On Nietzsche’s demon, a fur hat in Prague, and what Milan Kundera cut from his own record

“… what do I see from the window beyond hats and cloaks that might cover artificial machines, whose motions might be determined by springs? But I judge that there are human beings from these appearances, and thus I comprehend, by the faculty of judgment alone which is in the mind, what I believed I saw with my eyes.”

René Descartes, Meditations on First Philosophy, II (1641), translated by John Veitch

A novel about forgetting opens on a doctored photograph and gives its most famous sentence to a man who is trying to doctor his own past. This essay follows a fur hat through four hands: the state’s on a photograph, the novelist’s on the same photograph, the same man’s on his own shelf, and an heir’s on a dead man’s notebooks. It makes one guess, graded.

By Michael Cummins, Editor · September 24, 2026

It was snowing on the twenty-first of February, 1948, when Klement Gottwald came out onto the balcony of the Kinský Palace on Old Town Square in Prague and spoke to the crowd that had filled it. The photograph taken on that balcony shows him at the rail with the men of his party around him. Behind him, close enough to touch, is Vladimír Clementis, who within a month would be foreign minister, and beside Clementis, in the back row, the photographer Karel Hájek. Both Gottwald and Clementis wear fur hats against the weather, as Hana Píchová, following Jindřich Toman, has pointed out; the original print is in the City of Prague Museum. The picture was reproduced by the hundred thousand. It hung in classrooms and on posters, and a generation of children knew it the way children know a flag.

That is the photograph. What happened to it is the subject of the most famous opening in postwar Czech literature, and the opening is not quite what the photograph shows.

Introduction

This week, on 3 Quarks Daily, Sean Murphy asked why Milan Kundera still matters and answered with one sentence from The Book of Laughter and Forgetting, the one about the struggle of memory against forgetting, held up like a banner against the present. It is a good banner, and a reader can march under it in good faith. The novelist could not. In the novel the sentence has a speaker, and the speaker is in the middle of trying to rewrite his own past. Kundera gave the line about memory to a man with an eraser in his hand, and he knew it.

The question underneath is older than the novel, and it was first asked at a window. In the second of his Meditations, Descartes has been turning a piece of wax by the fire and finds that the word see has misled him. We say we see the wax, he writes, and we say we see the men passing in the street below, and in both cases the eye reports less than the sentence claims. From the window he sees hats and cloaks. That men are under them is something he judges, and could judge wrongly; the French edition he approved has him wondering whether they are spectres or feigned men. The default of the mind is to supply the man beneath the hat without checking. Everything in this essay turns on that default. The retoucher counts on it: leave the hat, remove the man, and the eye will finish the picture on its own. The novelist works it from the other side: put the hat on the wrong head and the eye, seeing a hat, is forced at last to ask whose. Kundera’s readers supplied Clementis under that hat for decades without looking at the print, until a scholar did. So, in its first draft, did this essay. The reader is the judge at the window, and the essay asks nothing more of him than Descartes did.

A second question, Nietzsche’s, waits in Section II: whether you could will your life again, all of it, unchanged. Most people would say yes to some of it and want the rest gone. The wish that a thing happened once and therefore never is close to what Kundera calls lightness, and the airbrush is that wish, mechanized. The balcony gives the wish its emblem, and it is Descartes’s: a hat on a head, and a wall where a man stood. René Magritte, who painted a bowler on a figure and turned him away from us, or set an apple where a face should be, painted the same object more often than anyone, and he will be in this essay by description, since his pictures cannot be.

The essay follows the hat through four hands. The state’s, which removed a man from the balcony and kept the hat beside him. The novelist’s, which may have moved that hat from one head to the other, and the story’s moral with it. The author’s, which removed an era from his own shelf and may have removed more. And an heir’s, which arranged a dead philosopher’s notebooks into a book he never wrote. Kundera had a theory that told the third hand from the fourth. The essay asks whether the theory survives the record.

It makes one guess, about a motive, and grades it where it is made. Everything else is the record, and where the record is a police file, it is treated as what it is.

I

The novel tells it this way. It was snowing and cold, and Gottwald was bareheaded, and Clementis, out of solicitude, took off his own fur hat and set it on the leader’s head. Four years later Clementis was charged with treason and hanged, and the propaganda section removed him from history and from every photograph. In the novel’s telling, all that is left of him in the picture is the wall behind the place he stood, and the hat.

