Tuesday, March 21, 2006

Professor Halivni and the Sealing of the Gemara - a new chronology

In Sidra #20, Professor Halivni just published a very interesting article proposing a new theory for the date when the Gemara was published in its present form. It is a summary of Halivni’s lifetime study of the Gemara, deconstructing it into its different layers. The article directly confronts the traditional theory proposed by R.Y.E.Halevi in Dorot Harishonim, which bases itself on the Iggeres Rav Shrira Gaon, the Seder Tanaim Veamoraim, Sefer Hakaballah by Ravad (the first) and other accepted sources.

Prof. Halivni suggested about 30 years ago that the “Stamaim” are the ones who wrote the discussions in the Gemara. They lived after the Amoraim, namely Ravina and Rav Ashi who were considered the editors of Shas. Halivni suggests that the original Gemara, at the time of Ravina and Rav Ashi (R&RA) was similar to the Mishna a compilation of rulings with very little records of discussions. That is how he understands the meaning of R&RA sof Hora’ah – in other words the last rulings. However the explanation and interpretation of these rulings began immediately after the demise of R&RA, and the Gemara as we know it took form. There was a formal transmission of the rulings through memorization from generation to generation, but the explanations and arguments were kept in a much more informal way. As time went on some were completely forgotten, others were partially lost and questions arose about what exactly was said. The Stamaim reconstructed whatever was forgotten and resolved most of the questions creating the discussions we find in the Gemara. The names of these great editors were never mentioned and they remained anonymous. Halivni also suggests that there were different schools involved in the process, explaining the different styles of the tractates. The analysis that Halivni used to arrive at these conclusions is mostly internal, linguistic and stylistic.

This article dates the Stamaim from about 550 until about 750. There was an immediate transition to the Sevoraim for a short period of about 50 years (750-800) followed by the Geonim. The Sevoraim still added short pieces to the Gemara but the Stamaim did the main work over a period of 200 years. That explains why we have this hiatus of 250 years after the death of Rav Ashi with no published Seforim until about 800. They were completely involved in putting together the Shakla Vetarya of the Gemara. It is only after that period that we find new publications by the Geonim such as R.Yehudai Gaon, Sheiltot and Halachot Gedolot.

This is completely different from the traditional theory, which is based on the Iggeret Rav Sherira, where R&RA sealed the Gemara with all its discussions, letting the Sevoraim just emend and add some short explanations. Halivni understands that the sof Hora’a was sealing the short rulings a generation or two after Rav Ashi, which are the same dates that Halevi proposed. However the Gemara continued evolving for over two hundred years, where these rulings were further analyzed and clarified, of course implying change. Only then was a final second sealing agreed to, and the Gemara as we know it published.

Halivni proves every position extensively. What I found the most daring is his disagreement with Rav Sherira who must have had a very reliable Masora. It is because of that mainly that he insists that his dating is only a conjecture that would explain many difficult Gemaros.

It is not clear to me how this new theory would affect Psak. There is a very interesting article in BDD #15 on the “Layers – Revadim method” by R. Neryah Gotel. He addresses the issue at length, with an interesting essay by R. Z.N.Goldberg on a sugya in Bava Kama and the consequences of the new approach. Although he finds the Derech interesting he shows that the traditional understanding of the Sugya is correct and therefore the Halacha remains unchanged.
However these new ways of looking at sources have a tendency over time, sometimes several generations, to enter the mainstream and influence Halachik rulings. More about this in another post.

31 comments:

  1. Another point worth bearing is that there is evidence that the Talmud remained studied as basically an oral text (although it did exist in writing too) well into the ge'onic period.

    >What I found the most daring is his disagreement with Rav Sherira who must have had a very reliable Masora.

    I would dismiss that objection out of hand. This isn't a matter of faith and masora (although I appreciate that others disagree). Bear in mind that the famous Aharon ben Moshe ben Asher wrote, in his Dikdukei Teamim, that the nekudot were mi-Sinai--this from someone who may well have been descended from the actual inventor of the points (oversimplification, obviously) some two centuries earlier!

    >It is not clear to me how this new theory would affect Psak.

