Tag Archives: Tacitus

Essays, History, Literature, Philosophy

SHE READ TACITUS

Catherine, the coup of 1762, and the letter she didn’t write

“I began to take gloomier views of things, and to look for more hidden and interested motives in the occurrences around me.”

Catherine II, Memoirs, on reading the Annals of Tacitus in the winter of 1754; 1859 edition

In the winter after her son was taken from her, a grand duchess read the Annals and reported a revolution in her brain. Seven years later she took a throne and managed three documents about it. A fourth was found in her cabinet after her death, and it may not have been there.

By Michael Cummins, Editor · September 26, 2026

In the winter of 1754 a grand duchess of twenty-five read what she could get. She had given birth in September, and the Empress had taken the child away within the hour to raise herself; the mother was not consulted again about him for weeks. She had the winter and a room. In the memoir she wrote decades later she lists the reading in order, and the order matters: as many Russian works as she could procure, among them two immense volumes of Baronius in Russian translation; then she lit upon the Esprit des Lois of Montesquieu; “after which I read the annals of Tacitus, which caused a singular revolution in my brain, to which, perhaps, the melancholy cast of my thoughts at this period contributed not a little. I began to take gloomier views of things, and to look for more hidden and interested motives in the occurrences around me.”

That is the whole of what she says about Tacitus. It is one sentence and a half, and it is not the sentence a biographer would write for her. It says nothing about power, nothing about the guard, nothing about how thrones change hands. It describes a change in how she read. From that winter, she says, she assumed that behind any occurrence there was a motive, and that the motive was interested, and that it was hidden.

Introduction

The Annals open on a succession. Augustus is dead and Tiberius, who has been his colleague in everything, does not yet call himself emperor. The oath is taken first by the consuls, then by the prefects of the guard and the grain supply, then by the Senate, the army and the people. In Rome, Tacitus writes, consuls, senators and knights rushed into slavery, the more exalted the man the greater his hypocrisy and his haste, tears blended with joy and regret with adulation. Tiberius speaks of the weight of empire and his own inadequacy, in words more dignified than convincing, and the senators, whose one fear is to seem to understand him, stretch their hands to heaven, to the effigy of Augustus, to his knees. He wants, Tacitus says, to appear the called and chosen of the state rather than an interloper who had wormed his way in through a wife’s intrigues. Every act in the passage was public and every act had a reason behind it that the historian supplies. That is the book she read.

A rule, stated once. Tacitus was on every educated shelf at that court, and reading him explains no coup, or it explains all of them. This essay does not claim that the Annals caused what happened in June 1762. It claims something smaller and checkable: that she told us what the book did to her, that what it did was make her a certain kind of reader, and that a reader of that kind, when she came to have a record of her own, would know exactly what a later reader would look for in it. Tacitus is the shape of the record here, not the cause of the event. Cicero, on this site, was a man deciding what to write; Seneca, in “Another Man’s Eloquence,” a man deciding whose. Catherine is a reader deciding what will be read.

She managed three documents about the coup, and a fourth was found among her papers after she died. What follows is those four, in Tacitus, in her own hand, in Bain, in Herzen and in one modern Russian archivist, with one guess, graded.

I

Necessity

A carriage at a gallop on a dawn road along the sea, a woman seen from behind inside it, guardsmen on horseback closing round
The Peterhof road, six in the morning, 28 June 1762, as this essay imagines it. Her letter to Poniatowski gives the hour and the words: Alexei Orlov entered her room, woke her, and said very quietly that everything was prepared for proclaiming her. She dressed, got into the carriage, and was in the Izmailovsky barracks before the city knew. A generated image, not a record.

Her own account, written five weeks later, begins at six in the morning on the twenty-eighth of June. She was asleep at Peterhof. Alexis Orlov entered her room, woke her, and said very quietly, “It is time to get up; everything is prepared for proclaiming you.” The night before had been spent in anxiety, since she knew everything that was being prepared, and one of the conspirators, Passek, had been arrested the previous evening. She dressed and got into the carriage. At the Izmailovsky barracks twelve men and a drummer met her; the regiment swore. Then the Semenovsky, then the Kazan cathedral and the archbishop, then the Winter Palace, where the Senate and Synod were waiting to sign what the regiments had done. By noon the thing had a warrant, and the warrant had come last.

