Manuscript of one of four lectures delivered at the Oxford Centre for Postgraduate Hebrew Studies, Trinity term (Spring) 1992.
THE NATURE OF THE BABYLONIAN TALMUD
OXFORD CENTRE FOR POSTGRADUATE HEBREW STUDIES
THE JACOBS LECTURES
The Oxford Centre for Postgraduate Hebrew Studies is pleased to announce a new annual lecture series in Rabbinic Thought. Rabbi Dr. Louis Jacobs, in whose honour these lectures have been established, will inaugurate the series with four lectures:
THE NATURE OF THE BABYLONIAN TALMUD
Week 1, Wednesday, 29 April: “The Babylonian Talmud: A Composite Work”.
Week 3, Wednesday, 13 May: “The Babylonian Talmud: A Literary Work”.
Week 5, Wednesday, 27 May: “The Babylonian Talmud: A Comprehensive Work”.
Week 6, Wednesday, 3 June: “The Babylonian Talmud: A Work Requiring Explication”
All lectures to be held at 5.00 p.m. in the Seminar Room at 45 St. Giles’, Oxford.
The Babylonian Talmud: A Composite Work
The Babylonian Talmud, in all the splendour of its over thirty folio volumes in the standard editions, gives the surface impression of a gigantic, unified work compiled at one time, like the great Encyclopedias and Dictionaries. The truth is, however, as every student of the Talmud knows, that the work is composite with material deriving from different periods. In the 20th century, the suggestion was made that a ‘Rainbow Talmud’ be published in which the various strata would be identified by special coloured type rather like the famous ‘Rainbow Bible’, though such a suggestion was never taken up because the cost of such an enterprise would be prohibitive, to say nothing of the difficulty of correctly assessing the different dates of the materials. For all that, in outline, at least, it is possible to detect the different strata with a fair degree of success.
For the nature and composition of the Babylonian Talmud see the article ‘Talmud’ by W. Bacher in the Jewish Encyclopedia, Vol. XII, pp. 1-27; the article ‘Talmud, Babylonian’ in Encyclopedia Judaica, Vol. 15, pp. 755-768; H. Albeck: Mevo ha-Talmud (‘Introduction to the Talmud’), Tel-Aviv, 1969; M. Mielziner: Introduction to the Talmud, 4th edition with new bibliography by Alexander Guttmann, New York, 1968; H. L. Strack and G. Stemberger: Introduction to the Talmud and Midrash, translated by Markus Bockmuehl, Edinburgh, 1991; J. Kaplan: The Redaction of the Babylonian Talmud, New York, 1933; J. N. Epstein: Mevuot le-Sifrut ha-Amoraim (Latin title: ‘Prolegomena Ad Litteras Amoraiticus’), ed. E. Z. Melammed, Jerusalem, 1972; J. Neusner, ed.: The Formation of the Babylonian Talmud: Studies in the Achievements of Nineteenth and Twentieth Century Historical Researches, Leiden, 1970; B. De Freiss: Mehkarim be-Sifrut ha-Talmud, Jerusalem, 1968; Shamma Friedman: Perek ha-Ishah Rabbah be-Bavli, Jerusalem, 1978, and the bibliography cited there; Judith Hauptman: Development of the Talmudic Sugya, Lanham, New York and London, 1988; H. Klein: Collected Talmudic Scientific Writings of Hyman Klein, Introduction by A. Goldberg, Jerusalem, 1979; S. M. Rubenstein: Le-Heker Siddur ha-Talmud, Kovno, 1952; J. Shachter: The Student’s Guide Through the Talmud, London, 1952; M. A. Tenenblatt: The Formation of the Babylonian Talmud (Heb.), Tel-Aviv, 1972; Abraham Weiss: Hithavut ha-Talmud Bishlemato (English title: ‘The Babylonian Talmud as a Literary Unit’), New York, 1943 and Le-Heker ha-Talmud (English title: ‘The Talmud in its Development’), New York, 1954; David Weiss-Halivni: Midrash, Mishnah and Gemara, Cambridge, Mass., 1986; David Kraemer: The Mind of the Talmud, OUP, 1990. (The name for a complete unit of the Talmud is sugya, from the root segi, ‘to go’, hence the layout of a passage, the way it ‘goes’).
