The editors of the “Ancient Jew Review” Website asked me to reflect on my scholarship on the stories of the Bavli and my “trilogy” of books, Talmudic Stories, Narrative Art, Composition, and Culture (1999), The Culture of the Babylonian Talmud (2003), and Stories of the Babylonian Talmud (2010), and I am grateful for this opportunity.
As I look back on this body of work that occupied me for over fifteen years, my first reflection is how important the role of serendipity is to my scholarship. I am not a literary critic, nor the son (or student) of a literary critic. As a graduate student at Columbia University in the late 1980s, I was not trained in literary or narrative theory, nor did I have much interest in it. I am not even sure if I knew anything about it as a method of analysis. Nor did I have any particular interest in Talmudic stories. I had read Jacob Neusner of course, and his work on Talmudic biography, and we all knew that such traditions could no longer be taken as history. But as a graduate student of David Weiss Halivni, and having studied with Shamma Friedman, I was focused on Talmudic source-, form- and redaction- criticism, and especially in the development of Talmudic law. I decided to write my dissertation on the history of the Festival of Sukkot because it entailed working with biblical and second temple sources in addition to rabbinic literature, and I calculated that a relatively broad field of expertise would make me more marketable. I was also encouraged to think about questions of ritual and religious change by the courses in comparative religion and religious theory in the Columbia Religion Department. After completing the project on Sukkot I planned to devote my scholarship to “hard core” Talmud source-criticism and the complexities of rabbinic law. Talmudic narratives were not even a blip on my scholarly radar screen.
After revising and publishing my dissertation, and authoring a few related articles on Sukkot material that could not easily be integrated in the book, I had had more than my fill of Sukkot, and promised myself I would never write another article on that topic. I happily returned to Talmudic source- and form-criticism, as per my plan. Part of the dissertation research had led me to study the development of the concept of the sukkah as a “temporary dwelling,” an abstract concept that develops only in Amoraic times and appears exclusively in the Bavli. I became very interested in the development of abstraction and abstract concepts in the Bavli, for which source-criticism and form-criticism are indispensable methods. Separating the Tannaitic, Amoraic, and Stammaitic layers reveals that the rabbis employ increasingly abstract formulations and concepts in their legal science. The study of legal abstraction and generalization became my next scholarly project, and it generated a few publications over the next year or so.
But throughout this time a fascinating Talmudic passage that I had encountered in my research on Sukkot kept festering in my mind. This was the long aggadic narrative at the beginning of Tractate Avodah Zarah 2a-3b about the gentiles and their reward in the world to come. The sukkah plays a small part in this aggadah: it is designated a “minor/easy commandment (mitsvah qalah)” and given to the gentiles as a test of their newfound commitment to Torah. This brief mention of the sukkah is of little significance to the understanding of the Festival of Sukkot itself, and I hardly referred to it in my dissertation. Of great interest to me, however, was the fact that the Hebrew aggadic narrative is constantly interrupted by Aramaic comments and digressions, clearly added by a later, editorial hand, which ask questions, raise difficulties, and juxtapose other sources. In this respect the aggadic passage resembles halakhic sugyot where an earlier layer has been supplemented by the Stammaim. Scholars long before me had been aware of these Aramaic additions, I soon discovered, and had already distinguished the core Hebrew aggadah from the Aramaic glosses. But I also noticed the Hebrew aggadah itself had a precise structure with several tri-partite units, not unlike halakhic sugyot, as Shamma Friedman had documented in his work on the structure of sugyot in the Bavli. My problem was that despite having identified the structure I did not really know what to do with this text—I had no method for analyzing it or thinking about it in a sophisticated way.
In the course of searching for a useful method I stumbled upon Yonah Fraenkel’s scholarship on Talmudic stories, which was becoming better known at that time. His structural and literary analyses, I felt, provided part of the answer, and could be applied fruitfully to the story in Tractate Avodah Zarah. Like many others, I was very excited by Fraenkel’s brilliant readings, literary sensibility, and meticulous attention to the text, and set about reading everything Fraenkel had written. I also realized I had to catch up on literary theory in general. Soon I felt I had something to say about this text and published my first article on a Talmudic narrative, and first effort at a literary analysis, in AJS Review (1996).
