« A.U.C. 746 = 8 B.C. »

The Augustan reform, which finally aligned the Roman calendar with the modern Julian calendar and set the period and phase of the intercalary cycle to that of the Julian calendar, was promulgated in this year. This result was determined by J. J. Scaliger, De Emendatione Temporum 159, 238 in 1583. He noted that Suetonius, Augustus 31.2, recorded that Sextilis was renamed "Augustus" at the same time, and that Censorinus 22.16 dates that event to the emperor's 20th year, the consulate of M. Censorinus and C. Asinius Gallus, i.e. A.U.C 746 = 8, as does Dio Cassius 55.6.6.

The story is, of course, not quite so cut and dried. There are indications that Sextilis may have been renamed earlier. Livy, Periochae 134, dates the event to A.U.C. 727 = 27. Macrobius, Saturnalia 1.12.35, quotes the actual senatusconsultum for the change, which justifies it in terms of Augustus' conquest of Egypt and the ending of the civil wars, possibly implying that it was passed not too long after these events. He further says that the motion was passed by a plebiscite under the tribune Sex. Pacuvius, and Dio Cassius 53.20.2-3 records Sex. Pacuvius as tribune in A.U.C. 727 = 27. Finally, K. Fitzler & O. Seeck, RE X (1917) 362 note the Egyptian inscription CIL III 6627, dated to Augustus in an unnamed year, which describes some miltary waterworks. Dio Cassius 51.18.1 says that Augustus began a program of clearing out the canals immediately after the conquest, hence Fitzler & Seeck suggest that CIL III 6627 is more likely to be close to 27 than to 8. They argue that the change actually took place in A.U.C. 727 = 27 and that the decree of A.U.C 746 = 8 merely acknowledged the plebiscite of A.U.C. 727 = 27.

The reference to "Sex[tilis]" in pOxy 61.4175, which dates from A.U.C. 730 = 24, clearly proves that the month was not actually renamed in A.U.C. 727 = 27, while CIL III 6627 is clearly too vague to be used as evidence.

A. B. Bosworth, HSCPh 86 (1982) 151, argued that the senatusconsultum quoted by Macrobius is actually somewhat subversive. He notes that two of the reasons given for choosing Sextilis are (a) that this was the month of his first consulate and (b) in this month the legions on the Janiculum gave their loyalty to him. In Bosworth's view both events were rather shady, and the honor given to Augustus by publicising them was somewhat satirical. He notes that Augustus himself tried to publicise September as the month of his first consulate (Velleius Paterculus 2.65.2 and the Feriale Cumanum), and that having a month renamed after him is the only major honour he received which he does not mention in the Res Gestae. Dio Cassius 55.6.6 lists the renaming towards the end of the year, amongst a discussion of actions involving Tiberius as consul-designate. Bosworth suggests that the senatusconsultum was introduced by Tiberius, on Augustus' instructions indeed, but that he disingenuously spun the wording to remind everyone of some of the more disreputable steps involved in Augustus' rise to power. This action was one step in the developing breach between Augustus and Tiberius which would shortly see Tiberius exile himself to Rhodes.

Its an engaging story, but I don't believe it. The weak points are still the evidence of Livy and the tribune Sex. Pacuvius. Bosworth suggests that Livy's reference merely proves that he wrote after 8, which noone doubts anyway -- but the point is that he listed the event in a book that covered 27. As to Pacuvius, he supposes that one of the unknown tribunes of 8 was another Sex. Pacuvius -- none other than the son of the tribune of 27. Certainly possible, but conveniently coincidental and something of a deus ex machina. (Because the motion for the name change was moved by Pacuvius, the suggestion of J. W. Rich, Cassius Dio: The Augustan Settlement, 224, that Livy's reference to the event under 27 must have been a digression, a forward reference to events of 8, also seems to me unlikely.)

It seems to me most likely that Fitzler & Seeck got the sequence of events correct but misunderstood which one caused the change to become operative. That is, the change was moved and approved in A.U.C. 727 = 27, but not actually implemented until Augustus, as pontifex maximus, issued the appropriate regulation, which he did in A.U.C 746 = 8. The motion was therefore unconnected with the impending breach with Tiberius. As to Bosworth's concerns, I have no doubt that Augustus later wanted to play down certain aspects of the story of his rise to power, but I also suspect that in 27 he was rather more concerned to remind people of the sources of that power and rather less concerned with the respectability of his reputation than he later became. As to why some sources date his first consulate to September A.U.C. 711 = 43 while others date it to Sextilis, this must remain a bit of a mystery.

