How Should Rabbinic Literature Be Read in the Modern World? 9781463211028

Through literary, historical, archaeological, and engendered readings, this collection of essays presents a multidiscipl

187 23 2MB

English Pages 226 Year 2006

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

How Should Rabbinic Literature Be Read in the Modern World?
 9781463211028

Citation preview

How Should Rabbinic Literature Be Read in the Modern World?

JUDAISM IN CONTEXT Volume 4

 Series Editors Naomi Koltun-Fromm, Lieve M. Teugels and Rivka Ulmer

How Should Rabbinic Literature Be Read in the Modern World?

EDITED BY MATTHEW KRAUS

GORGIAS PRESS 2006

First Gorgias Press Edition, 2006. Copyright © 2006 by Gorgias Press LLC. All rights reserved under International and Pan-American Copyright Conventions. Published in the United States of America by Gorgias Press LLC, New Jersey. ISBN 1-59333-214-9

GORGIAS PRESS 46 Orris Ave., Piscataway, NJ 08854 USA www.gorgiaspress.com Printed and bound in the United States of America.

Library of Congress Cataloging-in-Publication Data How should rabbinic literature be read in the modern world? / edited by Matthew Kraus. p. cm. -- (Judaism in context ; v. 4) Papers presented at a conference held Oct. 26-27, 2003 at Williams College entitled How Should Late Rabbinic Literature be Read in the Modern World? : Hermeneutical Limits and Possibilities. Includes bibliographical references. ISBN 1-59333-214-9 1. Rabbinical literature--History and criticism--Theory, etc.-Congresses. 2. Talmud--Hermeneutics--Congresses. 3. Aggada-History and criticism--Theory, etc.--Congresses. 4. Midrash--History and criticism--Theory, etc.--Congresses. I. Kraus, Matthew. II. Series. BM496.6.T49 2006 296.1'20601--dc22 2006011697

ACKNOWLEDGEMENTS

Matthew Kraus University of Cincinnati Acknowledgement of the polyvalence of rabbinic texts has correspondingly stimulated the application of numerous and varied methods in order to comprehend such a complex corpus of literature. Differences concerning the particular interpretation of rabbinic texts now must be viewed against differences in the interpretative framework applied to these texts. On October 26-27, 2003 a conference was held at Williams College which sought to address the implications of an emerging absence of a methodological consensus. Entitled, “How Should Late Antique Rabbinic Literature Be Read in the Modern World: Hermeneutical Limits and Possibilities,” the conference brought together scholars who approach rabbinic materials from several disciplines—history, literary studies, archaeology, gender-studies, and religious studies. The conference represented a modest attempt to move beyond the coexistence of various approaches to rabbinic literature to interactive dialogue between them. Each participant submitted a paper in advance that was circulated among the participants. Once we were gathered in the same room, each participant submitted to the thoughtful and trenchant comments of his/her colleagues. Occasionally, the discussions exposed the wide gaps generated by methodological difference. Surprisingly, however, the frequent moments of true dialogue underscored the sense of common ground and mutual interest. The subsequent conversations stimulated revisions of these essays (Charlotte Fonrobert, unfortunately, was unable to attend the conference because of an emergency, but we discussed her paper in absentia and she made revisions in response to the comments from the group). Thus, this collection of papers represents the impact of dialogue between those with methodological differences. In a sense, then, this book invites the reader to listen to this v

vi

HOW SHOULD RABBINIC LITERATURE BE READ?

dialogue and experience both the consonant and dissonant interaction between different approaches to interpreting rabbinic literature. Such a project would not be possible without generous contributions from numerous sources. I am especially grateful to Williams College, The Bronfman Fund for Jewish Studies, and the Oakley Center for the Study of the Humanities and Social Sciences for supporting and providing space for the Conference. Williams’ Dean of Faculty, Tom Kohut, and The Oakley Center’s director, Robert Kavanaugh, offered wonderful encouragement and advice. A special thanks goes to Marc Lee Raphael whose kindness and vision greatly facilitated this project. The people at Gorgias Press, George Kiraz, Lieve Teugels, and Ramakrishna Rithvik have been indispensable as they graciously and expertly turned a collection of conference papers into a book. The greatest credit, however, belongs to the participants in the conference whose seriousness, congeniality, and openness made real dialogue possible. Their patience and sage advice have nurtured this project at every stage. Matthew Kraus University of Cincinnati

CONTENTS Introduction (Richard S. Sarason) ...............................................1 What Difference Does the “Orality” of Rabbinic Writing Make for The Interpretation of Rabbinic Writings? (Martin S. Jaffee).............................................................................................11 A Bavli Sugya and Its Two Yerushalmi Parallels: Issues of Literary Relationship and Redaction (Alyssa Gray) ................35 The Semiotics of the Sexed Body in Early Halakhic Discourse (Charlotte E. Fonrobert).............................................................79 Closing the Circle: Yonah Fraenkel, the Talmudic Story, and Rabbinic History (Hillel I. Newman)..................................... 105 Context and Genre: Elements of a Literary Approach to the Rabbinic Narrative (Jeffrey L. Rubenstein)........................... 137 Ephraim E. Urbach and the Study of Judeo-Christian Dialogue in Late Antiquity Ȧ Some Preliminary Observations (Oded I. Irshai).......................................................................... 167 Archaeology and the Interpretation of Rabbinic Literature: Some Thoughts (Steven Fine)................................................. 199

vii

INTRODUCTION Richard S. Sarason Hebrew Union College/Jewish Institute of Religion, Cincinnati The academic study of classical rabbinic literature (dating from roughly the third through the early seventh centuries CE), in its literary, historical, cultural, and religious dimensions, has grown exponentially over the past generation. Many more scholars in North America, western Europe, and Israel have pursued this study than was previously the case, signaling a growing recognition in the academy of the central importance of this literature, once deemed foreign and not just a bit exotic, for a proper understanding both of the world of late antiquity generally and of the history and phenomenology of Judaism in all subsequent periods. Firmly established now in North American universities (as well as, of course, in modern rabbinical schools and colleges of higher Jewish learning), the academic study of classical rabbinic Judaism and its literature ultimately has benefited from exposure to, and participation in, the intellectual crosscurrents of the contemporary academy in the humanities and social sciences. Multiple analytical perspectives (including ongoing developments in literary theory, historiography, anthropology, and religious and cultural studies) continue to cast new light on an old literature and to suggest additional facets and broader connections. In general, contemporary scholarship in rabbinic literature tends to acknowledge, more than did its nineteenth- and early twentieth-century predecessors, both the complexity, multivocality, and multidimensionality of this literature—the thing observed—as well as the situatedness and cultural baggage of the scholarly observer. Neither the literary medium nor the act of reading can be viewed as neutral. This recognition correspondingly suggests the 1

2

HOW SHOULD RABBINIC LITERATURE BE READ?

