1. Four Inspectors, Four Thousand Mills

In 1833 the British Parliament passed a Factory Act. It provided that no child under nine could be put to work in a mill; that children between nine and thirteen could work no more than eight hours a day and forty-eight hours a week; and that young persons between thirteen and eighteen could work no more than twelve hours a day.

These numbers are compressions. What conditions looked like before the compression is something the statute never had to spell out — anyone reading it could work that out for themselves.

A law that sets out to forbid children under nine from entering a mill presupposes that children under nine were, in fact, already inside the mills. A law that limits children between nine and thirteen to eight hours a day presupposes that they had previously been working longer. A statute only ever writes down the stretch it means to stop; whatever it cannot stop, and whatever has already happened, is left inside the statute's silence.

To make these numbers stick, Parliament simultaneously established a system of factory inspection. There were four inspectors in all. The mills that needed watching at the time numbered roughly four thousand. Later parliamentary records would concede a fact that needed no dressing up: the Act was widely evaded.

Four against four thousand — the ratio alone is a verdict. It shows that the lawmakers knew exactly what needed regulating, and that the arrangement, as designed, was never meant to actually enforce it. The length of a day had been written down as a number, and whether that number could be made real depended on whether anyone went to look, how many went to look, and whether they could keep up with what there was to see. The moment a figure leaves the page, it has to be handed back into human hands.

This has happened more than once in the essays before this one. In Essay 3, the weight stamped on the face of a coin depended on someone willing to check the stamp. In Essay 5, the columns in a ledger depended on someone willing to look closely. In Essay 6, the characters printed on a note of jiaozi depended on a government still willing to honor them. A rule can be written as hard as you like. A rule cannot walk on its own.

Eleven years later, the Act of 1844 compressed the working day for women and for young workers between thirteen and eighteen to twelve hours, and wrote in some of the earliest health and safety provisions on the books, one of which required that machinery be fenced. The Ten Hours Act of 1847 compressed the statutory hours of women and young workers in textile mills still further, down to ten hours.

Three acts of legislation spanning fourteen years, all moving in the same direction, tightening notch by notch. None of them started from a principle. Each was squeezed out of a running argument: the millowners said that any further compression would shut the mills down; the doctors said that children's bones would not grow straight; the clergy said there was no time left to attend church; and the workers themselves stood at the gate and would not go in. What finally landed on the page, each time, was a handful of whole numbers.

The whole number is the shape the construct likes best. A childhood, a body still growing, the exact point at which a person standing beside a machine begins to make mistakes — none of these are whole numbers, and none of them ought to be. But legislating, inspecting, ruling in a courtroom on who has broken the law, all of that requires a whole number. The line between eight hours and twelve hours can be drawn anywhere at all and a reason can be found for drawing it there, and once it is drawn, the two sides of it become legal and illegal. Once the construct moves to act, it can only act with whole numbers.

Every round of legislation was a fight over the same thing: how long a day could be stretched. And every time the fight ended, the outcome had to come down as a number, written into a statute, and handed to someone to check.

Someone in London spent years reading the records the inspectors had written down, and copied long passages of them, wholesale, into a book he was writing. That book was published in 1867, under the title Capital, Volume One.

He was not copying those records as decoration. What he needed to prove required exactly those figures: how far a day can actually be stretched is not decided by technology, and it is not decided by morality — it is decided by force, and force shifts. Wherever the line of the working day gets drawn is a position each generation fights out anew; once fought out, it becomes a statute, and the statute is handed to someone to check.

In the ten essays before this one, the construct has always been the one keeping the books, and the remainder has always been the thing spilling over the edge. The construct's ways of dealing with that remainder have taken several forms — being absorbed into a larger construct, being renamed, being ignored, being folded into cost, having the boundary drawn in tighter around it. This time is different. This time, someone kept the books for the remainder instead.

He gave that spillover a name, pointed to exactly where it was produced, and gave it a quantity that could be calculated. This is the first time, in this whole line of essays, that someone did the accounting not for the construct but for the remainder, and did it with such rigor that people are still checking his arithmetic today.

And an account kept for the remainder is still an account.

2. The Hour That Wasn't His

The first thing Capital, Volume One does, in its opening pages, is drag a set of utterly unrelated things onto a single scale: linen, coats, gold, iron, bread.

