Although relatively abundant, the evidence for the Attic calendar is still patchy and often of contested interpretation. As it was obvious to them, no contemporary source sets out to describe the system as a whole. Further, during the period in question it underwent changes, not all perfectly understood. As such, any account given of it can only be a tentative reconstruction. Note that in this context the terms Athenian and Attic are largely interchangeable.
The divide between these neighbouring calendars perhaps reflected the traditional hostility between the two communities. Had the Boeotians been speakers of an Ionic dialect, like that spoken in Athens, there would have been overlap in the names of months. An example of this is the Ionian island of Delos, where the calendar shared four out of twelve month names with Athens, but not in the same places in the year. There, even though the island was under some degree of Athenian control from around 479 to 314 BC, the year started, as with the Boeotians, at midwinter.
The linking of the sun and the moon meant that the calendar was lunisolar. Twelve lunar months add up to about 354 days, eleven days or so shorter than the solar year. Under a purely lunar calendar, such as the Islamic one, the months creep backwards over the years with no relation between the months and the seasons. In Greece with its pronounced seasons this had to be prevented. By tying the start of their year to the solstice, the Athenians allowed the months to relate, with some elasticity, to the seasons.
This still left the problem that twelve lunar months fall eleven days short of the solar year. To make up for this, an extra month had to be inserted ("intercalated") about every third year, leading to a leap year of about 384 days. So normal years contained 12 lunar cycles and then when it was judged that the months had slid back enough, a year of 13 cycles was used to realign the lunar and solar years. This extra month was achieved by repeating an existing month. That is to say, the same month name was used twice in a row. Handbooks usually refer to the sixth month, Poseideon, as the month that was repeated, but months 1, 2, 6, 7 and 8 are all attested as being doubled. (Hannah 2005: 43)
Various cycles were in existence for working out exactly which years needed to take a thirteen month. A nineteen year cycle known as the Metonic cycle which was developed at Athens by the astronomers Meton and Euctemon (known to be active in 432 BC), could have been used to pattern the insertion of leap years so as to keep the lunar and solar years aligned with some accuracy. There is, however, no sign that any such system was in fact used at Athens, where the calendar seems to have been administered on an ad hoc basis.
The Athenian months were named after gods and festivals. In this the calendar differed from the Mesopotamian models that lie behind all Greek lunar calendars. In the Sumerian and Babylonian prototypes, for instance, the months were named after the main agricultural activity practised in that month. Many Athenian festivals did have links with different stages of the agricultural cycle, such as festivals of planting or harvest. This perhaps added to the need to keep lunar and solar calendars roughly aligned, though this was not always achieved. The year of farmers, however, was not the primary focus of the calendar.
Jane Ellen Harrison, in treating the Attic festivals in Prolegomena to the Study of Greek Religion (1903), noted at the outset that some, though not all of the festivals gave their name to the month in which they were celebrated, and that with the one exception of the Dionysia, none of the festivals were directly named for Olympians, or indeed, for any divinities (Harrison, p 30).
At Athens month six, Poseideon, took its name directly from the god Poseidon. More commonly, the god appears in the form of a cult title. (A cult title is the name or aspect under which a god was worshipped at a particular festival.) Examples are Maimakterion, named after Zeus ("the rager") and Metageitnion, after Apollo as helper of colonists.
Of all of the months, only the eighth, Anthesterion, was named directly after the major festival celebrated in its month, the Anthesteria. While the month-naming festivals of Pyanepsia, Thargelia and Skira were relatively important, some of the grandest celebrations in the life of the city are not recognised in the name of the month. Examples are the Great Dionysia held in Elaphebolion (month 9) and the Panathenaia are only indirectly recognised in Hekatombaion (month 1), named after the hekatombe, the sacrifice of a "hundred oxen" held on the final night of the Panathenaia. More often than not, the festival providing the month name is minor or obsolete. For instance, the second month, Metageitnion, is named after a cult title of the god Apollo, but there is no trace of a festival bearing the name. The same goes for months 5 and 6, Maimakterion and Poseideon.