It is one of the great openings, and its power is in the last detail, the residue the retouchers could not remove. But the residue is only a residue if the hat was his. If Píchová and Toman are right, and both men are wearing hats, then the hat on Gottwald’s head was Gottwald’s, and it stayed where it was for the plain reason that nobody needed to remove it. What the state cut from the picture was a man. What the novelist added to it was a gift. The story of forgetting begins with a small act of kindness that may not have happened, and the essay does not resolve it. It records that the record and the novel differ, and that the difference is a hat.

Something else is missing from the retouched print that the novel does not mention. The museum’s caption to the retouched print says that two people were removed from behind Gottwald, and names them: Clementis, and the photographer Karel Hájek. By then the state had already been through his flat. In June 1950, half a year before Clementis was arrested and two and a half years before the print was altered, the Interior Ministry took eighty thousand of Hájek’s negatives, his life’s work, away in a truck. He was not silenced; he went on photographing for the state’s magazines for another quarter of a century. But the state emptied the photographer’s flat before it emptied his photograph, and the frame that every schoolchild knew lost its witness as well as its victim, and kept its hat.

Then the novel does its second thing, which is the reason this essay exists. It is 1971, and Mirek says: “The struggle of man against power is the struggle of memory against forgetting.” He says it to explain what his friends call carelessness. He keeps a diary. He preserves his correspondence. He takes minutes of every meeting where they discuss what is to be done, and he keeps the minutes. He is the man Murphy’s banner is for.

And what he is doing, in the pages that follow the sentence, is trying to get back the letters he wrote years earlier to a woman named Zdena, so that he can destroy them. He does not want the letters because they are dangerous. He wants them because Zdena was ugly, and he cannot bear it to be known that he loved her. The narrator says it outright: Mirek wants to do to his own life what the Party did to the photograph. The man who keeps every record has found the one thing in it he would airbrush, and it is not a crime. It is a face.

That is the whole shape of the essay in the novel’s first pages. The sentence about memory is true. The man who says it is an eraser. Kundera wrote both, on purpose, and put them a few paragraphs apart.

II

The second question was asked by a demon. In The Joyful Wisdom, in a section Thomas Common headed “The Heaviest Burden,” Nietzsche imagines one creeping after you into your loneliest loneliness to say that this life, as you now live it and have lived it, “thou must live it once more, and also innumerable times,” with nothing new in it, every pain and every joy in the same series, down to this spider and this moonlight among the trees, and the demon himself. Would you gnash your teeth and curse him, or call him a god? The question the demon leaves is the one Common ends the section on: do you want this once more, and innumerable times?

Kundera opens The Unbearable Lightness of Being on that demon, grants that the idea is mysterious, and then says what it would weigh. If every second recurs, he says, we are nailed to eternity; every move carries an unbearable responsibility; and that is why Nietzsche called the thought the heaviest of burdens, das schwerste Gewicht, the same words as Common’s title. The opposite of the burden is lightness, and lightness is the condition of a life that happens once. Tomas says the German adage to himself, and the narrator glosses it: a thing that happens a single time has, for the purposes of weight, not happened. Einmal ist keinmal. Once is never.

Kundera is exact about what the lightness costs. In a world without return there is no verdict, because there is nothing to hold a thing still long enough to judge it; “everything is pardoned in advance and therefore everything cynically permitted.” He names Robespierre, who would be judged differently if he returned to cut off heads forever. He does not name the retoucher, but the retoucher is the same case. The airbrush is einmal ist keinmal with a tool in its hand: a way of making a man on a balcony have happened once, and therefore never.

The demon is the airbrush’s exact opposite, and this is the essay’s own reading rather than a source’s. What the demon insists on is not the great things but the unspeakably small ones, in his own list: the spider, the moonlight, this moment. Under eternal return nothing is too small to come back. The hat comes back. The letters to an ugly woman come back. Everything the wish for lightness would like to have happened only once returns in the same order, and the man who would airbrush his past is the man who cannot say yes to the demon. Descartes’s window and Nietzsche’s demon are the two ends of the same problem. One asks what you supplied under the hat. The other asks whether you could bear to see it again.