appropriateness of methodological self-consciousness on the scholar’s part, and a more tentative and humble stance in proposing explanations and interpretations. Indeed, “interpretation” here is the crucial concept, since all reading is now recognized to be an act of interpretation. While the goal of reading sine ira et studio generally remains in place, it is acknowledged that, in the study of human cultural products, there is no Archimedean point that will yield an objective view of the whole. Particularly is this the case in view of the daunting complexity of classical rabbinic literature, which must now be spelled out briefly. To begin with, the corpus of texts is the dense written precipitate of a much wider-ranging cultural activity: the generation, transmission, study, performance, and creative reworking of an immense body of traditions that originated in the social context of rabbinic study circles and study-houses in late antique and Byzantine Palestine and Sassanian Babylonia. The literature is collective and multivocal, rather than individually authored; it is thus rife with multiple and often conflicting perspectives. Recent scholarship has demonstrated time and again the crucial and creative role of redactors in the written deployment of these materials—arranging, shaping, juxtaposing, recontextualizing, reconfiguring, and ultimately reinterpreting them in the process. To complicate matters, the same traditions recur in slightly or grossly different forms throughout the corpus of texts. This “synoptic problem,” as we shall see in this volume, has been subject to a variety of interpretations, ranging from literary dependence to performative oral and/or written reenactment in different contexts. Historians, too, often have been stymied about the proper use of these materials for historical reconstruction. Whereas earlier scholarship tended rather naively to view rabbinic narratives as basically reliable (discounting, of course, its miracle stories and obvious exaggerations), more recent scholarship has shown over and over that rabbinic narratives are imaginative and didactic rather than journalistically reportorial, and that all rabbinic discourse is ideologically fraught. Indeed, the concept of culturally constructed “discourse(s)” that must be “deconstructed” to reveal their import is a hallmark of postmodern scholarship. Thus, there has been a turn in recent years to the examination of the cultural presuppositions embedded in rabbinic literature, and to its mythic worldview. Additionally, the advent of gender studies in the wake of contem-

INTRODUCTION

3

porary feminism has generated still further perspectives on the literature and its culture. The present volume of studies originates in a 2003 conference at Williams College addressing the topic, “How Should Late Antique Rabbinic Literature Be Read in the Modern World?: Hermeneutical Limits and Possibilities.” Reflected in this title is the recognition of difference and distance between late antiquity and modernity (or post-modernity), hence the invocation of hermeneutical limits to preclude unselfconscious anachronism. At the same time, the broadening of hermeneutical horizons in the contemporary academy, and the self-critical refinement of older methods, suggests the possibility of novel hermeneutical perspectives today that will enrich our understanding of these materials. In the remainder of this introduction, I will attempt to show how each of the studies below contributes to the ongoing scholarly colloquy on the interpretation of classical rabbinic texts, and how each may be seen to interact with the others. Martin S. Jaffee, University of Washington, has been in the forefront of contemporary studies on the oral component of rabbinic literature. While scholars have always paid lip service to this component (unavoidably, since the literature itself depicts the activity of the study house as involving recitation, memorization, and oral discourse about transmitted materials), nineteenth- and much twentieth-century scholarship, using philological methods developed in the field of classics, treated the materials textually. This meant not only examining manuscript traditions to account for written corruptions, glosses, and emendations over time, but also attempting to reconstruct an Urtext, that is, presuming the existence of a single original verbatim text-type from which all others flowed and/or diverged. Most scholars now view this quest as based on an anachronistic understanding of textuality in the ancient world, which was never that fixed. Similarly, earlier scholarship mostly viewed the synoptic relationships among rabbinic texts as a matter of citation and commentary or elaboration, of the familiarity of one text with another or with their common source. Jaffee suggests that some of these parallels instead reflect “alternative verbal performances” of the underlying traditions—longer or shorter versions generated out of mnemonic cues and stock phrases. In previous publications, Jaffee has proposed more and more refined models of how the oral and written registers interact in sev-

4

HOW SHOULD RABBINIC LITERATURE BE READ?

eral early rabbinic texts. Here he deals with the hermeneutical import of these models. In order to grasp fully both the meaning and the significance of early rabbinic texts and their synoptic relations to each other, he maintains, one must properly grasp the nature of the discourse that they embody. That discourse is fluid and dynamic, both creative and re-creative, making use of “a continuously elaborated thesaurus of mnemonic cues that encourage ever-new textual formulations.” A better understanding of the phenomenology of the texts and the nature of their production facilitates a better understanding of their content and meaning. Different notions of textuality, then, bear significant consequences for textual interpretation. Alyssa Gray, Hebrew Union College/Jewish Institute of Religion, has written at length about the synoptic relationships between the two Talmuds. Here, partially in methodological dialogue with the work of Martin Jaffee, she examines the relationship between a sugya (that is, the talmudic discussion of a unit of the Mishnah) in the Babylonian Talmud and its two (earlier) Palestinian parallels: the corresponding sugya in the Palestinian Talmud and an extensive parallel in the contemporary midrashic compilation Genesis Rabbah. Working with a high degree of articulate methodological selfawareness, Gray is able to show, according to a variety of wellreasoned criteria, that the Babylonian editor was familiar with the general outlines and some of the specific components of the Palestinian sugya and consciously reworked them. She also maintains that, between the two Palestinian versions of the sugya, the one in Genesis Rabbah is earlier. By referring to the type and distribution of differences between these two versions, she is able to claim that they do not simply represent two different oral performances of the same materials, but that the version in the Palestinian Talmud is a conscious improvement of that in Genesis Rabbah. Hence (and Jaffee would not disagree), the model of orality will not account for all kinds of variations among parallel texts; some indeed are better accounted for on the theory of diachronic development. At the same time, the Babylonian version of the sugya seems to display an awareness of both Palestinian versions and is “an interesting hybrid of the two.” Gray’s analysis also points to fluidity in the transmission of rabbinic materials, the re-presentation of these materials within a fixed structure when they move, in this case, from the Palestinian to the Babylonian Talmuds, and the conscious adaptation

INTRODUCTION

5

of the Palestinian materials in the Babylonian cultural context. Taken together, the papers by Jaffee and Gray demonstrate the complex layering of this literature and the complicated peregrinations of its constituent traditions, which resist unidimensional explanations. Charlotte Fonrobert, Stanford University, has written extensively about the construction of gender and the framing of genderrelated issues within the cultural discourse(s) of classical rabbinic literature. Gender in these texts, she maintains, must be thought of “in terms of a hermeneutical category.” Here she examines the rhetoric of gender duality, or “dimorphism” (that is, the determination that all sexual traits and behaviors must be construed as either “male” or “female,” with no continuum or middle ground between the two) in early rabbinic legal literature. She shows that the rabbinic treatment of middle cases (androgyny, sterility) in fact serves to bolster the norm of dimorphism, because in each case reproductive capacity rather than gender or hybridity is the ultimate focus of the Rabbis’ interest. Her contribution to this volume thus moves beyond the literary-constructive issues of the first two papers to deal with “how rabbinic texts have given and continue to give shape to Jewish culture,” how they “labor at constructing and shaping realities.” Fonrobert also insists that rabbinic cultural constructions and discourse must be viewed not as complete in themselves but always as taking place “within a complicated net of relationships to the cultural world that surrounds them,” such that “our thinking about rabbinic gender [constructs] gains in depth when rabbinic discourse is analyzed within this complex net of relationships.” For this reason, she examines intertexts in Greco-Roman culture (Aristotle, Second Sophistic, Stoics, Dio Chrysostom, Clement of Alexandria), pointing out similarities to, and differences from, rabbinic discourse. She suggests, in particular, that the rabbinic halakhic discourse about the androgyne echoes and recontextualizes that of the GrecoRoman physiognomic literature. The phenomenon of echoing and creative recontextualization is addressed in all of the studies in this volume, whether it be within the framework of the rabbinic textual corpus or within the enveloping framework of the intersecting cultures of late antiquity. Thus the texts and the discourses they reflect are never to be understood as static or hermetic.

6

HOW SHOULD RABBINIC LITERATURE BE READ?

