over in astronomical writings as a schematic, ideal year, and continued being
used in this context until the Seleucid period.^57
Astronomical sourcesSources such as the Astronomical Diaries are far more informative about the
calendar than administrative and economic documents, not least because their
dates can be astronomically identified with relatively high precision, and thus
converted if necessary to equivalent Julian dates. This enables us to determine
the exact relationship between the new moon and the beginning of the
Babylonian month. The dates in the Astronomical Diaries can be treated as
reliable, as they were meticulously redacted and would not have been prone to
error. For reasons explained above (near nn. 9–10), the calendar employed in
the Astronomical Diaries is highly unlikely to have differed from that which
was generally used in public life. Moreover, as we shall soon see, the informa-
tion in the Diaries is relevant, if only implicitly, to how in practice the calendar
month would have beenfixed.
The Astronomical Diaries reveal that in the sixth–first centuriesBCE, the
Babylonian calendar had evolved, to a certain extent at least, from new moon
sighting to new moon prediction. For in contrast to the neo-Assyrian period,
when as we have seen the month was almost entirely based on new moon
sightings, the Diaries often have the month beginning on an evening when the
new moon was not sighted but only predicted. Although, as mentioned above
(near n. 15), the Diaries are not explicit about how the beginning of the month
was set, they do provide, as a standard entry, the time-lag between sunset and
moonset on thefirst evening of the month. A new moon sighting is implicit
when the sunset–moonset lag is reported to have been observed (with phrases
such as‘I have watched’or‘measured’), or when the appearance of the moon
crescent is explicitly described as faint, bright, or whatever else. But in many
cases the moon is reported not to have been seen (‘I have not watched’or‘not
measured’), often because of‘clouds’or‘mist’; in these cases it may be assumed
that the sunset–moonset lag was predicted (or possibly sometimes postdicted)
(^57) Brack-Bernsen 2007. The 360-day year is attested in the form of an annual calendar in late
second-millennium astronomical works such as MUL.APIN and Enūma Anu Enlil, tablet XIV
(Neugebauer loc. cit.; Hunger and Pingree 1989; Brown 2000: 106–21, 249; Brack-Bernsen 2007:
93 – 8; Britton 2007: 117–19). This calendar, however, was only intended as theoretical or ideal;
this is evident e.g. in that dates are only given in multiples of 5 days (e.g. equinoxes and solstices
are given the standard but inaccurate date of the 15th of months 1, 4, 7, and 10). The ideal year is
generally believed to have served practical astronomical purposes as a model (‘grid’) or approxi-
mate calendar from which the actual dates of astronomical events could be inferred (Brack-
Bernsen loc. cit.); Brown (loc. cit.) favours a primarily divinatory or astrological purpose, because
of its gross imprecision in relation to the lunar calendar.
88 Calendars in Antiquity