The novel’s first image, after its pages of philosophy, is a man standing at the window of his flat, looking out. Kundera did not need Descartes to put him there. But that is where the essay’s two questions meet, and the rest of it is spent on what two hands did with what came back.

III

The author’s hand is the one the essay can date.

Milan Kundera published three books of verse before he published a line of fiction: a first collection in 1953, a long poem in 1955 in honour of Julius Fučík, the Communist journalist the Nazis had executed and the regime had made a saint, and a third collection in 1957. He joined the Party in 1948, was expelled in 1950, and was readmitted in 1956. The verse belongs to the years around that expulsion. Later he called it adolescent and would not have it reprinted; he also, as of 2008, had given no interview in a quarter of a century; and the editions of his work that he authorised begin, as if nothing came before them, with the stories of the 1960s. What he removed from the shelf was the verse of four years. The years around it are these: 1948, 1950, 1956. Section IV has one date to put beside them.

On his own theory this was not an airbrush. It was the exercise of a right. Testaments Betrayed, the book of essays he published in French in 1993, ends on Kafka and Max Brod: on the friend who was asked to burn the manuscripts and published them instead, and who then went on to publish the diaries and the letters and the drafts, and, in Kundera’s account, a Kafka of his own devising on top of them. Kundera does not quite say the novels should have burned. His charge is that Brod could not tell the work from the residue, and published both as if they were one thing, and that the residue then swallowed the work. The theory that comes out of this is exact rather than absolute: the work is what the author released; the author alone knows where it ends and the residue begins; and the executor who cannot tell the difference has betrayed a testament even when he has saved a masterpiece. It is the most uncompromising defence of the author’s scissors in modern criticism, and Kundera wrote it while using them. On its terms the unwanted verse, the unfinished pages and the interviews were never the work, and to keep them out of the canon was not to falsify the record but to draw the line the executor is supposed to draw, in advance, himself.

The theory cuts one way and not the other. It licenses the third hand in this essay and condemns the fourth. An author who prunes his own shelf is a gardener; an heir who arranges a dead man’s papers into a book is Brod, or worse. That is the line Kundera drew, and it is a real line: Elisabeth Förster-Nietzsche will be on the far side of it in Section V, and nothing that follows will move her back.

Kundera wrote his own window. In the same part of the book, under a title that comes from Stravinsky scolding a conductor for cutting his score (you are not in your own house here, my dear fellow), he looks across at a lit room opposite. A man enters, paces, runs his hand through his hair, realises he can be seen, and pulls the curtain. He was not counterfeiting money in there; he had nothing to hide but himself. Shame, Kundera says, is the instinct that requires the curtain, and insists that a letter addressed to A not be read by B; and when a regime broadcast a dissident’s private conversations as a radio serial, the country understood that “curtain-rippers are criminals.” He drew the curtain over his own drawer as well: none of his letters or journals, he declared, would be published after his death. That is the author’s right stated at its strongest, as privacy rather than property, and it is the version this essay has to test. Notice its terms. The letter addressed to A.

But notice what the theory needs in order to hold. It needs the author to be the only party with an interest in his own past. The gardener’s right over the garden is complete only if no one else was ever in it. Mirek believed that too, about his letters, and the novel does not let him keep it: the letters were written to somebody. Section IV is about a document Kundera did not write, in an archive he did not control, from the year the Party expelled him, and it is the one thing in his record the theory of Testaments Betrayed cannot reach, because it was never his to keep or cut.

IV

The date is the fourteenth of March, 1950.

On that afternoon, according to a half-page police report that surfaced in the archive of the Czech security services in the spring of 2008, a student named Milan Kundera, born on the first of April 1929 in Brno and resident at a hall of residence in Prague VII, presented himself at the local station and reported what a fellow student had told him. A young woman at the residence, Iva Militká, had met an acquaintance that day at Klárov, a man called Miroslav Dvořáček, who was believed to have deserted from the army and gone to Germany. “The said Dvořáček apparently left 1 case in her care,” the report says, and would return for it in the afternoon. Two policemen went to the residence, looked at the suitcase, and waited. Dvořáček, who had come back into the country the day before as a courier for the exile intelligence service, walked in at about eight in the evening and was arrested. The prosecutor asked for death. He was given twenty-two years and served almost fourteen.

That is what the report says. The essay states the rest at the record’s own confidence, which is low, because everything that follows is contested and the essay decides none of it.