Marx's claim is that these things can be exchanged not because of any physical resemblance between them, but because all of them can be reduced to the same social substance — abstract human labor. Value, on this account, does not depend on the natural properties of a thing; it depends on socially necessary labor time. He defines the phrase with some care: the time required to produce something under normal conditions of production, given the average degree of skill and intensity of labor prevailing in that society at that moment.

The weight of the definition sits on the word average, and the example he chooses squeezes out every ounce of that weight.

Suppose an English power loom cuts in half the time needed to weave a given length of yarn into cloth. A weaver still working by hand, even if he spends exactly as long at his loom as he always has, now has his hour counted as only half an hour of social labor, and the cloth he weaves has its value cut in half along with it.

This weaver has not grown lazy. He has not grown clumsy. He has not changed a single motion of his hands. His day is still twelve hours long, his back still bends the same way, he is still in the same room he has always worked in. What changed was a machine somewhere else. A scale standing outside him has revalued his hour as half an hour, and he has no standing even to object, because that scale does not ask after individuals.

There is a harder thing to see buried underneath this. What the weaver has lost is not money someone owes him — no one has shorted him, no one has broken a contract with him. What he has lost is his hour's allotment of social recognition, and that allotment was never his to hold onto in the first place. He believed he possessed a trade and a pair of hands. What he actually possessed was a ratio that could be revalued from outside at any time.

This lays the matter bare with great cleanness. The scale is not in the hands of the person doing the work. The scale sits on society's side of the ledger, and it can rewrite, at any moment, what a person's time is worth, for reasons that have nothing to do with that person.

Marx goes on to explain that this scale must eventually find a single, unified expression, and that expression is money. Price is nothing but the monetary expression of value; money converts a commodity's value into an independent, homogeneous form. He used a striking phrase for this: commodities become, in effect, social hieroglyphics — people first write them, through the value-form, into a set of exchangeable symbols, and are then forced to turn back and ask what those symbols are actually saying.

Scholars who study this closely have warned that reading the theory purely as an inventory of labor-hours misses more than half of it, because it is simultaneously a theory of money and a theory of social relations.

This connects directly to Essay 3. The coinage of Lydia, back then, gave the scale a portable shape, and once the scale existed, it did not manufacture any new differences in the world — it simply gave a place to differences that had previously had nowhere to stand. Abstract labor, as a scale, does exactly the same thing, except now its target is a living person's time: the same twelve hours in a day can be counted as twelve hours in one case and six in another, and that comparison could not even be formed before the scale existed. Every scale the construct adds does not add anything new to the world; it only puts another set of things into positions where they can be measured against each other.

And whatever overflows this scale is already written on the very same page. What enters the calculation is never the specific thickness of wear that a particular worker's life absorbs in that hour — it is the average, socially recognized, comparable time that a given kind of labor is supposed to take. Everything slower than the average, everything faster than the average, everything more damaging to the body, everything that takes years of accumulated skill to manage, everything that only holds together because someone else at home is quietly covering the gap, all of it has to be flattened before it can enter that scale at all. Flattened does not mean it disappears. It comes back somewhere else, in some other form.

Where it comes back is usually not on the books. It might come back in the silence a person carries after work. It might come back in a back that will not straighten out by the time he is forty. It might come back in the extra hours someone else in the household has to put in. All of this genuinely happens, all of it carries a cost, all of it can be described — none of it, however, folds into the phrase socially necessary labor time, because by definition that phrase is already computing an average.

The name Marx gave to this whole arrangement is commodity fetishism. He wrote that a definite social relation between men themselves assumes, in their eyes, the fantastic form of a relation between things.

That sentence is often read as saying that people are seeing an illusion, and read that way it comes out far too light. What he means is much heavier than that: the market has genuinely succeeded in writing social relations into price, into money, into the commodity-form, so that domination between people, dependence between people, the averaging-out of people, competition between people, really do operate on the surface as relations between things. Anonymous exchange is not fake. It has simply rewritten human relations into the grammar of objects.

3. What Is Actually Sold

The most decisive change Marx made to the labor theory of value he inherited was not to repeat, once more, that value comes from labor. It was to split labor from labor-power.

