Intro page | How to Read the Tables | The Roman Calendars | Sources | Analysis
Conversion table: (Excel) (HTML) (CSV) Fasti consulares: (Excel) (HTML)
This page provides a brief overview of the major sources used to prepare the conversion table.
The most important single source for Roman calendrics is Macrobius, Saturnalia. Several chapters of this work discuss the evolution of the republican and Julian calendar in great detail. Macrobius wrote in the fifth century AD, but he drew on a large number of earlier sources, mostly now lost, most of which he names. Many of these were directly contemporary with the usage he describes, and were written by people who, by virtue of their position and expertise, were speaking with great authority. Not all of Macrobius' statements are reliable. His description of the management of intercalation in the republican calendar is confused, and his description of the mismanagement of the Julian reform does not appear to be completely correct. However, Macrobius remains central to any study of the Roman calendar.
Several authors provide additional information, most notably Censorinus. Additionally, surviving contemporary calendars -- fasti -- provide much additional detail. The most important of these are the Fasti Maffeiani and the Fasti Antiates Maiores. The Fasti Maffeiani, when first discovered in the 16th century, was an almost complete Julian calendar from the Augustan epoch. Although many fragments have been lost, it can still be studied through reproductions made in the 16th century. The Fasti Antiates Maiores, discovered in the 1920s, are the only pre-Julian fasti yet known, dating from the 60s B.C. Although they are quite fragmentary, enough survives that the structure of the original fasti can be recovered, allowing us to supplement and correct Macrobius.
The primary modern study used here is A. K. Michels, The Calendar of the Roman Republic (Princeton, 1967). I strongly recommend this book to anyone interested in calendrics. It is well-researched, highly knowledgeable, full of useful detail, well written, and sane. Additional papers on particular topics have been addressed as appropriate in the analysis section.
Calendrical data is not sufficient to recover Julian equivalents of Roman dates before the Augustan reform of 8 B.C. took full effect. This is because the occurrence and lengths of intercalation were not finally fixed until that time. In order to recover them for the purposes of constructing conversion tables, it is necessary to study Roman chronology in some depth.
The fasti consulares allows a rough equivalence of Julian year and Roman years to be established; the source issues for these are discussed here. In order to establish an exact Julian date, however, it is necessary to establish more exact synchronisms between Roman and Julian dates. This problem has been studied in depth at intervals since the 16th century, most intensively in the late 19th century. The main guide for these pages, considered here -- surely not quite accurately as the "standard" view -- is P. Brind'Amour, Le calendrier romain: recherches chronologiques (Ottawa, 1983). Additional studies of individual periods have also been used, though I have focussed on 20th century studies rather than the 19th century ones. In several periods, however, I have drawn my own conclusions which I believe go some distance beyond previous studies.
The sources are different for different periods.
The Early Imperial Period (A.D. 8-60): The are two concerns in this period. The first is to confim that the Julian calendar in use was essentially the Julian calendar we use. The literary data used to do this is taken from an eclipse notice given by Suetonius combined with Dio Cassius, and a leap year notice in Ammianus Marcellinus. Additionally, there is contemporary data from a Pompeian graffito (CIL IV 4182), an Egyptian papyrus (pLond. 130) and a temple dedication (CIL VIII 6979) that confirms date alignment and intercalary phase lock.
The second concern is the nundinal cycle, which is important for Republican chronology. The main facts here are given by the same Pompeian graffito (CIL IV 4182) and an obscure notice in Dio Cassius. This issue has not received much study: G. Tibiletti, RSA 6/7 (1976/7) 27, and a brief but faulty discussion in Brind'Amour. The discussion here proposes a new solution to this problem.
Early Julian years (45 B.C.-A.D.8): The main ancient literary source is Macrobius, with some critical additional data in Dio Cassius. There has been general, though not quite universal, agreement on the seqence of early Julian leap years since Scaliger's work in 1583.
In the last century a number of contemporary sources have emerged from the period. By far the most important of these is also the most recently published (1999): pOxy 61.4175, an Egyptian astronomical papyrus. This gives lunar and planetary ephemeris data against Roman and Egyptian dates in late 24 B.C. and shows conclusively that Scaliger's model is wrong. The corrected model is given here.
The late Republic (86 - 46 B.C.): The main literary source for the period c. 60-45 B.C. is also contemporary: the letters and speeches of Cicero, with some additional material from Caesar. Dio Cassius, although writing two centuries later, is another key source, and Censorinus and Macrobius are also important. This period was probably the most intensively studied in the 19th century. Brind'Amour gives the best modern reconstruction, a key point of which was subsequently confirmed by a contemporary document. His relative chronology is accepted here, though it must be shifted by two days in light of the corrected model for the early Julian calendar.
The main sources for period 86-58 B.C. are two data items given by Macrobius and Plutarch, for 77 and 86 B.C. respectively, and the inscription CIL I2 2511, giving a lunar synchronism in 67 B.C. These items, combined with the inferred regulatory principles of the Lex Acilia, allow the precise chronology for these three decades to be accurately recovered. Brind'Amour discussed CIL I2 2511, but his conclusions require revision in light of the revised date for the start of the Julian calendar adopted here, and in light of a corrected reading of the date this inscription. The data presented by Macrobius and Plutarch has not previously been adequately addressed for its chronological value. Hence the reconstruction given here for this period is entirely new.