However, the matter of the Aramaic additions festered in my mind, as these were obviously comments by the Stammaim. What about the base Hebrew narrative itself? Did the Stammaim limit themselves to the Aramaic comments or also rework the narrative source they received, much as they reworked earlier halakhic traditions and organized them into sugyot? Could we see evidence of Stammaitic reworking of other stories? What was the role of the Stammaim in the production of Talmudic narrative in general?
By chance (serendipity again!) Tisha b’Av occurred as I was thinking about these questions. I was asked to lecture in an adult education venue and decided to teach the story of the destruction of the Jerusalem Temple in Gittin 55b-56a (what else do we all teach on Tisha b’Av?). I started working on that story, trying to analyze it as a literary composition, qua Fraenkel, but also as a “sugya,” by identifying its sources and distinguishing the role of the Stammaim, qua Halivni and Friedman. The redactional location in Tractate Gittin, as opposed to, say, the Talmud to Mishnah Taanit 4:6 which lists the destruction of the temples and other disasters that occurred on Tisha b’Av, also required explanation. Soon I turned to work on stories full time, abandoning my project on abstraction. This too was very fortunate, as Leib Moscovitz produced his magnum opus, Talmudic Reasoning, a few years later (2002), and did a much better job than I ever could have done. My study of Gittin 55b-56a led to work on other narratives and eventually to Talmudic Stories: Narrative Art, Composition, and Culture in 1999.
I offer this genealogy for two reasons (other than the narcissistic and self-indulgent pleasure we all feel when actually invited by others to write about ourselves.) First, to illustrate that the method I developed was not really innovative. It simply combined Fraenkel’s “close reading” approach, which takes narratives seriously as literature, with the source- and form-criticism of Halivni and Friedman (with some Ofra Meir and Jacob Neusner too). I have always considered myself a consumer of methods of analysis, not a producer. Second, it was the fact that I was very much involved in critical Talmud study with the work on abstraction, and attuned to the serious reworking of sources by the Stammaim, that led me to ask these questions about narratives.
To put this last point in slightly different terms: it seemed completely counterintuitive to me that the author-editors of the Talmud, whose interventions and reworkings profoundly impacted just about every halakhic sugya, would not have reworked stories too. In fact, because aggada has less authority than halakha—because there is less of an imperative to preserve statements verbatim, to “say a tradition in the name its originator” (lomar davar be-shem omro) with aggadic material—we can expect that the Stammaim would have felt even more free to rework their aggadic and narrative sources.
To understand Bavli narratives as deriving from the Stammaim has the great advantage of providing them a rough historical context. Neusner, Fraenkel and Shamma Friedman (in a trailblazing article, “La-aggadah ha-historit ba-talmud ha-bavli,” 1993) had taught us that Bavli stories are fictional compositions, hence the way that Tannaim are depicted in the Bavli, which may differ from depictions in other rabbinic documents, cannot be relied upon for historical purposes. Yet without a real theory of their provenance or dating it was difficult to provide a coherent account of how and why the stories had changed. The Stammaitic context offered the beginnings of a way to make sense of these changes. But of course more detailed study was necessary to appreciate how the culture of the Stammaim differed from other rabbinic cultures and why?
These were the questions I hoped to answer in The Culture of the Babylonian Talmud. The method was to compare systematically Bavli stories and other aggadic traditions with the parallels in the Palestinian Talmud (= Yerushalmi) and in the Palestinian Amoraic Midrashim and to identify motifs and themes that consistently appeared in the Bavli but not in the parallels. Study of those motifs hopefully would provide a window into the culture, social setting, and values of the Stammaim.
The picture that emerged was of a scholastic culture anchored in an institutionalized rabbinic academy. David Goodblatt had shown that Bavli Amoraim seem to have been organized in small-scale “disciple circles,” with a handful of students learning with one master. The same seems to have been the case in the Land of Israel throughout the Amoraic era. Many Bavli narratives, however, depict a bustling academy full of dozens of students and several leadership positions, in other words, an institution, much like the academies known from Geonic sources of the Islamic age. These large-scale academic institutions appear to have developed in the Stammaitic era, and the challenges, struggles, and ethos reflected in those narratives are those of the Stammaim. The academies were hierarchically organized with the “Head of the Academy” at the top. Status within the academy was determined by mastery of Torah, with the more proficient students in theory holding the top positions in the academic hierarchy and the coveted seats in the front of the academy. Lineage (yihus), however, played an important role in practice such that the highest positions seems to have been reserved for sages of exalted lineage and may even have been hereditary. The most respected form of Torah was dialectics, the ability to propound questions and respond to difficulties and challenges from other students. Sages who excelled in dialectics were honored and perhaps even ascended the academic hierarchy. To fail to answer a difficulty, on the other hand, resulted in a sage experiencing public shame, which could involve the loss of status. The academy was therefore an “agonistic” environment, with the sages engaging in verbal violence as they waged “the wars of Torah.”