The next two studies, like the first two, overlap and partially address each other. Here, the overriding issue is the appropriate framework within which to read rabbinic narratives: since “all texts exist in multiple contexts,” how do we determine which contexts are the most heremeneutically relevant or fruitful? What should constitute the “boundaries” of our reading? Hillel I. Newman, University of Haifa, was trained as a historian and takes up again the thorny issue of “literary” vs. “historical” readings of rabbinic narratives, aiming to reassert the validity of (methodologically self-conscious and careful) historical readings in the wake of the profound methodological skepticism of both Jacob Neusner and, addressed in this study, Jonah Fraenkel. These scholars, proceeding from rather different perspectives, have rejected the positivistic historicism of earlier scholarship as a fundamental misreading of the evidence: rabbinic narratives should be understood not as factual reportage but as didactic, imaginative, mythic, and perspectival religious-cultural constructions. They yield direct evidence about rabbinic culture and Weltanschauung rather than about historical events and persons. While Neusner’s approach is informed by the perspectives of academic religious studies, Fraenkel’s is informed primarily by literary criticism and romantic aesthetics. Influenced particularly by the school of New Criticism, which insists that literary texts be read as self-contained entities autonomous of the circumstances of their creation, Fraenkel reads rabbinic aggadic (that is, non-legal) narratives as “closed texts” that contain within themselves all the information and hermeneutical cues necessary for their proper interpretation. Newman regards this reliance on a single, somewhat dogmatic, hermeneutical approach as necessarily constraining and shows how alternative, contextualizing readings can enrich our understanding of these narratives. “Reading a story both in and out of isolation is a necessary exercise, which may just as easily yield complementary results as contradictory ones, all of which must be weighed against each other.” At a broader theoretical level, we may ask whether any single hermeneutical approach to these texts can claim to provide more than a partial insight—a perspective—into their dense web of significations. Newman, like Fonrobert, affirms the importance of comparative readings with external materials from the surrounding cultures; even if direct borrowing cannot be demonstrated, these external parallels still shed light on what was “in the air.” He asserts

INTRODUCTION

7

that “historical kernels” often can be found in rabbinic narratives, but concedes that any new historicism must take seriously the literary and anthropological critiques: “We are compelled in the circumstances to accustom ourselves to more elastic locutions which reflect in each case the range of conviction and contestability.” Newman thus shows how literary and historical concerns can, and must, be mutually informing. Jeffrey L. Rubenstein, New York University, has written extensively about rabbinic narratives, particularly those in the Babylonian Talmud, and here carries forward, quite independently, some of the threads of Newman’s essay. Even within the framework of a literary analysis of talmudic narratives, he asserts, the redactional context of the narrative must bear major hermeneutical weight. Like Alyssa Gray, Rubenstein underscores the creative activity of redactors, not only in deploying and juxtaposing, but also in reworking and effectively remaking, inherited traditions. Like Hillel Newman, he singles out the interpretive approach of Jonah Fraenkel as too arbitrarily restrictive—here, in ignoring the formative aspect of proximate redactional context (that is, within the bounds of the sugya) on individual talmudic narratives. But he also rejects the opposite approach, as illustrated in the textual readings of Yehudah Liebes, which synthetically views the entire set of talmudic traditions about, e.g., a particular Rabbi as relevant to the interpretation of each individual tradition irrespective of the deployment, rhetorical function, and contextual shaping of that tradition within a particular sugya. Rubenstein is also critical of other approaches that effectively decontextualize individual narratives. He notes that an excessive preoccupation with the genre of a particular story (common to both the folkloric and comparativist approaches) tends to focus overmuch on the story’s typical characteristics at the expense of its specific rhetorical functions in context. For Rubenstein, the sugya is the basic context of interpretation, within which individual traditions and narratives are to be construed. But he is also flexible in applying this insight: issues of context, genre, and theory of redaction should be treated as mutually informing and implicative. What is to be avoided, rather, is the totalization of any single perspective, best understood as providing only partial insight into complex textual phenomena.

8

HOW SHOULD RABBINIC LITERATURE BE READ?

Oded Irshai, Hebrew University of Jerusalem, like Hillel Newman, is a historian who works particularly on the history of Jews and Judaism in the Byzantine period, and the interaction between Judaism and Christianity at that time. Here he deals with the ways in which contemporary commitments can influence scholarly readings of the past. He contrasts the readings of Ephraim E. Urbach and Yitshak Fritz Baer, two prominent Israeli scholars of the previous generation, on the nature and degree of Jewish-Christian interaction that can be discerned from classical rabbinic literature. Urbach, a student of rabbinic literature whose textual acumen was formed in the traditional eastern European yeshiva and the philological-historically oriented Jüdisch-Theologisches Seminar in Breslau, minimized the impact of Greco-Roman and Byzantine Christian culture on rabbinic thought. The development of rabbinic religious culture and thought was to be viewed primarily in immanentist terms, as an internal Jewish phenomenon. By contrast, the historian Yitshak Fritz Baer, Urbach’s colleague and disputant at the Hebrew University whose German-Jewish upbringing had exposed him thoroughly to Christian culture and contemporary JewishChristian interaction, tended to retroject that high degree of interactivity into the medieval, Byzantine, and late antique periods. Irshai notes that this tendency in Baer’s writing virtually reversed itself in the late 1940’s, in the wake of the Holocaust. It is hardly coincidental that these two approaches to the interaction of Jewish religious culture with its environment in late antiquity and the medieval world precisely recapitulated the then-contemporary debate within cultural Zionist circles over the degree of autonomy of Jewish culture with regard to its cultural surroundings generally, with normative consequences for the developing Jewish national culture in Palestine (later, Israel). Irshai illustrates an extreme form of an ongoing interpretive crux: how does one discern, envision and represent complex, multidimensional religious-cultural influences, interactions, and interminglings—both witting and unwitting—when the evidence remains partial and sometimes one-sided? How does one avoid anachronism (let alone apologetics) when envisioning past worlds? These issues come to the fore as we widen the focus and expand the contexts in which rabbinic literature is to be read. The final essay in this volume introduces such an expanded context in regard to the material evidence from late antiquity un-

INTRODUCTION

9

covered over the past two centuries by near eastern archaeologists. Steven Fine, University of Cincinnati and Yeshiva University, is one of the few contemporary scholars of rabbinic literature and culture who has an expert mastery of this archaeological evidence and who is equally at home in both textual and material disciplines. Here he lists and illustrates those specific areas in which archaeology has contributed to the interpretation of rabbinic literature. These include the physical identification of artifacts and other realia mentioned in the literature; the linguistic contributions of the inscriptions and other epigraphy; and evidence for the physical setting of the social activities that generated much of the literature (remains of study houses and synagogues). Fine also stresses the interpretive significance of the illustrations of biblical stories and motifs found on the floor mosaics and painted walls of synagogues in the Land of Israel and in diaspora sites such as Dura Europos. These artistic interpretations constitute important and suggestive intertexts with those found in the midrashic literary corpora. Fine is equally quick to point out the difficulties in interpreting non-verbal representations, which often are not so specific in their association as written texts and should not be interpreted in a monistic fashion. (The latter cautionary remark in fact applies as well to the interpretation of the written texts, as virtually every essay in this volume makes clear.) “The essential project, then,” writes Fine, “is building the hermeneutical bridges between archaeology and rabbinic texts that allow for their creative interaction without allowing either type of source to dominate the other.” “Building hermeneutical bridges” indeed might be viewed as one of the major themes of this volume. The cautious, methodologically self-conscious juxtaposition and interweaving of multiple textual loci; of texts and a variety of contexts; of literary, historical, and religious-cultural perspectives and methodologies—all provide the contemporary scholar with fruitful lenses for the interpretation of what more and more is understood to be a dense, richly layered, multiform, and overdetermined (in the Freudian sense of being generated by multiple causal factors) literary corpus bearing witness to a complex and dynamic culture that produced and lies behind it. Under these circumstances, no single reading or interpretive lens will suffice to do justice to this rich complexity. As when attempting to view the multiple facets of a jewel, it is only by “turning it and turning it again” that we can register the cumulative impact of

10

HOW SHOULD RABBINIC LITERATURE BE READ?

the multiple partial insights provided by each of these varied perspectives.