The mid Republic (167 - 87 B.C.): This is the period most lacking in contemporary synchronistic data. For chronological purposes, it is virtually unstudied in the sources I have looked at for this reason. However, we have an absolute synchronism, long known from Livy, which fixes the chronology for 168 B.C. which allows us to limit the number of possible intercalations of the period.
There are two Greek calendrical synchronisms in this period which have not previously been studied for their implications for Roman calendrics. It turns out that the first of these, SIG3 674, is of considerable value, allowing us to reconstruct the principles on which intercalation was regulated under the Lex Acilia, and to fix the total number of intercalations and the distribution of 22:23 day intercalations in this period. This allows a considerable fraction of the period from 191 to 58 B.C. to be reconstructed precisely, and also enhances the value of any future calendrical synchronisms.
The reconstruction given here for this period is entirely new.
The early second century (191-168 B.C.): The primary source for this period is Livy, who gives two exact eclipse synchronisms in 190 and 168 B.C. These synchronisms are backed up by contemporary inscriptions from Greece which confirm that the rough Julian chronology implied by these synchronisms for sequence of events given by Livy, which is in turn largely derived from lost books of Polybius, is essentially correct.
Livian and Polybian chronology is a highly specialised cottage industry based on close analysis of the narrative. I do not pretend to have mastered it. In this section, I have in part used Brind'Amour, but have more closely been guided by several chronological papers on Livy written by V. M. Warrior, notably AJAH 6 (1981) 1, Chiron 18 (1988) 325, Latomus 50 (1991) 80, and in C. Deroux (ed.) Studies in Latin Literature and Roman History VI (Brussels, 1992) 119. However, I do not accept all her conclusions. P. S. Derow, Phoenix 27 (1973) 345 and P. Marchetti, BCH 100 (1976) 401 have useful observations although both papers are fundamentally flawed for different reasons.
This section presents a new analytical technique which is purely calendrical in nature. It is based on interactions between the dates Livy gives for Roman elections in this period and the 8-day nundinal cycle. This technique has been made possible by the adjustments to early Julian chronology necessitated by pOxy 61.4175, and has doubled the number of years whose exact chronology can be fixed.
Additionally, the regulatory principles of the Lex Acilia, which can be deduced from the data in this section and the previous section, allows the number of years whose exact chronology can be fixed to be increased still further.
The late third century (203-192 B.C.): Again, the primary source is Livy, with additional data from Polybius and some from Dio Cassius. From this point on, we are entirely reliant on literary sources.
For analysis of this decade I have primarily used Brind'Amour, with additional useful survey material in N. Prack, Der römischer Kalender (264-168 v. Chr.): Verlauf und Synchronisation (Sinzheim, 1996). J. Briscoe, A Commentary on Livy Books XXXIV-XXXVII (Oxford, 1981). P. Marchetti, AC 42 (1973) 473 and P. S. Derow, Phoenix 30 (1976) 265 also discuss this period, though each discussion contains serious problems in my view.
In this section I have introduced the analytical technique of estimating the number of intercalations since the last fixed point, 190 B.C., rather than estimating the number of intercalations beteen consecutive synchronisms. This variant is less prone to propagating errors and also allows one to fix the error bounds on the Roman side of the synchronistic equations.
The Second Punic War (218-203 B.C.): The primary sources are Livy and Polybius. The modern sources used are primarily Brind'Amour and Prack. P. S. Derow, Historia 30 (1976) 265 also discuss this period with some useful comments, although the paper's approach is fundamentally flawed in my view.
For analytic purposes, this section is based on the technique of baselining intercalary estimates on 190 B.C. introduced in the previous section.
The mid third century (262-219 B.C.): This is the period of the First Punic War (264-242 B.C.), for which we have a lot of chronological data, and the interval between the Punic wars, for which we have much less. Our primary source is Polybius, who in turn was drawing on now-lost Roman annalistic historians. Additionally, Dio Cassius supplies some key material, and an important datum is known only through the fourth century Livian epitomiser Eutropius.
Brind'Amour is not strong for this period. I have relied principally on a paper by M. G. Morgan, Chiron 7 (1977) 89, with some additional material from Prack's survey.
For analytic purposes, the technique of baselining intercalary estimates on 190 B.C. is even more critical to this period, in part because of the growth in the error limits to estimates, and in part because almost all the synchronisms we can derive are relative to the start of the consular year. It is disputed whether this started on Kal. Mai. or Id. Mart. The key evidence is given by the Fasti triumphales. This indicates that a number of consuls celebrated triumphs between Id. Mart. and Id. Apr., which would be after the end of their term of office if the consular year ended on prid. Id. Mart.
The error bounds on seasonal synchronisms baselined on the eclipse of 190 B.C. exceed the size of an intercalation around the start of the First Punic War. It is not possible to use this technique to estimate intercalations before this date. Before 300 B.C. even the accuracy of the consular fasti becomes questionable, making it difficult to determine the Julian year of an event. In the absence of other astronomical, synchronistic or calendrical crosschecks -- the only possibility curently available is the eclipse of Ennius -- it is pointless to try to estimate exact conversions any earlier.
Website © Chris Bennett, 2001-2011 -- All rights reserved