The shift in the form of rabbinic learning from disciple circles to large scale academies, similar to the Geonic academies, may be related to the decision not to continue preserving attributions but to formulate traditions in an anonymous manner. The academy is a corporate body, an institution with many sages and students, with an identity of its own. Of the sages of Pumbedita and Sura in Geonic times, we only know a few names other than those who were Heads of the Academy. The same seems to be the case of the Stammaitic era, when attributions dwindle and then disappear, but generations of anonymous editors continue to work. This explanation is speculative, given the lack of available evidence. Nevertheless, it seems preferable to me than the leading alternative, the “lachrymose” theory of change that would attribute the shift to persecutions in late Amoraic times that disrupted academic life such that the sages felt they could no longer continue in the ways of the past. Continuities between Stammaitic and Geonic culture, in general, are a fertile subject for further exploration.
The esteem of dialectics reflected in narratives provides a nice parallel to the dialectical quality of the Stammaitic stratum of halakhic sugyot, and contributes to the evidence of Stammaitic authorship. So too does the depiction of the highly developed academy for the reasons delineated above. The other motifs that appear in narratives and in the Stammaitic halakhic stratum, but rarely in Amoraic dicta, also support this claim. This phenomenon is analogous to the presence of abstract terms and concepts exclusively in the Stammaitic stratum, a consideration that provides good evidence of the post-Amoraic provenance of such passages.
However, the claim that these narratives were of Stammaitic, not Amoraic, provenance, deserves additional support. This issue has been among the main points of criticism of my work, namely, that there is not always sufficient evidence to attribute a narrative to the Stammaim—that I have not reckoned seriously enough with the possibility of Amoraic “authorship.” Granted the Bavli versions of many stories differ substantially from those of the Yerushalmi and of other Amoraic Midrashim from the Land of Israel, how do we know that the Bavli Stammaim, rather than Bavli Amoraim, bear responsibility for those changes?
This is certainly a fair concern, and methodological caution is always warranted. Certainly it is harder to identify the authorship of narrative material because the formal criteria we employ in the study of halakhic discourse typically cannot be applied to narratives. Unlike halakhic sugyot, narratives are not composed of attributed statements of the Amoraim and anonymous dialectical discussion of the Stammaim, which are relatively easy to separate. In the case of halakhic sugyot, the Hebrew of baraitot and of many Amoraic meimrot usually can be distinguished from the Aramaic of the Stammaitic stratum. But this is not the case with narratives, where Hebrew and Aramaic sometimes mix in messy and complex ways.
In an essay “Criteria of Stammaitic Intervention in Aggada” (2005), and in Stories of the Babylonian Talmud (2010), I tried to give this issue detailed attention. The Stammaitic provenance of many narratives, in my opinion, can be recognized by various additional criteria. Most important is the propensity of the composers of narratives to supplement the sources they received, as reflected in the parallel versions, by transferring material from elsewhere in the Bavli. This is a Stammaitic technique par excellence for composing halakhic sugyot, and is hardly attested among Amoraim. The fact that many stories share connections with the proximate halakhah or with the Mishnah, connections sometimes absent from the parallels, also points to the hands of the Stammaim, as they are responsible for the redactional context. Sometimes there are other indications of late provenance specific to a given story, such as a great plethora of manuscript variants.
Another factor that must be considered in assessing the claim to Stammaitic provenance of Bavli narratives is the relevance of the stories in the Yerushalmi, as some critics have pointed out. Because we find some highly-developed narratives in the Yerushalmi, which clearly date to the Amoraic period, we need not consider all lengthy Bavli narratives to be Stammaitic.
