Jul 22, 2025

The Formation and Redaction of the Babylonian Talmud

 

Once again, I refer to the invaluable history of Rabbi Isaac Halevy, who divided the development of the Talmud into 4 distinct stages: (1) the collection of the teaching of the Tannaim which pertaining to each individual section of the Mishnah; (2) the collection of the teaching of the Amoraim by Rava and Abbaye; (3) the editorial work of Rav Ashi; (4) the final work of Rav Ashi’s court as his death in completing the editing of the Talmud till the death of Ravina II. 


The common and widely received view is that two figures that were primarily responsible for the compilation of the Talmud are Rav Ashi and Ravina, the former of whom taught at the Jewish yeshiva in Sura (specifically, his yeshiva was in the town of Mata Mehasya), and had the longest tenure there than any other figure of Babylonian Jewry (Iggeret Rav Sherira Gaon, VII. 5). It is important to note that the official “sealing” of the Bavli took place in AD 500, 73 years after Rav Ashi’s death. The Gemara itself places these two figures as marking the end of the Amoraic period: “Rabbi Yehuda HaNasi and Rabbi Natan are the end of the Mishna, i.e., the last of the tannaim, the redactors of the Mishna. Rav Ashi and Ravina are the end of instruction, i.e., the end of the period of the amoraim, the redacting of the Talmud, which occurred after the period of the tannaim. And your mnemonic to remember that Rav Ashi and Ravina redacted the Talmud is the verse: “Until I entered into the sanctuary of God, and considered [avina] their end” (Psalms 73:17). The sanctuary alludes to Rav Ashi, while the term avina alludes to Ravina, which is a contraction of Rav Avina. The phrase: Their end, is interpreted as a reference to the redacting of the Talmud.” (Bava Metzia 86a). The linking together of Rav Ashi with Judah Ha-Nasi may also form a fairly strong indirect proof for the former’s leading role in creating the Talmud:—“R. Adda bar Ahavah said: “I also say, that since the days of Rabbi [Yehudah Hanasi] until Rav Ashi, we have not found Torah [learning] and greatness [in secular matters combined] in the same person.” (Sanhedrin 36a)


Sherira Gaon maintains that Rav Ashi and Ravina were simply the final figures in a long work of compiling the Gemara, and not the first to do so; contra Rashi, as we shall see further below.


According to Maimonides (Introduction to the Talmud [New York: Judaica Press, 1975], pg. 149), Rav Ashi had 4 goals in his redactional work:


[1]. To explain the statements of the Mishnah by recording the differing interpretations of the Tannaim, the arguments one rabbi made against another, and what the correct opinion was decided to be.


[2]. To record the final decision, especially when there was a disagreement present amongst either the Tannaim or the Amoraim (Introduction to the Talmud: Histories, Personalities, and Background [Mesorah Publications, 2019], pg. 136)


[3]. To the record the new applications derived from the Mishnah by both the Tannaim and the Amoraim. This also reveals some of the basic presuppositions operative within rabbinical law. Rav Ashi also recorded all of the Gezayros and Takonnos enacted up to his time. Thus many things not explicitly included in the Mishnah are set forth by the Talmud, and how they were derived therefrom.


[4]. To record the Aggadah fitting for each topic in its particular chapter (perek). 


And again, Rashi summarizes this work: “Until the time of Rav Ashi and Ravina, all the Amoraim before them did not have a systematically ordered Gemara. Rather, when a question was raised regarding the rationale of a Mishnah in the study hall, or a question arose concerning a legal case—whether in monetary law or matters of forbidden and permitted actions—each scholar would state his reasoning. Rav Ashi and Ravina compiled the teachings and discussions of the earlier Amoraim and arranged them according to the order of the tractates, placing each one alongside the appropriate and relevant Mishnah. They formulated the objections that required answering (kushiyot), and the resolutions that were fitting to explain (terutzim), either themselves or with the Amoraim of their generation. They established all of this in what we now call the Gemara—with standard expressions such as "eitivei", "metivei", "raminei", "ibaya lehu", and their associated answers—comprising what had been left behind by their predecessors or stated before them. The objections and resolutions that had been said earlier by others were not previously fixed within the structure of the tractates or alongside the Mishnaic text arranged by Rabbi Yehudah HaNasi. Rav Ashi and Ravina came and incorporated them into an organized form.” (Rashi on Bava Metzia 86a)


