Forsythe - Time in Roman religion : one thousand years of religious history

In citing ancient sources throughout the text and notes I generally follow those employed in the third edition of The Oxford Classical Dictionary , edited by Simon Hornblower and Antony Spawforth (Oxford University Press 1996). In the bibliography at the end of the book I employ the following abbreviations in citing journals.

One of our single most important sources of information on Roman religion and its temporal associations is the Roman calendar; and thanks to the survival of numerous epigraphic versions, it has received much modern scholarly attention. This calendar, however, was the one revised by Julius Caesar (hence, its name the Julian calendar), consisting of 365 days with an additional intercalary day inserted every four years. It was established as Romes new calendar as of January 1, 45 B.C. and continued to be the calendar of the Roman Empire and on through the Middle Ages until revised slightly by Pope Gregory XIII in 1582.

Julius Caesars reformed calendar replaced a lunisolar one of 355 days with an intercalary month of 22 or 23 days inserted every two years or so; and this lunisolar calendar, which modern scholars usually termed the pre-Julian calendar, had been the Roman states method of marking the passage of time during the previous four hundred years or so of the Republic. Nevertheless, ancient testimonia and the structure of the calendar itself suggest that the pre-Julian system was the third stage in an evolutionary process of early Roman time reckoning. During the last two centuries B.C. Roman antiquarians were of the opinion that Romulus, as part of his role as Romes first king and city founder, had established a calendar, but one of only ten months; and that his successor, Numa Pompilius, to whom the later ancient tradition ascribed virtually all of Romes religious institutions, reformed Romulus scheme by adding two months.

According to Macrobius ( Sat . I. 12.3) and Censorinus (20.3), Romulus ten-month calendar numbered exactly 304 days, and the two ancient writers further specify which of the months had 30, and which had 31 days. This overly precise scheme is certainly unhistorical and is simply the product of later antiquarian reconstruction. Alternatively, rather than viewing Romulus calendar as having consisted of ten months of roughly 30 days each to correspond to the lunar cycle, A. K. Michels (1949 330) as offered a different explanation, according to which the ten periods would not have been lunar cycles but periods of varying length and marked by different natural or astronomical phenomena that extended over the entire circuit of the year. A natural calendar of this sort would have been perfectly adequate for the early Roman farmer, and we possess an example of such a calendar in Hesiods Works and Days (11.383617). Similarly, in his own treatise on agriculture, Varro ( De Re Rust . I.2736) describes a farmers natural calendar consisting of eight periods. Thus, there are two plausible ways in which we can account for the Roman antiquarian tradition of an initial ten-month (or ten-period) calendar. Furthermore, as suggested by Hesiod and Varro, it could have been used by the early iron-age inhabitants of the site of Rome, scattered among several different hilltop villages, before they had coalesced into an organized community.

The second phase of the Roman calendar is revealed by the three dividing days of the months (kalends, nones, and ides) preserved in both the pre-Julian and Julian calendars: for these days, as the later Romans well understood, had originally marked the appearance of the new moon, the moons first quarter, and the full moon in a calendar that was based upon the lunar cycle. Roman adoption of a lunar calendar can be seen as part of the larger pattern of state formation, which, as the result of Phoenician and Greek influence, began to transform central Tyrrhenian Italy during the seventh century B.C. (see Drews 1981).

Thus, six of the Roman month-names (Quinctilis, Sextilis, September, October, November, and December) may hearken back to the earliest phase of calendrical reform in Rome.

Exactly when the pre-Julian calendar replaced the lunar one is unknown, but in attempting to date this innovation modern scholarly speculation has ranged from the sixth century (generally taken to be the Etruscan phase of Romes later regal period) to the late fourth century and coinciding with the curule aedileship of Cn. Flavius in 304 B.C.to the nones indicates that the Romans of c.400 B.C. were using the pre-Julian calendar that was divorced from the lunar cycle.

The organization of the pre-Julian calendar clearly shows that it was the product of experts who were guided throughout by certain basic principles. The Roman calendrical year originally began with March, as Ovid ( Fasti III.135366) makes abundantly clear, and as is also shown by the sequence of months QuinctilisDecember that take their names from the numbers 510. January and February were initially the eleventh and twelfth months of the year. Macrobius ( Sat. I .13.5) and Censorinus (20.4) inform us that the framers of the calendar regarded odd numbers as more auspicious than even ones, probably because the latter seemed inherently unstable since they are all divisible by two and can therefore be cut in half as it were. Consequently, all the months except February were assigned an odd number of days, roughly corresponding to the lunar cycle of twenty-nine and a half days: January, April, June, August (Sextilis), September, November, and December were given 29 days, whereas March, May, July (Quinctilis), and October were composed of 31 days. Thus, the pre-Julian calendar comprised 355 days, itself being an odd number. Februarys even number of days and terminal position in the calendar were designed to express the months overall character as a somber period devoted to the worship of the dead. In addition, the three dividing days of each month were also assigned to odd-numbered days: the kalends was, of course, on Day 1; the ides fell on Day 13 of the shorter months and on Day 15 of the longer ones; and the nones, which preceded the ides by nine days (counting inclusively as the Romans always did), occurred either on Day 5 or 7. In the epigraphic calendars, these days were always displayed in large letters and abbreviated as KAL, NON, and EID.