The earliest material in the Talmud consists of the numerous Biblical verses quoted throughout the work, introduced generally by dikhtiv (‘it is written’) or de-amar kera (‘the verse says’) or she-neemar (‘as it is said’) or rahmana amar (‘the All-merciful says’). There are more than eight thousand quotes from Scriptural verses in the Babylonian Talmud.
For the Biblical verses quoted in the Babylonian Talmud see the Index volume in the English translation of the Talmud under the editorship of Dr I. Epstein, Soncino Press, London, 1952, and the work by M. D. Rubin: Kol ha-Mikraot she-be-Talmud Bavli, Jerusalem, 1985.
Next comes the material from the Tannaitic period, the first two centuries of the current era the Mishnah, the Tosefta, the Halakhic Midrashim (Mekhilta, Sifra and Sifre) and the many other Baraitot from the same period. A major part of the Babylonian Talmud consists of commentaries and discussions on these works of the Tannaim, especially on the Mishnah, by the Babylonian (and Palestinian) teachers, from the beginning of the third to the end of the fifth centuries, known as the Amoraim. A problem with regard to the terms Tannaim and Amoraim is that these terms are employed in the Talmud in two quite different senses. There are frequent references to two functionaries, one of whom (the Tanna) rehearsed Baraitot before his teachers, the other (the Amora) whose task it was to make explicit the cryptic sayings of a sage. But these two terms are also used in a different sense. A Tanna is, in this sense, a teacher belonging to the Mishnaic (hence Tannaitic) period and an Amora a teacher of the later period, hence the term Amoraim i.e. those who expounded the earlier Tannaitic material. The plausible suggestion has been made that originally both sets of teachers (those of the Mishnaic and the post-Mishnaic periods) were called simply hakhamim (‘sages’) or talmidey hakhamim (‘disciples of sages’) and it was not until the latest stages of Talmudic redaction, of which more will be said presently, that that the sages of the first two centuries were called ‘Tannaim’ and the sages of the later centuries ‘Amoraim’, both terms being adopted from the earlier usage in the sense of the two functionaries.
For the transformation of the terms see A. S. Amir: Mosedot ve-Toarim be-Sifrut haTalmud (English title: ‘Institutions and Titles in the Talmudic Literature’), Jerusalem, 1977; see especially pp. 11 -17 on the Tannaim and pp. 176-177 on the Amoraim. Amir (p. 191) suggests, on grounds of greater accuracy, instead of speaking of the Tannaitic and Amoraic periods or of the Tannaim and the Amoraim, we should rather speak of the period of the Sages of the Mishnah and the period of the Sages of the Talmuds (the Babylonian and the Palestinian). This suggestion has not been widely accepted by scholars since, after all, the terms Tannaim and Amoraim do appear in the second sense in the framework of the Talmud, at least, as Amir admits. For examples of the term Tanna as rehearser of Baraitot see Berakhot 5a; 12a; 57a; Shabbat 66b; Sotah 22a; and in the other sense of Tannaitic teacher Berakhot 2a; 9a; 24a; Shabbat 53b; Betzah 19a; Makkot 2b; Gittin 32a. Examples of Amora in the sense of ‘expounder’ of the words of a sage are: Yoma 20b; Ketubot 106a; Sotah 40a; Hullin 15a; and in the sense of Amora as a later teacher Shabbat 112b and see especially Megillah 15b, ‘like all the Tannaim and all the Amoraim’. The interchange of terms is so frequent that possibly, contra Amir, they were used in both senses from the beginning.