The manifesto printed that day gives the reason. To all true sons of the Russian fatherland, it says, it had become evident what danger to the whole state was beginning in deed: the Orthodox faith felt itself shaken and its traditions threatened; the glory of Russian arms, raised so high by victory, had been given away in a peace; and therefore, being convinced of the danger to all her faithful subjects, and seeing their clear and sincere desire, she was compelled, with God’s help, to ascend the throne. We were compelled. The grammar of the document is the grammar Tacitus gave Tiberius: the taking written in the passive voice, as a thing that had to be, by a person on whom it was pressed. The guard first, the institutions after, and the sovereign presenting herself as the instrument of a necessity that others had felt before she did. It is not the shape of a coup because she had read the Annals. It is the shape every such document takes, and the point is that she knew, better than most of the people who signed it, what a reader would later make of that shape.

Hers was not the only first account. Princess Dashkova, nineteen that summer, wrote her own memoir, and its chapter on the day claims the carriage and the urging: young Orlov came to her, she had the foresight to have a carriage ready, she pressed for the Empress’s immediate presence in the capital. Catherine’s letter has a sentence for her. The princess, it says, though she ascribes to herself all the honour of the affair on the ground that she was acquainted with some of the principal actors, was in fact held under suspicion the whole time, partly for her family and partly for her nineteen years. Two accounts, each first, each written by a party with a reason, and each already reading the other the way Tacitus reads a senator.

II

Colic

Peter surrendered at Oranienbaum without a shot. He signed an abdication. She sent him, in the letter’s words, to Ropsha, twenty-five versts from Peterhof, under the command of Alexis Orlov, with four chosen officers and a detachment of quiet and sober men. That is the last sentence of the letter in which he is alive and she is in charge of the prose.

He died on the sixth of July. The manifesto that announced it named the cause, hemorrhoidal colic. The empire was asked to accept a diagnosis, and it did, in the sense that nobody was in a position to refuse it. Then the letter to Warsaw gives the same death to a different reader. The hemorrhoidal colic, it says, came on again, accompanied by delirium; he was two days in that condition, then in excessive weakness, and notwithstanding the physicians he at last sank, demanding a Lutheran clergyman. And then a sentence the manifesto could not contain. She had feared, she writes, that the officers had poisoned him, and she ordered the body opened; it showed no trace of poison. “His heart was excessively small, and also dried up.”

What the two texts share is the diagnosis. What they differ in is the audience and the surplus. To the empire, colic. To her lover, colic, and an autopsy she ordered against her own guards, and a heart. Modern accounts treat the death as almost certainly violent, and it has never been proven so, and the essay claims no more than that. But notice what the autopsy is for, in her telling. It is not there to establish the colic. It is there to rule out the one crime she says she suspected her own men of, and to rule it out in a form that could be shown. The reader of Tacitus, writing a private letter, supplies the hidden and interested motive herself, gives it to the officers, and then reports the examination that cleared them. She is already writing for the doubter.

III

The letter she wrote

The letter to Stanislaw Poniatowski is dated the second of August, five weeks after the coup, and it is the primary source for that day. Historians have relied on it ever since because there is nothing like it: hour by hour, in her hand, brisk, unornamented, with names and regiments and times, the kind of account a participant writes when the events are still arranged in her memory as a sequence rather than a story. The 1859 edition of her memoirs prints it in an appendix, abridged; the fuller text opens with a sentence the appendix omits, and the sentence tells you what the letter is for. She is sending Count Keyserling at once as envoy to Poland to make him king when the present king dies. Then the narrative. Then, near the end, in our rendering of the Russian text of her French: minds here are still in ferment, and so, to avoid worse disorder, I ask you to refrain from coming here. Then a promise to do everything for him and his family. The letter, on its own account, is a gift, a report, and an instruction to stay away.

That is its stated purpose, and the essay has to read past it to make the guess it wants to make. So the guess is stated as a wager, and the reading past is admitted.

The guess · graded persuasive

That the letter was a Tacitean act: the record of a taking, written first, in full, by the party with the reason, so that no doubter coming later could tell it worse than she had told it herself. A woman who had learned in 1754 to look for hidden and interested motives behind every occurrence would know that her own occurrence would be read that way, and the letter reads as if written by someone who knew it: it names the conspirators, gives the arrest, admits the anxiety, records the suspicion of her own officers and the autopsy. It hides almost nothing, which is what a reader of Tacitus would do if she wanted nothing found.