WHAT DIFFERENCE DOES THE “ORALITY” OF RABBINIC WRITING MAKE FOR THE INTERPRETATION OF RABBINIC WRITINGS? Martin S. Jaffee University of Washington The title of this essay is more or less a quote from a question posed to me by a colleague at a conference sometime around 1995. At the time I had been some 5 years into the line of thought that led to my recent book on rabbinic oral tradition.1 That book, I am afraid, does not really answer—or even properly raise—the hermeneutical query that troubled my colleague. There I was more concerned to illustrate the oral dimension of rabbinic text-making, rather than to show how that orality might affect our interpretations of textual meaning. Here, by contrast, I hope at last to begin a reply to my colleague’s rather gentle insinuation that, once all we have to say about rabbinic orality is said and done, our business of interpreting rabbinic literature will be rather unaffected. After all, the material substratum of our interpretive work remains the printed rabbinic text, more or less well edited; or, in the best of all possible worlds, the manuscript copies (or their photo reproductions) that survived the Middle Ages. We may indeed feel confident that a living tradition of oral transmission stands at some greater or lesser distance behind such written texts—but we must ask: in what ways might sensitivity to this oral dimension of the text affect the way in which scholars approach the task of reading them?

1

M. Jaffee, Torah in the Mouth: Writing and Oral Tradition in Palestinian Judaism, 200 BCE-400 CE (Oxford: 2001).

11

HOW SHOULD RABBINIC LITERATURE BE READ?

12

The question is complicated by the fact that, in the study of the rabbinic literary corpus, “orality” is not a “one size fits all” proposition. The oralist study of rabbinic literature has not yet established a typological range of “oral text-types” against which we might plot the different rabbinic compilations and genres on a range from “high orality” to “residual orality” to “writing that imitates orality.”2 There are as yet no firm criteria by which to answer the question of whether, say, the Mishnah exhibits a greater degree of dependence upon an “oral-performative background” than the Tosefta. 3 And the issue only grows more difficult as the generic traits of the rabbinic literature itself increase in complexity. We have little more than impressionistic judgments about the “oral background” that supports the written materials we find in, for example, the various sorts of midrashic responses to Scripture, the amoraic memrot that clearly circulated in close proximity to tannaitic traditions, or the extraordinary orchestration of diverse textual layers and editorial voices characteristic of the full-blown talmudic sugya.4 2

An excellent start has been made on this issue in biblical studies in the work of S. Niditch, Oral World and Written Word: Ancient Israelite Literature (Westminster John Knox: 1996), pp. 8-38. See also J. Foley’s model of the “spectrum of oral poetry,” which includes texts classified, on the basis of internal formulary criteria as well as empirical criteria, on a continuum from “oral performance” (orally composed and transmitted), “voiced texts” (which may begin in writing but issue into oral performance), “voices from the past” (which may have begun in orality but survive now primarily in written exemplars), and “written oral poems.” A full discussion of this schema is in J. Foley, How to Read an Oral Poem (University of Illinois: 2002). 3 Early in his career, J. Neusner was rather more confident about his ability to distinguish the Mishnah as a text formulated for oral preservation from the Tosefta as a text formulated with written transmission in mind. See, for example, “Oral Tradition and Oral Torah: Defining the Problematic,” in F. Talmadge, ed., Studies in Jewish Folklore (Association for Jewish Studies: 1980), p. 257. 4 See, for example, my own speculative discussion based upon circumstantial comparison with the use of writing in Greco-Roman rhetorical education, in Jaffee, Torah in the Mouth, pp. 100-140. For a different angle on the impact of Byzantine literary culture on the composition of rabbinic texts, see P. Mandel, “Between Byzantium and Islam: The Transmission of a Jewish Book in the Byzantine and Early Islamic Periods,” in Y. El-

“ORALITY” OF RABBINIC WRITING

13

In what follows I make no pretense to solve any of these broad typological problems. I hope only to make a modest contribution towards illustrating how such typological studies might procede with regard to the most basic question at hand. That is, I attempt to develop a strategy for presenting and commenting upon rabbinic writing so as to highlight the interplay within rabbinic compositions of “traditional” textual material that bears signs of oral-performative transmission and the more individualized “authorial” compositional work of scribal composers who, as participants themselves in the transmission of orally-managed textual traditions, see little distinction between the work of transcription and that of composition and editing. I hope to show how interpretive sensitivity to the interplay of collective transmission and individual redaction might help us to appreciate not only the kind of “orality” exemplified in rabbinic writing, but also they way this type of “orality” stands behind characteristic modes of rabbinic thought and reflection. I am currently conducting some experiments along these lines at various sites in the rabbinic corpus at which tractates of the Mishnah intersect with comparable material in the Tosefta and the Talmuds. For the present purpose I wish to offer an example from perhaps the most well-known work of rabbinic literature, the tractate Avot, which has since medieval times circulated in manuscripts and printed editions of the Mishnah. For an oralist, in particular, it is impossible to study Avot without paying close attention to its closest rabbinic cousin, the compilation of Avot-like traditions known as Avot d’Rabbi Natan (hereafter, ARN). I do not try to determine here whether, in its relationship to the Mishnah, ARN is more like a “tosefta” or a “talmud.” 5 Nor am I particularly conman and I. Gershoni, eds., Transmitting Jewish Traditions: Orality, Textuality & Cultural Diffusion (Yale University: 2000), pp. 74-106. Although an earlier version of the material in Torah in the Mouth appeared in the Elman/Gershoni volume, I did not see Mandel’s essay until my own volume was in press. 5 An excellent, comprehensive discussion of the textual tradition of ARN, as well as its relationship to Mishnah Avot is now available from M. Kister, Studies in Avot de-Rabbi Nathan: Text, Redaction, and Interpretation (Hebrew; Hebrew University/Yad Itzhak ben Tzvi: 1998). Despite some no-longer plausible arguments to root the texts of Mishnah Avot and ARN in the Hasmonean period, L. Finkelstein’s Hebrew study, Introduction

14

HOW SHOULD RABBINIC LITERATURE BE READ?

cerned with the question of whether materials in ARN are “commentaries” on those in the Mishnah. Indeed, I suspect that the very term “commentary,” which has such powerful resonances to classical writing practices, may be rather misleading as a key to tracking the oral dimension of rabbinic interpretive texts. Rather, I hope to reconceive the intersections and interpenetrations of parallel textual transmissions as signs that point not so much to textual “commentary” or “editing” but to textual “performance.” That is to say: in considering parallels within and among diverse compilations in the rabbinic corpus, I hope to emphasize dimensions of the texts that point its contemporary readers towards the social world of interpretive textual performance as the matrix of textual development and, indeed, the fundamental purpose of rabbinic text-making.6 So, ultimately, I hope to illustrate some hermeneutical pay-off above and beyond the technical question of how, within a single rabbinic composition, it is possible to distinguish various sorts of transmissional modes. But first, by way of introduction, let me offer some thoughts on the general question of how we might construe the “orality” of the rabbinic text-making process for the purpose of shaping our own interpretive discourses.

