in the early Achaemenid period), these decisions were based more than ever
on the astronomical expertise that would have been offered by the astrologers.
Administrative and economic dated documents are also useful as evidence
for the Babylonian calendar in this period. In the next few pages, however,
I shall argue that their datings are more likely to reflect scribal conventions
than how the calendar was actually reckoned.
Administrative and economic documentsThe information that administrative and economic documents are able to
provide about the Babylonian month is limited.^44 Unlike astronomical texts,
these documents cannot be astronomically dated; their Babylonian dates
cannot be converted, therefore, into precise Julian calendar equivalents, and
their relationship to the new moon cannot be precisely determined. These
sources can only tell us that certain months were 30 days long, by providing
dates on the 30th of the month.
Evidence of 30-day months in administrative and economic sources was
used, however, by Peter Huber to determine the accuracy of the Babylonian
month in relation to the new moon. Modern astronomy enabled him to
establish whether any given month should have been 30 days long had it
been accurately based onfirst visibility of the new moon. He concluded that as
much as 33% of the Babylonian months attested in the administrative and
economic documents deviated from the new moon, and began either one day
too early or one day too late. This considerable inaccuracy contrasted, how-
ever, with his analysis of astronomical sources, according to which only
between 5% and 5.5% of Babylonian months began one day earlier or later
thanfirst visibility of the new moon.^45
The contrast between the evidence of administrative/economic and astro-
nomical sources—which Huber did not attempt to explain—is problematic.
As argued above (near n. 10), it is unlikely that different calendars were used in
administrative/economic documents and in the Astronomical Diaries. The
contrast cannot be explained as reflecting different historical periods: for
although the administrative/economic sources that Huber studied are gener-
ally earlier than his astronomical sources (the former dating mainly from the
mid-seventh to mid-fifth centuries, and the latter from thefifth to second
centuriesBCE), both are reasonably well represented in all periods.^46 The
(^44) The terms‘administrative’and‘economic’are not clearly distinguished and can often be
used for the same document.
(^45) Huber (1982) 25–9. A different analysis of the astronomical sources yields the higher
result of approximately 8% (see below).
(^46) See ibid. 26, 51–5. The last economic text in Huber’s list is in fact quite late: an accounting
tablet dated 30 Aiaru 219SE(= 93BCE), of which the month would have been one day off
84 Calendars in Antiquity