The narrative art and compositional techniques of the stories of the Yerushalmi have yet to receive comprehensive treatment, despite Catherine Hezser’s superb book on the stories in Yerushalmi Neziqin. Further study should be able to help us understand better the differences between stories in the two Talmuds. However, I don’t think the presence of a few long and complex stories in the Yerushalmi or in the Amoraic Midrashim is evidence against my claims, as I agree that not all Bavli stories should be assigned to the Stammaim. Indeed, in the concluding chapter of Stories of the Babylonian Talmud I presented a fairly extensive Bavli narrative of Amoraic provenance, as indicated by the absence of the markers of Stammaitic composition. Just as the criteria discussed above evidence the Stammaitic provenance of narratives, so their absence points to an Amoraic composition that the Stammaim did not rework.
The issue of distinguishing Amoraic from Stammaitic narratives raises the question of how similar or distinct were the rabbinic cultures of the Amoraim and Stammaim. What are the lines of continuity as opposed to discontinuity? A few scholars have pointed out that some of the elements I claim to characterize the culture of the Stammaim are attested among the Amoraim, including the importance of lineage or yihus and the fear of shame.
This point is also well-taken. Certainly cultural change is a slow and uneven process. Just about all historical periodization is artificial and the reality much more complex, and the terms we use are partially for heuristic purposes. Renaissance culture differs from that of the “Middle Ages,” and that is why we use the term, but obviously the Renaissance also has roots in medieval tradition, and it is difficult to say exactly when one ended and the other began. The same is true of the relationship between the Stammaitic and Amoraic periods. So I did not intend to draw a rigid line in the sand between Amoraic and Stammaitic culture; if I gave the impression of a clean historical break I am happy to clarify my stance now. The fact that some Amoraim of the seventh, eighth and perhaps even later generations are mentioned by name in the Talmud suggests a transition that took several generations to complete. Nevertheless, even where critics have pointed out lines of continuity, there may still be differences worth thinking about. Lineage, yihus, as any classicist knows, was critically important in the Roman world (and probably in most ancient civilizations of the Levant), even if not to the same extreme as in Sasanian culture, where it almost became a type of caste system. For this reason we find considerations of yihus relevant in the Yerushalmi and among the Babylonian Amoraim. But in Stammaitic times yihus becomes a critical factor in the criteria for academic leadership, such as for the position of Head of the Academy, and this function of lineage is hardly attested in the Yerushalmi or in Amoraic sources within the Bavli. Likewise fear of shame appears in almost all rabbinic documents in one form of another—it may be a characteristic of “Mediterranean Culture” of even a universal human experience. Only within Stammaitic sources, however, do we see a horror of “academic” shame, of public humiliation due to the inability to parry an objection, and with possible consequences for a sage’s position in the academic hierarchy.
In considering these issues, I keep returning to the point above. Study of sugyot involving abstract terms and concepts, and really analysis of just about every halakhic sugya in the Bavli, point to the prominent, pervasive, and relentless reworking of traditions by the Stammaim. Yaakov Zussman, among the greatest Talmud scholars of our generation, if I understand him correctly, even considers it unlikely that we can distinguish the original Amoraic statements from the reworkings of later editors due to the extent of editorial activity in the Bavli. With stories, the same should be the case, and most likely to an even greater degree. It strikes me as extremely unlikely that the Stammaim left most of the narratives they inherited untouched. The real question is to refine the methods of determining exactly how the Stammaim reworked their narrative sources, for what purposes, and what this tells us about their culture.
For all the work of the past fifteen years, the study of Bavli stories is still in its infancy. I routinely receive emails asking if there is any scholarship on a certain story. Sometimes there is a study or two, often there is nothing. Compared to the scholarship on any chapter of Bible or line in the Dead Sea Scrolls, not to mention the prodigious body of research on any passage in the Christian Scriptures, rabbinic stories are unexplored territory. We still have a great deal of work to do, and continued study should help clarify some of the questions above.
Another scholarly desideratum is to set Bavli stories in the context of Pahlavi and Syriac literature of the Sasanian world. At some point we have to get outside of the Bavli to the ambient culture fully to appreciate Bavli narratives, as is true of the study of any other text. Greco-Roman literature and literary models have been used to great benefit in the study of narrative sources in the Yerushalmi and in the Amoraic Midrashim, and even for some Bavli stories, which evince Hellenistic influences. But many Bavli stories feature motifs, images and narrative devices which we do not fully understand. Greater comparative study hopefully will help clarify these obscurities.