The Rumour That Grew a Roof

Three witnesses, nine questions

This page reads three ancient narratives of the fire. Tacitus tells it in the Annals; Suetonius in his Life of Nero; and Dio in a book we read in the abridgment made in the eleventh century by Xiphilinus, a monk of Constantinople (For Books LXI–LXXX our chief authority is Xiphilinus, as Cary puts it). They are not the only ones: Furneaux, after naming the writers who charged Nero with the fire, adds that a similar view is taken by the author of the ‘Octavia’ (831–833), and is followed by Orosius (7, 7); while Sulp. Severus (see on c. 40, 3) more closely follows Tacitus. Pick a question. Each column lights the words that answer it, in the Latin or Greek with a public-domain English translation under it, and says how that writer holds it.

All nine at once, as the columns tag them:

• Tacitus: 4 stated as fact, 2 reported as rumour, 1 left open, 2 silent; 0 of the 3 details of the performance stated as fact.
• Suetonius: 6 stated as fact, 3 silent; 3 of the 3 details of the performance stated as fact.
• Dio: 6 stated as fact, 3 silent; 3 of the 3 details of the performance stated as fact.

Read down the columns. The earliest writer, Tacitus, keeps the cause open, puts Nero at Antium, and gives the song, and the stage it was sung on, as rumour. The two later writers state as fact every detail of the performance, where he stood, what he wore and what he sang, which Tacitus either reports as rumour or does not give. Furneaux’s commentary on the Annals, read here in the second edition revised by Pelham and Fisher (1907), puts it in one line: Suet. (Ner. 38) and Dio (62. 18, 1), who affirm as a fact what Tacitus here gives as a rumour. On the cause he wrote: We owe it to Tacitus that any doubt is left on the subject. He adds that Tacitus himself at least inclines to charge Nero with the second outbreak (c. 40, 3).

Two cells need a closer look. Dio’s roof first. The Greek is ὁ Νέρων ἔς τε τὸ ἄκρον τοῦ παλατίου, which Cary renders Nero ascended to the roof of the palace, and his note on it reads τοῦ παλατίου R. Steph., τὰ παλάτια VC.. The note’s first words, τοῦ παλατίου, of the palace, are the reading printed by R. Stephanus; where they stand, the two manuscripts V and C have τὰ παλάτια. So the top, τὸ ἄκρον, which Cary renders roof, is in the manuscripts; what differs is the form of the word for palace. Second, how long it burned: Church and Brodribb print At last, after five days, an end was put to the conflagration at the foot of the Esquiline hill where the Latin has sexto, sixth (this page’s gloss): Sexto demum die apud imas Esquilias finis incendio factus.

Two dismissals, one ladder

There are two quick ways to dismiss the story, and they point in opposite directions. One says: Nero was at Antium, so it is a myth. The other says: three historians agree, so it happened. Both treat the three accounts as equal witnesses standing side by side. They are not. Here each stands where it belongs, by when it was written and by how it reaches us.

1. AD 64 The fire. Some, Tacitus says, noted the day it began: fuere qui adnotarent xiiii Kal. Sextilis principium incendii huius ortum (Some persons observed that the beginning of this conflagration was on the 19th of July). The year of the fire.
2. Before AD 79 Pliny the Elder, Natural History. Pliny mentions Nero’s fires in passing; the three editions are below. He died in the eruption of Vesuvius, which Cary’s margin to Dio dates A.D. 79. Within 15 years of the fire.
3. AD 115 or 116 Tacitus, Annals. Furneaux: The date of completion and publication of his various works rests on internal evidence. And: This last work must have been completed when the conquests of Trajan had reached their highest point, in 868 or 869, A.D. 115 or 116. 51 or 52 years after the fire.
4. AD 120 Suetonius, Lives of the Caesars. Rolfe: the Lives of the Caesars, published in 120. 56 years after the fire.
5. AD 200 to 222 Dio, Roman History. Cary, from Dio’s own account of his years of work: The period of these labours may be roughly estimated as the years 200–222. 136 to 158 years after the fire.
6. The ninth century Oldest copy of Suetonius. Rolfe on Suetonius’s manuscripts: The codex Memmianus of the ninth century. Our oldest and best manuscript. No year given, only the century.
7. The latter half of the eleventh century Oldest copy of Tacitus, Annals 11-16. Furneaux: the second Medicean manuscript, a manuscript in Lombard characters, generally assigned to the latter half of the eleventh century, which contains all that we have of these later Books of the Annals. Of the other copies: Many other manuscripts of this portion of the works of Tacitus exist, but none of them can claim any earlier date than the middle of the 15th century. The theory that they come from another old manuscript has been adopted by some editors, notably by Walther. Furneaux goes on: Recent editors, however, have generally held that all the other MSS. are ultimately derived from the Medicean. No year given, only the century.
8. AD 1071 to 1078 Xiphilinus abridges Dio. Cary: For Books LXI–LXXX our chief authority is Xiphilinus, Xiphilinus, a monk of Constantinople, who made an abridgment of Books XXXVI–LXXX at the request of the emperor Michael VII. Ducas (1071–78); his task was very carelessly performed. Epitomes like his, Cary adds, while they often repeat entire sentences of Dio verbatim, or nearly so (as may readily be seen by comparing extant portions of the histories with Zonaras or Xiphilinus), must, nevertheless, be regarded as essentially paraphrases. Cary prints the chapters on the fire under the running head EPITOME OF BOOK LXII and ends them Xiph. 166, 17–169, 10 R. St.. 1007 to 1014 years after the fire.
9. The fifteenth century The two manuscripts the other copies derive from. Cary: of the epitome’s manuscripts, all the rest are derived from one or the other of two fifteenth century Mss., Vaticanus 145 and Coislinianus 320. No year given, only the century.
10. AD 1551 First printed edition of the epitome. Cary’s list of editions: 1551 R. Stephanus. Editio princeps of Xiphilinus. Cary prints his reading of the word for palace, not the manuscripts’; the note: τοῦ παλατίου R. Steph., τὰ παλάτια VC. 1487 years after the fire.

