According to the scholarly communis opinio, our sources give the length of the Republican Rome’s fourth-century solitudo magistratuum/ἀναρχία as four or five years with the exception of Diodorus, according to whom it lasted one year.Footnote ¹ The aim of this paper is to demonstrate that he says nothing of the kind and that his chaotic Roman chronology of the period strongly suggests that in the source he followed there were also several years without eponymous magistrates.

The main texts are as follows:

Scholars usually opt for the one-year version, in spite of Diodorus’ poor reputation as a historian of Rome in general and as her chronographer in particular. As succinctly expressed by Stephen Oakley, the author of the most recent exhaustive treatment of the question,Footnote ² ‘only D.S. with an anarchy of one year approaches the brevity required. We shall see that this version of D.S. is likely to be the oldest and to reflect some genuine tradition’ (Oakley [n. 2], 105).

Now, the two pillars of any serious attempt at reconstructing Rome’s fourth-century chronology are: (1) Polybius’ synchronism of the Gallic sack of the City with the Peace of Antalkidas and the destruction of Rhegion by Dionysius (the Athenian year 387/386: Polyb. 1.6.1–2); (2) the rejection of the four dictator years in the ‘Varronian’ chronology represented by the Fasti Capitolini (333, 325, 309, 301) as a historical and constitutional absurdity, which must be ignored. Yet, once we reject the dictator years, the Varronian chronology is the same as in Livy, and both match Polybius’ synchronism. Does it mean that there is no problem with the Roman chronology? No, says Oakley, for ‘86 colleges of consuls and consular tribunes were needed between the sack and 300, or 85 if, as seems most likely, there really was a period of anarchy; but it seems that only 81 were available … Both the dictator years and the extended anarchy seem to have been artificial devices designed to make up for the shortage of eponymous magistrates. For 81 colleges and one year of anarchy leave a shortfall of four years, and this may be met either by the insertion of four dictator years or by the extension of the anarchy to 5 years. Varro unwisely incorporated both; L., more sensibly, ignored the recently invented dictator years and chose to extend the anarchy’ (Oakley [n. 2], 105). And the conclusion: ‘It thus follows that for the period before 366 (or 370) we cannot synchronize with confidence events in Greek and Roman history’ (Oakley [n. 2], 106).

The ‘shortage of magistrates’, however, is entirely of our making: if we do not arbitrarily reduce the solitudo magistratuum from five years to one, there is no ‘shortfall of four years’. It is also difficult to believe that Varro, the most learned of the Romans, got so baffled by the primary-school arithmetic required to make up for the said shortfall that he did his homework twice without being aware of the error (which the less informed and often careless Livy did not commit). More importantly, Polybius’ synchronism implies that his date of the Gallic sack––compatible with Livy’s chronology which did include the five-year-long solitudo magistratuum––figured already in his Roman source, Fabius Pictor.Footnote ³ In other words, Livy did not have to make any choice: this choice had already been made two hundred years earlier. The addition of the four phoney dictator years must therefore have been an outcome of speculations with an entirely different purpose. What this purpose was is an open question, but I do not think that, ‘since the dictator years appear to have been confined largely or entirely to chronographic works, it is natural to look for an explanation in this area’ (Drummond [n. 3], 567–8). Making the Romans get rid of their tyrants in the same year as the Athenians did is as good a guess as stretching the count post reges exactos to match that of Cn. Flavius’ post [aedem] Capitolinam dedicatam (actually better chronologically). Be that as it may, the ‘shortage of eponymous magistrates’ can be safely ruled out: that shortage had already been accounted for by the solitudo magistratuum. The sceptics’ only argument, other than conviction that the anarchy could not last so long, is Diodorus’ report.

Or is it? Strictly speaking, Diodorus’ two references to anarchy in Rome, not only 15.75.1 but also 15.61.1 (369/368, Varronian 377), are not parts of the narrative but of the formula with which he starts describing events of successive years: ἐπ᾿ ἄρχοντος δ᾿ Ἀθήνησι or the like plus a given year’s archon’s name, followed by ἐν Ῥώμῃ or similarly plus the names of that year’s consuls or consular tribunes; in other words, they are yearly eponymic dates. For 369/368 the formula is more developed, because Diodorus first notes that the year started without magistrates, gives the reason of contention, mentions the στάσις and the subsequent ἀναρχία, and only then records the election of consular tribunes, the year’s eponymsFootnote ⁴ . Then, after 368/367 (Varronian 376), under 367/366 there is our notice: κατὰ μὲν τὴν Ῥώμην ἀναρχία διά τινας πολιτικὰς στάσϵις ἐγένϵτο. Which Varronian year it stands for is, however, anybody’s guess, for though in the subsequent year (366/365) the dating formula returns to normal (15.76.1), it enumerates the tribunes of 370. One thing is certain: this single year is Diodorus’ counterpart of the five years of solitudo magistratuum in Livy (Varronian 375–371). The question is what happened to the remaining four?