    It wouldn't, any more than the recognition that earlier authorities didn't always have access to all the things that later authorities did. For example, the assumption of the rishonim is that the reason we theoretically pasken like the Bavli over the Yerushalmi is because halakhah kebasrai. And the reason for halakhah kebasrai is because it is assumed that the later authorities had access to the same information as the earlier ones. The rishonim assumed that the Babylonian Amoraim had a Talmud Yerushalmi in front of them. We know this isn't true, but it doesn't affect pesak (although perhaps it should...?).

    ReplyDelete
  2. >However these new ways of looking at sources have a tendency over time, sometimes several generations, to enter the mainstream and influence Halachik rulings.

    In this I agree! There is definitely a trickle down (or trickle up?) affect.

    ReplyDelete
  3. >I would dismiss that objection out of hand

    I was not talking from the point of view of masora but RSG was close to the Stamaim according to this dating less than 150 years and already he missed them altogether?
    I understand your point with Ben Asher but that was tampering with Torah shebiksav which requires Halacha lemoshe for legitiomization

    ReplyDelete
  4. >It wouldn't, any more than the recognition that earlier authorities didn't always have access to all the things that later authorities did

    That is the point of the article by Gotel but more subtle. I will try to summarize in another post.

    ReplyDelete
  5. >I understand your point with Ben Asher but that was tampering with Torah shebiksav which requires Halacha lemoshe for legitiomization

    Do you think he didn't believe it himself?

    ReplyDelete
  6. >I was not talking from the point of view of masora but RSG was close to the Stamaim according to this dating less than 150 years and already he missed them altogether?

    He could have missed them, if they were indeed anonymous. Perhaps he conflated the Savoraim with them. Perhaps Halivni's wrong in his dating, and perhaps the stammaim wore two hats and they were identical with either amoraim or savoraim. But maybe not. If you view the stammaim as the glue or cement which holds the Talmud together--in fact, how long did it take you to realize that *someone* is saying all those things in the Gemara? I am younger than you, R. David, but I was certainly learning Gemara for awhile before this became apparent (although I expect that if I'd begun learning Gemara as an adult it would have been fairly obvious right away and I'd wonder who was talking).

    ReplyDelete
  7. >Do you think he didn't believe it himself?

    Do you think Moshe believed it when Niskarera daato in Menochos?

    ReplyDelete
  8. >I am younger than you, R. David, but I was certainly learning Gemara for awhile before this became apparent (although I expect that if I'd begun learning Gemara as an adult it would have been fairly obvious right away and I'd wonder who was talking).

    I learned Rashbam in Bava Basra 95 (or 96) the last one on the page. Of course the baalei tosfos knew all that. It is the datingthat is novel and even more the argument that Chasimas hatalmud was similar to Chasimas Hamishna! In that article Halivni discusses the similarities and differences between the two. In fact he argues that most discussions between lets say Abayeh verovo are reconstructs. (The Rashbam above seems to be saying the same about R.Yochana and R.Yeshoua ben Levi).

    Dont be impressed about my bekius. I just learned that Rashbma yesterday afternoon and I already am not sure of the page. But it is almost the same every daf.

    ReplyDelete
  9. That's no parrallel. That's an Aggadah. You'd have a question if Moshe wrote it himself.

    We know that the traditional view came to be that the nekudot were mi-Sinai (or mi-Ezra), to the extent that when R. Eliyahu Bachur said otherwise it was shocking! And for two hundred years serious scholarly opinion, both inside and outside Jewry, continued believing that he was wrong. You published a great article on that... ;)

    I realize that we can't readily admit evidence from the 15th century about what Ben Asher believed in the 10th, but we do have his own words and they attest to the belief that the nekudot were Sinaitic. Now, it may seem strange to know that not only was he an insider in the whole Massoretic zach, but he was also, perhaps, its greatest exemplar. So how the heck could he think that, barely two hundred years after they were invented--and invented by his physical forebears, perhaps (among others)?

    The answer is to be found in The Invention of Tradition. In short, its rather simple for even relatively recent innovations to be conflated with that which is from time immemorial. The proof is that it happens.