Read against the ladder, the agreement of the three is not three witnesses agreeing. The nearest of the three to the fire calls the song a rumour. The two who make it fact wrote later, and the latest, with the people leaping into the flames, the citharode’s gear and the palace roof, comes to us through an eleventh-century abridgment. Gainsford, who defends the story as ancient, still says: Cassius Dio can’t be regarded as independent of the earlier two.

The alibi covers the start

The other dismissal, that Nero was at Antium, is Tacitus’s own clause, and it is true of the start: Eo in tempore Nero Antii agens non ante in urbem regressus est quam domui eius, qua Palatium et Maecenatis hortos continuaverat, ignis propinquaret. (Nero at this time was at Antium, and did not return to Rome until the fire approached his house, which he had built to connect the palace with the gardens of Mæcenas.). The same sentence brings him back while the city burned, and the next one says what happened then: neque tamen sisti potuit quin et Palatium et domus et cuncta circum haurirentur. (It could not, however, be stopped from devouring the palace, the house, and everything around it.)

It began, Tacitus says, on a day some people noted: fuere qui adnotarent xiiii Kal. Sextilis principium incendii huius ortum. Counting a.d. XIIII Kal. Sextiles back from the first of August, both ends counted, gives the 19th of July, which is what Church and Brodribb print: Some persons observed that the beginning of this conflagration was on the 19th of July. The first phase ended Sexto demum die apud imas Esquilias finis incendio factus, and then rursum grassatus ignis patulis magis urbis locis: But before people had laid aside their fears, the flames returned. Tacitus gives no length for that second outbreak. Suetonius gives For six days and seven nights destruction raged.

A stone gives a total. Domitian dedicated a fire-altar and the precinct around it, and one of the stones that marked out the precinct (Dessau’s ILS 4914) says the vow was made quando urbs per novem dies arsit Neronianis temporibus, when the city burned for nine days in Nero’s time (this page’s translation). Dessau notes that this stone was semel descripta, copied once, and that Similes lapides etiam in aliis urbis partibus fuisse videntur, that is, similar stones seem to have stood in other parts of the city (this page’s translation). Furneaux calls it a votive inscription, not now extant. Dessau ties its nine days to Tacitus’s second outbreak: Scilicet cum ‘sexto die finis incendio factus’ esset, mox denuo erupit. Nipperdey, Furneaux reports, reconciled the two by supposing that the second outbreak, here mentioned by Tacitus, lasted three days: 9 − 6 = 3. That is a supposition, and the page draws it as one.

So the alibi holds for the beginning and no further. By Tacitus’s account Nero came back while the fire was still advancing on his house; it then consumed the house, and it burned on, to the sixth day and then again. On which day he came back, Tacitus does not say.

The debunk, checked

Pages written to correct the story make claims of their own about the sources. Here are seven sentences from three of them, History.com, Live Science and The Phrase Finder, each beside the words it describes. Open one.

Of these 7 sentences, 1 matches the ancient text, 1 gives one count of several, and 5 do not match it.