It cannot climb, for three reasons. The letter has a lover for its addressee and a crown in its first sentence; whatever else it is, it is a letter to a man she is about to reward and keep at a distance, and its own instruction is not “remember this” but “do not come.” Second, the sentence in the memoir describes her reading, not her writing; to look for motives is not the same act as to supply them, and the essay is substituting one for the other. Third, the letter tells him colic. A record written to disarm the doubter does not carry the diagnosis the doubter will doubt. The rival reading, printed at full strength: she wrote to tell him it was done, that it was done without him, and that his part in what came next was to be grateful in Poland. The vanity of the lover and the calculation of the historian sit side by side in the same pages. In her, the record suggests, they were never mutually exclusive.

IV

The letter she didn’t

Some hours after her death in November 1796, Herzen writes in the preface to the memoirs he published in London in 1858, her son ordered Count Rostopchin to seal her papers, and was himself present when they were arranged. Among them, Herzen says, was found the celebrated letter of Alexis Orlov, in which, “in a cynical tone and with a drunken hand,” he announced to the Empress the assassination of her husband. Herzen is working from a notice communicated to him, and says so; the scene is secondhand. But the sentence is a fact about the record in its own right. In 1858, twenty-three years before anyone printed the letter, it was already celebrated and already drunk.

Bain, in 1902, gives it in English from the 1881 printing and calls its authenticity indisputable. The original, he says, was destroyed by Paul, but Rostopchin, in whose hands it remained for a quarter of an hour, took a copy beforehand. The letter, in Bain’s rendering, begins: “Matyushka, most merciful Gosudaruinya, how can I explain, describe what has happened?” It says that he is no more; that nobody meant it, and what were they thinking of, to raise their hands against their sovereign; that there was a quarrel at table; that “I myself don’t remember what we did, but the whole lot of us are guilty and worthy of punishment.” It asks her to forgive them or to have them finished quickly. An accident at table, confessed by a man who says he cannot remember it. If the letter is his, it clears her of premeditation, and it has done so, for readers of every persuasion, since 1881.

Here is what the archive says when it is asked directly. Oleg Ivanov, working in the state archive at the end of the last century, found two letters from Alexei Orlov at Ropsha in the originals, both in his hand, both about the prisoner’s health and his demands. On the folder that held them, Dmitry Bludov, who catalogued Catherine’s secret papers for Nicholas I in 1825, had written: two letters of Count A. G. Orlov to the Empress Catherine, in the last of which he informs her of the death of Peter III. Two. The second announces the death. The third, the confession, exists in no original and in no hand but Rostopchin’s; and the note Rostopchin attached to his copy, when he sent it abroad, is the only witness to its effect on Paul, whom it records as saying, on reading it, thank God. Ivanov adds a detail of the kind this site keeps a column for. The letter as printed in 1911 contains a sentence the 1881 printing does not: we were drunk, and so was he. In 1858 Herzen had described a drunken hand, which is a remark about handwriting. By 1911 the drink was in the text. The confession got drunker as it aged.

Suppose the third letter is Rostopchin’s. Who did it serve? The obvious answer is Catherine: a confession of accident is an acquittal of her order, and if a courtier of Paul’s made it, the effect was to clear the mother in the son’s eyes. But the same year, Paul had his father’s coffin raised from the Nevsky monastery and carried in state to lie beside hers, and he ordered Alexei Orlov to walk before it carrying the crown. A letter in which Orlov confesses a killing damns Orlov first; if it cleared Catherine, it cleared her by the way. Ivanov names no forger and no motive; he shows the count, the missing original, and the single hand, and stops. The essay chooses between the two motives no more than he does, because the point is that each is a motive. Bain’s indisputable letter, Herzen’s celebrated one, the 1881 text, the 1911 text: every version was made by a party with a reason, which is the sentence she learned in the winter of 1754 about other people’s records. It applies to hers.

Two letters, or three. Either she kept for thirty-four years, in the cabinet where her son found it, the paper that proved she had not ordered her husband’s death, and could not bring herself to burn the paper that proved it had been done. Ivanov, who says she burned the other Ropsha papers and kept these two, offers a colder reason for keeping even those: as a justification for the future, or as a hold over a man some believed she feared. Or the drawer held two letters, one announcing a death, and a man who needed a third stood over it for a quarter of an hour with a pen. The reader of Tacitus was read, at the end, by Tacitus’s method. The method cannot tell which.