The split was forced on him. If what a worker sells on the market is labor itself, then it becomes very hard to explain how a capitalist could go on profiting, continuously, in a world where everyone trades at equivalent value. If the answer is that the capitalist simply pays less than he owes, that is fraud, that is breach of contract, that is a question of one person's character, and it has nothing to do with the system as such. What Marx wanted to explain was precisely the case in which there was no fraud at all.

This is a very severe form of self-restriction. He does not write about the millowners who shortchange wages, or the ones who tamper with the scale, or the corner-cutting and the violent evictions. All of that existed, and existed everywhere. But if an argument leans on any of it, the only conclusion it can reach is that some people are bad. The conclusion Marx wanted was that even if everyone behaved with perfect propriety, the outcome would still come out the same.

So he swapped the commodity. What the worker sells is not labor, it is labor-power, the capacity to work over a given stretch of time. This commodity carries two separate prices. Its value is fixed by whatever necessities of life are needed to maintain and reproduce that capacity. Its use-value, however, is capable of creating, in the very process of being used, a new value greater than its own. Surplus value is produced exactly in the gap between these two.

His own example is easy to follow. Suppose a worker's daily necessities are worth three shillings, and producing those necessities takes, on average, six hours of labor. Then this worker's labor-power, for that day, has a value of three shillings. The capitalist buys a day of his labor-power for three shillings, and has not underpaid him by a single coin.

The trouble is that what has been bought is not six hours, it is the right to use that capacity for an entire day. Once the worker has put in six hours, he has already reproduced the value of his own three shillings. But he still has another six hours ahead of him. Whatever value those six hours create does not belong to him.

Marx calls this a trick the capitalist carries out entirely inside the rules of equivalent exchange. What matters is the form the trick takes: no contract is broken, no law is violated, and on paper, the books balance perfectly. Wherever the imbalance actually lies, it is not on the paper.

In an 1865 address to a workers' organization, he laid this distinction out even more bluntly, and in passing drew a line that mattered. What the worker sells, he said, is not directly his labor but his labor-power. And immediately afterward: if a person were allowed to sell that capacity for any indefinite period whatever, slavery would be immediately restored.

That sentence is not merely a technical proposition. In Essay 8, the line ran clean through a person, someone pressed head to foot into the plane of price, entered onto an insurance policy, counted off in tonnage. In the wage relation, Marx points out, what stands between the two is a term: what is sold is a time-bounded right of command, not the sale of the whole person outright. What keeps the wage relation from becoming that other thing is not goodwill, and it is not a moral standard, it is that limit on time. Take the limit away, and the boundary disappears immediately.

This line holds, but it is thin. It separates two things using nothing but a single time parameter, and a time parameter can always be stretched: sixteen hours a day, seven days a week, a debt that never gets paid off, a contract nobody can walk away from, a place of residence nobody is allowed to leave. Every notch it gets stretched, the line grows a little thinner. The point of that sentence is not that the other thing is safely in the past, it is that what separates it from the present case is only a quantity, and a quantity can be adjusted.

Precisely because of this, labor-power as a commodity can never, by its nature, line up cleanly with the whole person. What the contract transfers is the right to use a capacity, and the capacity happens to live inside a person who grows tired, falls ill, gives birth, refuses, goes on strike, and dies.

Marx has a line of quite extraordinary coldness: the use-value of labor-power, that is, labor itself, belongs no more to its seller than the use-value of oil, once sold, still belongs to the merchant who sold it. What the contract transfers is the use-value; the pain, the exhaustion, the aging, the humiliation, all happen in the process by which that use-value is actually consumed.

The work of pricing keeps subdividing further. Time-wages cut a day into hours; piece-wages settle labor by the piece. Marx eventually puts it bluntly: piece-wages are nothing but a converted form of time-wages, and time-wages are themselves nothing but a converted form of the value or price of labor-power. What all of this accomplishes is plain enough: it turns a person's working capacity into an item that can be priced, settled, compared, and supervised exactly the way oil, cotton, coal, and machinery are priced, settled, compared, and supervised.

4. The Historical and Moral Element

Pushed this far, one question has to be answered: how, exactly, is the value of labor-power determined?

Following the logic of the system, the answer ought to run like this: calculate the necessities of life required to maintain and reproduce this capacity, add up the value of those necessities, and that sum is the value of labor-power. Everything in the book up to this point has been derived this same way, from commodity to value, from value to money, from money to labor-power.

At this point, he cannot push the derivation any further.