The report is registered as 624/1950. It carries the name and the date of birth; it lacks, by later accounts, the identification number and the signature such reports ordinarily bore. Respekt, the Prague weekly that published it on the thirteenth of October 2008 under the title “Milan Kundera’s denunciation,” presented it as the key to a fifty-eight-year mystery. The security-services archive said the next day that the document was not a forgery, and said nothing else. Kundera telephoned the Czech press agency and denied it entirely. Two days later a literary historian, Zdeněk Pešat, said that the student who told the story, Miroslav Dlask, had told him years before that he, Dlask, was the one who went to the police. Dlask had died in the 1990s. He had married Militká, and by her account he admitted only one thing before he died: that he had mentioned the visit to Kundera. So the woman with the suitcase, the man who is said to have said he reported it, and the researcher who found the report are one family. That is not an accusation of anyone. It is the shape of the record, and a reader weighing the counter-account has to see it. A later reader of the file, Karen von Kunes, pointed out what the report itself records: that what was reported was a suitcase, not a man.

Descartes would have recognised the shape of that defence, and its limit. The report carries two things: an object left in a room, and a belief about the man who owned it, that he had deserted and gone to Germany. To say what was on the floor is to say what one saw. To say what the owner was is to supply the man under the hat, and the file does not say who first supplied him: the woman who mentioned the visit, the friend who repeated it, whoever walked to the station, or the officer who typed it up. The window cuts both ways. What it can establish is only that by the time the report was filed, someone had done the judging, and the state acted on the judgment with two policemen and a wait.

The man who found the file says how he came to it. Adam Hradilek, the researcher at the Institute for the Study of Totalitarian Regimes who wrote the article with Petr Třešňák, says in his own epilogue that he came to the story through a cousin, that Militká was one of his relatives, and that the search had been given a little spice by a family connection before the report appeared. He also says he pondered for a long time whether he had a moral right to publish it. The essay records both statements because he made both. What the file can and cannot establish is a question the archive itself declined to answer, and the essay is not better placed than the archive.

Here is the date the earlier section held a place for. Kundera joined the Party in 1948. In 1949 he and a friend exchanged letters critical of a Party official, the letters were intercepted, and Kundera was expelled; Respekt places the expulsion just before March 1950. The verse he later cut from his shelf runs from 1953 to 1957. The file sits in the middle of the era he removed.

The essay’s one guess, graded persuasive. That he cut the era because of what was in it. The grade is the lowest one, and it is given because the coincidence of the dates is not evidence of motive, because he stated a different motive (the verse was adolescent), because the file names a stranger’s arrest and not a poem, and because his own theory in Testaments Betrayed would have licensed the same cut for an author with nothing to hide. Nothing in the record contradicts the guess and nothing in the record supports it. It is the guess a reader at Descartes’s window makes, and the window is where it stays.

What the record keeps is an object. A fur hat on a balcony that nobody needed to remove, and a suitcase on a floor in Prague VII that two policemen looked at and left where it was, waiting for the man who would come back for it.

V

The fourth hand is the simplest, and Kundera has already told us what to call it.

In the summer of 1885 Nietzsche wrote down a title for a book he meant to write, The Will to Power, and for three years he filled notebooks toward it. At the end of August 1888 he set the plan aside for another. In January 1889 he collapsed in Turin and did not write again. His sister, Elisabeth, took control of the papers, founded an archive, and in 1901, with Nietzsche’s friend Peter Gast, published a book under the abandoned title, assembled from the notebooks; in 1906 she published a larger one, and for half a century the world read it as his last and systematic word. The scholars who finally edited the notebooks from the manuscripts, Giorgio Colli and Mazzino Montinari, found that the words were Nietzsche’s and nothing else was: the order was the editors’, the omissions were the editors’, and single notes had been cut apart and sewn to other notes without any mark to say so. Montinari’s objection was not that the sentences were false. It was that the editors had chosen to publish them “in precisely this way, which is neither correct nor appropriate for the Nachlass.” A book the author had put down was picked up by an heir and stood on its feet as if it had walked there.

It was a cento with no label. On this site’s rule that is the whole offence: not that a sentence was invented, but that an arrangement was passed off as an author’s. And on Kundera’s rule it is Brod’s offence exactly, the executor who cannot tell the residue from the work and publishes the residue in the work’s clothes. A hat and a cloak set upright on a heap of notes, and the world at its window judging there was a man inside.