By this page’s reading, three of the five slips run in Nero’s favour: he hurried back, in two of them, when Tacitus says he did not come back until the fire approached his house; and Tacitus doubted the rumour, when he only reports it. Two run against him: a song about Rome, when Tacitus says Troy; and an instrument in Suetonius, who gives him none.

An earlier witness, in three editions

One writer is earlier than all three. Pliny the Elder died in the eruption of Vesuvius: his nephew’s letter giving an account of my uncle's death says of the cloud that it was found afterwards to come from Mount Vesuvius, and Cary’s margin to Dio dates the eruption A.D. 79. In the Natural History Pliny tells how long the lotus trees at the orator Crassus’s house lived: until Nero’s fires. Whether he adds that Nero burned the city depends on the edition you open.

The clause is quibus cremavit urbem, by which he burned the city (this page’s translation). Bostock and Riley’s English of 1855 follows a text that has it: they were in existence down to the period when the Emperor Nero set fire to the City, one hundred and eighty years after the time of Crassus. Furneaux’s commentary quotes it too: Pliny also (N. H. 17. 1, 1, 5) speaks of ‘Neronis principis incendia quibus cremavit urbem’. This page does not judge between the editors; it shows what each one printed, read from the page images. This page did not read Mayhoff’s introduction, so it does not say what the D and v of his note stand for.

Where the fiddle came from

Barrett, who traced the English phrase in 2021, starts from this: The image of Nero actually playing an instrument during the Great Fire is not, in fact, found in ancient sources. Gainsford reads Dio the other way: in his rendering Nero put on his gear for kithara performances, and he writes, We have to substitute an ancient concert kithara for the modern violin. The instrument arrives in English, and it changes on the way. Each entry says whether its words were read here in a printed text, read in a modern text, or only reported.

1. reported by Barrett, not read here Cooper’s dictionary, 1548. Barrett: the first English reference is in Cooper’s revision of Elyot’s Latin–English dictionary, where Nero is said to play a harp during the conflagration. The instrument: a harp.
2. read here in a modern text Henry VI, Part 1, act 1, scene 4. Gainsford dates the play Henry VI part 1 (ca. 1591-1595). Talbot, over the dying Salisbury: Plantagenet, I will; and, like thee, Nero,
   Play on the lute, beholding the towns burn. The instrument: a lute.
3. read here in print The Tragedy of Nero, printed 1624, act 3, scene 3. At the theatre, not at the fire: Petronius has seen Nero In Mynstrills habit stand before the Iudges, and Seneca answers, I have a Fidler heard him, let me not
   See him a Player. Bullen’s note on the first word reads 4to. I'd.: the quarto has I’d. A fiddler on the stage, as an insult.
4. read here in print The same play, scene 4: the fire. Soft Musique. Enter Nero above alone with a Timbrell. Nero speaks: I, now my Troy lookes beautious in her flames; The instrument at the fire: a timbrel.
5. reported by Barrett, not read here George Daniel, Trinarchodia. Barrett: In 1649 the royalist poet George Daniel applied the term ‘fiddle’, citing Trinarchodia. The Several Reigns of Richard II, of Henry IV and of Henry V (London, 1649). The line was not read here. Grosart’s edition of Daniel’s poems, printed in 1878, says on its title page that it is edited from The Original MSS. in the British Museum: HITHERTO UNPRINTED.
6. read here in print Silas Titus, House of Commons, The 17th of November, 1680. Arguing that the House should not take up the state of Tangier: we may be reflected on, as Nero was, for being playing on his Fiddle when Rome was on Fire. Printed in A Collection of Debates in 1725. Barrett says the familiar form was apparently coined in a 1680 English parliamentary speech by Silius Titus; the print spells him Silas.

The first wording read here that puts a fiddle in Nero’s hands while Rome burns is therefore the speech of 1680, as printed in 1725. The phrase, in its exact form or near it, keeps being found earlier, which is why every earliest is only the earliest found so far: Gainsford, in 2017, said a speech of 1820 was what this appears to be the earliest occurrence of the exact phrasing, ‘Nero fiddled while Rome burned’; Barrett, in 2021, quotes the preface of The Noble Cricketers of 1778, That Nero fiddled while Rome burned.; and Tréguer, in 2026, gives as the earliest allusive use he had found a near form, from The Public Advertiser of 15 March 1770: Nero fiddled whilst Rome was burning. None of those three was read here.

The search runs over the quotations shipped with this page, not over whole books. The whole-chapter version of the same test is described in the check below; it runs only where the full chapters are kept, inside the site’s own repository, so a reader’s run cannot make it.