Rabbinic Literature

The corpus, its genres, and the transmission of halakhah from the Talmud to the Shulchan Aruch

Working handout · compiled 8 September 2026
Claude Opus 5 guided by David Schapiro

1 · What counts as rabbinic literature

The term is ambiguous, and the ambiguity is not trivial — the boundary you choose determines the whole map.

In the narrow sense, sifrut Chazal means the corpus produced by the sages from roughly the destruction of the Second Temple to the closing of the Talmud: Mishnah, Tosefta, halakhic midrash, the two Talmuds, the classical aggadic midrashim. In the broad sense it continues through the Geonim, Rishonim and Acharonim to the present day. German-language scholarship (Strack–Stemberger, Einleitung in Talmud und Midrasch) uses the narrow sense; yeshiva usage and most English survey courses use the broad one. This handout treats the narrow corpus in sections 3–4 and the broad continuation in section 7.

What is not rabbinic literature, despite frequent conflation: Qumran, Philo, Josephus, the Apocrypha and Pseudepigrapha (Second Temple, not rabbinic); piyyut (liturgical poetry, an adjacent Palestinian development); medieval philosophy; and Kabbalah. The Zohar imitates midrashic form and Aramaic diction, which is precisely what makes its thirteenth-century Castilian provenance a matter of literary detection rather than genre classification.

2 · Periodisation

The periodisation itself rests on a primary source: the Iggeret of Rav Sherira Gaon (987 CE), a responsum to the community of Kairouan reconstructing the chain of transmission. Virtually every modern account of who the Savoraim were, and when the Talmud closed, leans on it. It is not a neutral chronicle — Sherira is defending Babylonian authority against Palestinian claims — and reading it as such is a standing methodological problem.

3 · The tannaitic corpus

The structural insight that makes this layer intelligible: the same body of law was recorded twice, in two incompatible arrangements. The Mishnah and Tosefta organise material topically — apodictic law, stated without scriptural anchor. The halakhic midrashim organise the same material sequentially, running verse by verse from Exodus 12 to the end of Deuteronomy. Neither is a commentary on the other; they are rival redactional strategies, and the tension between them — why does the Mishnah so rarely cite a verse? — is a founding question of the field.

• Mishnah — six orders (Zeraim, Moed, Nashim, Nezikin, Kodashim, Toharot), 63 tractates, in Mishnaic Hebrew, redacted by Rabbi Yehudah ha-Nasi around 200 CE. Whether he produced a written text or an authoritative oral recension is unresolved and consequential.
• Tosefta — follows the Mishnah's order tractate for tractate but runs roughly four times its length; traditionally ascribed to R. Chiyya and R. Hoshaya. Modern scholarship (Friedman, Hauptman) has argued that in many parallels the Tosefta preserves the earlier form and the Mishnah is the abridgement — the reverse of the traditional assumption.
• The two schools — David Zvi Hoffmann's 1887 thesis, still the working framework, divides the halakhic midrashim between the academy of R. Yishmael (Mekhilta de-R. Yishmael on Exodus, Sifrei Bamidbar, Midrash Tannaim on Deuteronomy; the thirteen middot; דיברה תורה כלשון בני אדם) and that of R. Akiva (Sifra on Leviticus, Sifrei Devarim, Mekhilta de-Rashbi, Sifrei Zuta; ribbui and mi'ut, every particle exegetically loaded). The division is terminological and attributional; it has been sharpened rather than overturned since.
• Baraitot — tannaitic traditions excluded from the Mishnah, surviving only as citations in the two Talmuds. Their reliability as genuinely tannaitic, especially in the Bavli, is a live source-critical question.

4 · The amoraic corpus

The amoraic period splits geographically before it splits generically: two academies, two languages, two Talmuds — and alongside them the aggadic midrash, which is Palestinian almost throughout.