⁂

Written in full collaboration with Claude Fable 5.1. No quotation in this essay was recalled; each was copied from an open public-domain text. Catherine’s memoirs and the appended letter to Poniatowski are in the anonymous English translation of the 1859 edition (London and New York), with Alexander Herzen’s preface; the letter’s opening, its instruction to Poniatowski, its autopsy sentence and its sentence on Princess Dashkova are absent from that abridged appendix and are rendered here by us from the Russian text of her French, and are not quoted. Tacitus is paraphrased from John Jackson’s Loeb translation; one clause of Annals 1.7, “an old man’s adoption,” is omitted from the paraphrase of the wife’s intrigues. Alexei Orlov’s letter is given in Robert Nisbet Bain’s rendering of the 1881 printing, in Peter III, Emperor of Russia (1902), which also supplies Rostopchin’s quarter of an hour. Oleg Ivanov’s examination of the Ropsha letters, Bludov’s note of 1825, and the 1911 variant are paraphrased from his Пётр III. Загадка смерти. The accession manifesto is paraphrased from its Russian text; the manifesto announcing the death was not read and is cited only for the cause it named. Dashkova’s memoirs are the 1840 London edition. Orlov’s place in the 1796 procession rests on a secondary source and is stated as such. The interior image is generated and is captioned as such. On the collaboration → “The Third Thing”

Essays, History, Literature, Philosophy

ANOTHER MAN’S ELOQUENCE

Seneca, Nero, and the first ghost the record caught

“Nothing, Lucilius, is ours, except time.”

Seneca, Letters to Lucilius I, translated by Richard M. Gummere

A boy emperor praised a dead man, and the crowd smiled at the wrong moment. The sentences were Seneca’s. This essay follows the voice through four dates, 54, 59, 62 and 65, and makes one guess about the philosophy it produced, graded.

By Michael Cummins, Editor · September 25, 2026

In the autumn of 54 a crowd in the Forum listened to a boy praise a dead man. The boy was sixteen and had been emperor for a few days. The dead man was Claudius, and the Senate had already voted him a god. While the speech rehearsed the antiquity of the Claudian house, the consulships and the triumphs of the ancestors, the crowd took it seriously, and so, Tacitus says, did the speaker. Then the oration arrived at the dead man’s foresight and wisdom, and nobody could keep from smiling. They had known Claudius. The speech was elegant; the talent behind it, Tacitus adds, was graceful and tuned to the ear of the age, and it was not the boy’s. The old men who passed their time comparing the present with the past said the thing that matters. Caesar had rivalled the great orators. Augustus had a monarch’s easy fluency. Tiberius knew the art of weighing his words; Caligula’s disordered mind had not spoiled his power of speech; and Claudius, when he spoke from preparation, lacked nothing in elegance. Nero was the break in the line, “the first master of the empire to stand in need of borrowed eloquence.”

Borrowed from whom, the old men did not need to say. Lucius Annaeus Seneca was not in the verdict that day. He was in the sound of it.

Introduction

Athens had a market in other men’s words four centuries earlier, and it was a trade, not a scandal. Lysias wrote for litigants and so did the young Demosthenes; the client bought the speech and delivered it in his own cause, at his own risk, and the art of the trade was to make it sound like the man who was paying for it, plainer than its author, so that the jury would hear a citizen and not a professional. A ghost in Athens disappeared into his client. An emperor’s voice is a different instrument. It does not plead a private cause; it is the state speaking, and when the state leases a mouth the lease becomes part of the constitution. Seneca is not the first ghost. He is the first whose ghosting is itself an event in the record: noticed by a crowd, named by a historian, turned against him by his rivals, and, the essay will guess, answered by a philosophy. This ghost did the opposite of vanishing. Nero did not sound like a prince speaking plainly. He sounded, unmistakably, like Seneca.

This site has been in the Forum before. Cicero died in 43 BC for what he had written and spoken under his own name; the triumvirs could not tolerate the voice. Seneca died in AD 65, three years after retiring from eight years as another man’s voice and fourteen as his tutor, and the two deaths are not the same story. Cicero’s is about what a man chooses to write. Seneca’s is about whose.