The reason is not hard to state. What belongs in the basket called necessities of life depends on what a given society believes a person minimally needs in order to live. How many loaves of bread. How often meat is eaten. Whether the dwelling needs a window. Whether the children need schooling. Whether there is room to lie down when someone falls ill. None of this can be derived from the conditions of production. All of it is the outcome of a group of people arguing it out, and different countries argue it out differently, and the same country argues it out differently in different eras.

So, in the chapter dealing with the value of labor-power, he sets down a genuinely strange sentence. This value, he writes, contains not only the physiological minimum required to sustain a worker's body, but also a historical and moral element.

The historical and moral element is a term that cannot be derived from the system at all. What it means is that what counts as a livable life is not something biology can compute. It is bound up with custom, with struggle, with social standards, with levels of education, with how many family members are quietly backstopping the household without pay, and with what kind of life the people of that country generally regard as a life worth living. Every one of these things shifts over time, and every one of them is arrived at through people arguing, not through calculation.

Marx did not sidestep it, and he did not pretend he could derive it after all. He wrote it into the definition, admitted that it belonged there, and kept going.

This move needs to be kept separate from concealment. He did not bury that term in a footnote, and he did not dress it up with a technical-sounding name to make it look as though it had been derived. He used the words historical and moral, words that sit uncomfortably in a book aiming to be a work of science, and knowing exactly how uncomfortable they were, he wrote them down anyway.

This is where the remainder sits on Marx's own books. It is not that he failed to see it. He saw it, named it, set it down in plain sight, and did not fold it into the equation.

What is interesting is that, on this very point, his opponent had made a move heading in the exact opposite direction but shaped exactly the same way. Essay 10 described how Menger wrote the possibility of human misjudgment directly into his definition of value, and then wrote straight past it, the thinkers who came after him usually handled that gap by simply assuming judgment was correct. Marx, for his part, wrote a whole society's collective judgment about what makes a livable life directly into his definition of the value of labor-power, and then also wrote straight past it, the thinkers who came after him usually handled that term by giving it a fixed basket. Two systems, launched from opposite directions, each left behind, at the single most load-bearing point in its foundation, a term that could neither be derived nor removed.

Dignity, then, does not come in through the front door. It is smuggled back into the theory through the back, as the historical and moral remainder that cannot be cleared out of the determination of value.

Whatever comes in through the back door shares one feature: it gets in, but it never gets a seat. It lives in the definition, not in the derivation; it shapes the outcome without ever taking part in the calculation. Wherever the calculation actually proceeds, that term is simply treated as a given constant, even though it happens to be the least constant term in the entire book. When the construct's own derivation runs up against something it cannot derive, its solution is to acknowledge the thing, name it, and then fix it in place, so that the steps that follow can keep moving.

This is neither an oversight on Marx's part nor a triumph, scoring it as a point for either side misses what it is. That term has to be there, and the reason is entirely real: a wage is never a transaction with just one person's physiology. It is a transaction with what an entire group of people believes counts as a livable day. And that second matter can only be settled by custom, by struggle, and by the age one happens to live in, settled once, and then settled again by the next generation.

Precisely because of this, the wage figure is not what it looks like. It looks like a price, sitting on the same table as the price of cotton, the price of coal, the price of machinery. But underneath it presses something else entirely: the minimum standard of human life a society is currently willing to recognize. None of the other figures on that price table carry this extra layer. Cotton does not rise in price because people have decided it deserves a better life.

One more thing is worth noticing inside that term, something pointed to but never given a column of its own: the labor performed inside the household that never enters the books at all. It appears in this book only as a condition for labor-power to be reproduced, never as labor in its own right.

5. The Credit the Worker Extends

The wage relation looks like the most anonymous of all exchanges. It has a contract, a timesheet, a settlement day, a unit price; the two sides need not know each other, need not have any history together, need not trust one another at all. The kind of token that, in Essay 6, absolutely required mutual recognition in order to hold its value seems to have become entirely unnecessary here.

The whole benefit of a contract lies exactly here. It lets two strangers complete a transaction without either one investigating the other's background, without needing a guarantor to vouch for them, without waiting until some relationship has had time to form. This is a genuine liberation, and especially so for anyone previously shut out by their origins. Essay 10 showed how the phrase one ratio across the whole market smashed the old practice of pricing by pedigree; the wage contract does the same work, it strips a person out of their background and leaves only how much work they can do.