    All that said, who knows, maybe Ben Asher was touting the Massoretic party line--but I doubt it.

    If I'm not wrong, then it seems like it would have some bearing on how R. Sherira could have some mistaken details about even a relatively recent process. Remember, a couple of centuries isn't really a small amount of time--it only seems to be so compared with the grand sweep of history.

    ReplyDelete
  10. >I learned Rashbam in Bava Basra 95 (or 96) the last one on the page. Of course the baalei tosfos knew all that.

    No, I wasn't saying that "stamma de-Gemara" is a modern invention, only that the desire to analyze it with fewer or different assumptions is.

    ReplyDelete
  11. >You'd have a question if Moshe wrote it himself.

    No but the gemoro portrays Moshe accepting the idea even though it is a fiction. it is just a way of saying that it is legitimate.

    However i still have a hard time accepting that RSG did not have some kind of oral or written record of his own predecessors two hundred years back.It is possible that for some reason he did not want to disclose that as it would weaken the authority just like Halacha lemoshe grants authority.

    ReplyDelete
  12. "The rishonim assumed that the Babylonian Amoraim had a Talmud Yerushalmi in front of them. We know this isn't true, but it doesn't affect pesak (although perhaps it should...?)."

    But it IS true that they had access to the shitas of the yerushalmi; there was a lot of travel back and forth between bavel and E"Y, as is clear from scores and scores of gemaras. The rishonim didn't think they had an actual yerushalmi open in front of them; that's not their theory.

    ReplyDelete
  13. "I would dismiss that objection out of hand."

    If you read Igeres R Sherira, it is clear that he had a lot of info about what was going on in yeshivas 100 years earlier. He writes confidently about minhagim in R Saadya's yeshiva (sura) on all sorts of things, nusach on birkas keriyas shema, etc. Proofs from other people without that level of access aren't dispositive, because R Sherira had detailed info on all kinds of things. This is in no way a matter of "faith" in mesora and it's a bit silly IMO to dismiss this objection out of hand. To put it differently, even Halivni doesn't appear to dismiss this objectoin out of hand; it's a scholarly objection, not a "frum" objection.

    ReplyDelete
  14. "If you view the stammaim as the glue or cement which holds the Talmud together--in fact, how long did it take you to realize that *someone* is saying all those things in the Gemara?"

    R Sherira noticed it.

    ReplyDelete
  15. "However i still have a hard time accepting that RSG did not have some kind of oral or written record of his own predecessors two hundred years back.It is possible that for some reason he did not want to disclose that as it would weaken the authority just like Halacha lemoshe grants authority."


    Given the level of detail of RSG's knowledge about amoraim, and what was going on in both yeshivas, and what various fights were about and the like, it's almost impossible to conceive that there was a whole editing process that he was oblivious to.
    I guess positing that it was a plot or coverup is a different story, but I find that difficult too.

    "This is completely different from the traditional theory, which is based on the Iggeret Rav Sherira, where R&RA sealed the Gemara with all its discussions, letting the Sevoraim just emend and add some short explanations."

    R Halevi's position is quite a bit more sophisticated than this "Traditional theory." He agrees that the stamaim were at least a dor after Ravina/Rav Ashi, and he thinks Ravina is ravina achron, and etc. He didn't miss any of these issues.

    ReplyDelete
  16. >He didn't miss any of these issues.

    I agree that he also held that after R&RA there was additions to the Gemara. However I don't recall that he identified Stamaim . He extended the final seal to RavinaII and Rava Tosefa'ah but immediately after them came the Sevoraim with their emmendations.
    If I recall correctly Rishonim talkj about additions from R.Yehuda Gaon to gemara too. Halivni's chronology would place him as among the Sevoraim.Halevi would consider it a geonic illegitimate intercalation that cannot be relied upon. (i am not sure I read halevi correctly, it has been a while)

    ReplyDelete
  17. "there was additions to the Gemara. However I don't recall that he identified Stamaim"

    i think he says they *were* the stamaim.