We know the now prevailing answer to this: these four years are as non-existent as the dictator years, with the inescapable gloomy corollary that Polybius’ synchronism is wrong, so that, for example, the sack of Rome, if datable at all, should be dated to 383/382 rather than to 387/386. To the equally inescapable question of how so great a mistake could have been made one might retort that, if Roman intellectuals were able to concoct such a monstrosity as the dictator years, their capacity to blunder and invent was unlimited. This, though, does not quite settle the problem, for it was a Greek who recorded the synchronism which in our sources goes together with the five/four-year-long anarchy. But the real objection is the above-mentioned fact that Diodorus––unlike Livy’s eaque solitudo magistratuum … per quinquennium urbem tenuit, and Zonaras’/Cassius Dio’s ἐπὶ τέσσαρσιν ἔτϵσιν ἀναρχίαν γϵνέσθαι τῷ δήμῳ––does not say how long the anarchy lasted; he only dates one year with a record of its occurrence.

Now, this date is but a consequence of a chronological error, one of the many Diodorus committed, though greater than most (see Table at the end of this Note). In Books 13 and 14, embracing the period ending with the Peace of Antalkidas and the destruction of Rhegion, he uses Polybius’ synchronism of the Greek and Roman chronology in which the Gauls sacked Rome in the same Athenian year 387/386; and this gives a difference of three/four years to the Varronian chronology. Yet, whereas in that year (Varronian 390) the difference still has this length (14.110.1), in the next (Varronian 389), described already in Book 15 (15.2.1), it grows to eight/nine years, because at the year’s beginning the author lists for the second time the consular tribunes of the Varronian year 394 (first enumerated in 14.97.1) and keeps this ‘synchronism’, starting with repeating in the subsequent yearly dates the colleges of 393–390,Footnote ⁵ up to the aforementioned year 376, Athenian 368/367. Then, after reporting the anarchy, he almost returns to Polybius’ synchronism, this time calibrated to four/five years of difference, with a jump forward to the Varronian year 370, Athenian 366/365.Footnote ⁶ It is in this chronological doublet and the resulting five-year-backward shift of the Roman chronology that the four missing years of Livy’s solitudo magistratuum were lost.

Naturally, one may speculate what brought this doublet about. Andrew Drummond believes that we have to do with a manipulation: ‘… although he appears to accept the Greek synchronism which located the Gallic Sack in Ol. 98.2 (387/6 b.c.), his list of consuls has been … blatantly manipulated to fit it’Footnote ⁷ . But why should Diodorus, always clumsy in chronological matters and basically uninterested in the early fourth-century Rome,Footnote ⁸ ‘manipulate’ the Roman chronology of Book 15 backwards, starting almost from the book’s end (in our editions from Chapter 75 of the book’s ninety five)? Just to make the year of the Roman ἀναρχία––to him plainly the same date as any––fall at the chronologically accurate (more or less) place at the price of an unbelievable mess in the preceding fourteen years? The only reasonable explanation for doubling the colleges of 394–390 is an accidental skip back on the list of (for a Greek) constantly recurring similar exotic names, made at the book’s beginning.Footnote ⁹

If so, then the view that Diodorus preserved a tradition which limited the solitudo magistratuum to one year is groundless. To the contrary, the fact that he was able to make a sudden jump ahead from 376 minus one to 370 suggests—nay, indicates—that in the source, or sources, he followed there were several years with no eponyms. We read about one of these unnamed years in Pliny the Elder’s note about the temple of Iuno Lucina, anno, qui fuit sine magistratibus, CCCLXXIX urbis … condita (‘founded in the year which was without magistrates, the City’s three hundred seventy ninth’ [Varronian 375], HN 16.235). But in an eponymic system of chronology the years sine magistratibus in a certain sense could not exist: for example the chronograph of 354, in his list of the consuls from 509 b.c. to a.d. 354 representing the ‘official’ Roman chronology which included both the dictator years and the five-year-long solitudo magistratuum, in order to account for the latter was obliged to put in five fictitious pairs of eponyms (Degrassi [n. 1], 396–7).Footnote ¹⁰ To Diodorus they offered space to correct his Roman chronology (Perl [n. 2], 113).

All this leads to another, more general conclusion: on the evidence of our sources the ancients agreed that the anarchy caused by the conflict over the Licinio-Sextian rogatio lasted several years (four? five?). Of course, they may have been wrong about it, just as they may have been wrong about the pan-Mediterranean synchronism recorded by Polybius and everything concerning the Early Republic. But modern scholars who reject these claims can base their rejection on nothing but their inability to imagine how it was possible.

Table: Diodorus’ yearly dates and their ‘Varronian’ equivalents (Degrassi [n. 1], 386–97)