others, the nominal intention—but in practice no effort was made to enforce it
precisely. Statements associating the conjunction with the last day of the
month and the new moon with thefirst day of the next month—which, as
mentioned above, are arguably astronomically incompatible—represent
Graeco-Roman rationalizations or simplifications of a calendar that was
actually somewhatfluid and undefined.^16
Political controlWhat precisely defined the beginning of the new month was thus not the
appearance of the new moon—with respect to which the month may have
been somewhatflexible—but rather political decisions. It is evident from our
sources that in ancient Greece, the calendar was controlled by political autho-
rities: in every city, it is the local magistrates (the eponymous archon at
Athens, ephors at Sparta, etc.) who decided when the months began and
whether the year should be intercalated. This is why the Athenian calendar
(which we also call‘festival calendar’) is commonly designated, in epigraphic
sources, as‘according to the archon’; hence its alternative modern designation,
‘archontic’.^17 Political authority established and maintained the calendrical
consensus; the process of calendar reckoning was thus a fundamentally politi-
cal activity.^18
The calendar at the city of Abdera appears, in this context, as an exception
that proved the rule. Stratonicus (early fourth centuryBCE) is said to have seen,
(^16) Various attempts have been made to establish the precise beginning of the Greek month on
the basis of specifically dated events, e.g. the date of Alexander’s death (e.g. Grzybek 1990: 15–6,
29 – 35, 52–60), although the inference remains speculative and inconclusive (A. Jones 1997: 164).
Depuydt (1996a) infers from the date of Meton’s observation of the summer solstice on 13
Skirophorion in 432BCE, which is given in some sources as equivalent to Egyptian 21 Phamenoth,
that the Athenian month began on the day of conjunction (see also Bowen and Goldstein 1988:
64; Hannah 2005: 54–7). But even if this is correct for that particular month, it does not
necessarily reflect general practice, which may have been variable and, as I argue, undefined.
Moreover, Depuydt does not take into the account the possibility of suppression of days (on
which see presently below); the suppression of one or two days before Skirophorion in 432BCE
could easily have caused this month to commence on the day of conjunction. Furthermore,
Depuydt’s argument is based on the assumption (following Bowen and Goldstein; see also
Hannah 2005: 57–8) that the date of Meton’s observation was given according to the Athenian
civil calendar; however, since the sources of this date are not earlier than thefirst centuryBCE
(Diodorus 12. 36 and the Miletusparapegma), it is equally possible that it was retrospectively
calculated according to a schematic, astronomical calendar attributed to Meton and based on the
day of conjunction (even though there is no evidence that such a calendar was instituted as early
as Meton’s time: see}3 and n. 88 below, Bickerman 1968: 100–1 n. 35, and in more detail Samuel
1972: 44–6, A. Jones 2000a: 150–1).
(^17) Evidence for the role of the ephors at Sparta is in Plutarch,Agis and Cleomenes, 16 (Samuel
1972: 93 18 – 4).
As rightly sensed, though from a more narrow perspective, by Loraux (2002) 171–90.
Calendars of AncientGreece 29