Varronian Chronology

The Romans did not have a common era like we have. Instead, they called their years after the two supreme magistrates, the consuls. The year that corresponds to our 59 BCE was known to them as 'the year in which Gaius Julius Caesar and Marcus Calpurnius Bibulus were consuls'. As long as one had a list of magistrates, one could date all past events. Using a similar list, we are able to convert Roman year names to our year numbers.

Originally, the pontifex maximus was responsible for the maintenance of the official version of this list, which was published at the end of the second century BCE. However, there are two problems, which are usually ignored by modern historians.

1. The Roman year did not start on 1 January, but on 1 September (in the fifth century) or 1 July (in the fourth century). A Roman year name should after conversion have two elements (e.g., '300/299'), not one ('300'). If it has only one element, it is almost certainly inaccurate.
2. The list seems to be incomplete. Probably, four couples of consuls are missing. This is the main problem.

To return to the main problem: it was already recognized in Antiquity, and three solutions were proposed. The first one was to accept a period of anarchy of several years in the 360's. This was not a bad solution, because it is certain that there was at least one year in which the tribunes Licinius and Sextius forbade the election of magistrates. The hypothesis that this period lasted longer than one year can not be falsified, and this may be the reason why the great Roman historian Titus Livy (59 BCE - 17 CE) accepted this solution.

The second solution was to introduce three fictional couples of magistrates following one year of anarchy in the 360's. The names of the six fictional consuls are Papirius and Vivius, Sacraviensis and Caeliomontanus, Priscus and Cominius, and are mentioned in a document from Late Antiquity known as the Chronicle of the year 354.

The third solution is that of the Roman antiquarian Marcus Terentius Varro (116-27), a contemporary of Julius Caesar. He introduced four years in the second half of our fourth century BCE in which Rome was ruled by dictators. These years are, in the Varronian chronology, 333, 324, 309 and 301.

There is not one single argument to corroborate this constitutional absurdity, which is just an all too transparent attempt to invent historical precedents for the annual dictatorship of Julius Caesar in 48 BCE. In fact, there are indications that falsify Varro's idea:

• Pliny the Elder writes that the career of the popular Roman general general Marcus Valerius Corvus lasted 46 years, but it lasts 50 years in Varro's chronology. The difference disappears when the 'dictator years' are ignored. It also means that the treaty between Rome and Carthage, usually dated to Varronian year 348, was in fact concluded in 344/343 BCE, exactly at the moment when the Carthaginians are known to have intervened in a civil war on Syracuse and needed support in Italy.
• Livy writes that in the Varronian year 351, the Romans and the inhabitants of Tarquinii concluded a peace treaty for forty years. War broke out again in the Varronian year 308, which is indeed the fortieth year if we ignore the dictator years.

• The dating of the invasion of the Gauls in 390; the real date can be deduced from the Greek historian Polybius of Megalopolis and is 387/386 BCE (World History, 1.6.1-2). There is independent confirmation from Diodorus of Sicily. (The Roman army was defeated on 18 July. As we have already seen above, this can not be identified with 'our' 18 July, but we can be almost certain that the invasion took place in the summer/autumn of 387, although 386 can not be excluded.)
• The dating of the expulsion of Rome's last king, Tarquin the Proud, in 510. Varro must have liked this date, because it offers a nice synchronism with the expulsion of the tyrant Hippias from Athens. (And it must have been appreciated more generally. Writing a century later, Pliny the Elder was also aware of it.). However, it is simply wrong. A fragment from the city chronicle of the Greek town Cumae, quoted by the Greek scholar Dionysius of Halicarnassus (Roman Antiquities, 5.36), proves that the battle of Aricia in the second or third year of the Roman republic, took place in 505/504; this means that the republic was founded in either 507/506 or 506/505 (more...).
• The dating of the foundation of Rome in 754/753. Varro obtained this year by multiplying the number of known kings, seven, by 35 regnal years. It must be noted that there are still Italian scholars who ignore the radiocarbondates and maintain that the hill top villages of Rome were united -the equivalent of the founding of a city- in the middle of the eighth century, which is a century too late.

Livy is better. His History of Rome since its foundation contains the most complete list of magistrates available. We can use it to convert Roman year names into our year numbers, and can see where the four missing couples of consuls must be placed.

To start with, in Livy's History, the career of Marcus Valerius Corvus lasts 46 years and the peace with Tarquinii lasts forty years. This means that we can rely on his list of magistrates for the years between 347/346 (351, according to Varro) and 299/298. This may even be corroborated, if Livy's statement in 7.28, that in the consulship of Gaius Marcius Rutilus and Titus Manlius Torquatus the sky turned dark, is a reference to a solar eclipse; in Livy's chronology, this year corresponds to our 340/339 and that is correct, because a eclipse took place in the early morning of 15 September 340. (Varro and his modern epigones want us to believe that this happened in 344.)

To sum up: we can be certain about Livy's datings in the years after 347/346 BCE. All Varronian dates can be converted to our era.

However, there is no external check for the preceding period. The invasion of the Gauls offers the point where we can restart our analysis. As we have already seen, Polybius states that this happened in 387/386. Livy mentions 34 groups of magistrates and 5 years of anarchy for the years until 347/346, which means that his relative chronology is one year too short.

Counting backwards from 387/386, the problems disappear. Livy and Varro are more or less in agreement on the colleges of magistrates. Using 387/386 as starting point (instead of Varro's 390), we arrive at 506/505 as the first year of the Roman republic, which is in line with the date mentioned above.

This suggests that the list of magistrates for the fifth century is reliable, and that we can count backwards from 387/386 BCE. Two arguments may corroborate this.

1. Using this system, the year in which Marcus Fabius Vibulanus, Marcus Folius Flaccinator and Lucius Sergius Fidenas were military tribunes with consular powers, is identical to 430/429 BCE (and not Varro's 433). The plague that is mentioned by Livy for this year, is perhaps identical to the one described by the Athenian historian Thucydides under the year 429 BCE.
2. The Sicilian general Dionysius became sole ruler of Syracuse in 409 BCE; until then, Syracuse had been a democracy. According to Varro's chronology, the Romans bought grain from the tyrant in 411, which is impossible. According to Livy, it was 408/407, which fits excellently.

It is possible to be a little bit more precise about the location of the missing colleges. Fabius Pictor states that the first plebeian was consul in the twenty-second year after the invasion of the Gauls. These events took place in 390 (Varro) and 366 (Varro). In fact, only nineteen colleges are known for this period, which narrows the location of the missing colleges to the interval between 390-366 (Varro).

Finally, it must be noted that Livy is not the only one who uses an extended anarchy. Varro used it too and must have found it in sources he could not ignore. Moreover, the three couples of magistrates in the Chronicle of 354 CE (above), must have been invented to fill an awkward gap in a consular list. This suggests that the tradition of an extended period of anarchy in the 360's was well established. In fact, the easiest solution to all chronological problems is simply to accept Livy's list as it is, including a long period of anarchy and an uncertainty margin of about one year. This does not solve all problems, but results in a far better chronology than the common Varronian chronology.