WHAT IS THE “ORALITY” OF RABBINIC WRITING? The question of the orality of rabbinic writings is important because it forces us to realize that we are not quite sure what the making of a rabbinic text represents as a literary act. And if we are uncertain about what written rabbinic texts are meant by their creato the Treatises ‘Abot and ‘Abot of Rabbi Nathan (Jewish Theological Seminary of America: 1950), remains useful in discussing the complex relationships of the Mishnah and the two recensions of ARN. Unfortunately for the present study, Finkelstein offers no comments on the specific texts discussed below. For recent revisionist discussions of the relationship between the Mishnah and Tosefta, both as redacted texts and in terms of their constituent materials, see the essays collected in H. Fox and T. Meacham, eds., Introducing Tosefta: Textual, Intratextual, and Intertextual Studies (KTAV: 1999) and the recent study of S. Friedman, Tosefta Atiqta : Pesah Rishon. Synoptic Parallels of Mishna and Tosefta Analyzed with a Methodological Introduction (Hebrew; Bar-Ilan University: 2002). 6 I offer a richer theoretical exploration of oral text-interpretive tradition in early Judaism, including the formative centuries of the rabbinic movement, in Jaffee, Torah in the Mouth, pp. 9-10.

“ORALITY” OF RABBINIC WRITING

15

tors to be doing as written vehicles for meaning, we remain as well unclear how to read them as literary artifacts. Few academic scholars, after all, want to argue that rabbinic manuscripts represent stenographic reproductions of a collectively mediated tradition transmitted exclusively by word of mouth until, for whatever reason, a scribe copied it down as he received it. Were this model convincing, we would know how to read these texts—as so many discrete “words of tradition” enjoying a timeless literary life essentially unaffected by those who transmit them. That is, we would read them pretty much as they have been read by countless generations of rabbinic Torah scholars. In which case, scholars trained in the European university traditions of textual scholarship would be redundant as interpreters of the text, bringing to them only some particular philological advances that might illumine this and that word transmitted by tradition. Similarly, I am familiar with no one who holds that any rabbinic work is the product of a single scribe’s creative authorship, in roughly the same way that a contemporary scholarly or literary work is the direct creation of the author. Were this model convincing, we would also know how to read these texts. That is, we would read them as if they represented a unified, conscious message directed from an authorial mind to a certain, historically-delimited, audience of hearers or readers. In which case, we would be guilty of not one, but two, hermeneutical sins: first, the anachronism of confusing the nature of written communication in the age of print and the electronic text with the nature of written communication in cultures of the manuscript;7 the second, of assuming that historical textual meaning is in some sense identical with and exhausted by the supposed intentions of authors.8

7 Despite its breathless, romantic infatuation with “orality,” W. Ong’s Orality and Literacy: The Technologizing of the Word (Routledge: 1982) remains a crucial resource for distinguishing among the various technologies of verbal communication. 8 This is no place to summarize the enormous literature on the problem of authorial intention generated by E.D. Hirsch, Validity in Interpretation (Yale University: 1967). But for a fair representation of crucial points of view, including most importantly that of H-G Gadamer, see D. Hoy, The Critical Circle: Literature, History, and Philosophical Hermeneutics (University of California: 1982).

HOW SHOULD RABBINIC LITERATURE BE READ?

16

Unfortunately, neither of these positions on the nature of the rabbinic text is tenable. It is obvious that someone sat down with a pen, ink, and some sort of treated surface to produce the extant versions of compilations such as the Mishnah, Tosefta, the Midrashim, and the Talmudim. But to say this is to assert almost nothing about the process that produced our texts. It seems beyond doubt that the scribes responsible for the surviving medieval manuscript copies of classic rabbinic works copied them with relative care from written exemplars.9 But these texts, emerging in the tenth or eleventh century and later from the dark tunnel that connects the high middle ages to late antiquity, tell us rather little about the textual practices that stand at the beginning of the tradition that yields the medieval manuscript copy. What sorts of textual material, for example, did scribes who produced the earliest written copies of the Mishnah or the Talmuds have in front of their eyes as they wrote? Did they feel constrained, as did medieval scribes, to copy such texts more or less as they received them? There is, of course, nothing for rabbinic literature analogous to the Dead Sea Scrolls which have been so valuable in enabling biblical scholars to peek in on earlier and divergent versions of the canonical texts of the Hebrew Bible. Without the ability to study at first hand the actual products of rabbinic scribes of late antiquity, we are, figuratively speaking, in the dark regarding the means by which they transformed oral tradition into textual compilations as well as the purposes that guided them in their work. Happily, this darkness is not total. As we grope slowly about, we become familiar with some dimly-perceived objects that gradually cohere into something approximating a navigable pattern. Both of these objects are direct creations of those who produced the 9

A remarkably comprehensive collection of essays on the role of medieval scribes in the transmission of Hebrew literature is available in P. Alexander and A. Samely, eds., Artifact and Text: The Re-Creation of Jewish Literature in Medieval Hebrew Manuscripts (Bulletin of the John Rylands University Library of Manchester: 1993). Most relevant to the present discussion is the essay in that volume of M. Beit-Arie, “Transmission of Texts by Scribes and Copyists: Unconscious and Critical Interferences,” pp. 3352. For a comprehensive study of the theory and practice of textual emendation in the history of rabbinic literature from antiquity into the age of print, see Y. Spiegel, Chapters in the History of the Jewish Book: Scholars and Their Annotations (Hebrew; Bar-Ilan University: 1996).

“ORALITY” OF RABBINIC WRITING

17

texts. One concerns the way the social world of rabbinic learning is represented in the texts through various reports about the process of textual learning among rabbinic disciples. The second concerns the technical aspect of the composition of rabbinic texts, namely, the loosely-structured, agglutinative patterns by which virtually all surviving rabbinic texts are compiled. In my view, both the presentation of rabbinic learning as a series of discrete “learning paragraphs” and the representation of rabbinic learning as a process of memorization-for-performance suggest the essential “performativity” of the rabbinic textual tradition. If we can rely at all on the depictions of rabbinic learning that pepper rabbinic texts, we can affirm that the learning of tradition involved memorization of highly formulaic texts and the ability to criticize them and expound them in a formal disquisition before various types of audiences—the single teacher, the immediate study group under a given sage, the favored study-partner who served as both critic and enabler of one’s rendition of the tradition, the community of text-memorizers who constituted the “critical apparatus” against which any written text-version could be measured. Rabbinic writing, as best we can tell, retrieves these performative texts from the primary performance setting of face-to-face study, where the text lived through the dialogue of embodied discourse, to the secondary performance setting of the manuscript, where a rendition of the text is stored as the basis for further performative presentations. Moreover, the written manuscripts seem themselves to serve as models for future performance; scripts that provide the mnemonic foundations for returning the text from the written surface back to the oral/aural milieu that is the preferred context of study and teaching. On the basis of this hypothetical model of the performative character of rabbinic text-making, it is helpful to imagine early rabbinic writing as a kind of “textprocessing” activity in which written texts (viewed now with the eye) and remembered texts (recalled with the ear) could mix together in the literary activity of the scribe as he vocalized texts in the course of composing his own version of the received discourse. If we understand written textual production as a variant of oralperformative textual transmission, then we can imagine the early rabbinic text-makers to have written, so to speak, with their ears, not so much copying the texts they transmitted, but rehear(s)ing

18

HOW SHOULD RABBINIC LITERATURE BE READ?