    ReplyDelete
  18. >But it IS true that they had access to the shitas of the yerushalmi; there was a lot of travel back and forth between bavel and E"Y, as is clear from scores and scores of gemaras. The rishonim didn't think they had an actual yerushalmi open in front of them; that's not their theory.

    I didn't say they thought they had an actual Yerushalmi. But its been demonstrated that the assumption that they knew and therefore took into account all of shittas of the Amorai E"Y is not true. No biggie.

    ReplyDelete
  19. >If you read Igeres R Sherira, it is clear that he had a lot of info about what was going on in yeshivas 100 years earlier. He writes confidently about minhagim in R Saadya's yeshiva (sura) on all sorts of things, nusach on birkas keriyas shema, etc. Proofs from other people without that level of access aren't dispositive, because R Sherira had detailed info on all kinds of things. This is in no way a matter of "faith" in mesora and it's a bit silly IMO to dismiss this objection out of hand. To put it differently, even Halivni doesn't appear to dismiss this objectoin out of hand; it's a scholarly objection, not a "frum" objection.

    I didn't say I would dismiss that objection out of hand because R. Sherira couldn't have known what was flying a hundred years ago before him. Of course he could have--but he also might not have. I dismissed Reb David's considetation that a theory grounded in textual analysis is daring simply because R. Sherira "must have had a very reliable Masora." Naturally Iggeret R. Sherira is essential studying to understand how the Talmud came to be and no one dismisses it.

    ReplyDelete
  20. "I didn't say they thought they had an actual Yerushalmi. But its been demonstrated that the assumption that they knew and therefore took into account all of shittas of the Amorai E"Y is not true. No biggie."

    Who demonstrated this? Pls give sources. It is quite clear that there was steady cross pollination between teh two communities, and extensive traveling between them.

    ReplyDelete
  21. This discussion about Yerushalmi and bavli knowing each other is old going back to the Rishonim. I dont have Halivni's article here now but i think he addresses it trying to prove the Bavli did not know or consider Yerushalmi. will come back tonight.

    ReplyDelete
  22. "I didn't say I would dismiss that objection out of hand because R. Sherira couldn't have known what was flying a hundred years ago before him. Of course he could have--but he also might not have. I dismissed Reb David's considetation that a theory grounded in textual analysis is daring simply because R. Sherira "must have had a very reliable Masora." Naturally Iggeret R. Sherira is essential studying to understand how the Talmud came to be and no one dismisses it."

    We *know* he had a reliable mesora. This is your original comment:

    ">What I found the most daring is his disagreement with Rav Sherira who must have had a very reliable Masora.

    I would dismiss that objection out of hand. This isn't a matter of faith and masora (although I appreciate that others disagree). Bear in mind that the famous Aharon ben Moshe ben Asher wrote, in his Dikdukei Teamim, that the nekudot were mi-Sinai--this from someone who may well have been descended from the actual inventor of the points (oversimplification, obviously) some two centuries earlier!"

    We aren't discussing whether R Sherira's letter is essential studying; we are discussing whether R Sherira's knowlege base is analogous to Ben Asher's. The cases are different. With Ben Asher, there is some familial connection, but there is no institutional memory; the academies were institutions independent of specific gaonim. In addition, the kind of details that R Sherira provides in the igeres, such as where everyone sat in a given period, argues rather strongly against him not knowing what the work of the academy *was*. When you say "Of course he could have [known what was flying]--but he also might not have" - the idea that it's just as likely he didn't know as that he did is contradicted by the level of detailed institutional knowledge he demonstrates. This doesn't prove he knew how the gemara was finalized, but it makes daring to posit otherwise.

    ReplyDelete
  23. >This doesn't prove he knew how the gemara was finalized, but it makes daring to posit otherwise.

    It is even harder to accept that R.Sherira did not know about 200 years of Stamaim in different yeshivos as Halivni posits. The only way i can see it is that RSG purposely covered up - this is also not very likely.They were anshei Emes.

    The thing that really is interesting is having a gemoro like a mishna, with rulings only, discussions memorized and all added afterwards. It would be interesting if someone would reorganize the gemara accordingly. I wonder if it would help clarify many murky issues.

    ReplyDelete
  24. how does halivni deal with this?

