The transmission of Torah

Zwinglius' "Carnival" roundup included a pointer to an upcoming Hila Dayfani paper. And now it's on academia.edu: The Scope of the Transmission of the Pentateuch in the Second Temple Period.

This deals with scribal errors. This 2TP precedes the Masoretic checksums and synagogues which froze the text of Genesis in time for Qumran and other texts in time for Masada and Wadi Murabba'at (and Peshitta, and Vulgate, and . . .). If you know Imperial Aramaic or the postBiblical Hebrew which its script often transcribes, you know you need glasses and a kickstart of coffee. Earlier generations didn't much have access to either. Telling one set of blocky characters from others is... well, it's like Arabic.

The most complete texts we got of earliest Hebrew are the Torah of the Masoretes (MT) and the Torah of the Samaritans ("SPentateuch" for her but I'm going with ST here). Famously, they differ. Some differences involve a text which harmonised Torah books with one another; these harmonies are first seen in Qumran so not (yet) Samiri. Other differences boost the Samarian temple which was on Gerizim up north, near Shechem which the locals now call Neapolis. But others look like honest differences in opinion on what the consonants looked like. That is: somebody goofed.

Dayfani doesn't care who goofed. She's not considering any texts outside MT / ST - so, no Jubilees or Peshitta or 4QDt (pdf). She's looking at the distribution of MT / ST goofs as to suss out when those two goofers quit consulting one another. She identifies 121 instances of consonantal dissimilarity.

First, this points to an era in Samaritan piety when they were still using the Imperial Aramaic script rather than their own script, which was truly Hebrew. But we kind of already guessed at this period, given the use of true Hebrew in some Qumran scraps and the existence of proto-ST harmonies there. Still, nice to have the confirmation.

More on-point is that archaic scripts differed over time to make graphic blunders different over time. Take Y and W, where they look more similar in Herodian times. This includes the ST write-o hkw [sic] at Genesis 27:36; correctly hky in MT and indeed for both 29:15. By contrast - I didn't know this - the scripts for B and M look wholly different today and in Herodian times, but under the last Ptolemies they could be mutually confused. Orthographic evolution for 200-0 BC is where Qumran excels.

Sometimes a difficult reading encouraged a tendentious rewrite. For a D/R interchange: Exodus 23:17 / 34:23, where the h'dn "face" of God is effaced in ST h'rwn. (D/R gets ridiculous still later, in Syriac.) Dayfani calls these "complex".

The vast majority of interchanges are Y/W, mostly "complex"; followed by D/R, mostly noncomplex. Dayfani concludes: The MT and SP thus appear to evince limited scribal activity from the third century BCE. This increased after the middle of the second century BCE, reaching its peak in the middle of the first century BCE.

Alternatively: the Torah community crossed sectarian bounds, such that most Torahs were similar - until the middle of the second century BC. The Maccabean / Hasmonean rebellion, then, posed a break. With the Jerusalem Temple in Hasmonean hands, under the Samaritan Temple roof formed a separate scribal shelter.