The compilation of the Gemara was also a work undertaken in the context of the kallah, the large Talmudic assemblies held semi-annually. It was supported by the Sasanian King Yazdegerd I, known for his friendly disposition towards the Jews. According to Hai Gaon, Rav Ashi held the position as head of the Academy of Sura for 60 years. It generally took six months to study each tractate (mesechta), which adds up to a total of 30 years for the 60 tractates by these Babylonian amoraim. This same process was repeated again for the next 30 years after that— “The first time Rav Ashi taught this matter he said to us: The first lender acquires the property. The last time Rav Ashi taught this matter he said to us: The lenders divide the property between them. And the halakha is that they divide the property between them.” (Bava Batra 157b). Some of the Tosafists fantasize vainly that a pillar of fire descended from the heavens during such Talmudic lectures (Tosafot on Berachot 17b). 


The final completion of the Talmud was done by Ravina II, who died in (AD 475), though Sherira Gaon places it at the year 500 (which, as noted above, is the year in which the Talmud was officially closed and the period of the Savoraim began). This year is also corroborated in the anonymous text Seder Olam Zutta. While an exact date may not be in order, we know that the Talmud was sealed right around the time of the closing of the 5th century, or the beginning of the 6th. Nachmanides dates it to “400 years after the Destruction [of the second temple].”  During medieval times, the dating of the Talmud was a matter of some dispute, and there were absurd claims given about it by both Jews and their opponents. In the Disputation of Paris (1240), Rabbi Yechiel ben Joseph asserted that the Talmud was by then “1,500 years old”, which would place it in the 3rd century BC. His Christian opponent Nicholas Donin placed it in the later medieval period. 


Unlike Neusner, Hermann Strack argues strongly against the notion of a uniform or homogenous redaction of the Gemara (Introduction to the Talmud and Midrash, pg. 195). 5 tractates – Nedarim, Nazir, Meilah, Keritoth, and Tamid — have significant linguistic and grammatical differences from the rest of the Talmud; this is something which was already noticed by the Jews of the Middle Ages (E.E. Urbach, The Tosafists: Their History, Writings, and Methods [Jerusalem, 1955], pg. 561). It is possible that some of these differences may be due to peculiarities of Aramaic according to the different parts of Babylonia in which the Gemara was formed – Sura, Nehardea, and Pumbedita. 


When it comes to the baraitot which are referenced in the Talmud, there is no difficulty whatsoever when the Bavli and the Jerusalem Talmud both record the attributed saying as such. However, when a saying is reported as a baraita only in the Babylonian Talmud, and not in the Yerushalmi (such as the story of Rabbi Eliezer ben Hyrcanus in Bava Metzia 59b, and the Jerusalem Talmud, Moed Katan 3:1, 81c), this does raise some interesting questions. Why is there this difference? What does this tell us about the accuracy of rabbinic transmission, or the lack thereof? Is it possible that certain halakhic sayings were passed down as a baraita only for polemical purposes without real historical foundation in the figures to whom they are attributed? 


