The Many Greek Calendars: Time-Reckoning in an Urban Network
Already in the second millennium BCE polities in the regions around the Aegean Sea had reached a complexity that made coordination in the form of an ordered sequence of named months necessary. These are attested for the Mycean period of the second half of this millennium. Such ‘calendars’ were far from rule-based systems that allowed for the fixing of meetings or shared activities by day months or even years in advance. As in many regions around the Mediterranean Sea, time-reckoning was based on lunar months (meis in Greek, from which the Latin mensis/es and modern English month/s developed). Usually the lack of moonlight or the very first appearance of a crescent (‘new moon’, noumênía in Greek ) were used to start counting the days of a month. Such an observation was (and is) far from easy and precise, if sight was limited by weather conditions or mountains. Hence, consensus about the start of a new month was a matter of local discussion and decision. Important events such as assemblies or festivals would be announced for an advanced phase of the month or tied to the full moon rather than the new moon. Arrogating the authority to define the beginning of a month must have been a contested issue and an expression of the location of supreme authority at a given moment in time.
An important event, if not the most important event, offered the name for the month. Thus, festival names, whether related to specific deities or to opening or concluding moments of agricultural activities, formed the basis for the months of the many different local Greek calendars. This is evident in many later data and we have to assume that it was true from early on, even if some of the festivals were not continued and survived only in the form of the name of a month. Festivals developed later might or might not change the name of their month. In Athens, the paramount festival of the Panathenaia never changed the name of the month Hekatombaion; perhaps the Athenians were reluctant to change the name of the very first month of their year. From Mycean documents the following names are securely interpreted as names of months: a-ka-[..]-ri-.-jo-jo me-no (KN = texts from Knossos = Oa 745), a-ma-ko-to me-no (KN Fp (1) 14), de-u-ki-jo-jo me-no (KN Fp (1) 1), di-wi-jo-jo me-no (KN Fp (1) 5), ka-ra-e-ri-jo(-jo) me-no (KN Fp (1) 6, 7, 15, 18, Gg 7369, M (1) 1645), ki-ri-ti-jo-jo (PY = texts from Pylos = Es 650), pa-ki-ja-ni-jo-jo me-no (PY Fr 1224), ra-pa-to me-no (KN Fp (1) 13), sa-pa-nu-wo me-no (KN X 999) and wo-de-wi-jo (-jo) me-no (KN Fp (1) 16,48, Ga 953+955, V (2) 280). We do not know how to exactly pronounce these names written in the early script of Linear B; it is not before the Greek poet Hesiod at the end of the eighth century BCE that we read a name of a month in alphabetic and thus phonetically reliable script: Lênaiôn (Works and days 504). In many instances, names of months were changed, above all to honour gods and kings of the Hellenistic period from the late fourth century BCE onwards or to demonstrate loyalties to Roman emperors from the very late first century BCE onwards. Now, occasionally, even calendars based on the names of the ‘twelve gods’ (always varying in composition) appeared.
In the short run, however, sequences of names of months were stable and allowed orientation and coordination of activities. In classical Athens for example the sequence was Hekatombaion, Metageitnion, Boëdromion, Pyanopsion, Maimaktêrion, Posideon, Gamêlion, Anthestêrion, Elaphêbolion, Mounychion, Thargêlion and Skirophorion. Such sequences were different from place to place in principle. And yet, shared traditions of gods and festivals, migrations and real or imagined shared ancestry made for traditions of names of months, even if the idea that the Greek calendar had been centrally introduced by the priests of the Oracle of Delphi has long been disproved. Thus, the around two hundred different names of Greek months known so far allow to identify three traditions of naming based in the Ionian (todays west coast of Turkey) and Attic region, North-West and Dorian Greece (Peloponessos) and in the central Thessalian and Boiotian area, Lesbos included.
The average length of the period from new moon to new moon, the so-called lunation, 29.5306 days, was captured by counting from start on up to 29 or 30 days. Days ‘10’ and ‘20’ were marked by more elaborate cult, in some places, one started to count down to the next month from day 20 onwards, maybe simply dropping the last day in the so-called ‘hollow months’ of 29 days. In order to allow for appointments by day even for the next months to follow, usually Greek cities and their hinterlands, the so-called poleis, used conventional instead of empirical months: Months of 29 and 30 days followed one another. Occasionally, if seen necessary by the authority, an additional day was ‘intercalated’. The necessity was as much the obvious difference with the visible phases of the moon as some political convenience to gain a day of a month for some business or to postpone an approaching date by a day (or more).
The same held true for an even more pressing problem, the coordination of the lunar months with the solar year, the course of the sun and its meteorological consequences, quite marked on the northern shores of the Mediterranean Sea. To relate the lunations to the average length of the solar year, 365.2422 days, is the classical problem faced by all societies developing a workable, long-term stable luni-solar calendar, allowing both, to use the lunar months as visible indications of the progress of time and to keep a stable relationship to the course of the sun in the long run. The Greek calendar never dropped one of these interests. Again, the easiest ways to deal with the problem, was empirical intercalation, now not of a day, but of a full month. The most convenient, even if not trivial, points of references were the solstices, easier to identify than the equinoxes. In Athens, the additional month was intercalated after the Posideon, that is to say, after the winter solstice. Solstices, too, were used to define the start of the year; the autumn equinox became popular as start of the year in Greek cities only during the Roman imperial period from the end of the first century BCE onwards. Now, the birthday of the emperor Augustus, the 23rd September in the Julian calendar, could be identified (and one may say cosmicised) as an astronomical and dynastic date, thus honouring the military and political overlord.