    It also leaves you with problems historically. The impression from R Sherira, Rambam, Halevi etc is that the immediate acceptance of the gemara was due to the stature of beis dino shel r ashi. But if one posits all this work of anonymous stamaim laboring in obscurity, then it's universal dissemination and acceptance is puzzling. How did it get disseminated and universally accepted, was it sent out only with the authority of the academies and no specific person? If so, how did that escape R Sherira's attention, and how did it leave no historical imprint? I'm not saying any of this is dispositive, but it seems to weigh against this scenario - how does Halivni deal with this question?

    ReplyDelete
  25. Halivni says that Rav Ashi ubeis dino edited a set of rulings, memros just like a mishna with the two or three opinions listed. The stamaim then filled in the gaps and created the discussion.

    I will address this more from home.

    ReplyDelete
  26. Halivni proposes that there were two separate sealings, that happened unoficially and naturally, just because at the first sealing the ability to make rulings stopped and at the second sealing the ability to write the shakla vetarya diminished. He does not explain what caused these declines.

    He also posits that when the stamaim finished their work, there were no formed Masechtot just an unorganized collection of discussions. There was then editors that gathered all these up and organized them by subject.They were also spread among different batei midrash, expalining stylistic differences in masechtot. That explains why there are discussions that do not fit well in a specific sugya, sometimes contradictory with similar discussions elsewhere. Once the Stamaim finished their work, the later editors did not dare tamper with the text.

    Unless I missed it, Halivni does not address the issue of what gave the gemoro its authority and acceptance.

    He does address the issue of RSG masora. he conjectures that the archives and other data available would record names and events but would skip over work that was done anonymously. Furthermore he suggests that the Geonim wanted to legitimize Rav Ashi and Ravina as final Psak therefore glossing over the stamaim.I still think that this is a problem that needs to be resolved to make Halivni's theory acceptable.

    ReplyDelete
  27. weird. there are a lot of loose ends here, I guess one would have to see the article.
    thanks for the summary.

    ReplyDelete
  28. If you email me I will scan the article and send it to you.

    ReplyDelete
  29. that's very kind of you. thanks!

    ReplyDelete
  30. Who were the stammaim of yerushalmi? Although it is less oganized than bavli it is crtainly not just notes and it is accepted that its writing was cut short and already it is a sophisticated product.
    avakesh at avakesh.com

    ReplyDelete
  31. I sent your question to my friend Heshey Zelcer and here is his response:

    I discuss this on p. 55 in my “A Guide to the Jerusalem Talmud”



    The Stammoth in Bavli and Yerusahalmi. Both the Bavli and Yerushalmi contain anonymous, argumentative “give and take” text which are referred to as stammoth. While the stammoth of the Bavli are rich, multitiered, and complex, those of the Yerushalmi tend to be simple and narrow in focus, responding only to the question at hand. In fact, the older parts of the Yerushalmi, Baba Kama, Baba Metzia, and Baba Bathra (generally referred to as Yerushalmi Nezikin) contain almost no stammoth. According to Halivni, the stammoth of the Yerushalmi differ not only in that they are shorter and less complex, but also in what they represent. He believes that the stammoth of the Yerushalmi are by and large the conversations of the last generation of Amoraim recorded in the Yerushalmi. The editors of the Yerushalmi decided to record their own ideas without specific authorship, while those of their teachers, they quoted attributionally. Had the Yerushalmi continued for another generation, the stammoth would have been identified by the following generation of Amoraim, and the succeeding generation would have had their ideas expressed as stammoth. Halivni (142) expresses this as follows:



    When the editors of the Yerushalmi decided to close the development of the text and to “freeze” it into a book, they recorded what went on in the academy at that particular time. The ideas of the then “present speakers” were therefore recorded anonymously (hence the existence of stammoth in the Yerushalmi). Had the editors allowed the text to develop further, the then “present speakers’” ideas would have been transmitted by their disciples attributionally, in their names. By arresting the further development of the text, the editors created an anonymous ring around the attributional material. But the ring was thin and not always significant; it was the result of a technical decision to record the “present speakers’” ideas anonymously.

    ReplyDelete