The calendars of the Ionian cities of Asia Minor (along the western coastline modern Turkey) often share month names with Athens. For instance, at Miletos four of the same month names were in use, namely Thargelion, Metageitnion, Boedromion and Pyanepsion, and the last of these even occupied the same position as month four in both communities. Traditionally, these Ionian cities were founded by colonists from Attica (perhaps around 1050 BC). It may be then that the Athenian month names refer to a festival schedule some hundreds of years out of date.
Athenian festivals were divided between the 80 or so annually recurring celebrations and a set of monthly holy days clustered around the beginning of each month. These were often the birthdays of gods, the Greeks thinking of birthdays as a monthly rather than a yearly recurrence. Every month days 1-4 and 5-8 were all sacred to particular gods or divine entities, amounting to some 60 days a year:
Monthly and the annual festivals were not usually allowed to fall on the same days. This means that every festival month had an opening phase with exactly recurrent practices and celebrations, while in the body of each month there was a unique schedule of festival days.
A parallel function of this calendar was the positioning of the perhaps 15 or so forbidden days on which business should not be transacted.
Each month was divided into three phases of ten days associated with the waxing moon, the full moon and the waning moon. The naming of the days was complex. The first day of the month was simply noumenia or new moon, a name used in virtually every Greek calendar. From there the days were numbered up to the 20th day. For the final third of the month the numbering turned around to do a countdown from ten to the last day. Only the middle phase had numbers for the days running higher than 10 and even these were often phrased as "the third over ten" and so forth. In the wings of the month, the numbered days ran 2-10 and then 10-2. Days in these sections were distinguished from each other by adding the participle "rising" and "waning" to the month name. In the centre of the month with its unambiguous numbering there was no need for this, though later the term "of the middling month" was used. The final day of the month was called hena kai nea, "the old and the new". Peculiar to Athens, this name presents the day as bridging the two moons or months. Elsewhere in Greece this day was usually called the thirtieth.
Rather than thinking of the month as a simple duration of thirty days, the three-part numbering scheme focuses on the moon itself. In particular the waning days 10-2 and the waxing days 2-10 frame the crucial moment where the moon vanishes and then reappears.
A date under this scheme might be "the third (day) of Thargelion waning," meaning day 28 of the month Thargelion.
|Moon waxing||Moon full||Moon waning|
|New Moon||11th||later 10th|
|2nd rising||12th||9th waning|
|3rd rising||13th||8th waning|
|4th rising||14th||7th waning|
|5th rising||15th||6th waning|
|6th rising||16th||5th waning|
|7th rising||17th||4th waning|
|8th rising||18th||3rd waning|
|9th rising||19th||2nd waning|
|10th rising||earlier 10th||Old and New|
To summarise the days with special names.
This strange juxtapositioning of the two days called the tenth, the earlier and the later, further highlighted the shift into the moon's waning phase.
When a month was to last 29 instead of 30 days (a 'hollow' month), the last day of the month ("the old and new") was pulled back by one day. That is to say, the "2nd day of the waning month" (day 29 in straight sequence) was renamed as month's end.
This decimal ordering extended to the creation of a supplementary calendar with ten months. Each year each tribe contributed 50 members to the council of 500 (boule) that played an important role in the administration of the city. For one tenth of the year each tribal fifty was on duty, with a third of them in the council chamber at all times as an executive committee for the state. Their period of office was known as a 'prytany' or state month.
In the 5th century this calendar was sun-based using a year of 365 or 366 days and paying no attention at all to the phases of the moon. One likely arrangement is that the ten prytanies were divided between six months of 37 days followed by four months of 36 days. This would be parallel to the arrangement in the 4th century, given below.
From several synchronised datings that survive it is evident that the political and the festival years did not have to begin or end on the same days. The political new year is attested 15 days out either way from the start of the festival year. This system is known from the 420s; whether it had been in place from the beginning of the ten month system is not clear.