them in memory as they rendered them visibly in new configurations on the page.10 There is, in this appreciation of the performativity of the classic rabbinic text, an important—and to some degree, sobering— implication for textual interpretation. If the written text is, as it were, snatched out of the air in the process of composition, much of its meaning remains in the air—where it is necessarily lost to us, unless the happy existence of a parallel formulation in some other written compilation or a manuscript variant enables us to use the fuller text as a commentary on the more spare one. That is to say, the oralist textual scholar is in a sense engaged in an effort to “restage” the oral-performative milieu of the textual tradition by comparing various written testimonies to the shaping of the texts and enabling the contemporary reader of the finished texts to “rehear” them in their multiple versions. This reperformance is achieved through translational and exegetical work that seeks to replenish the text on the basis of other performances retrieved from the frozen texts of manuscript variants and alternate compositional variations. We should not imagine ourselves to be retrieving any authoritative original version of a text or tradition; rather, it is almost the opposite. We may at best offer a fictionalized representation or reconstruction of the multi-tonal quality of the living tradition. That is, on the basis of written survivals of oral-traditional material in performance, we seek to reconstruct the echoes lost from the tradition as it was transformed into manuscript for preservation and transmission as written word. This appreciation of the multi-tonal or pluriform quality of the constituent traditions of the rabbinic text bear implications for the ways in which we view the anthological compositions in which discrete traditions are preserved. Some of these, like the extant mishnaic “tractates,” for example, ultimately reached a high degree of editorial “closure” rather well before they entered the so-called “dark tunnel” of the early middle ages, from which they eventually emerged as formal literary works collected together in great codices like other medieval books. Even before entering that tunnel, scribal 10 I believe that Steven Fraade has produced the most important theoretical model of this process of performative rabbinic textual production in his influential study, From Tradition to Commentary: Torah and its Interpretation in the Midrash Sifre to Deuteronomy (State University of New York: 1991), e.g., pp. 17-20.

“ORALITY” OF RABBINIC WRITING

19

additions and changes were resisted in favor of a more or less “canonical” version. The closure of these “canonical” versions was signaled by the fact that the transmission of accumulated interpretive tradition concerning the texts was preserved beyond the boundary of the manuscript versions, in new compilations. The tractates of the Tosefta and the two Talmuds represent this process for the mishnaic tradition in particular. With regard to the present discussion of Mishnah Avot, of course, Avot d’Rabbi Natan stands as the analog of the Toseftan/talmudic interpretive textual repository. But even in light of the virtual canonization of certain written textual recensions of all such tractates, these recensions themselves never quite achieved the “finished” character associated even in antiquity with single-authored works associated with genres such as philosophy, epic poetry, or historiography. Rabbinic texts stemming from late antiquity were never worked into self-contained discourses that used the convention of an authorial “voice” to establish the lines of intelligible communication between “author” and “audience.” There is no clear “logos” in the classical rabbinic compilations. That is, a reader encounters numerous textual voices within a given composition, many of which appear to be in conversation with each other, but none has final authority to impose an overall determination of the text’s meaning upon the reader. Instead, the literature remains open-ended in its literary style, anticipating and even insisting that the anthologies on various topics and so on would continually be re-performed in successive generations of study circles presided over by masters of the tradition in a host of institutional and cultural settings. Every rabbinic compilation, at least in theory, can be reduced to the rendering of traditional elements drawn upon by a scribal editor who links them into chains of discourse through his own literary intervention into the thesaurus of the tradition. Entire compilations are built up associatively from such repeated copyings. Principles of association can vary from genre to genre of rabbinic text, from scribal circle to scribal circle, and even within different sections of what is now transmitted as a single compilation.11 11 For various recent perspectives on the nature of rabbinic textual composition, see the following representative discussions: P. Schaefer, “Research into Rabbinic Literature: An Attempt to Define the Status

20

HOW SHOULD RABBINIC LITERATURE BE READ?

In my view, the hermeneutical challenge of the “oralist” conception of the performative character of the rabbinic text is to develop contemporary reading strategies that highlight the degree to which a given text depends upon, recalls, and interfaces with other texts that are not present in the particular composition before the reader. In a word, the “oralist” approach to the rabbinic text is a variant of the intertextualist approach to literary interpretation combined as well with a kind of “audience-response” sensibility.12 We understand a rabbinic text best, I hold, when we can locate it synoptically in the range of textual elements that remained “in the air” when the scribe reduced part of the performative tradition to a fixed textual rendering. I would be very hesitant to suggest that such a synoptic picture would represent the entire tradition at a given period or moment in the life of the tradition. Too much has been lost to believe that even the most assiduous mining of extant textual sources can reconstruct the full life of discourse from which rabbinic textual compilers drew. At best we can create a representation of the life of the textual tradition that evokes in a pale way the fullness of meaning that would have been available to participants in the learning communities out of which our manuscripts have emerged over many centuries. With this in mind, let me now offer a modest illustrative example of the kind of interpretive strategy I seek to apply.

Quaestionis,” JJS 37 (1986), pp. 139-152, C. Milikowsky, “The Status Quaestionis of Research in Rabbinic Literature,” JJS 39 (1988), pp. 201211, Schaefer’s rejoinder to Milikowsky in JJS 40 (1989), pp. 89-94, E. Segal, “Anthological Dimensions of the Babylonian Talmud,” Prooftexts 17 (1977), pp. 35-61, and P. Mandel, “Between Byzantium and Islam,” pp. 74-106. 12 It should be recalled that the sort of intertextuality that obtains between conventional literary works is not a matter of one work quoting another. Rather, intertextuality denotes the presence in a literary work of echoes of the literary tradition within which a given author takes up his or her literary discourse. The entire inherited tradition, as it were, is evoked in the discourse of one of its concrete representations. In this sense “literary intertextuality” is much like the kind of intertextuality I am positing for oral literary traditions. It is the stock of generic forms, discursive patterns, topical themes that the formulator of a given oral-traditional speech act assumes in the audience as a condition for understanding and participation in textualized meaning.

“ORALITY” OF RABBINIC WRITING

21

MNEMONICS, PERFORMANCE, AND THE INSCRIPTION OF TEXTUAL TRADITION One contribution that an oralist hermeneutic can offer is to highlight ways in which text-makers situated themselves in the interstices between their own writing and the body of tradition in which they moved and which, in turn, moved through them. This process of scribal self-insertion into the tradition, which results in a significant alteration in the direction and reference of the oral-traditional material, is particularly evident where traditions known from the Mishnah are adopted and adapted in extra-mishnaic (or, if you will, post-mishnaic) literary settings. Below I try to construct a triangulated conversation between three quite discrete samples of rabbinic text. 13 The first text is Mishnah Avot 5:1, which introduces a sub-collection of material (5:1-15) organized around the mnemonic principle of number (“With ten Divine Utterances …,” “With ten trials was Abraham tested…,” “There are seven traits of the clod and seven of the Sage…,” “There are four types of human character…,” etc.). In context, Mishnah Avot 5:1 is represented as traditional wisdom, anonymously conveyed. The second selection, from Mishnah Tractate Sanhedrin 4:5, is extracted from a tractate on judicial procedure. In context, it represents a judicial charge to witnesses in direct discourse—a “quotation from life,” as it were, rather than a traditionally-mediated source of knowledge. As I shall argue, an important intertextual element shared by these two very different texts demonstrates their common embeddedness in a single continuum of mnemonically-structured oral-performative tradition. This intertextual element provides for us a context within which to interpret the activity of a third scribal creation, a text from ARN, version A, chapter 31.14 This text knows each mishnaic text in its extant version. But, I claim, it knows as well the mnemonic underpinnings of the orally-performed tradition upon which each written mishnaic text relies. That is, while Mishnah Avot and 13

For the Mishnah I use the edition of C. Albeck, and for ARN the recent reissue of S. Schechter’s edition with an introduction and corrections by M. Kister. 14 Another important player in this intertextual conversation is ARN, B:36. In order to keep our analysis simple, I will allude to ARN, B’s text in my notes to the translation of ARN, A.