• Yerushalmi — Galilean Aramaic and Hebrew, terse, elliptical, badly transmitted (essentially one complete manuscript, Leiden 1289; editio princeps Venice 1523). It covers Zeraim in full, which the Bavli does not, plus Moed, Nashim and Nezikin; it lacks Kodashim entirely, and of Toharot has only the opening chapters of Niddah, breaking off mid-sentence. Lieberman argued that the Nezikin section is an independent, earlier recension from Caesarea. Its neglect relative to the Bavli is a fact of reception, not of quality.
• Bavli — Eastern Aramaic, roughly 2,711 folios in the Vilna edition, gemara on some 37 tractates. The decisive modern insight (Halivni, Friedman) is stratigraphic: beneath the attributed amoraic statements lies an unattributed dialectical voice, the stam, which supplies almost all the argumentation, is later than the amoraim it arranges, and constitutes the bulk of the text. Once you see it, the Bavli stops looking like a transcript and starts looking like a composition.
• "Midrash Rabbah" is a bibliographic accident. It is ten unrelated works bound together by a Venice printer in 1545. Bereshit Rabbah and Vayikra Rabbah are fifth-century Palestinian; Shemot Rabbah is composite and partly eleventh-century; the first half of Bamidbar Rabbah is twelfth-century, from the school of Moshe ha-Darshan. Citing "Midrash Rabbah" as a single authority is a category error that survives in a great deal of homiletic writing.
• Structural principle of the homiletical midrashim — Pesikta de-Rav Kahana is arranged by the festival and special-Sabbath lectionary, not by the Torah's sequence. The calendar, not the scroll, supplies the order. Tanchuma-Yelammedenu is identified by its opening halakhic question, yelammedenu rabbenu.
• Targumim sit on the border: Onkelos, Pseudo-Jonathan, Neofiti, the Fragment Targums. Translation and midrash at once, usually treated as an adjacent corpus rather than a rabbinic genre proper.

5 · The anatomy of a daf

Nothing about the physical page is ancient. Its shape is a printer's decision of the 1520s that hardened into the way the text is thought about.

Daniel Bomberg's Venice edition (1520–1523), the first complete printed Talmud, fixed the foliation and the basic disposition of the page. The Romm brothers' Vilna edition of 1880–1886 added apparatus and became the standard, but it kept Bomberg's pagination. That is why a citation like Bava Metzia 59b works across every edition, why the Shas runs to roughly 2,711 folios, and why the daf — a leaf, with two sides, amud alef and amud beit — rather than the page is the unit of reference.

The invariant worth memorising is not left and right, which swap between the two sides of a leaf, but proximity to the spine: Rashi always occupies the column nearest the binding, Tosafot always the column farthest from it. The centre carries the Mishnah at the head of each chapter and then the Gemara, set in a distinctive square type, with Rashi's column in the semi-cursive font that printers later called "Rashi script" — a typeface he never saw, named after him because it was used to set his commentary.

The margins are a sixteenth-century Italian addition. Ein Mishpat, Mesorat ha-Shas and Torah Or are all the work of one man, R. Joshua Boaz ben Simon Baruch (d. 1557), and they do three different jobs: point from a talmudic ruling to where it lands in the codes, point to parallel discussions elsewhere in the Talmud, and identify the biblical verses being quoted. Later editions crowded in more — Hagahot ha-Bach (the same Yoel Sirkis who wrote on the Tur), Gilyon ha-Shas of R. Akiva Eiger, Rabbeinu Chananel, Hagahot ha-Rashash, and on many tractates newly recovered rishonim that the Romm press printed in the outer margins.

Two things follow that are worth saying to a class. First, the page is an argument about authority made in typography. Rashi and Tosafot get full columns; the codes get a thin marginal cross-reference and nothing more. The Rif, the Mordechai, the Ran and the Rosh are not on the daf at all — they are bound in at the back of the tractate. A student reading the page absorbs a hierarchy in which the Talmud and its immediate commentators are the text and the codes are an index to it. That is precisely the priority the Maharshal defended and Karo's method inverted.

Second, the numbering. Every tractate begins at daf 2, never 1, because the title leaf counts as folio 1 — a bibliographic accident that has attracted homiletic explanation ever since, the best-known being that one never finishes beginning to learn. Moderate confidence on the bibliographic account, high confidence that the homiletics came afterwards.

6 · Genres that cut across periods

Responsa are the largest of these by volume — hundreds of thousands of items — and by far the richest documentary source for the social and economic history of Jewish communities, which is precisely why they were the last to be systematically indexed.

The geonic bridge deserves naming separately: the She'iltot of Rav Achai, Halakhot Pesukot and Halakhot Gedolot (the first freestanding codes), Seder Rav Amram Gaon (the first complete prayer book), and the Iggeret of Rav Sherira Gaon.

7 · From the Talmud to the Shulchan Aruch

This is the stretch most often misrepresented, and the misrepresentation has a single shape: drawing it as a chain.

