can’t think aloud on demand said so in a medium that let them take their time, and they were heard in it by more people than any seminar holds. A reader who would have sat silent through a discussion could write two careful sentences, reread them and post them, and find them recommended by strangers. The thread wasn’t a seminar, and nothing in it proves how any of those readers would have fared in Cruz’s room. But the quiet half didn’t have to be imagined. It described itself, at length and in its own words, in the one form of answering the op-ed proposed to set aside.

In a dusk-lit seminar room, a standing student argues while the others turn toward her; at the far end of the table a quiet student writes in a notebook whose glow is the room’s warm light
The other half, made for this essay.

The professions that rely most on spoken examination have had to deal with this. In 1975 R. M. Harden and three colleagues described a new format for testing medical students, the objective structured clinical examination. It was designed, they wrote, “to avoid many of the disadvantages of the traditional clinical examination,” and to be “more objective.” That paper concerns clinical exams in medicine, not seminars in philosophy, and it shouldn’t be made to say more. It does show that a field built on spoken examination found reasons to structure it.

There’s another way to read Cruz’s result, and it takes his evidence more seriously than his conclusion does. His students didn’t arrive at the oral exam empty-handed. Each had already produced a written argument, with whatever help his guidelines allowed, and had turned it in to be examined. Cruz reads their fluency aloud as proof that writing was never where the thinking was. It could just as well be proof of the opposite: that the paper, however it was made, gave each student something definite to defend, and the exam measured how well they had come to know it. A student who has fixed a claim on a page, seen it in order and been told it will be pressed walks in with a map. The oral exam may be testing what the writing built. Nothing in one semester’s results can decide between the two readings, and Cruz doesn’t claim they can. The second asks nothing of him except that the forty-five minutes be credited to the paper as well as to the talk.

Cruz’s evidence is one class, last semester, at Williams. He says as much himself: there will be challenges of scale, and not every institution can adapt. The question his half raises isn’t whether the oral exam works for the students who shone in it. It’s what happens to the ones who didn’t, when the page, the one place they could be heard, is the thing being set aside.

III

The Thesis That Improved

Cruz’s first reply in the thread went further than his essay had. Writing, he wrote, “is itself a technology for cognitive offloading.” It lets us keep memory, attention and concentration out in the world rather than in our heads, which is enormously useful but may mean we exercise some capacities less. He suspected that “ease of assessment is a major driver” of the humanities’ reliance on it. Then he asked a question the op-ed hadn’t: “What kinds of thinking should we be willing to offload,” and which do we want students to keep practicing for themselves?

That’s a better question than the one his essay asks. “Writing isn’t as important as we think” invites an argument about ranking, speech against the page. His reply’s question is about trade-offs, and it applies as much to the machine as to the pen. It’s also the question his readers took hardest. Several objected that writing doesn’t offload thought at all, it builds it. The objection has a well-known ancestor in Walter Ong’s 1986 essay, whose title makes a stronger claim still: “Writing Is a Technology that Restructures Thought.” One reader brought it into the thread. Another, signing as gnowxela, in New York, put it in a sentence of his own: “Text gives complex abstract webs of ideas visible manipulable form.”

Cruz answered him. Our cognitive technologies, he wrote, are likely to have both expected and unanticipated consequences. And he agreed: writing is “no small thing,” and may be “on the short list of profound ways” we have increased our ability to handle abstraction. What he’d meant, he said, was that other paths run alongside it and deserve a reckoning.

Set his headline beside that sentence. In the op-ed, writing isn’t as important as we think. A few hours of written questioning later, it’s on the short list of the most profound tools the species has for thinking in abstractions, and the argument has become one about what else should be practiced alongside it. That’s a thesis getting better under questioning, which is exactly what Cruz says conversation is for. He deserves full credit for it. He read the objections, answered them in his own name, and gave ground where the ground was theirs. His reply was itself a second pass: the argument of the op-ed, read back as if someone else had written it, and answered. It is the thing he asks his students to give up, done well and in public. Most authors of guest essays do none of these things. But the conversation that improved his argument was conducted in writing. It could be reread before it was answered, and quoted back at the writer. It could wait for his reply without losing a word, and it’s still there for anyone to check what he conceded.