22

HOW SHOULD RABBINIC LITERATURE BE READ?

Mishnah Sanhedrin themselves depend upon an oral-peformative milieu, the ostensibly “interpretive” or “commentarial” text of ARN reveals the same pattern of dependency. Its scribal creator is not so much “commenting upon” or “interpreting” the received tradition in the way a textual interpreter stands apart from and turns attention upon the interpreted text as a “work.” Rather, our scribe in ARN is rehearsing and extending his received texts. He does so by both recalling the language of earlier material and reconfiguring it in new structures of thought. Thus, he is at one and the same time: 1. familiar with the oral-performative tradition that preceded the composition of tractates Avot and Sanhedrin; 2. aware of how the formal written compilation of these tractates intervened in that tradition and shaped it to the purposes of the composing scribes; and, finally, 3. a participant in the work of intervention in shaping the materials in his own direction.15 Here, then, are the texts. For purposes of this illustration, we begin with a brief comparison of Mishnah Avot 5:1 and its close, but significantly different, parallel in ARN, A:31. Italicized materials represent close textual correspondences that we will later exploit for our own analysis of the oral-performative dimension of the texts. After making a few observations we shall than move to Avot’s close intertext at Mishnah Sanhedrin 4:5: Mishnah Avot 5:1 A. With ten Divine Utterances was the world created. B. And what does the Scriptural Teaching [of Genesis 1:1ff.] propose? Could it not have been created by a single Utterance? C. Rather, this is to punish the wicked who destroy a world that was created by 10 Divine Utterances, and to grant a

15 In formulating the issue in this manner, I differ significantly from the approach of M. Kister, who is working with a much more “scriptural” model of rabbinic intertextuality, in which later texts are viewed as “commentaries” on earlier texts. This leads him to express disappointment that “we cannot view ARN as an early and reliable interpreter (parshan) of Tractate Avot” capable of bringing readers to “the original meaning (mashma`ut hameqorit) of the sayings in Avot.” See his discussion of ARN, A:31, m. Avot 5:1, and m. Sanh. 4:5 in Kister, Studies in Avot de-Rabbi Natan, p. 138.

“ORALITY” OF RABBINIC WRITING

23

good reward to the righteous who sustain a world that was created by 10 Divine Utterances.

Avot D’Rabbi Natan, A:31 A. With ten Divine Utterances was the world created.16 B. Now, why do the world’s inhabitants need to know this?17 C. Rather, this is to teach you that one who performs a single commandment,18 or one who observes a single Sabbath, or one who saves a single [Jewish]19 life is acknowledged by Scripture as one who sustained an entire world that had been created by ten Divine Utterances. 16 ARN, B:36 cites the entirety of m. Avot 5:1A-B, and answers the question as follows: “Rather, to punish the wicked who destroy the world.” Do we have here an “original” version of m. Avot 5:1 of which the present m. Avot 5:1 is an expansion? Or, does the Mishnah text of ARN, B:36 “condense” the text of the extant Mishnah? Or, perhaps, is ARN, B:36 drawing upon m. Sanhedrin 4:5C (see below). To complicate matters, ARN, B:36 then re-cites m. Avot 5:1A and adds a list of scriptural verses that constitute the ten utterances. It is likely that this inclusion, which disrupts the flow of ideas, is an interpolation grounded in Genesis Rabbah 17:1 and/or b. Rosh Hashanah 32a. But final judgments of the relationships of the two recensions of ARN to each other and Mishnah Avot, in this and many other cases, must await the completion of M. Kister’s proposed critical edition of ARN. At that point it will also be possible to test the applicability of the fruitful model for comparing the transmission of different recensions of rabbinic texts proposed by P. Mandel, “Between Byzantium and Islam,” p. 99, with regard to the textual tradition of Lamentations Rabbati. 17 The text of ARN, B:36 is somewhat garbled at this point. See Kister, Studies in Avot de-Rabbi Natan, p. 220, n. 506. 18 The theme of the decisive impact of “a single commandment” (mitzvah ehat), is itself something of an oral-performative hook in rabbinic texts. See for example, Mek. Yish., p. 157; b. Sot. 36; b. Qid. 39b. 19 The word “Jewish” (miyisrael) does not appear in the first edition of ARN. On the basis of mss readings, Schechter includes it within brackets in his own edition of ARN, A and ARN, B. Insofar as the topic concerns the performance of commandments the readings that include miyisrael are appropriate. See n. 22 below for discussion of the relation of this text to its parallel at m. Sanhedrin 4:5. I wish at this point to acknowledge the help of my teacher and friend, Prof. Richard Sarason, in sorting out the textual issues here and, especially, in pointing out the significance of important comparative material overlooked in the original draft of this essay.

24

HOW SHOULD RABBINIC LITERATURE BE READ? And one who commits a single transgression, or one who desecrates a single Sabbath, or one who destroys a single [Jewish] life is acknowledged by Scripture as one who destroyed an entire world that had been created by ten Divine Utterances.20 D. For thus have we found regarding Cain who murdered Abel his brother, as it is said: “The voice of your brother’s bloods cries out to me!” (Gen. 4:10) He spilled one blood, but many bloods are spoken of here. Rather, this teaches that the blood of his children, his grandchildren, his great grandchildren, and all his descendants until the end of all generations that would have proceeded from him, all of them arose and cried out to the Blessed Holy One. E. Thus you learn that a single person is as valuable as the entirety of Creation!

Traditions known only from amoraic compilations are clear that the “ten Utterances” are the various repetitions of “and God said” that can be found in the Genesis creation account (Genesis Rabbah 17:1, ARN, B:36, b. Rosh Hashanah 32a). Each of the texts before us takes this for granted, and each makes the same fundamental point: that the multiplicity of divine speech-acts that produced the world increases both the value of the world and, thus, the responsibility of its inhabitants. For they now have the knowledge of their capacity to contribute to the good repair of a world precious to God or to its destruction. Compositionally, there are obvious correspondences between the passages in their extant scribal form. The assertion of a fact drawn from scriptural study (Mishnah Avot 5:1A/ARN, A:31A), followed by a question about the scriptural intention (5:1B/A:31B) and an answer (5:1C/A:31C) is a common rhetorical form found in midrashic exegesis. ARN’s further elaborations (A:31D-E) have no parallel in Avot, but they do conform to stylistic traits common in midrashic formulations. We shall attend to them more closely later, after our discussion of Mishnah Sanhedrin 4:5. For now it is sufficient to ask: how do we account for the overlapping material 20 ARN, B:36 at this point produces only the italicized text, but reverses their sequence. I read this as neither an “improvement” of ARN, A:31 nor an “abbreviation.” It is simply another possible rendition of a conventional idea.

