Toward a Standard Text
By the end of the second century of the Common Era, the rabbinic movement faced a problem that the priests of the destroyed Temple had never had to confront in quite the same form. The scrolls of the Torah, the…
Biblical Narrative
By the end of the second century of the Common Era, the rabbinic movement faced a problem that the priests of the destroyed Temple had never had to confront in quite the same form. The scrolls of the Torah, the Prophets, and the Writings circulated in many hands, in many copies, in many minor variants. With the Sanhedrin reconstituted at Yavneh and then in the Galilee, and with Rabbi Judah the Prince compiling the Mishnah at Beit She'arim and Tzippori around 200 CE, a parallel project of textual stabilization was quietly under way: the fixing of a single authoritative consonantal text of Scripture.
The rabbinic literature preserves memories of this effort. The Talmud (Yerushalmi Ta'anit 4:2; Sifrei Devarim 356) recalls three Torah scrolls kept in the Temple court that were collated against one another, with the majority reading adopted in each disputed place — a procedure of reconciliation by triangulation. The professional scribes of the period, the soferim, were credited with eighteen tikkunei soferim (scribal corrections) and with itturei soferim (scribal omissions). They counted letters: the middle letter of the Torah, the middle word, the middle verse — a discipline that turned every scroll into its own auditor.
What emerged was the proto-Masoretic text: the consonantal skeleton on which, seven centuries later, the families of Ben Asher and Ben Naftali in Tiberias would hang the elaborate vowel points, cantillation marks, and marginal notes that became the Masorah proper. The vowels were still unwritten. The reading tradition was still oral. But the consonants were now fixed — letter by letter, line by line — with a rigor that no other ancient text in the Mediterranean world ever received.
The cost of this stabilization was the silencing of alternative streams. The Septuagint's Hebrew Vorlage, the Samaritan Pentateuch, and the wildly varied scrolls from Qumran represent the textual diversity of the late Second Temple period. After 200 CE, that diversity narrows almost overnight. The biblical scrolls recovered from Wadi Murabba'at and Nahal Hever, copied during and just after the Bar Kokhba revolt, already match the medieval Masoretic consonants letter for letter.
Make a fence around the Torah.Pirkei Avot 1:1
Archaeology · History · Genetics
Before the discoveries at the Dead Sea, the oldest complete Hebrew Bible manuscripts known to scholarship were the Aleppo Codex (c. 930 CE) and the Leningrad Codex (1008 CE), separated from the world of Rabbi Judah the Prince by some eight hundred years of unbroken transmission. The skeptical assumption — that such a long copying chain must have introduced massive corruption — was reasonable. It was also wrong. The biblical fragments from Wadi Murabba'at and Nahal Hever, dated paleographically to c. 100–135 CE, demonstrated the proto-Masoretic text already in essentially its medieval form.
Emanuel Tov, in his Textual Criticism of the Hebrew Bible (3rd ed., 2012), classifies the Qumran biblical scrolls into four groups: proto-Masoretic, proto-Samaritan, those reflecting the Septuagint's Vorlage, and a non-aligned category. The Murabba'at and Nahal Hever scrolls, by contrast, are uniformly proto-Masoretic. Something happened between c. 70 CE and c. 135 CE — a deliberate selection of one text-type and the abandonment of the others — and the rabbinic centers of the early Tannaitic period are the most plausible agents of that selection.
Shemaryahu Talmon, writing in The Cambridge History of the Bible (vol. 1), argued that the proto-Masoretic text was originally the Pharisaic-rabbinic recension, one of several living traditions in the late Second Temple period. The destruction of the Temple in 70 CE eliminated the priestly custodians of rival recensions; the failed revolts of 66–73 and 132–135 dispersed or destroyed alternative communities. The rabbinic text-type survived because the rabbinic community survived, and as it survived it standardized.
The proto-Masoretic period also saw the codification of which books belonged to the canon at all. The discussions preserved in m. Yadayim 3:5 — whether Song of Songs and Ecclesiastes 'render the hands impure,' i.e. are sacred Scripture — record the late stages of a process largely complete by the time of Rabbi Akiva. The closure of the canon and the closure of the text were two faces of the same scribal-rabbinic project.
The biblical scrolls from Murabba'at and Nahal Hever already exhibit the consonantal text of the medieval Masoretes, almost letter for letter.Emanuel Tov, Textual Criticism of the Hebrew Bible (3rd ed., 2012)