7.1 Two streams, braided twice

7.2 Three rival answers to "what should a code delete?"

The three great codes are not three stages of one project. The Talmud has four features a codifier can keep or discard, and each of the three made a different set of cuts.

Read that table and the standard story collapses. The Rambam's code is the most radical of the three, and it is not the ancestor of the Shulchan Aruch. The Tur is. Karo's book inherits the Tur's four divisions, the Tur's restriction to law operative in exile, and the Tur's chapter numbering. Which produces the most under-noticed consequence of the whole development: because the Tur dropped Temple and agricultural law, and the Shulchan Aruch followed it, the Mishneh Torah remains the only comprehensive code of halakhah ever written. Everything after it codifies the practicable remainder.

7.3 The detailed flow

The edges do not all mean the same thing:

• Talmud → Rif / Rambam / Rosh are three genuinely independent reworkings, not a sequence. The Rif and the Rosh both keep the Talmud's tractate order and abridge sugya by sugya; the Rosh's contribution is that he pours Tosafist material into Alfasi's format. The Rambam alone breaks the container.
• Rif → Ran and Mordechai is containment, not influence. The Ran's principal halakhic work is a commentary on the Rif; the Mordechai is arranged following the Rif's order. Both are printed on the Rif's pages in the standard Shas. They are not parallel authorities to Alfasi — they are growths on him, one Spanish and one Ashkenazi.
• Rif / Rambam / Rosh → Tur is not ordinary citation. The Tur's method is to state his father's ruling and then array the alternatives, so the three pillars are already the visible skeleton of the Tur before Karo ever names them. Karo did not choose that triad freely; the book he chose to comment on had largely chosen it for him.
• Tur → Beit Yosef is the load-bearing edge. Karo's project is source-critical: take every ruling in the Tur, trace it to its talmudic origin, follow it through the Rishonim, adjudicate. Thirty-two years, 1522 to 1542. Everything below is downstream of that labour.
• Tur → Darkhei Moshe → Mappah is the exact structural mirror, and it is almost never said out loud: Isserles wrote his own full commentary on the Tur, then abridged it into glosses, precisely as Karo wrote the Beit Yosef and abridged it into the Shulchan Aruch. The book on your desk is two commentaries on the Tur, each reduced to its skeleton, interleaved by a printer.

7.4 The Tur is the Rosh's son

Yaakov ben Asher was born in Cologne around 1269 and brought to Toledo as a young man when his father fled Germany. The relationship is structural, not merely biographical. Before writing the Turim, Yaakov produced the Kitzur Piskei ha-Rosh, an abridgement of his father's rulings, printed after each tractate in the standard Shas. He edited his father's legal corpus first, then built his own book on top of it. Inside the Tur, the Rosh's position is not one opinion among several — it is the default, stated first, the baseline against which Rif and Rambam are measured.

Run that forward. Karo takes the Tur as his base text and adopts its three-authority frame as his decision rule. The Ashkenazi vote in the majority calculus that produces the Shulchan Aruch therefore reaches Karo pre-weighted, through a son's editorial treatment of his own father.

The migration pattern doubles the irony. The Rosh's family fled Germany for Toledo in 1303–1305, carrying Franco-German learning into Spain. Karo's family fled Toledo in 1492. The Sephardi code rests on a Tur written by German refugees in Castile and was completed by a Castilian refugee in Ottoman Safed. The two streams in Fig. 4 are cleaner as a schema than they were as history: they were braided by expulsion, twice, in the same city.

He is called the Ba'al ha-Turim, and the gematria-laden Torah commentary printed under that name in most Chumashim is also his — the abridged introduction to a much longer and much more sober peshat commentary that almost nobody reads.

7.5 The mechanism of decision

Karo states his rule explicitly in the introduction to the Beit Yosef. He decides by majority among three pillars — Rif, Rambam, Rosh — where two outweigh one; where a question is not treated by all three, he turns to Ramban, Rashba, Ran, Mordechai and others; and he reserves the right to follow settled Sephardi practice where it is unambiguous. Two of his three pillars are Sephardi and the third is Ashkenazi-trained. The method is therefore not "Karo ruled what he thought right" but a declared, mechanical, majoritarian algorithm over a fixed canon — whose output is Sephardi by arithmetic rather than by intention.