While the year of the reform can be established, the exact date it was promulgated is unknown. If Suetonius is correct, it took place at the same time as Sextilis was renamed August, which on Bosworth's analysis was towards the end of the year. But, as noted above, there are problems with Bosworth's reasoning and no need to accept his date.

The analysis of the imperial nundinal cycle suggests that this was reformed at the same time as the leap year cycle, and probably as part of the same decree. On the model developed here for the nundinal reform, the earliest leap year affected by it must have been that of A.U.C 746 = 8. This requires that the reform was promulgated in the first few weeks of the year, in time to take effect immediately on the nundinal cycle. This date is consistent with the account of Dio Cassius 55.5, which says that Augustus spent only a little time in Rome that year before heading off to Gaul and Germany, where he spent most of the year.


Reform of the month lengths

The main purpose of the Augustan reform was to replace the triennial leap year cycle, which had been in operation since A.U.C. 713 = 41, with the correct quadrennial leap year cycle, and to realign the calendar. It is widely believed, largely because the story has been propagated in many successive editions of the Encyclopedia Brittanica, that the reform also changed the lengths of several months in the calendar. According to R. Lamont, Popular Astronomy 27 (1919) 583, this story originates with the medieval computist Johannes de Sacrobosco. In his De Anni Ratione, published in 1232, Sacrobosco claimed that the Caesarian months had the following lengths:

31, 29(30), 31, 30, 31, 30, 31, 30, 31, 30, 31, 30

while the Augustan reform changed the month lengths to their modern values as follows:

31, 28(29), 31, 30, 31, 30, 31, 31, 30, 31, 30, 31

Sacrobosco asserted that, when Sextilis was renamed Augustus, its length had to be changed from 30 days to 31 days, since Augustus (the emperor) was "envious because his month was shorter than the month Iulius". The lengths of later months and of Februarius were changed accordingly, as shown in blue.

Sacrobosco gave no source for this account, and I have not been able to trace any earlier reference; it was not known to Bede in the 8th century. Nevertheless, while it could have originated with him, I doubt it. He incorrectly describes the pre-Julian calendar as consisting of alternating 29 and 30 day months, starting in Martius, so that the last month of the year was a 30-day Februarius. Caesar allegedly added 2 days to each of the odd (29-day) months, but reduced Februarius from 30 to 29 days, except in leap years, so that the last odd month, Ianuarius, would be 31 days long. This description views the pre-Julian calendar as a simple lunar calendar of 354 days, which appears to show the influence of a Greek or Arabic source. This influence might be only intellectual, in that these sources used lunar calendars and Sacrobosco, along with contemporary scholars such as Robert Grosseteste, were newly exposed to Greek and Arab astronomical and calendrical ideas -- i.e. Sacrobosco may have simply assumed a lunar structure for the pre-Julian calendar. However, it may also be that some such source gave exactly such a story.

Sacrobosco's story may be compared with Dio's garbled account of the Julian reform (Dio Cassius 43.26), which also assumes that the pre-Julian year was lunar in character, and which also appears to assume a pre-Julian year starting in Martius, since he believes that only 67 days were inserted into A.U.C. 708 = 46 rather than 90. Dio cannot be directly the source of Sacrobosco's account. According to him, Caesar created the Julian calendar by starting with the Egyptian calendar and distributing the 5 epagomenal days, together with 2 days taken from one month (i.e. Februarius), over 7 months to create 7 months of 31 days. That is, Dio attributes to Caesar, not to Augustus, the Julian calendar that we know; by contrast, Sacrobosco's Caesarian calendar has only 6 months of 31 days. Nevertheless, Dio's account shows that alternate accounts of the origins of the Julian calendar were circulating in the Greek East as early as the third century, accounts which may have originated as descriptions of the Julian reform of the lunar Greek calendars of the East. It seems to me perfectly possible that Sacrobosco somehow tapped into this tradition.

If you know more, please email me!

Whatever its source, Sacrobosco's account is provably false. It is contradicted not only by Macrobius Saturnalia 1.14.7 (5th century) and Censorinus 20.9 (3rd century), but also by contemporary evidence. Three items can be adduced, two of which are perhaps more certain than the third:


The Asian Calendrical Reform

Direct evidence for this year is given by OGIS 458 = iPriene 105, which reformed the Asian calendar. This decree was issued by the Greek cities of Asia in response to a proposal of the proconsul of the Asian province, Paullus Fabius Maximus, in early Ianuarius of an unspecified year. It reformed the calendar of the Asian province to one that was locked to the Roman civil calendar. As is usually understood, the decree not only specifies the triennial intercalary cycle but also states that the first leap day occurs in the year of the reform (U. Laffi, SCO 16 (1967) 1 at 28 n. 13).