However in 407 BC the two calendars were synchronised to start and end on the same days. Hereafter as described in the 4th century Constitution of the Athenians the civic year was arranged as follows:
In years where an extra month was intercalated into the festival calendar, the political months were probably lengthened to 39 and 38 days, a method that would have maintained the balance between the tribes. Evidence, however, is lacking.
These political months had no name, but were numbered and given in conjunction with the name of the presiding tribe (which, as determined by lot at the expiry of their predecessors's term, gave no clue as to the time of year). The days too were numbered and here with a straightforward sequence, running from 1 to the total number of days for that month.
One of the main roles of the civic calendar was to position the four assembly meetings that were to be held each prytany. Where possible, assembly meetings were not held on festival days, including the monthly festival days clustered at the start of each month. As a result, the meetings were bunched slightly toward the end of the month and made to dodge especially the larger festivals.
A date under this calendar might run "the 33rd day in the 3th prytany, that of the tribe Erechtheis." This is the style used in Athenian state documents (surviving only as inscriptions). Sometimes, however, a dating in terms of the festival calendar is added as well.
There is clear evidence that this was practised later. In Athens in 271 BC just before the Great Dionysia four days were inserted between Elaphebolion 9 and 10, putting the calendar on hold. Presumably this was to gain extra rehearsal time for the festival with its performances of tragedy and comedy. A similar story comes from the 5th century, but at Argos: the Argives, launching a punitive expedition in the shadow of the holy month of Karneios when fighting was banned, decided to freeze the calendar to get some extra days of war in. However their allies rejected this rearrangement and went home.
Aristophanes' Clouds, a comedy from 423 BC, contains a speech where a complaint is brought from the moon: the Athenians have been playing round with the months, "running them up and down" so that human activity and the divine order are completely out of kilter. "When you should be holding sacrifices, instead you are torturing and judging. A situation is known to have applied in the 2nd century BC where the festival calendar was so out of sync with the actual cycles of the moon that the lunisolar date was sometimes given under two headings, one "according to the god", meaning apparently the moon, and the other "according to the archon", meaning the festival calendar itself.
By contrast, the Attic calendar had little interest in ordering the sequence of years. As in most Greek cities, the name of one of the yearly magistrates, at Athens known as the eponymous archon, was used to identify the year in relation to others. That is to say, the sequence of years was matched to a list of names that could be consulted. Instead of citing a numbered year, one could locate a year in time by saying that some event occurred "when X. was archon." This did allow the years to be ordered back in time for a number of generations into the past, but there was no way of dating forward beyond ordinary human reckoning (as in expressions such as "Ten years from now").
There was for instance no use of a century divided into decades. A four year cycle was important which must have helped structure a sense of the passing years: at Athens the festival of the Panathenaia was celebrated on a grander scale every fourth year as the Great Panathenaia. But this was not used as the basis of a dating system.
As both narrowly local and cyclical in focus then, the calendar did not provide a means for dating events in a panhellenically (internationally) comprehensible way. A dating system using the four-yearly Olympiads was devised by the Greek Sicilian historian Timaeus (born c.350 BC) as a tool for the historical research, but it was probably never important on a local level.
The seasons were not viewed by Greeks as dividing the year into four even blocks, but rather spring and autumn were shorter tail sections of the overarching seasons, Summer and Winter. These divisions could formalised by using star risings or settings in relation to the equinoxes: so for instance, winter is defined in one medical text as the period between the setting of the Pleiades and the spring.
The older tradition as seen in Hesiod's Works and Days was extended by astronomical research to the creation of star calendars known as parapegmas. These were stone or wooden tablets listing a sequence of astronomical events, each with a peg hole beside it. Lines of bare peg holes were used to count the 'empty days' between what were taken as the significant celestial events. Often set up in town squares (agora), these tablets put the progression of the year on public display.
This system will have been fundamental to an individual's sense of the advancing year, but it barely intersected with the festival or state calendars. These were more civic in character and required managing to maintain their coherence with the year of the seasons. The seasonal and sidereal calendar, on the other hand, was immune to interference. So, Thucydides can date by the rising of the star Arcturus without having to wade into the confusion of disconnected city-state calendars.