But in discussing how wages actually get paid, Marx quietly dismantles this whole impression.

He points out that in capitalist countries, labor-power is typically paid for only after it has already done its work. The worker works first; only at the end of the month or the end of the week does he see any money. Which is to say: the worker has, in effect, already extended the use-value of his labor-power to the capitalist on credit. He states this without any ambiguity at all: everywhere, it is the worker who extends credit to the capitalist.

This single line turns the wage relation completely over. On its surface it is the most purely cash-based, the most standardized-by-contract, the least dependent on anyone's character or connections of all exchanges. Yet in how it actually operates, there is always a gap between working and being paid, and inside that gap there is delay, there is default, there is the capitalist going bankrupt, there is the worker who never sees his money. The most impersonal-looking relation of all has, buried inside it, a stretch of time that depends entirely on trusting the other side to make good.

And the risk across that stretch of time is not shared equally. What the capitalist takes on credit is work already completed, work that has already become goods sitting in a warehouse. What the worker extends on credit is strength already spent, and once spent, it is gone, and there is no getting it back. If the mill fails, the goods can still be sold off to settle its debts; the worker's exertion from those days settles nothing, and appears on no such sheet. So this is not simply a stretch of mutual waiting. It is a stretch of time in which only one side has put up any collateral at all.

The worker, then, is calculated on one side as a capacity to be settled by the hour, while on the other side he is forced to bear, as a continuously existing person, the entire uncertainty of that stretch of time. What the contract records is a capacity. What is actually staked is a person. On paper these two things are written as though they were one and the same. In reality they never have been.

Essay 6, in discussing jiaozi, observed that stripping the metal away from money is the same thing as pressing its entire weight onto personhood instead. Here the direction runs the opposite way: the wage relation is straining with everything it has to drive personhood out of the transaction, replacing it with hours, piece-counts, unit prices, and settlement days. But it cannot be driven out completely. Trust has not been eliminated. It has simply withdrawn into the gap before settlement, into the part of the arrangement the contract never writes down.

Neither of these two media has ever simply replaced the other. Not once, not anywhere.

Counting back through the series, this same line has always been there. In Essay 1, in the age before ledgers, a debt of gratitude had to be backstopped by both sides remembering it. In Essay 2, a debt recorded on a clay tablet had to be backstopped by the temple's own hierarchy. Here, a wage has to be backstopped by a contract and by a mill that is still open for business the following week. Each time, the medium grows colder, cold enough, eventually, to erase every visible trace of personhood, and yet every single time, the cold end of it still connects to a stretch of road that only gets walked because somebody keeps faith. The construct can shorten that stretch of road again and again. It cannot remove it.

And that gap is bridged by the worker's own body. He has to keep showing up, his body has to keep holding out, simply to make it to settlement day; he has no other collateral to put up against that stretch of time. He is reduced, on one hand, to a capacity settled by the hour, and forced, on the other hand, to use himself, a continuously existing person, as the guarantee that backs the whole contract.

Put the two ends of this together and something becomes clear: the wage relation is, on one side, the closest thing to pure price among all forms of exchange, and on the other side, the form of exchange most dependent of all on a person's continuous, bodily presence. It does not choose between these two things. It pushes both of them to their absolute limit at once.

6. Where the Young Marx Went

In 1844, a twenty-six-year-old Marx wrote a set of notebooks in Paris. One passage from them has since been quoted more than any other: the worker, therefore, only feels himself outside his work, and in his work feels outside himself. He feels at home when he is not working, and when he is working he does not feel at home.

The same notebooks contain a line that sits even closer to the wage relation itself. What the worker receives, he writes, is only the portion absolutely indispensable to keep him in existence, and that portion gives him an identity not as a human being but as a worker. Further on he puts it more starkly still: if capital no longer needs him, he has no existence as a human being at all, only as a worker.

These notebooks are usually summarized as a fourfold separation: the separation of a person from his own product, the separation of a person from the activity of laboring itself, the separation of a person from his own species-being, and the separation of person from person. The more the worker produces, the poorer he himself becomes; the product turns back on him as an alien power and dominates him; labor collapses from a free activity into a bare means of staying alive; and once both the product and the activity have been alienated, the relation between one person and another is rewritten as domination and subordination.