Some scholars such as David Weiss Halvini argue that the anonymous sayings in the Talmud (stammaim) come after the Amoraim, and therefore the compiling of the Bavli was a much longer process than traditionally thought (Halvini, The Formation of the Babylonian Talmud, trans. Jeffrey L. Rubenstein [Oxford University Press, 2013]). Some of the medieval Rishonim attributed these anonymous sayings to Rav Ashi himself, or at least believed they carried equal legal authority (Tosafot and Rabbi Asher ben Jehiel [abbreviated as “Rosh”], commenting on Shabbat 9b).  Halvini focuses especially on the contradictions one may find in the Talmud, and the difficulty of this reality with the traditional view of the Talmud’s sealing. If Rav Ashi and Ravina did indeed edit the Talmud, why didn’t they attempt to reconcile these conflicting passages? Why do the medieval commentators speak of “contradictory sugyot” (Tosafot on Menachot 58b)? Scholars such as Yehonatan Etz-Chayim have compiled numerous such examples from the tractates of Seder Nezikin (see Halvini, pg. 230n10). The medieval Jews generally resort to forced explanations in order to reconcile these Talmudic contradictions, where the two passages being examined are simply relegated to “there are some who say…” (ika de’amri) and “another version” (lishna aharina), while the Gemara itself makes no such reconciliation! If there was such a clean redaction of it, why do these lishna aharina still appear? I now wish to record here some of Halvini’s analytical observations, albeit in summary form.


[1]. Rabbi Eleazar Ha-Kappar said “A Nazarite is called a sinner.” Should this statement be taken as referring only to an impure Nazarite or also to a pure Nazarite? According to Nazir 3a, this statement is only in reference to a ritually impure Nazarite. However, the same tractate later quotes the full statement of R. Eleazar and says that “The Gemara raises a difficulty with Rabbi Elazar HaKappar’s dictum: But this verse, labeling the nazirite a sinner, is written with regard to an impure nazirite, and we are saying that even a pure nazirite is a sinner. The Gemara answers: Rabbi Elazar HaKappar holds that a pure nazirite is also a sinner. And this is the reason that the statement that a nazirite is a sinner is written in reference to an impure nazirite rather than a pure one: Since he repeated his sin, as his impurity causes him to start his naziriteship again, he thereby deprives himself for a longer period. He should have taken extra care to prevent this from happening.” (Nazir 19a). So, which is it? What is the true meaning of R. Eleazar’s statement? Elsewhere, the Talmud objects that “If so then the Nazarite’s sin-offering should also require libations, since it is not on account of sin?” (Sotah 15a). The Gemara responds “He holds like R. Eleazar Ha-Kappar who said that a Nazarite too is a sinner.” Clearly, the Talmud is saying here that a pure Nazarite is also a sinner. Why didn’t Rav Ashi or Ravina II provide a reconciliation here?


[2]. In Chullin 95a, we find two succeeding passages which are both anonymous and contradictory. In the beginning it says, “Rav says: Once meat is obscured from sight and unsupervised, it is forbidden, as one must be concerned that it was exchanged for non-kosher meat.” The Talmud debates this ruling, and makes the following conclusion at the bottom of the same folio: “The Gemara comments: And this statement of Rav, that meat that was obscured from sight is forbidden, was not stated explicitly by Rav. Rather, it was stated that he held this opinion based on an inference from a different statement of Rav.” However, this is a mistaken inference and does not provide any proof, since Rav’s statement was specifically in regard to the “market of the idolaters” (Chullin 95b). The Talmud then asks “How then did Rav ever eat meat?” David Helvini comments “whoever asked this question either was not aware that Rav never stated the prohibition explicitly, or he knew that Rav never stated the prohibition explicitly and that it was derived from an inference, yet did not know that the inference was problematic.” (The Formation of the Babylonian Talmud, pg. 78). Why are these two contradictory statements juxtaposed in immediate succession to one another?


[3]. If Rav Ashi and Ravina were the primary editors of the Talmud, then why do we read of various disagreements between their two respective successors, Rav Aha and Ravina II, usually including terminology such “one said x….and one said y” and not knowing which sage stated each ruling, and how to decide the genuine halakhah? How did the editors not know what took place in their own academies? If they did know, why didn’t they record it? We have a recorded instance of such a disagreement taking place between Rav Ashi and Ravina I - “Rav Ashi holds like Reish Lakish…Ravina holds like Rabbi Yohanan…Rav Aha and Ravina disagreed: one said [concerning the punishment of a certain sin] ‘by extirpation’ (karet), and one said ‘a positive commandment.’ He who said ‘extirpation’ holds like Reish Lakish; he who said ‘a positive commandment’ holds like Rabbi Yohanan.” (Yevamot 11a). Clearly, Ravina is the one who stated “a positive commandment”, and therefore the Ravina that disagrees with Rav Ashi is not the same Ravina that disagrees with Rav Aha. 