Isserles operates on a different principle: hilkheta ke-batra'i, the law follows the later authorities, combined with the near-decisive weight he gives to established Ashkenazi minhag. Note what this means. The two halves of the book you hold are not two opinions within one system. They are two different theories of what makes a ruling binding — canonical majority versus local custom plus recency. Nobody reconciled them. A printer bound them together.

Two consequences worth putting in front of a class. An arithmetic one: because the Rif and the Rosh work in the same format on the same sugyot, they agree with each other far more often than either does with the Rambam, so a great many "majorities" are in practice Rif-plus-Rosh outvoting Maimonides. A jurisprudential one: this is a rule about counting authorities, not evaluating arguments. It is the opposite of how the Bavli itself decides anything, and that inversion is the real intellectual event between 1180 and 1565.

7.6 It was fiercely contested

• Maharshal (Shlomo Luria, d. 1573), in Yam shel Shlomo, refused the enterprise entirely: rulings must be argued back to the Talmud each time, not looked up.
• R. Chayim ben Betzalel, the Maharal's brother, wrote Vikkuach Mayim Chayim specifically against Isserles — the objection being that a young Polish scholar had no business fixing custom for communities not his own.
• The Maharal attacked codification as such: law severed from its reasoning produces decisors who cannot reason.
• Mordechai Yaffe's Levush was the deliberate middle path — code plus reasoning — and it lost comprehensively, on grounds that had little to do with its merits.

What settled the matter was print. Manuscript codes could not impose uniformity; a Venice or Krakow edition could. The Shulchan Aruch is the first halakhic work whose authority is partly a function of publishing technology.

7.7 Canonisation by annotation

• Sma (Yehoshua Falk, on Choshen Mishpat) and Bach (Yoel Sirkis, on the Tur) — the first generation, still arguing from the Tur outward rather than treating Karo as the base text.
• Taz (David ha-Levi Segal) and Shach (Shabtai ha-Kohen), both on Yoreh De'ah from 1646 — the decisive pair. Once two major commentaries take the Shulchan Aruch as the text worth commenting on, the question of whether it should have been written is over.
• Magen Avraham (Avraham Gombiner, published 1692) on Orach Chayim — dense, terse, and for two centuries the operative Ashkenazi authority on daily practice.
• Bi'ur ha-Gra (Eliyahu of Vilna) — the internal counterweight, and the refutation of the idea that a code overrides the Talmud. The Gra routes every ruling back to its talmudic source and rules against Karo wherever the sources demand it. A code that can be overturned by better argument is not what "code" means in a civil-law sense, and the comparison to Napoleonic codification — a commonplace of German-language treatments — breaks here.
• Mishnah Berurah (Yisrael Meir Kagan, 1884–1907) on Orach Chayim only, and Arukh ha-Shulchan (Yechiel Michel Epstein, roughly contemporaneous) on all four divisions. Opposite methods: the Mishnah Berurah adjudicates among later authorities and tends to stringency; the Arukh ha-Shulchan reconstructs each law from the Talmud forward and tends to validate existing practice. That the first won the Ashkenazi world and the second did not is a fact about twentieth-century yeshiva culture, not about their relative quality.

8 · Glossary of terms

Periods and persons

Chazal

Acronym for Chachameinu Zichronam Livrachah, "our sages of blessed memory" — the Tannaim and Amoraim collectively.

Tanna (pl. Tannaim)

Aramaic "repeater." Sages of c. 10–220 CE, quoted in the Mishnah, Tosefta and baraitot.

Amora (pl. Amoraim)

"Speaker, expounder." Sages of c. 220–500 CE, whose discussions form the gemara.

Savora (pl. Savoraim)

"Reasoner." Post-amoraic Babylonian sages, c. 500–650, credited with editorial work on the Bavli.

Stam / Stammaim

The anonymous dialectical voice of the Bavli, and (in Halivni's terminology) its authors. Later than the amoraim it arranges, and the bulk of the text.

Gaon (pl. Geonim)

Heads of the Babylonian academies of Sura and Pumbedita, c. 650–1050.

Rishonim

"The first ones." Authorities c. 1050–1500, from the Rif to the eve of the Shulchan Aruch.

Acharonim

"The later ones." Authorities from roughly the Shulchan Aruch onward.

Works and genres

Mishnah

The topically arranged code of tannaitic law, six orders, 63 tractates, redacted c. 200 CE.