These notebooks were written between April and August of 1844. They were first published in 1932.

A great deal happened in between. Marx finished Capital, Volume One, and died. Engels put together Volumes Two and Three from his papers, and died as well. The party built in Marx's name became a major party in several countries, and in one country his name was used to remake the entire system. All of this happened while those notebooks were still sitting in a box.

Add it up and the gap comes to eighty-eight years. Which means that for nearly fifty years after Marx's own death, everyone who read Marx was reading a Marx who had never said any of these things. Once 1932 arrived, the young Marx, full of talk of alienation and species-being, suddenly stood in front of everyone at once, and only then did a question become, for the first time, a genuinely systematic one: is there rupture or continuity between the early Marx and the mature Marx?

One school pushed rupture as far as it would go. The French philosopher Louis Althusser proposed what he called an epistemological break, arguing that the Marx writing before 1845 still remained at the level of ideology, and only afterward turned to a scientific critique of political economy, and that the concept of alienation presupposes some unified human essence, a residue that had not yet been cleared away.

Another school argued for continuity. Researchers who have worked through this thread in recent years hold that alienation was never abandoned, it was rewritten, carried over from the vocabulary of ethics and philosophy into commodity fetishism, abstract labor, the wage form, and capital's consumption of labor-power: a colder, more institutional kind of analysis. A major encyclopedia of philosophy notes, under this very heading, that the view holding Marx spoke of alienation only in his early years is by now very hard to square with the textual evidence.

Material can be marshaled on both sides, and neither side's case is weak. The continuity camp can match the texts line for line: in 1844, labor is said to be external, coerced; in Capital, the wage-form is said to produce an illusion in which, although only part of the labor is actually paid, the whole of it appears to have been paid. In 1844, the worker grows poorer the more he produces; Capital translates this into the claim that the value of labor-power is fixed by what is needed to reproduce it, while its use-value can, over a longer working day, go on producing additional value beyond that. In 1844, human relations are said to become deformed within alienation; Capital renders this same deformation as human relations appearing in the form of relations between things. The vocabulary has changed. The skeleton underneath is visibly the same one.

The rupture camp is not arguing from nothing either. In the later texts, the concept of human essence genuinely no longer carries any weight in the argument; the derivation now runs on the value-form, on labor-power as a commodity, on the laws of accumulation, not on any judgment about what a human being ought to be.

There is no need to settle this dispute here. What is worth treating as an event in its own right is the controversy itself. A stack of papers sat inside a box for eighty-eight years, and once it came out, an entire field of study had no choice but to go back and redraw the boundaries of its own author, deciding all over again which parts counted and which parts were to be treated as not yet mature. A generation's reading list can be struck from the record. The paper itself cannot be.

The effect of this on how the work gets read is entirely real. Every interpretation formed before 1932, no matter how rigorous, was built on incomplete material, and some of those interpretations had, by then, already hardened into institutions, been written into textbooks, been built into organizations. The text arrived late; the interpretation had arrived early. Once that order is set, the late-arriving text stops being a supplement. It becomes a problem.

7. The Column That Never Balances

Marx's own book has places that will not balance either, and more than one of them. There are at least three separate disputes, and they are often tangled together.

The first concerns value and price. The very title of Chapter 9 of Capital, Volume Three, announces it: The Formation of a General Rate of Profit (Average Rate of Profit) and the Transformation of the Values of Commodities into Prices of Production. In this chapter Marx concedes that commodities in any specific branch of production need not sell at their own individual value, some sell above their value, and by the same proportion, others sell below it, and it is precisely this deviation that allows a general rate of profit to form in the first place. The deviation is not the exception. It is the norm.

This volume is a posthumous compilation. Engels assembled it from manuscripts written between 1863 and 1883, and it did not appear until 1894. The controversy, from the very start, has carried the shadow of textual history: no one can be entirely sure which steps Marx himself had finalized and which steps his editor smoothed over.

The heaviest early artillery came from the Austrian economist Eugen von Böhm-Bawerk. In the 1890s he declared that Volume Three's account of average profit and prices of production could not be reconciled with the theory of value laid out in Volume One. The standard twentieth-century version of the critique is usually traced back to a scholar writing around 1906 and 1907: Marx converted outputs into prices of production but failed to convert inputs into prices of production with the same consistency, so that the resulting system of equations, solved under conditions of simultaneous determination, simply fails to close. This reading eventually became close to the textbook-standard account.