It is clear that the sayings of particular Amoraim are presented in the Talmud as part of the sugya in which their teachings are discussed. The conventional term for the framework of the sugya is stama de-gemara (‘the anonymous part of the Gemara’) that is, the framework into which the opinions of the various teachers are inserted. Another conventional term, for those who provided the framework, is mesaderey ha-Gemara (‘those who gave order to the Gemara’ or the editors of the Talmud). David Weiss-Halivni has coined the useful term Stamaim (‘The Anonymous Ones’) to denote the authors of the framework. It must be realised that this framework is provided for all the material in the Talmud and it uses throughout the same vocabulary and has throughout the same style, structure and form, suggesting that it all came from the same hands. Judging by the Talmudic evidence, the final editors of the Talmud managed successfully to conceal their identity. Nowhere in the Talmud is there any definite statement about the process of redaction and how it was done and by whom. In the middle ages, however, on the basis of a somewhat obscure Talmudic reference, the editors of the Talmud were said to be the fifth century Amoraim Rav Ashi (d. 428) and Ravina (d. 499). The Talmudic statement in question is in tractate Bava Metzia 86a. Here it is said: ‘Rabbi [Judah the Prince] and R. Nathan were the end of the Mishnah. Rav Ashi and Ravina were the end of teaching [sof horaah]’. It is far from certain what ‘the end of teaching’ means but, on the analogy of the statement about Rabbi Judah the Prince and R. Nathan, it was concluded that just as Rabbi Judah the Prince was the editor of the Mishnah, Rav Ashi and Ravina were the editors of the Babylonian Talmud. This cannot possibly be taken at its face value. The framework embraces the opinions of Rav Ashi and Ravina themselves in exactly the same style as the rest of the framework of the Talmud. This very passage—even if it means what it is said to mean, that these two Amoraim were the editors of the Talmud—must have been supplied by those who came after Rav Ashi and Ravina. Consequently, it is now generally recognised that in the framework we have yet another and very late stratum of the Babylonian Talmud.
The major mediaeval source for the question of the redaction of the Talmud is the famous Letter of Sherira Gaon (d. 1006), the Gaon of Pumbedita from 968-1006, in reply to a question on this very problem sent to him by Jacob b. Nissim of Kairovan in 981. The best edition of the Letter in that of B. Lewin: Iggeret de-Rav Sherira Gaon, Jerusalem, 1972 (photo-copy of the Haifa edition of 1921). R. Sherira (Lewin ed, p. 69) accepts that the passage in Bava Metzia does mean that Rav Ashi and Ravina were the editors of a Babylonian Talmud but this probably refers to the core material of our present version not to the present version itself. The whole question of redaction has been discussed at length in the modern works mentioned above, especially by Kaplan, Epstein, Abraham Weiss, David Weiss-Halivni, Neusner, Strack/Stemberger, and Kraemer but I here pursue my own views.
We have then three strata in the Babylonian Talmud: 1) that of the Tannaim; 2) the actual teachings of the Amoraim whose names are given; 3) the late framework into which the other two strata are incorporated. From Sherira onwards it has been noted that a fourth stratum can be detected, that of the Savoraim (rabbanan savorai), the ‘Expounders’ or ‘Theorisers’. Sherira states that he has a tradition to the effect that the opening passage in tractate Kiddushin is a Savoraic addition and other mediaeval commentators have declared the opening passage in tractate Bava Metzia to be Savoraic. The style in these two passages is, in fact, different from the usual style of the framework as can easily be seen when the two are compared. Abraham Weiss has gone so far as to argue that practically every opening passage of the Talmudic tractates is Savoraic. The Savoraim were also responsible for other additions to the completed Babylonian Talmud such as mnemonics at the beginning of some passage and the final rulings at the end of lengthy debates in the form vehilkheta (‘the ruling is’). The point to be noted here is that the Savoraic additions are precisely that; they have been added to an already complete work and in a style different from that of the framework. Investigation has thus detected four strata in the present text of the Babylonian Talmud: 1) that of the Tannaim; 2) that of the named Amoraim; 3) the all-pervading framework; 4) the Savoraic additions.
For the Savoraim see B. Lewin: Rabbanan Savorai ve-Talmudam, Jerusalem, 1937; Strack/Stemberger, pp. 222-225; Abraham Weiss, Kaplan, and article ‘Savora, Savoraim’ in Encyclopedia Judaica, Vol. 14, pp. 920-921. Sherira’s reference to the Savoraim is in the Lewin edition, p.71.
In some circles, nowadays, under the influence of Kaplan, there is a tendency to apportion a good deal of the framework material to the Savoraim but this is to ignore the clear stylistic differences between the framework and the acknowledged Savoraic material. There are difficulties in deciding how much of the framework material is original with those who provided it, the Stamaim, and how much is the work of the Amoraim recorded by name in the framework material but there can be no doubt that there are four distinct strata in the Babylonian Talmud, dating from different periods: 1) The Tannaitic material; 2) the actual words of the named Amoraim; 3) the Stamaim, those who provided the framework to the whole; 4) the Savoraim. The framework, comprising a major part of the Talmudic material, was, in all probability, added after Ravina, which means that the Babylonian Talmud assumed its present form not earlier than the beginning of the sixth century.