So the theory of Testaments Betrayed holds here without strain, and the line it drew in Section III is a real line: the author who put the plan down had the right to, and the sister who picked it up did not. That is the mirror, and it is not an equivalence. The third hand cut what was its own; the fourth arranged what was not.

But the sister’s case is simple for a reason the theory does not mention. There was no one else in the garden. The notebooks were Nietzsche’s and only Nietzsche’s; no third party had written in them, no stranger had been arrested on their account, no file about them lay in an archive he did not control. The right that Kundera claimed for himself on his shelf was complete over the verse, for the same reason. It was not complete over the year the verse belongs to, because Section IV showed one document from that year that was never his to keep or cut, and that came back anyway. The theory tells the third hand from the fourth by asking who holds the scissors. The file asked a different question, which is who else was in the room, and that is the question the demon asks too.

The approved photograph is still in the archive of the Czech press agency, and it is the one the country knew. The crowd, the snow, the balcony rail, the men of the party in a row; where two of them stood, a length of wall. Gottwald at the rail with a fur hat on his head. Whose hat it was, the print does not say, and those who have looked at the original say it was always his. The man behind him was hanged before dawn on the third of December 1952, after the man in the hat, by the standard accounts, had refused him clemency. Then the record did something the novel does not mention. Ten years later Clementis was rehabilitated, and the original photograph was permitted again, by the same state that had cut him out of it. The museum in Prague holds both prints. The man came back into the picture by the hand that removed him, and the hat had never moved.

Descartes said that from a window we see hats and cloaks, and that the men are supplied by judgment. The retoucher was counting on it. The novelist was counting on it too, from the other side, and may have moved the hat so that we would look. Both of them, and the sister with her scissors, and the author with his, were working the same default: the eye that finishes the picture and moves on.

The demon’s question was whether you could bear to have it all come back, in the same order, down to the small things. A hat. A suitcase. A face you would rather not have loved. Nobody in this essay said yes. The record did not ask them. It kept the objects, and left the judging to whoever comes next to the window.

A hat on a head, and a wall where a man stood.

⁂

Written in full collaboration with the machine, and the ledger requires the name be exact: drafted and edited with Claude Fable 5.1. No quotation in this essay was recalled: each was copied from an open text and checked against it. Descartes is in John Veitch’s translation; Nietzsche in Thomas Common’s; Kundera’s The Book of Laughter and Forgetting in Aaron Asher’s, The Unbearable Lightness of Being in Michael Henry Heim’s, and Testaments Betrayed in Linda Asher’s as printed in The New York Review of Books in 1995, one sentence from each; the 1950 police report as printed in English by Respekt in 2008; Montinari from The Malahat Review, 1972; Hana Píchová on the hats from The Art of Memory in Exile (2002), as quoted by Johannes Lichtman in the Los Angeles Review of Books (2015). The two prints are in the City of Prague Museum, MMP H 085 795 and MMP H 085 778. The hat that changes hands is Kundera’s; the photograph is Karel Hájek’s. On the collaboration → “The Third Thing”

On Cicero at Astura, the other father, and what a sceptic does with a wish he cannot root out

“… the corpses of so many towns lie in helpless ruin …”

Servius Sulpicius Rufus to Cicero, on the death of Tullia, 45 BCE. Ad Familiares 4.5, translated by Evelyn Shuckburgh

In the spring of 45 BCE the most famous advocate in Rome hid in a wood, wrote that nothing can be affirmed, and shopped for land on which to make his daughter a god. What a mind holds beyond what it writes can only be guessed. This essay makes two guesses and grades both.

By Michael Cummins, Editor · September 23, 2026

Early in the day he goes into the wood. It is dense and wild, he tells his friend, and he does not come out of it until evening. The house behind him stands on the coast at Astura, right in the sea, within sight of Antium and the headland of Circeii. He is sixty-one. He has been consul, exile, governor, and the most famous advocate in Rome, and none of that goes into the wood with him. Books go. He says his only conversation now is with them, and that even this is broken by weeping, which he fights for as long as he can and is not yet equal to. It is March of 45 BCE. His daughter has been dead a matter of weeks; the sources do not agree on how many. Each evening he comes out of the trees and writes to Atticus, and in the first of those letters, in its last clause, he mentions that he would like to buy something: a hiding-place, a refuge for his sorrow.