Greek astronomers, partly based on the long-term systematic observations of astronomical phenomena in Assyria, respectively Babylonia, were more and more aware that the relationship of the lunar and the solar year was a stable, even if complex one. Probably already by the end of the sixth or the beginning of the fifth century BCE a rule was established, to produce a relatively neat correspondence by adding three months of 30 days in the course of eight years of 12 lunar months each, for instance in year 3, 5 and 8 of that period (‘octaëteris’). By such a system the remaining difference of a surplus of one and a half day would have been a theoretical rather than a practical problem, given the dealing with the lengths of months.
Even more precise observations led to a system of 235 months in the course of 19 years, allowing for a higher congruity of the conventional and the real lunar months as much as of the 19 civil and solar years. 125 of these months would have been months of 30 days, 110 months of 29 days. Evidently, this system presupposed a stable use of conventional rather than empirical months, a major advance in terms of practicality. Following such a system observation of the month was replaced by following rules or rather tables. This so-called ‘Metonic cycle’ was named after the mathematician Meton, who was said by the later historian Diodorus to have started this reckoning from 27 June 432 BCE onwards, together with Meton’s pupil Euktemon. With further advances in mathematics and astronomy, for instance by Callipus, who reckoned a new type of cycles from 330 BCE onwards, the systems gained a precision that allowed the precise prediction of solar and lunar eclipses, otherwise seen as major and bad omens. And yet, these systems never made it into daily calendar pragmatics. Like the names of months, the practices of intercalation of days in the same manner as of full months, remained a local matter and were regulated by short term political or fiscal considerations rather than available astronomic knowledge. Calendars were used to ascertain political authority rather than a public status of mathematical expertise. Other experts, however, built on such knowledge. In 1900 archaeologists discovered a heap of filigrain mechanical parts in a shipwreck off the island of Antikythera. This ‘Antikythera mechanism’ proved to be a machine working on the base of the Metonic cycle. With the help of a differential gear and more than thirty cogwheels an external ring with the names of the month of the Greek lunar year is synchronised with the solar year. It allows to trace the apparent course of the sun through the zodiacus on an internal ring and to translate it into civic dates. Again, practical observation was replaced by a rule-based mechanism.
The Greek calendar lacked a clearly articulated weekly structure. The permanently repeated rhythm of seven days of the Jewish week remained unknown before the slow rise of the planetary week for astrological purposes during the Hellenistic period and the Roman reception of this week for the Christian empire of the fourth century CE and later. Likewise the recurrent Roman week of eight days, more or less restricted to the city of Rome, was never present. Nevertheless, the months were internally articulated in a sequence of days, in particular in the first half, dedicated to different deities. In Athens, the day after the appearance of the New moon, that is the second day of the month, was celebrated as the day of the Agathos daimon, the ‘good tutelary spirit’. The third day of each month was celebrated as the birthday of the goddess Athena, the sixth as birthday of Artemis and the seventh of Apollon. The fourth day had a special relationship with Herakles, Hermes, Aphrodite and Eros. In a similar manner, the eight day was related to Poseidon and Theseus, before the full moon witnessed the biggest festival of each month as mentioned before. For practical purposes market days tied to the monthly rhythm were carried out in many places of importance.
Whereas the complex rhythms of the mathematicians’ proposals for ever more perfect intercalary cycles were of course fixed in writing, the permanent recurring series of months, the easy internal structure of the month and the flexible practice of intercalation at the whimsy of political authorities did not need or even defied a written fixation. Graphical representations of the type developed in Rome (see this blog post for details) were unknown in Greece. The so-called ‘sacrificial calendars’ known for example from Athens were not tables of all days of the year, but lists of those days that featured ritual activities. The information about cults were different from place to place. The Athenian copies are exceptional in focusing on the regulation of financing and distributing the burden of costs for sacrifices among the sub-groups of the city, the ‘demes’. Indications of the persons, in particular the priests responsible for the actual performance of the rituals, were lacking from these lists. Nevertheless, such lists frequently contained complex information and regulations, which invited fixing in writing as much as public display in order to make people aware of their own and others’ duties.
Greeks were aware of their many different calendars. Colonies were asked to continue the calendars of their founding cities, in order to facilitate political and economic communication, the definition of periods of loans and interests for instance. Due to the growing territorial reach of the kingdoms of the Hellenistic period, lists featuring equations of different calendars were produced in order to allow translation from one calendar systems into the other, local into Macedonian calendars for instance. In the course of these processes even the luni-solar character of local calendars was given up in order to achieve an easy relationship with the Roman purely solar calendar with its year of 365 days and the intercalation of just one day every four years. Counting days only up to thirty (and talking of ‘thirty a second time’) enabled to keep up the appearance of a lunar calendar with its length of 29 and 30 days. In the widening urban network of the Imperial period, of the 1st century BCE and the following centuries, the empire did not abolish local urban calendars, but it made them translateable.
— Jörg Rüpke
Jörg Rüpke is a scholar of religion, focusing on the ancient world. He is the co-spokesperson of the KFG “Religion and Urbanity: Reciprocal Formations”.