The page has failures of its own, and the thread had them too. Once a quotation starts to travel, nobody is present to vouch for it, and the replies that arrive early and sound sure gather the recommendations that decide what later readers see first.

IV

The Second Pass

Cruz’s proposal deserves to be stated in his terms. He wants a university in which ideas are expressed in real time, and thought unfolds as debate, oral defense or group discussion. He asks for courage on both sides: from faculty, to demand “active, participatory mastery,” and from students, to “abandon the safe recursion of revision.”

That phrase is where his argument and this essay part most sharply. In Cruz’s account, revision is a refuge: the second pass lets a student hide behind a polished page. What he wants instead is thought that happens once, in the room, under pressure. There’s a real gain in that. A student who can’t revise can’t pretend, and the examiner sees the mind at work rather than its best draft.

But the second pass is also where a thinker finds out that the first was wrong. Revision is the one moment a writer is asked to read what she has made as if someone else had made it, and to answer it. It’s slower than debate and less exciting than defense, and nobody is watching. It is also, whatever else it is, a form of taking something back.

V

The Partner Who Recants

The chatbot answers. Ask it a question and it doesn’t keep a solemn silence. It will answer for as long as you keep asking. On Socrates’ first objection, the one Cruz relied on, the machine passes. The third objection is harder. Thamus warned of students who “will appear to be omniscient and will generally know nothing.” A fluent answer isn’t knowledge. On a bad day, that describes a student who performs well in an oral exam as much as it describes a machine. A reader signing as Alain, a physician who teaches residents in Manizales, Colombia, put the difficulty in one sentence: “an AI that tends to agree with its user is not a Socratic partner.”

The Phaedrus shows what a Socratic partner does, and it isn’t simply disagreeing. Phaedrus arrives in love with a speech by Lysias arguing that a boy should favor a man who doesn’t love him over one who does. Pressed to do better, Socrates takes the same side and makes a speech of his own. He covers his head to do it: “I will veil my face and gallop through the discourse as fast as I can, for if I see you I shall feel ashamed.” When he finishes, Phaedrus wants the other half, a matching speech on the advantages of the non-lover: “I thought that you were only half-way.” Socrates declines, says there has been enough of both, and gets up to cross the river home. Then comes the moment the dialogue turns on. “As I was about to cross the stream the usual sign was given to me,—that sign which always forbids, but never bids.” He turns back and calls what he has just said foolish, “to a certain extent, impious.” Then he recants: “I will be wiser than either Stesichorus or Homer, in that I am going to make my recantation.”

Nobody argued him out of it. His listener wanted more of the same. The check came after the work was done, when he was already on his way home, and it came from inside him, unasked. That’s the answer the thread couldn’t give Cruz and the machine doesn’t give anyone. It’s often absent in a student with a deadline too. It’s the answer a mind gives to itself against what it has just produced. A machine that replies is new. A machine that could stop itself at the stream would be something else again. Whether one can be built, and whether any student would choose the one that did, the dialogue doesn’t say.

⁂

Joe Cruz’s guest essay, “I’m a College Professor. Writing Isn’t as Important as We Think,” appeared in The New York Times on 29 September 2026; his replies are quoted as he published them in the comments beneath it. The comment thread was captured on 30 September 2026, and the comment total and recommendation counts given here are as they stood then. Readers who commented are identified only by the handle and place they gave, quoted briefly and only for their own arguments; no one is quoted in order to be corrected. One reader disputes Cruz’s description of a study his essay links; the study could not be recovered for checking. Plato’s Phaedrus is quoted in Benjamin Jowett’s translation, and every quotation from it was confirmed in at least two online texts of that translation (Monadnock, the University of Northern Iowa’s reading text, williamwolff.org and Plato in Depth). R. M. Harden, M. Stevenson, W. W. Downie and G. M. Wilson, “Assessment of clinical competence using objective structured examination,” British Medical Journal (22 February 1975), is quoted from its abstract via PubMed Central. Walter J. Ong’s “Writing Is a Technology that Restructures Thought” was published in 1986; James Britton’s “Shaping at the Point of Utterance” in 1980. The header and interior images were generated with Gemini for this essay; neither is a historical illustration or depicts any real person. Drafted with Claude Opus 5.5; literary editing by Claude Opus 5; copy editing by Gemini.