Hold that clause.

Introduction

This month, in Aeon, Massimo Pigliucci opened an essay on Cicero’s scepticism with a scene: early 45, a villa beside the Lucrine Lake, the Academica being finished while a life comes apart. It is a good scene, and true of the dialogue, whose second edition Cicero set at Varro’s villa by that lake. About the desk it is incomplete. The record of that spring and summer puts the writer elsewhere: at Astura, at a villa of Atticus, at Tusculum, at Arpinum, where he revised the book in the rain. In March, with the Academica still ahead of him, he was here, in the trees by day and at a table by night, and the evidence of where his mind was is unusually thick, because he wrote to Atticus almost every day and the letters survive.

Pigliucci’s subject is the method: Cicero’s probabilia, the things a mind may follow and can scarcely affirm. This essay’s subject is the mind that needed the method that spring. What was in it, beyond what he was writing?

The letters answer part of that. The rest can only be guessed, and a guess about a sceptic ought to be made by the sceptic’s own rules. Carneades, whose arguments Cicero would turn into Latin before the summer was out, graded impressions, as the tradition records his ladder, by the scrutiny they had survived: the merely persuasive; the undiverted, which nothing else one knows contradicts; the examined. There are two guesses in what follows. Each carries its grade in plain sight, and neither is offered as fact. Where the essay invents, it says so and fences the invention.

I

What survives of those weeks is one side of a correspondence. Atticus’s letters are lost. Cicero’s come almost daily from the seventh of March, written for one reader, and from them a day can be rebuilt.

The letters of that spring mourn the daughter and do not describe her. Tullia was in her early thirties. She had been married three times, and had divorced the third husband, Dolabella, while carrying his child. The boy was born in January at her father’s house in Rome. She died about a month later. The child was still alive in March, when Cicero defends what his will does for his grandson. He did not live long.

His mourning had begun, before Astura, in a library. At Atticus’s house he read everything anyone had written on the lessening of grief, and found it was not enough. So he did, in his own words, “what certainly no one ever did before me”: he tried to console himself by writing a book. At Astura he writes all day. He does not claim that it works. Writing and study, he reports, do not soften the feeling; they distract it. What he is trying to recover is composure, “not of heart, yet, if possible, of countenance.”

The book needed examples: Romans who had buried their children and gone on. Here the letters show something no summary of his philosophy would predict. He checks them. Did Caepio die in the shipwreck before his father’s death, or after it? Did Rutilia die in her son’s lifetime, or survive him? Did Clodia outlive hers? Atticus, who kept a chronicle, is told which named people to ask, and is asked again when the answer is slow. “These facts,” Cicero explains, “affect the book.” A few days later he calls such questions “my own triflings,” and sends two more. A father weeks bereaved will not cite another parent’s loss until he knows which of them died first. Grief, at Astura, checks its citations.

The rest of his life arrives by the same courier. There is Terentia’s dowry to repay, and her complaint about the witnesses to his will. There is a son to be kept at Athens on the rents of two city properties. And there is a wife. Publilia, in her middle teens by the modern estimates and married to him not long before, writes begging to be allowed to come with her mother. He tells Atticus about it in a passage he marks as written in his own hand. He has answered that he wishes to be alone. He does not think the letter is her own composition. He concludes that the one way of avoiding the visit is “to fly away,” and he asks Atticus to find out the last day he can stay without being caught.

Then there is a letter that holds the whole season on one sheet. Atticus has been urging him back toward the forum. He replies that he is ruined, that he has known it a long while and confesses it only now, “when I have lost the one thing that bound me to life.” A few lines on, to show that he is “not prostrate,” he changes the subject. Atticus’s chronicle records the year Carneades came to Rome on the famous embassy. What was the reason for it? “It was about Oropus I think, but am not certain.”

It is the kind of question a book on the Academy would need answered. It sits inside the mourning letter, a few lines below the sentence about his life, and it is already phrased the way that book will teach a mind to hold anything. I think, but am not certain.

II

The clause can be released now. The man who in early March wanted to buy a hiding-place spends the rest of the spring trying to buy its opposite.