The counterattacks run at least three lines deep. One holds that there is no need to demand a one-to-one correspondence, for every single commodity, between labor value and money price, so long as the proportional relationship between total profit and unpaid labor time still holds, the argument for exploitation stands regardless. A second holds that Marx was never working with two parallel systems, a value-system running alongside a price-system, but with a single analysis of total social capital that presupposes capital advanced in money form from the outset, and that the so-called transformation problem is, to a considerable degree, an artifact manufactured by later readings of him. A third holds that price and value determine each other across historical time rather than being solved simultaneously at a single instant, and that the logical contradiction is something a static model has imposed on Marx from the outside.

Which of these is correct is not for this essay to decide. But the reason this controversy has dragged on for so long is worth stating plainly: it is not only a matter of whether the equations check out. What it exposes is that Marx needed to say two things at once that are extremely difficult to say at once: that the surplus is real, that it comes from living labor and not from any trick of buying and selling; and that the surface arrangement by which this surplus gets distributed will systematically distort its own origin. In other words, what he wanted was a quantity that is simultaneously real and necessarily obscured.

A quantity that is necessarily obscured will never find, anywhere on the books, the column that corresponds to it.

The same point can be turned around: any quantity that can find its own corresponding column on the books will never run into this particular problem. The transformation dispute has dragged on for more than a century precisely because the quantity Marx was dealing with is, by his own account, one that cannot be shown directly. He had to build an algorithm for something that is, by definition, invisible, and the only way to test that algorithm is to see whether the results it produces line up with the prices that actually are visible.

The second dispute is with marginalism. Essay 10 described what those three men did around 1871. Set the two approaches side by side and something more interesting than who was right comes into view: they were not asking the same question.

What Marx wants to ask is why, in a society generally organized around commodity exchange, utterly dissimilar kinds of labor can be reduced to a single social scale at all, and why profit can never come from simple buying low and selling high. What marginalism wants to ask is how, under conditions of scarcity, a single person compares his own subjective satisfaction at the margin, and how that comparison gives rise to demand, exchange ratios, and equilibrium prices. Some later readers have simply concluded that the two sides were, at times, answering different questions entirely.

The boundary drawn in Essay 10 can therefore be restated. What gets shut outside that boundary is not only justice, institutions, and time, it is a still deeper question underneath all of those: where commensurability itself comes from. Marginalism starts from commensurability already in hand; it assumes that different things can already be placed on the same axis and compared, and then studies how the comparing is done. What Marx is asking is exactly the step before that one: how that commensurability got built in the first place, what keeps it standing, and on whom its cost falls.

The third dispute sits inside Marxism itself, and runs deepest of all. Is value the abstract labor that already exists at the moment of production, or does it only become real once it has passed through the social recognition of market exchange? One school holds that abstract labor and value are not already sitting, ready-made, inside the product, simply waiting to be expressed, value has to win genuine social recognition through exchange before it counts as value at all, and is therefore not a natural property of things but a historically specific social relation. The other school insists that the opening chapter is describing abstract human labor already expended in production, and that exchange does nothing more than express that labor as a money price; it does not create something that did not exist before.

This last dispute looks, on the surface, the most academic of the three, but its consequences are the most immediate. It determines exactly how fetishism ought to be read: as a veil, with an already-existing labor-content sitting pressed underneath it, or as the claim that value can only ever exist in this social form to begin with, so that lifting the veil reveals no other, more real quantity waiting there beneath it.

8. Two Pens

One last matter remains unaccounted for: where did that discovered surplus eventually get connected to.

The passage most often reached for as evidence is the preface Marx wrote in 1859. It contains a line quoted more times than almost any other he wrote: no social order ever perishes before all the productive forces for which there is room in it have been developed. Immediately afterward, the Asiatic, ancient, feudal, and modern bourgeois modes of production are laid out as a sequence of epochs.

Read only these two passages and it is very easy to come away with a strongly staged, linear reading: every social formation has its own historical task, and only once its productive forces are fully developed does it give way to a higher one. One school of twentieth-century scholarship built precisely in this direction, reconstructing the whole of historical materialism with the development of the productive forces as its core.