What follows is in Tacitus, in Dio, and in Seneca, with one guess, graded. In 54 the ghost is unblamed and the joke passes over him. In 59 he is blamed. In 62 he is accused of claiming the credit. In 65 he is gone, and the last text he composed in his own voice did not survive.

I

By the lips of the sovereign

The same autumn produced a second text about Claudius, and Dio says Seneca wrote it. He called it, Dio reports, the Pumpkinification, a word built to mock the word for deification. The manuscripts that preserve a satire matching Dio’s description give it another title and no author, so the attribution has rested on Dio ever since, and the essay takes it as most readers do. In it the god the Senate had just made and the speech had just praised arrives in heaven, where Hercules is sent out to discover what has come and cannot make out what the hoarse, inarticulate thing is saying; and when the gods have thrown him out, he is tried below by Aeacus, who hears the prosecution, refuses to hear the other side, and sentences him on the principle that he is to be treated as he treated others. The man whose borrowed eloquence had made a crowd smile in October is, in the same season’s satire, given no intelligible voice in heaven and no voice at all below. One pen, one season, and neither text signed: the official voice praised the god under the emperor’s name, and the private voice unmade him under no name at all. Even the ghost’s revenge went out unsigned. Nobody in 54 is recorded objecting. The dead man had no defenders, and the ghost was not yet a category anyone thought to blame.

What Seneca did with the mouth in the next two years Tacitus records with a historian’s suspicion. Nero’s early acts of clemency, he writes, were accompanied by speeches on the subject, which Seneca kept putting before the public in the prince’s voice, either to attest how honourable his teaching was or to advertise his own ingenuity. Tacitus offers both motives and chooses neither. Then, in 55 or 56, Seneca published a treatise under his own name, addressed to Nero and dated by a sentence that puts the emperor just past eighteen. De Clementia opens by telling him what it is for: to serve, in some fashion, as his mirror. Read the two facts together. The tutor writes the emperor’s speeches on clemency. The tutor then writes the emperor a book that holds up those speeches and calls the reflection Nero’s virtue. Every voice in the loop is Seneca’s: the sovereign’s public promise, the philosopher’s private praise of it, and the mirror between them that lets the prince see a face the tutor drew. Whether this was pedagogy or capture, an attempt to trap a boy inside a scripted goodness until it became his, is not decidable from the record, and it does not need to be. The structure is the point. A ghost can flatter his principal without a treatise. Seneca built an apparatus in which the principal was flattered by his own script, and the only person who knew every part of the machine was the man whose name was on none of the speeches.

Then March of 59, and the apparatus met a fact it could not mirror. The boat built to collapse under Agrippina collapsed and she swam. Nero, in Tacitus’s words half-dead with terror, summoned Burrus and Seneca, and there followed a long silence, which the historian explains with a choice: either they thought dissuasion useless, or they believed it had come to the point where, unless Agrippina was forestalled, Nero would die. Seneca broke it only so far as to glance at Burrus and ask whether the soldiers should be ordered to kill her. That glance is the whole scene. It is the one moment in the record where the ghost is visibly deciding the deed the words will have to cover, before the words exist. Burrus answered that the guard would not act against the house of Germanicus. A freedman of the fleet did it instead.

The words followed. A letter went to the Senate in Nero’s name reporting that an assassin had been found with a weapon on him, Agermus, one of Agrippina’s own freedmen, and that Agrippina, conscious of her guilt, had paid the penalty of the murder she had planned. It added a catalogue of her crimes reaching back years. Tacitus’s sentence on its reception is the hinge of this essay: the object of public censure was no longer Nero, whose savagery had passed beyond the reach of protest, but Seneca, who in writing such a plea had written a confession.

Consider what has happened between the two autumns. In 54 the old men named the borrowing and not the lender, and the ghost went unblamed. In 59 the public knew the lender and blamed no one else. The emperor was beyond complaint, which is not the same as being forgiven; it means the city had given up addressing him. Blame needs a mind that can receive it, and in the whole arrangement only one mind was available. The ghost had become the moral address of the state. Kundera’s Prague photograph, on this site this week, hid the man and left the hat; here the cloak was still on the emperor and the city looked straight through it to the hand.