He calls it a fanum, a shrine. The letters give it a designer, Cluatius, whose plan he likes; a contract for columns with Apella of Chios; and a deadline, this summer. What it lacks is ground. He considers the island at Arpinum and rejects it as too out of the way: it would diminish the honour of the departed. He considers gardens across the Tiber, a cramped lot at Ostia, the grounds of Drusus, of Lamia, of Silius, of Scapula, of Clodia. He does not want plate or fine furnishings or a view, he tells Atticus. He wants this. And of the site he asks one thing, again and again: “what I want is a frequented spot.”

The law is in the way. Spend past a fixed sum on a monument and the same sum is owed to the treasury. The fine is not what troubles him. “I am anxious to avoid any likeness to a tomb,” he writes, so as to come as near as he can “to an apotheosis.” If Atticus can think of a way round the statute, he will take it.

What the letters shop for, the lost book performed. The Consolatio does not survive. But some three and a half centuries later the Christian writer Lactantius, arguing that the pagan gods were only the honoured dead, called Cicero as a witness and quoted him speaking to his daughter: with the approval of the gods he will place her, “the best and most learned of all women,” in their assembly. Lactantius adds that someone may say Cicero raved from grief, and answers that the passage shows constancy and judgment. He needed his witness sane. So the sentence in which a father deifies his child comes down to us because a prosecutor of the practice entered it in evidence.

Was the shrine a breach of his scepticism, or its purest case? The Academic may follow what persuades him and must not affirm what he cannot know. Even the fragment is a promise and not a claim: I will place you, I will consecrate. Nowhere in the letters of that spring does he tell Atticus that Tullia is a god. He says the wish cannot be rooted out of his heart. He calls these “foolish ideas of mine,” and confesses them to be so. He says he knows he is infatuated, and asks Atticus to go on indulging the aberration. By the tradition’s ladder the impression never climbed past the first rung: it persuaded him, and by his own account it failed examination. He followed it anyway, with contracts. It is the one thing that spring he neither affirms nor lets go.

III

There was another father.

In 336, when word came that Philip of Macedon was dead, Demosthenes appeared in public in a garland and white clothing and made thank-offerings. Six years later, prosecuting the proposal to crown him, Aeschines told the jury what week that had been. It was “the seventh day after the death of his daughter,” and the rites of mourning were not yet complete. Then comes the sentence that has outlived the case: he had lost “the first and only one who ever called him father.” Aeschines adds at once that he does not reproach the misfortune. He is only probing the character. A bad father, he explains, can never be a safe guide to the people.

Cicero knew the passage. Against Ctesiphon and the speech that answered it, On the Crown, were the pair he had proposed to put into Latin. And within months of Tullia’s death, in the third book of the Tusculans, the book on grief, he cites it: Aeschines attacks Demosthenes for having sacrificed septimo die, on the seventh day, after his daughter’s death. The day matches Aeschines. Plutarch, who names Aeschines as his source, makes it six. Cicero’s point is that an orator can do anything with the prejudice that good men must grieve as hard as they can. His next sentence is about those who in sorrow seek out solitude, and he quotes Homer’s Bellerophon, eating his own heart, keeping clear of the tracks of men.

In an Athenian court a father was taunted for grieving too little. Rome was murmuring about one who grieved too much; Atticus had warned him what the mourning was costing his influence. The tradition pairs these two as orators. In the book on grief, Cicero paired them as fathers.

Speculation, graded persuasive. The Bellerophon sentence is a self-portrait. The man who wrote it had spent March going into a wood early in the day and avoiding everyone he knew. Nothing proves he meant himself.

Speculation, graded undiverted. The other father was already with him at Astura. No letter to Atticus from that year names Demosthenes. In Books XII to XVI of those letters the name appears once, in 44, in a remark about a speech of Brutus. So nothing in the record turns the guess aside. Silence does not support it either: the higher grade records a test survived, not a stronger case. The guess was run against every letter in those books, where the self-portrait admits no such test. Nothing examines it.

Demosthenes answered the speech and not the taunt. On the Crown is long, and his daughter is not in it. What he says there about mourning concerns the city’s dead: that after the defeat the people chose him, and not Aeschines, to speak over them, and that the families held the funeral feast at his house. What he says of parents is that the Athenians of old each felt he had come into being “not for his father and his mother alone, but also for his country.” Plutarch, centuries later, made the defence he did not make himself: to bear such a loss serenely is no sign of a cold nature.