Since Fabius was consul in A.U.C. 743 = 11, since proconsulates were awarded following a consulate, since proconsulates ran from summer to summer, and since the Augustan reform suspending the triennial intercalation was promulgated in A.U.C. 746 = 8, his proconsulate must have been in 10/9 or 9/8, i.e. the leap year of decree must have been in A.U.C. 745 = 9 or A.U.C. 746 = 8. It has been suggested (e.g. R. K. Sherk, Roman Documents from the Greek East: Senatus consulta and epistulae to the age of Augustus, 336, M.-T. Raepsaet-Charlier, Phoenix 39 (1985) 292) that triennial intercalations may have continued in the provinces even after they had been suspended in Rome. This is extremely unlikely, for the reason given by U. Laffi, SCO 16 (1967) 1 at 32. OGIS 458 clearly states that the first day of the year in the reformed Asian calendar was the birthday of Augustus, a.d. IX Kal. Oct. in the Roman civil calendar. The new Asian calendar took effect part way through the Asian year, in Ianuarius, but very shortly before an intercalation. If there was no corresponding intercalation in Rome, then the reform would have failed to meet its stated goal on the very first day of its first full year of operation!

The standard reconstruction of the triennial cycle requires dating the inscription to A.U.C. 745 = 9. However, there are two factors which favour dating it instead to A.U.C. 746 = 8.

Thus, on the usually accepted interpretation, OGIS 458 shows that the Augustan reform was promulgated too late for news of its actual contents to reach the provinces, and that the last triennial leap year was A.U.C. 746 = 8. This analysis is accepted here.

B. A. Buxton & R. Hannah, in C. Deroux (ed.) Studies in Latin Literature and Roman History XII 290, have recently challenged Laffi's interpretation of this inscription in an interesting article (my thanks to Robert Hannah for a copy). In their view, rather than stating that there is an intercalary Xandikos of 32 days in the year of the decree, the intended meaning is that Xandikos shall be 32 days in each intercalary year. Further, rather than specifying that intercalation shall taken place every third year, the intended meaning is that intercalation shall commence on the day after 14 Peritios in the third year following promulgation of the decree (counting inclusively). In their view, Fabius was proconsul in 7/6, the decree was passed in 7, and the first intercalation was intended to occur on the Julian cycle starting in 5 B.C.

This dating is part of a larger thesis, that certain special coinages of Augustan proconsuls are related to the proclamation as Augustus' heirs of his grandsons Gaius and Lucius, and that these events were in reaction to the split between Augustus and Tiberius in 6 B.C. Buxton and Hannah's thesis requires that all the proconsuls involved were in office in 6 BC or the immediately following years. Since Fabius was one of them, this requires that his proconsulate was in 7/6 or later.

In specific support of their proposal on the date of OGIS 458, Buxton and Hannah make the following circumstantial arguments:

Buxton and Hannah also make the following evidentiary arguments:

The last point highlights what I regard as the decisive argument against Buxton and Hannah's proposal: it does not track the Augustan reform. According to their analysis of the Peritios date, Fabius designed the Asian reform so that the proclaimed alignment would give correct Julian dates from the start. That is, by introducing quadrennial intercalation in 5 B.C., their proposed interpretation of the reform aligns the Asian calendar, not to the Roman calendar but to the true Julian calendar -- despite the fact that the decree explicitly equates the changeover date, 1 Dystros (new) = 15 Peritios (old) to the Roman civil date a.d. IX Kal. Feb., clearly showing that the Roman/Asian alignment was supposed to be correct from the start. On their model, the Roman and Julian calendars were three days apart on that date.

The Julian calendar would continue to slip against the Roman calendar until the latter started its own quadrennial leap year cycle. That means that the calendrical alignments given in the decree would not be correct against the Roman calendar for another 4 (or 8) years. In other words, Asia would be celebrating Augustus' birthday 1 (or 2, in their scheme) days late in the first four years of operation of the new calendar.

I do not believe that this is what Fabius intended to happen.

Website © Chris Bennett, 2001-2011 -- All rights reserved