According to this interpretation, the rising of Sothis would have played a
critical, regulatory role in the lunar calendar. But the Ebers document does not
hint at any procedure of intercalation. This opens the possibility that unlike
Greek and Babylonian calendars, the synchronism of the lunar monthswp-
rnpt,th
%
y,etc. with the rising of Sothis was achieved through other procedural
means. The rule outlined above would have obviated, in fact, the need for any
formal procedure of intercalation. As explained above, the lunar months were
named in the Ebers document with reference to day 9 of the civil calendar that
occurred within them; but sometimes (once every two or three years), there
would have been a lunar month without any civil day 9 within it.^65 This
unnamable, interstitial month was effectively intercalary, though not as the
result of any deliberate procedure of intercalation, but only as the result of an
anomaly in the calendar’s rule.
How this unnamed month would have been referred to is unknown; it is not
explicitly featured in the Ebers calendar, and there is no evidence of a
‘nameless’lunar month in any Egyptian source.^66 The insertion of this extra
month may well have led to uncertainty about the names of subsequent
months, at least until the next rising of Sothis arrived. Uncertainty about
month-names as a result of informal intercalation is a well-attested phenome-
non in other non-modern cultures;^67 it is all the more plausible in ancient
Egypt where lunar month-names were probably not that important in
public life.^68 The sequence of Egyptian named lunar months, in spite of the
wp-rnpt. See further ibid. 75–6, 80–3, and n. 39, for a convincing refutation of other interpreta-
tions. In a later article, Depuydt (2008) argues thatwp-rnptis not a lunar month-name, but
rather a feast-day name (whereas the other names in the text, starting fromth
%
y, still designate
lunar months).
(^65) This would occur when a 29-day lunar month began on day 10 of one civil month and
finished on day 8 of the next: no lunar month-name would be available for it, since it did not
include day 9 of any civil month. An unallocated or unnamed lunar month of this kind was even
more likely to occur within the 35-day interval from IV Shemu to I Akhet.
(^66) It could have been given the next lunar month-name on the list, but this would only have
postponed the problem. Depuydt (1996b:69–70) suggests a formal system wherebywp-rnptwas
repeated either at the beginning or at the end of an intercalated year; but there is no evidence for
this in the Ebers document, and indeed, this system would run counter to the synchronism put
forward in this document. 67
See Turton and Ruggles (1978), on the calendrical practices of the Mursi of south-western
Ethiopia (also in Gell 1992: 300–5): the difficulty of reconciling the lunar calendar with the
seasonal year frequently leads them to disagreement or uncertainty regarding the current lunar
month; these ambiguities are only resolved retroactively at a much later stage. As a result, the
occasional addition of a thirteenth lunar month is alwaysex post facto; it is not formally,
deliberately, or even knowingly intercalated—much as I am arguing, in fact, in the context of
the Ebers calendar (see further Chapter 1, n. 20). A similar case might be the Dakota of North
America, whom Nilsson (1920) 241 cites as having often heated debates about which month it is
(but other examples ibid. 246 are, in my view, less clear). For a further example, see next n.
(^68) Similarly, Malinowski (1927) 209–15 points out that the frequent uncertainty of Trobriand
islanders regarding the current lunar month reflects the fact that their dominant calendar is not
their lunar but their seasonal one; see also Leach (1950).
148 Calendars in Antiquity