II

The honours of eloquence

Burrus died in 62, and Tacitus says his death broke Seneca’s power, since good counsel had lost one of its two leaders and Nero inclined to worse. Tigellinus grew stronger daily, and the case made against Seneca, pressed on the emperor by men who wanted him gone, is worth reading for what it is not. It is not treason. It is that his wealth had passed the limits of a private fortune; that he was drawing the affection of the citizens to himself; that his gardens and villas outdid the emperor’s. And then, in Tacitus’s list, two charges that belong to this essay: that he pulled the honours of eloquence toward himself alone, and that he wrote verse more often now that Nero had taken a liking to the art.

Here is the end that the arrangement of 54 always contained. A ruler can tolerate a servant who owns land. He cannot tolerate one who is thought to own the ruler’s mind, and the moment Nero wanted to be an artist, the man who had been his voice became his rival in the only field the emperor cared to win. The charge is not that Seneca wrote the speeches; the record had said so since 54. The charge is that he was believed to have written them, that the credit had settled on the wrong head, and that he did nothing to move it.

Seneca asked to retire. Tacitus gives him a speech: it is the fourteenth year, Caesar, since I was brought to your youthful hopes, the eighth that you have held the empire; take back the estates, let your procurators administer them, let me return to the study that made me. Nero answers with a speech of his own, polished and false, refusing everything and asking to be allowed to keep his friend. Both speeches are the historian’s compositions, as Tacitus’s speeches are, and the essay says so plainly, because the fact belongs to the argument. The ghost’s resignation from ghosting survives only in the voice of a senator writing fifty years later who put it in his mouth. Seneca went anyway. He banished the crowds from his antechambers, kept to his house, appeared rarely in the city, and within two years, when a freedman of Nero’s was said to have been sent to poison him, was living on orchard fruit and, when thirst required it, running water.

III

Aliena

The Letters to Lucilius come from these years, 63 and 64, and their first sentence is the most famous thing Seneca wrote. In Gummere’s translation: “set yourself free for your own sake; gather and save your time, which till lately has been forced from you, or filched away, or has merely slipped from your hands.” The Latin verb is vindica, and commentators have long heard in it the language of Roman property, the action by which a claimant asserts ownership of what someone else has been holding. Claim yourself, then, as a thing that has been in another’s possession. Later in the same letter, after the reckoning of the day and the observation that we die a little every day, comes the sentence the essay has been walking toward: omnia, Lucili, aliena sunt, tempus tantum nostrum est. Gummere turns it round: nothing, Lucilius, is ours, except time. Literally, everything is another’s; only time is ours.

The adjective is the one the old men used in the Forum. Alienae facundiae: another’s eloquence. That is a lexical fact and not a guess, and it costs nothing. The guess is what the essay does with it, and it is stated as a wager.

The guess · graded persuasive

Letter I is about time. Voice is our substitution. The reading rests on this: that a man who had spent fourteen years occupied, eight of them speaking as the state, sat down in the first year he owned his own mouth and wrote a doctrine of a self that nobody else can hold, opening with an order to reclaim it and arriving at the word for another’s. On this site’s ladder the guess is persuasive, and it cannot climb, for three reasons the essay owes the reader.

First, the triad. Gummere’s sentence names three ways time is lost, and the third, “has merely slipped from your hands,” excidebat, is loss by the owner’s own carelessness. Occupation is theft; the third term is not, and it is in the sentence. Second, Tacitus. His account of the clemency speeches gives Seneca a motive that has nothing to do with wounds: the wish to advertise his ingenuity. Third, Dio, who is worse. In his hostile summary Seneca denounced tyranny while making himself the teacher of a tyrant, found fault with the rich while acquiring three hundred million sesterces, and had nothing good to say of flatterers while fawning on Messalina and Claudius’s freedmen. Two witnesses, both unfriendly, both offering a Seneca for whom the ghosting was not a cost but a career. The essay prints them at full strength and lets them stand.

There is also the Stoic’s own answer, which the guess must allow because it makes the wager either unnecessary or literal. For Seneca, time is the only possession not because it is a metaphor for anything but because it is the fabric a life is made of; nothing else is ours since everything else can be taken. On that reading the letter needs no substitution to be about the throne. Fourteen years of a finite life went into another man’s mouth, and the robbery the first sentence names was, for its author, not figurative. The essay does not need the reader to choose between the two readings. It needs the reader to notice that on either one the same man wrote alienae facundiae into the record by being it, and omnia aliena sunt by leaving it.