Based on these types of considerations, we should keep in mind when speaking of the Talmud being “sealed” in AD 500 that this may be a bit of an anachronistic misnomer, and the true story very well may be more complicated than we might think prima facie. For starters, the statement that “Rav Ashi and Ravina” were “the end of hora’ah” could simply refer to the completion of apodictic law, rather than to the dialectical argumentation of the Gemara. The word “talmud” itself can refer to general rabbinic teaching, as distinct from halakhah and aggadah proper: רִבִּי זְעִירָא בְשֵׁם שְׁמוּאֵל אֵין לְמֵידִין לֹא מִן הַהֲלָכוֹת וְלֹא מִן הַהַגָּדוֹת וְלֹא מִן הַתּוֹסָפוֹת אֶלָּא מִן הַתַּלְמוּד; “Rabbi Zeïra in the name of Samuel: One makes inferences neither from practices (halakhah), nor from homiletics (haggadot), nor from extraneous sources (tosafot), but only from study (ha-talmud).” (Jerusalem Talmud, Peah 2:6, 17a) – The rabbis known as the Savoraim (who lasted for a relatively short period; Sherira Gaon traces it as AD 500-589, while R. Abraham ibn Daud marks the end of this period a century later at AD 689) did make some glosses and additions into the text of the Talmud itself; a fact which suggests that perhaps they could be considered as final redactors. Some tractates begin with Savoriac sugyot, and such additions are found throughout Bava Metzia (Hermann Strack, Introduction to the Talmud and Midrash, pg. 206). Perhaps this is why Sherira Gaon says of Rabbi Yose (a Savora) that “the ‘end of hor’ah’ occurred during his days, and the Talmud was completed”? According to Rabbi Abraham ben David (AD 1125-1198), Rabbi Yose, a Savora, lead the yeshiva in Pumbedita for 38 years (Iggeret Rav Sherira Gaon, VII.7)


“The Sages of the first generation after the Amoraim are known as the Rabbanan Savorai. Although the Gemara had been finalized and sealed, several explanatory discussions of the Rabbanan Savorai were later incorporated into the Gemara as well. For this reason, the names of several Rabbanan Savorai are occasionally mentioned in various discussions in the Gemara. These statements do not include any new teachings not mentioned by the Amoraim. Rather, these statements were incorporated into the Gemara merely to aid in understanding the words of the Amoraim (Iggeret Rav Sherira Gaon, VI.10).” (Introduction to the Talmud: Histories, Personalities, and Background [Mesorah Publications, 2019], pgs. 137-138)


As seen in this quote from the book by Mesorah Publications (the same has also published their own edition of both the Mishnah and Talmud), information like this is usually freely admitted within Jewish circles today, and I do not mean to wield it as an unjust polemical weapon against them. I do however ask: is the notion of an oral Torah going all the way back to Sinai really consistent with such realities? Are the glosses and later additions of Savoraim and Geonim also part of this oral Torah? What about the Rishonim, and the derived laws they make? What bearing does this type of information have on the claim of Maimonides that “From the two Talmuds, the Tosefta, Sifra, and Sifrei, from all of these, comes the knowledge of what is forbidden and what is permitted, what is tamei and what is tahor, who is liable and who is exempt, what is invalid and what is valid, as it had been recorded by one man from the mouth of the other, from the mouth of Moshe Rabbeinu who heard it from Sinai.” (Maimonides, Introduction to Yad HaChazakah)

No comments:

Four Arguments Against the Oral Torah

  #1 - The written Torah says without any hesitation or second-guessing that all that was spoken to Moses by the Lord on Mount Sinai was wr...