At the other end of the scale, there is no need to reject the supposition of Abraham Weiss and others that earlier portions of anonymous material were used by the final editors, the Stamaim. This may well be the case but it was all apparently recast in a sustained framework; the final editors using this earlier material as they used the actual words of the Amoraim to produce carefully organised units. For instance, the opening passage of tractate Bava Kama (2a-5b) discusses the question of main categories of damage and their derivatives. The main categories are called avot (‘fathers’) and their derivatives toledot (‘offspring’). Here some of the toledot seem already to have been delineated and described in a language different from that used in the rest of the passage (mai hi as distinct from mai ninhu). Yet a careful study of the whole passage demonstrates that it has all been reworked so as to provide a sustained and contrived narrative.
This passage is quoted by Weiss in his Le-Heker ha-Talmud, pp. 98-100 and see the discussion in my Studies in Talmudic and Methodology, London, 1961, pp. 132-151, where I have taken issue with Weiss’s contention that we have here a double process of redaction. It is not without significance that the usual expression for the derivatives of contamination (tumah) mentioned in the sugya, is not toledot but rishon or sheni or vlad (see e.g. Mishnah Pesahim 1: 6). It would seem that in our passage the term toledot is used throughout so as to make it fit in with the other toledot (those of ‘damages’ and ‘Sabbath’) presented and discussed.
There is a good deal of material stemming from Palestine in the Babylonian Talmud. The leading early third century Palestinian teachers, R. Johanan and Resh Lakish, for instance, are mentioned very frequently in the Babylonian Talmud though they never visited Babylon. But this fact does not permit the conclusion that there are units of shaped Palestinian material in the Babylonian Talmud only that the opinions of the Palestinian teachers were used and recast by the Babylonian editors as they used the opinions of the Babylonian Amoraim for the same purpose of providing a sustained narrative. Even if the editors of the Babylonian Talmud had copies of the earlier Palestinian Talmud (a very dubious proposition in any event), they never refer to these. In fact it can be shown that occasionally, at least, the Babylonian editors were not averse to putting their own opinions into the mouths of the Palestinian teachers they quote, all for the purpose of creating the sugyot in their own contrived style.
In tractate Megillah 10b, R. Johanan is quoted as saying that when Israel crossed the Sea and the Egyptians were drowning God declared to the angels: ‘The work of My hands are drowning in the sea and you wish to sing’ i.e. God cannot rejoice and allow His angels to sing when His creatures, the Egyptians, are drowning. But, as J. Heinemann (Aggadot ve-Toledotehen, Jerusalem. 1974, pp. 175-179) has shown, in Palestinian sources R. Johanan refers not to the Egyptians but to the Israelites as ‘The work of My hands’. According to Heinemann, the Babylonian editors used the saying of R. Johanan in an Introduction to the Book of Esther (as the passage states) but, living in the lands where the events of the book took place, were more circumspect so that the ‘Babylonian’ R. Johanan is made to say that God is aggrieved by the destruction of Israel’s foes. Another example is the different treatment of a saying by the third century Palestinian teacher, R. Eleazar, in the Palestinian Talmud, Ketubot 1: 1 (24d) and in the Babylonian Talmud Ketubot 8b-9a. In the Palestinian version R. Eleazar says ‘She is forbidden to him’ whereas in the Babylonian version R. Eleazar says ‘He is believed’. This is because the Babylonian Talmud is concerned in the passage with the question of belief in the man’s testimony so that R. Eleazar is made to say what the particular sugya requires him to say. There has been much discussion on whether or not the editors of the Babylonian Talmud possessed a copy or copies of the Palestinian Talmud, see Y. J. Greenwald: Harau Mesaderey ha-Bavli et ha-Yerushalmi, New York, 1954; Strack/Stemberger, pp. 218-219; and J. N. Epstein: Mevuot, pp.290f. Epstein (p. 291) rightly concludes that there was no text of the Palestinian Talmud in the hands of the editors of the Babylonian Talmud and no earlier version of the text of the Babylonian Talmud in the hands of the editors of the Palestinian Talmud. For the nanotey (‘descenders’), the Palestinian scholars who conveyed the teachings of their native land to the Babylonians, see the Letter of Sherira Gaon, ed. Lewin, p. 61 and Strack/Stemberger, pp. 198 and 219.