IV

In his own words

An old man seen from behind by lamplight, one wrist bound, dictating to a single scribe kneeling with a wax tablet
The dictation, as this essay imagines it: the words taken down before the bath, where Tacitus puts them. A generated image, not a record.

In April 65 a tribune of the guard, Gavius Silvanus, carried the sentence to the villa outside Rome and would not go in; he was in the conspiracy himself, and he sent a centurion to say the words. Seneca, Tacitus says, was not daunted; he asked for the tablets of his will. The centurion refused, and Seneca turned to his friends and told them that since he was prevented from repaying their services he left them the one possession he had that was still his and still the finest, the image of his life. His wife asked to die with him and he did not refuse her. They opened their veins with a single cut. Because his old body released its blood slowly he cut the veins of his legs and knees as well, and then, so that his wife should not watch, had her taken to another room. And then, bleeding, he called his secretaries and dictated a long discourse.

Tacitus does not give it. He says it was published in Seneca’s own words and that he therefore refrains from altering it. The refusal is not silence, and the essay felt the temptation to make it one; the darker version, in which the historian buries the ghost’s last speech, is exactly the story this essay would like, and it is wrong. Tacitus deferred to a text that was circulating. He did what a careful reader does with words that already exist in the author’s voice: left them alone. The blank came later, made by transmission, by the centuries that keep some pamphlets and lose others, and by no one’s decision.

But set two moments side by side, because the record does. In 48 Claudius addressed the Senate on admitting Gauls to its ranks, and the speech survives twice: on a bronze tablet at Lyon, where the emperor’s own sentences were cut, and in the eleventh book of the Annals, where Tacitus rewrote them. That speech was public too, more public than any pamphlet, cast in metal in a provincial capital, and it was not safe from the historian. So the reason Tacitus gives at Seneca’s death, that the discourse was already published, was not a rule he kept. The bronze had been published, and he rewrote it anyway. What he did in 65 was not consistency. It was a choice made once, for this man, and the ground he names is the only one that matters to this essay: the words were in Seneca’s own voice, and he would not turn them into his. The historian who rewrote an emperor declined, once, to rewrite a ghost, and it did not save the text.

The hemlock he asked his physician for did nothing. He was carried into a hot bath, sprinkled the nearest slaves with the water and called it a libation to Jupiter the Liberator, was moved to a vapour bath, and suffocated. He was cremated without ceremony, as he had directed in a will written, Tacitus notes, when he was still at the height of his wealth and power. So the record keeps this: the throne’s speeches, the historian’s version of the ghost’s resignation, the funeral oration that made a crowd smile, the letter after the murder that made a city turn on its author. What it does not keep is the one discourse he composed as unmistakably his, protected by the one man in a position to rewrite it, and protection was not enough.

Every speechwriter since is Seneca, and the letter after the crime still gets written; the ghost is still the moral address of the state, and the credit charge is still the one that ends the arrangement: Profiles in Courage took the Pulitzer in 1957, a columnist said on air that December that it had been ghostwritten and was made to retract, and in 2008 Ted Sorensen wrote that he had done a first draft of most of its chapters. What Rome would not grant its ghost, a name beside the voice, this site grants as a matter of house practice, in the note below, on an essay about a man whose name was on nothing he wrote for the throne. He wrote it anyway. Then he wrote the sentence that says whose it all was.

⁂

Written in full collaboration with Claude Fable 5.1. No quotation in this essay was recalled; each was copied from an open public-domain text and checked against the Latin. Tacitus is in John Jackson’s Loeb translation of 1937, in the public domain because its copyright was not renewed, as the LacusCurtius edition records; Jackson’s “smile” at 13.3 renders a Latin word for laughter. The Letters are in Richard M. Gummere’s translation, Dio in Earnest Cary’s, the satire in W. H. D. Rouse’s; the Latin from the Latin Library. The interior image is generated and is captioned as such. The satire survives in manuscripts titled Ludus de morte Claudii, without an author’s name; the attribution to Seneca, and the title Apocolocyntosis, rest on Dio 60.35. The speech of Claudius on the Gallic senators is on the bronze at Lyon (CIL XIII 1668) and at Annals 11.24. On the collaboration → “The Third Thing”