Tosefta

"Addition." A parallel tannaitic collection following the Mishnah's order, roughly four times longer.

Baraita (pl. baraitot)

"Outside." A tannaitic tradition not included in the Mishnah, preserved as a citation in the Talmuds.

Gemara

The amoraic discussion built on the Mishnah. Mishnah + gemara = Talmud.

Sugya

A discrete talmudic discussion; the basic unit of study.

Daf

A folio (two sides, a and b) of the standard Vilna Talmud pagination.

Shas

Acronym for shishah sidrei, "six orders" — a standard name for the Talmud as a whole.

Midrash halakhah

Verse-by-verse legal exegesis of Exodus–Deuteronomy: Mekhilta, Sifra, Sifrei.

Midrash aggadah

Non-legal exegesis and homily: the Rabbah collections, Pesikta, Tanchuma.

Petichta

"Opening." A homiletical proem starting from a distant verse and arriving at the lectionary text.

Targum

Aramaic translation-cum-paraphrase of Scripture: Onkelos, Pseudo-Jonathan, Neofiti.

Chiddushim

"Novellae." Analytic reconstruction of a sugya rather than a running commentary.

She'elot u-Teshuvot (Shu"t)

Responsa; case law conducted by correspondence.

Nosei kelim

"Arms-bearers." The standard commentaries printed around a code — Shach, Taz, Magen Avraham and the rest.

Legal vocabulary

Halakhah / aggadah

Legal material / narrative, ethical and theological material.

Psak

A binding ruling; posek, one who issues it.

Minhag

Established custom, which in the Ashkenazi tradition can be decisive against a contrary textual ruling.

Klalei ha-psak

The meta-rules for deciding between disputing authorities.

Hilkheta ke-batra'i

"The law follows the later authorities." A post-talmudic decision rule, central to Isserles.

Lechatchilah / bedi'eved

Ab initio / after the fact — the standard two-tier structure of a ruling.

The four divisions of the Tur and Shulchan Aruch

9 · Who's who

Acronymic names, expanded. Dates are approximate where sources disagree.

Tannaim and Amoraim

Geonim

Rishonim — the Sephardi line

Rishonim — the Ashkenazi line

The early modern codifiers and their opponents

The nosei kelim

Modern scholars cited

10 · Confidence notes and further reading

Confidence

• High — the periodisation; the structure and dating of the Mishnah and the two Talmuds; the Tur's structural paternity of the Shulchan Aruch; Karo's three-pillar rule, which is explicit in his own introduction; the Ran's dates and his dependence on the Rif; the Darkhei Moshe / Mappah parallel to Beit Yosef / Shulchan Aruch; the existence and character of the opposition to codification; the role of print.
• Moderate — the school-attribution of the halakhic midrashim (Hoffmann's framework holds but has been complicated); the Rif-plus-Rosh coalition pattern, which is widely observed and easy to demonstrate on samples but which I know of no systematic count for; publication years for the Mappah (Krakow 1578–1580), the Taz and Shach (both 1646) and the Magen Avraham (1692), where editions and reprints muddy the record; the Rosh's dates in Toledo; the Tur's completion c. 1340.
• Low — redaction dates for Shemot Rabbah, Devarim Rabbah, Tanchuma and Pirkei de-Rabbi Eliezer, each disputed by a century or more; the dating of Mekhilta de-Rashbi and Sifrei Zuta, reconstructed largely from Genizah fragments and later citations; how much chronological distance separates the stam from the latest named amoraim; any confident account of why the Levush failed.

Verify all publication years against a catalogue before printing them on a source sheet for students.

Further reading

• H. L. Strack and Günter Stemberger, Einleitung in Talmud und Midrasch (C. H. Beck) — the reference that will be assumed in a German faculty.
• Chanoch Albeck, Mavo la-Talmudim — for the two Talmuds.
• David Weiss Halivni, The Formation of the Babylonian Talmud — for the stam.
• Yaakov Sussman, on the orality of the Oral Torah — for the transmission question.
• Menachem Elon, Ha-Mishpat ha-Ivri / Jewish Law: History, Sources, Principles — for the codificatory chain treated as legal history.

A teaching exercise

The sharpest single exercise for section 7: put Rambam, Hilkhot Shabbat 1:1 beside Tur and Shulchan Aruch, Orach Chayim 242 and let the room notice that one text tells you nothing about where the law came from and the other tells you almost nothing else.