Another school immediately pointed to the trouble hiding inside that reading. Buried within it is a kind of functionalism: some institution survives because it promotes the productive forces, and yet no one can specify the actual mechanism by which that survival occurs. History ends up looking as though it has a purpose, even though that purpose belongs to no one in particular. Reach for an analogy with biological evolution and the theory becomes more contingent, but its long-range predictive power weakens accordingly; insist on keeping that predictive power strong, and the suspicion of teleology only thickens.

The material that pushes back against that straight line comes from Marx himself, and it is not light material.

In 1877, he wrote to the editorial board of a Russian journal to protest that someone had taken his historical sketch of how capitalism arose in Western Europe and rewritten it into a universal marching order that every nation was bound to follow. He left almost no room for misreading: turning that account of what happened in Western Europe into a universal path imposed by fate on every people, he wrote, would be doing him too much honor while at the same time entirely misunderstanding him. He pointed, too, to the example of the Roman plebeians who were stripped of their land, noting that similar events, dropped into different historical settings, lead to entirely different outcomes, which meant that no master key taken from outside history could ever unlock every one of them at once.

In 1881, he wrote to the Russian revolutionary Vera Zasulich, along with several drafts he never sent, and put the matter even more plainly: the Russian rural commune was not fated to pass through some Western-European ordeal of capitalism before it could move toward any other social form; whatever historical necessity might mean, he restricted it explicitly to the countries of Western Europe alone.

One line of research seizes on this body of late material and argues that the later Marx had already moved beyond the single-track developmental scheme, away from a Eurocentric road and toward a more global, more multi-linear view of history. Other researchers resist calling this a full reversal; their disagreement usually is not about whether the late texts show this kind of openness, but about whether that openness amounts to a revision and deepening of the old framework or a rupture at its root.

Put all of this together, and the reading that stays closest to the material is probably this: Marx worked with two pens. One liked to draw great arrows, sentences with direction, with stages, with an endpoint. The other kept getting forced back by concrete material into bending the arrow, running into the Russian commune, into the colonies, into paths that were not Western at all, it would bend the line already drawn, and sometimes break it outright. Both pens were his. Both left their marks. Neither one ever crossed the other out.

The arrow itself is worth pausing over on its own terms.

It is simply another way of balancing a ledger. If an account cannot be made to add up no matter how it is calculated in the present, then connect it to a future instead: it does not balance now only because the day has not yet come; once that day arrives, every item that failed to add up will be settled all at once. An endpoint is a way of making an imbalance bearable, and, at the same time, a way of making that imbalance look merely temporary.

Taken this far, the account kept for the remainder turns, once again, into a construct that wants to close. Its way of handling its own remainder is to push the day of settlement further into the future.

And Marx himself, twice, bent that arrow back with his own hand.

Both times, he bent it in the face of concrete material. Once when someone tried to use his book to explain a different country. Once when someone came to ask what should be done about the Russian village commune. He did not trim the material to protect his scheme. He revised the scheme instead. This is not a habit that has been carried forward very often by those who came after him.

His most solidly standing achievement, in fact, needs no arrow at all. A wage contract looks as though it settles the transaction, but it only ever prices a single capacity, it never settles accounts for a person's entire life activity, for the wearing-down of a body, for the uncompensated care given at home, for the particular taste of being treated as a line-item expense. This failure to settle is not the fault of some especially bad employer, cannot be canceled by a raise, and will not vanish on its own on some future day. It is simply what is left over, necessarily, whenever a person is treated as an item that can be settled at all.

He himself, in his own definition, left a name for this leftover part. He called it the historical and moral element: a group of people's shared judgment about what kind of day counts as a life worth living. That judgment carries no unit, cannot be added up, gets renegotiated by every generation, and will change again once it has been renegotiated. It cannot enter the equation, and it cannot be removed from it either, so it simply stays there, at the edge of the ledger.

A ledger can fold hours of work into hours, fold hours into shillings, fold shillings into a rate of profit, folding cleanly, layer after layer. What it cannot fold away is the person sitting in front of the machine. That person is not a column, and will not become one. He grows tired, falls ill, walks off, fails to show up, talks it over with the others, stops work together with them. Every time something like this happens, the figures on the books have to be adjusted all over again, and once they have been adjusted, that column underneath is still empty.

The ledger has not yet balanced. It is still being kept.