and predicted on the same day, then obviously the month began on that day. If
the new moon was not sighted on either the 30th or the 31st (of the outgoing
month), e.g. because of persistent bad weather, then the beginning of the
month was based entirely on a prediction, and could be either the 30th or the
31st.^66 The problem arose only in cases of conflict between sighting and pre/
postdiction: typically, when the new moon was predicted (or postdicted) on
the 30th, but onlyfirst sighted on the 31st.^67 Conflicts of this kind are attested
in the Diaries, but inconsistently resolved: in some cases, the month begins
with the prediction on the 30th,^68 and in others, with thefirst sighting on the
31st.^69 This inconsistency is attested throughout the period of the Astronomi-
cal Diaries—sometimes even within the very same year^70 —and therefore does
not reflect any historical change. The inconsistency cannot be explained,
moreover, on the basis of a calendrical rule: the conflicts were not resolved,
for example, in such a way as to achieve an alternation of 29-day and 30-day
months (typical of schematic lunar calendars).^71
The inconsistent resolution, in the Diaries, of conflicts between predictions
and sightings can only be the reflection of informal,ad hocprocedures.Wedo
not know who had the authority of deciding when to begin the month. In the
neo-Assyrian period, it would have been the king: indeed the letters of
astrologers to the Assyrian kings, cited above, deal exactly with the problem
of the new moon that should have been seen on the 30th but because of bad
weather was only sighted on the 31st; according to these letters, it is the king
who is expected to decide when the month begins.^72 We cannot assume that
the situation remained unchanged until the end of the Astronomical Diaries
(^66) For instance, on 4 November 274BCEthe month began with a non-sighting on the 30th
(which, as it happens, was an early prediction); the moon was still not sighted on the second
evening, which is recorded in the Diary as‘very overcast’(Sachs and Hunger 1988–2006: i.
33867 – 9).
It is difficult tofind evidence of the reverse, i.e. conflicts between sighting on the 30th and
pre/postdiction on the 31st, and to establish how they would have been resolved. But this
scenario must have been very rare, because of the general tendency in the Astronomical Diaries
(as seen above) towards early predictions on the one hand, and late sightings on the other.
(^68) Even though the new moon was sighted on the next evening, as is explicitly reported in a
number of cases in the Diaries: see Stern (2008) 40–1 n. 36 (cases ranging from 322 to 137BCE).
(^69) This would seem to apply to months beginning with a late new moon sighting on the 31st
(listed ibid. 40 n. 30, with cases ranging from 303 to 87BCE): in these cases, presumably, the new
moon was correctly predicted for the 30th, but the prediction was ignored. 70
See ibid. 41 n. 37 (cases ranging from 329 to 87BCE).
(^71) Ibid. n. 38. In one case where the month began on the 30th of the outgoing month, 19 May
376 BCE, it is reported that the sky was misty and the moon not seen (i.e. in Babylon), but that it
was seen in Borsippa (Sachs and Hunger 1988–2006: i. 130–1, and comment p. 138). This
suggests, perhaps, that a problem regarding the beginning of the month was resolved on the basis
of a new moon sighting in another city (much in the same way as is attested in the neo-Assyrian
period: see above, near n. 31). This case, however, is exceptional in the Diaries and its significance
remains unclear.
(^72) See above, the letters of Bullut
:u and Adad-šumu-usur (note that these sources call‘29th’
and‘30th’what the Astronomical Diaries call‘30th’and‘31st’: see above, n. 28).
The Babylonian Calendar 91