So the two fathers can be set in one room only by arrangement.

IV

Beyond the record three questions stay open, and each can be argued both ways from the same letters.

Was it grief, or the performance of grief? He manages the look of it. He arranges sworn excuses for his absence from public duties, and would rather be thought away “in virtue of the augural law” than because of sorrow. When Atticus, by Cicero’s account, reports that people are talking, and fears for their influence, Cicero answers like an advocate: his critics could not read as much as he has written, and no caller has found him unwilling to talk. That is a man tending a reputation. But a man who is only performing does not tell his one confidant that the writing fails, that the shrine is an aberration, that he would not lessen the sorrow if he could. His own distinction covers it: a countenance he could compose, and a heart he could not.

Did Servius’s substitution fail? Servius told him to mourn the Republic and not the girl, who had at least outlived it. Cicero’s answer is exact. While she lived, home consoled him for the forum; now neither can console him for the other. Yet everything he went on to write that year was written for Romans, in a Latin he was building for the purpose. The consolation failed in the spring. Whether it worked by autumn, the letters do not say.

And why was the shrine never built? Walter Englert has argued that by the Tusculans he had reasoned himself out of it: a monument, like Artemisia’s, only prolongs the grief. Shuckburgh thought the money ran out, with Terentia’s dowry to repay. The archive holds verdicts as well as facts, and verdicts do not reconcile the way dates do. Even the facts here are loose. Three authorities give three answers to where she died: at Tusculum; at Rome or else Tusculum; in the house of Lentulus.

One thing was not in his mind, because it had not happened. The speeches against Antony, which he would name after Demosthenes’ against Philip, were eighteen months away. Plutarch paired the two men partly for how they ended. Demosthenes took sanctuary in Poseidon’s temple on the island of Calauria, and Antipater’s man came across in small boats to fetch him out. When the proscriptions reached Cicero, Plutarch says, the first place he made for was Astura, to take ship. He was killed near Caieta, in a litter that was being carried “through the wooded and shady walks towards the sea.”

A wood, again, and the sea.

At evening he came out of the wood and wrote to Atticus about columns.

That is nearly the last the record shows of the wish: a contract to settle with a man from Chios, a summer deadline, a site not yet found. The shrine was never built. The book he wrote to console himself, the thing he said no one had ever done before him, is lost but for the sentences other men quoted. What he wanted for her was a place people would come to, through the endless generations he counted on, and he did not get it.

He got something else. In June of 1345 Francesco Petrarch, in Verona, came upon the letters to Atticus where he least expected them, lost for centuries, and read them through to the end. He had expected the sage of the treatises. He heard a man “give voice to many laments” and waver in his opinions, and what he did next he had never done before. He wrote back. His letter reproaches and loves in one breath. It is uttered, it says, “not without tears.” It wishes its reader had spent his last years quietly in the country. It says farewell forever.

Cicero would have known the form. In an earlier book he had written a dead man into a living man’s dream: Africanus, appearing in a shape the dreamer knew only “from his picture,” to promise those who served their country a place in heaven. Petrarch knew his dead man only from his letters. After Petrarch, Machiavelli, on his small property outside Florence, would put on court dress each evening to question the men of old, and reports that “they in their benignity answer me.” After Machiavelli, anyone who opens the book.

The frequented spot was the correspondence, and the visitors come for him. Neither that letter nor Petrarch’s second to Cicero names her. The first visitor we know to have answered came nearly fourteen centuries late, and left his address.

“Written in the land of the living … in Verona … in the thirteen hundred and forty-fifth year from the birth of that God whom thou never knewest.”

⁂

Written in full collaboration with the machine, and the ledger requires the name be exact: drafted and edited with Claude Fable 5.1. No quotation in this essay was recalled; each was copied from an open public-domain text. The letters are in Evelyn Shuckburgh’s translation; the Tusculans and the Dream of Scipio in C. D. Yonge’s, who renders septimo die loosely as “within seven days,” where the essay follows the Latin; Aeschines in Charles Darwin Adams’s; Demosthenes in A. W. Pickard-Cambridge’s; Plutarch in Bernadotte Perrin’s; Lactantius in William Fletcher’s; Petrarch in Mario Cosenza’s; Machiavelli in W. K. Marriott’s. On the collaboration → “The Third Thing”