A number of modern scholars have purported to detect in the Babylonian Talmud editorial work from different Babylonian centres and not only in different tractates but even in the same tractate, explaining in this way the existence of duplicate sugyot and especially contradictory sugyot. In many instances, however, it is possible to explain the duplicates and contradictory passages as due not to different editors but to the same editors shaping the material they had in different ways. The point to be noted in this connection is that the style of the framework is the same throughout and since there is no internal evidence of a redaction in different places the existence of duplicates and contradictions can adequately be explained without the hypothesis of editing in different centres. Attention has been called, however, to the different style in the framework in tractates Nedarim, Nazir, Me’ilah, Keritut and Temurah, which would seem to suggest that these tractates have been edited in different places. Yet the differences are slight, the use of some different words and the like, and, again this needs to be stressed, the overall style is the same for the framework of all the tractates in the Talmud, including these five. This very close similarity of style throughout has largely been overlooked in modern scholarship so that, at the most, even if different centres were at work there is a unifying pattern which they all followed.
The hypothesis of redaction in different centres has been advanced especially by J. N. Epstein: Mevuot and by N. Aminoah: The Redaction of the Tractate Kiddushin in the Babylonian Talmud (Heb.), Tel-Aviv, 1977. See for example Aminoah’s analysis (pp. 163-166) of the sugya in Kiddushin 50b-51a.The whole passage can be explained quite adequately on the grounds that the final editors put it all into its present form in a contrived manner and there is no need to invoke the notion of dual redaction. On the five different tractates see: Strack/Stemberger, pp. 213-2l5. In support of my contention that the differences can be explained without the hypothesis of redaction in different centres, Rosh ha-Shanah 34a should be compared with Avodah Zarah 43a-b and Sanhedrin 52b compared with Avodah Zarah 11 a (where the style in both is exactly the same but the material has been arranged so as to arrive at different conclusions). If different editorial hands were at work why should they have used exactly the same style? Cf. Sanhedrin 27a where two different versions have been woven into a single narrative and see Bava Kama 72a-b where the same basic material receives further elaboration. To be noted in this connection is the way the Talmud itself accounts for duplication e.g. in Kiddushin 50a. A good illustration of different editorial aims in the presentation of the same material is provided when the three passages in Berakhot 20a; Ta’anit 24a-b and Sanhedrin 106b are compared. In all three passages Rava seeks to account for the fact that while R. Judah’s prayers for rain were answered immediately his, Rava’s, prayers were not. It cannot be because R. Judah had greater proficiency in learning since Rava was convinced that his ability in learning was greater. In all three passages the material is presented in more or less the same way but in each Rava is made to give a different reply. In Berakhot R. Judah’s superiority is said to be because he exhibited a greater willingness to sacrifice his well-being for his religion. In Ta’anit Rava’s failure is said to be due to the unworthiness of his generation. In Sanhedrin Rava attributes his failure to the fact that the All-merciful desires the heart i.e. R. Judah was more sincere than Rava. It would be absurd to explain the difference on the grounds that the three passages were edited in three different centres. The differences are obviously due to the three different concerns in these passages. In the Berakhot passage the concern is with self-sacrifice, as the whole passage demonstrates. In Ta’anit the whole section, of which this is a part, deals with fasting when the rains fail to come and with the merit required both by the one who prays for the community and the community itself if the prayers are to be answered. In the Sanhedrin passage the theme is that of unworthy but learned men like Doag and Ahitofel, men who knew a good deal but had no ‘heart’ i.e. it does not follow that all learned men are sincere and Rava proves this from his own experience, believing that he was less sincere, though more learned, than R. Judah.
Even after the Savoraim, some few additions were made to the text of the Babylonian Talmud. Rav Hai Gaon (B. Lewi: Otzar ha-Geonim to Bava Kama, Jerusalem, 1943, p. 37) remarked long ago that here and there marginal notes were incorporated into the text and it is well-known that some of Rashi’s notes managed to slip into the text from the margins. But this is not really a matter of additions to the Talmud but belongs rather to the different question of the history of the Talmudic text.