Ketubah of Antinoopolis, Letter of Resh Galuta and Aramaic Tombstone Inscriptions from Zoar Or What Was the Original Molad Calendar of Hillel Bar Yehuda?
Main Article Content
Abstract
The paper describes the major reform of the Jewish calendar in 358/9 CE by Hillel Bar Yehuda, the historicity of which was recently questioned by Sacha Stern. This requires a separate assessment of the solar part (the intercalation cycle) and the lunar part (the timing of the mean lunisolar conjunctions) of the calendar that emerged after 358/9.
As for the solar part, evidence from several Christian sources from the late 4th century proves the intercalation cycle in the 360s differed from the modern one perhaps only in year 16. The dating of the 5th century Ketubah from the Egyptian city Antinoopolis implies that the difference disappeared by 417 CE.
New evidence from the Aramaic tombstone inscriptions from Zoar, a locality in Jordan, shows that year 9, not 8 as in the modern cycle, was intercalated from 372 to 467 or even 542. We conjecture that the original calendar of Hillel Bar Yehuda followed the cycle GUĤADZT counted from Tishri 311 BCE. We also conjecture that year 9, together with years 6 and 17, ceased to be intercalary during Emperor Justinian’s reign (527–565) though years 6 and 17 recovered intercalary status sometime after 823.
As for the lunar part, the letter of Resh Galuta from 835/6 implies that the calendar of Hillel Bar Yehuda was a Nisan-based Molad calendar. The Julian date for the Passover in 387, given by two Christian authors, implies the presence of the rule Molad Zaqen for Rosh Ĥodesh Nisan at 18 hours and suggests that Molad Nisan was at least 19.5 minutes later than the modern one associated with Molad BaHaRaD.
We tested this Nisan-based Molad calendar with the sequences of (30–29)-day months, the rules LO B-D-U PESAĤ and Molad Zaqen and two variable months, Ĥeshvan and Kislev, against available historical data: the dates on 18 Aramaic tombstone inscriptions from Zoar from 392–526 and two dates from Iggeret of Sherira Gaon—and found a perfect agreement except for two cases which imply the rule LO U PESAĤ was absent in the original calendar of Hillel Bar Yehuda and was introduced after 506, requiring a third variable month, the role played by Tevet as seen from the letter of Resh Galuta.
The letter of Resh Galuta implies that the Nisan-based calendar lasted for almost 500 years. We conjecture that 823–836 was the time of transition to the modern Tishri-based calendar. We also give the reasons why Hillel Bar Yehuda’s name was all but forgotten.
The reference epoch (the first Molad Nisan) of that calendar and the length of the calendar month will be established in a subsequent paper. This would allow us to pinpoint, with some exceptions, the Julian dates behind the Jewish dates from 360–836.