“ORALITY” OF RABBINIC WRITING

25

shared by Avot and ARN as well as the significant textual divergencies? Many scholars have observed that the literary relationship between Mishnah Avot and the two recensions of ARN is highly complex. While most acknowledge that ARN in either recension post-dates the compilation of the Mishnah and Mishnah Avot, many have argued that the version of Mishnah Avot known to ARN is an earlier version of the tractate than ultimately came to be circulated with the Mishnah as a whole in the medieval mishnaic and talmudic manuscripts. We will not be able to take a stand on this larger question here.21 All that we can note is that both passages are built upon a common formal structure which constitutes a widespread generic trope in midrashic text-making. But it is hardly likely that either the scribe of Avot or that of ARN was selfconsciously perusing the text of this companion document while contemplating the shape of his own version of the meaning of the ten Utterances. From the perspective of an “oralist” analysis, the core material shared by the two passages represents written transformations of mnemonically managed oral-performative material. There are two intertwined mnemonic dimensions. First is the abstract form that would be recognized by text-performers as the most basic rhetorical pattern for transmitting a type of exegetical tradition: Assertion of a Scriptural fact Query (mah talmud lomar [Avot]/ mah tzarikh ba’ei ha‘olam bekhakh [ARN])22 Reply (‘el’a) + explanation of Scriptural fact.

The second mnemonic dimension is substantive and verbal, rather than merely formal. This concerns the actual formulaic phrases that can be mobilized into specific discursive propositions in the course of midrashic performance. In the present case these are: With 10 Divine Utterances was the world created Query

21 The issue is usefully summarized in G. Stemberger, Introduction to the Talmud and Midrash, 2nd ed (T&T Clark: 1995), pp. 226-228. 22 See Sifre Dt. 37, Sifre Nu. Zuta 27:12, and Mekhilta Rashbi, Va’era 6:2 and the discussion of Kister, Studies in Avot de-Rabbi Natan, p. 220, n. 506.

26

HOW SHOULD RABBINIC LITERATURE BE READ? Rather + Those who sustain/destroy a world created with 10 Divine Utterances

What I imagine here is that the scribe of each of our texts was fully versed both in the repertoire of abstract forms that underlie the present texts as well as the specifics of the “10 Divine Utterances” theme. Each scribe would have been exposed to multiple versions of this performative tradition in the course of his circulation through rabbinic study circles. Each employed this common mnemonic element of the text-performative tradition in creating his own scribal version. Neither scribe needs to have “read” the written work of the other; rather each, working independently of the other, is dependent upon the oral-performative work of the antecedent tradition for his own creative literary intervention into its unfolding. Thus it would be a misconstruction of the literary relationships to construe them in the model of “dependence,” that is, to wonder if the editor of ARN “expanded” a written text known from Mishnah Avot or, to the contrary, if the mishnaic pericope before us is an “abbreviation” or “condensation” of ARN. Rather, I would propose that these two texts be viewed within the model of “filiation,” that is, each is generated by the same underlying structure without any necessary historical “dependence” upon the other. We are now in a position to approach the third text that participates in the oral-performative tradition that nourishes our scribes. Here I will offer two altered type-faces to represent two quite separate sets of textual correspondences. Italics will represent points at which we find clear verbal correspondences between Mishnah Sanhedrin 4:5 and Mishnah Avot 5:1; bold-face text will represent verbal correspondences between the Mishnah Sanhedrin 4:5 and ARN A:31. Mishnah Sanhedrin 4:5 A. How do they charge the witnesses regarding capital testimony? They would convene them and charge them as follows: “Would you perhaps speak on the basis of speculation, or on the basis of rumor, from second-hand testimony, or on the grounds that ‘we heard from a reliable witness?’” Or—

“ORALITY” OF RABBINIC WRITING

27

“Don’t you know that ultimately we will submit you to examination and verification?” Know well that civil suits are not comparable to capital cases, for in civil suits a person renders his money and he is absolved, while in capital cases, his blood and the blood of his offspring hang in the balance until the end of the world. B. For so do we find regarding Cain who killed his brother, since it is said: “your brother’s bloods cry out.” (Gen. 4:10) It does not say “your brother’s blood,” but rather “bloods”—that is, his blood and the blood of his offspring … C. For this reason Adam was created as a single individual, in order to teach you that anyone who destroys a single23 life is acknowledged by Scripture as one who destroyed an entire world; and anyone who sustains a single life is acknowledged by Scripture as one who sustains an entire world.24 D. And, furthermore, for the sake of peace among people, so that no one will say to his neighbor, “My father is greater than yours.” And also to testify to the greatness of the Blessed Holy One, for a person can stamp many coins with a single mold, and each is identical to the next, but the King of Kings of Kings, the Blessed Holy One, stamps each person with the mold of the original Adam, but none of them is identical.

23 Most printed texts of the Mishnah include miyisrael at this point, and thus agree with the manuscripts of ARN, A and ARN, B used by Schechter. After examining all available manuscripts and editions of Mishnah Sanhedrin, ARN, A/B and parallels in medieval citations, E. Urbach concluded that the intrusion of miyisrael is a relatively late interpolation. See E. Urbach, “‘Whoever Sustains a Single Life . . .’: Textual Vicissitudes, the Impact of Censors, and the Matter of Printing,” in E. Urbach, From the World of the Sages (Hebrew; Jerusalem: 1988), pp. 561-577, esp. 269-273. Cf. Avi Sagi, Judaism: Between Religion and Morality (Hebrew; Hakibbutz Hameuchad: 1998), p. 378, n. 16. I am indebted to my dear friend and guide, Prof. Yaakov Elman, for pointing me towards Sagi’s rich book. 24 While t. Sanhedrin 8:4 offers many supplements to this mishnaic passage it is inexplicably silent on the issue of the value of a single (Jewish) soul.

HOW SHOULD RABBINIC LITERATURE BE READ?

28

E. For this reason, each and every person must say: “for me was the world created!” …

The Mishnah’s judges (4:5A) are reminding the witnesses that when an innocent person is executed, the result is to cut off not only that life but all lives that the victim might have spawned. The thought is demonstrated exegetically (4:5B) by recalling the odd locution of Gen. 4:10, which, indeed, does refer to the “bloods” (Hebrew: damim) of Abel. The point yields a second regarding the larger purpose behind the biblical depiction of the creation of humanity through a single prototype: a single human life is equivalent to an entire cosmos, just as the entirety of the human race was contained in its single forebear (4:5C). This explanation of the singularity of the human creature, appropriate to the theme of the value of even the accused murderer’s life, is now extended to a new topic: from the singularity of Adam we learn also about the equality of nations and the uniqueness of every human individual (4:5D). The passage ends with a concluding coda summarizing the point of the whole (4:5E). Let us now attend to some of the obvious intertextual phenomena exhibited here, first in connection with Mishnah Avot 5:1: Mishnah Avot 5:1C Rather, this is to punish the wicked who destroy a world that was created by 10 Divine Utterances, and to grant a good reward to the righteous who sustain a world that was created by 10 Divine Utterances.

Mishnah Sanhedrin 4:5C For this reason, Adam was created as a single individual, in order to teach you that anyone who destroys a single life is acknowledged by Scripture as one who destroyed an entire25 world; and anyone who sustains a single life is acknowledged by Scripture as one who sustains an entire world.

The italicized material here highlights the degree to which the intertexts that link Mishnah Avot 5:1 and Mishnah Sanhedrin 4:5 have no particular meaning beyond the textual constructions within 25

note.

On the significance of the term “entire” (malei) see the following

“ORALITY” OF RABBINIC WRITING

29

which they are woven. In the first mishnah, sustenance or destruction of the world is thematized within the problematic of a world interpenetrated by divine speech; in the second, the over-riding issue is the creation of Adam as an individual being within whom all future humans have their origins. The formulators of such mishnahs are not relying on citations of each other’s work; rather, they are “citing” a traditional trope that can be marshaled for virtually any purpose that comes to the mind of the textual performer. What is “quoted” is a text that lies behind both mishnahs in the oral milieu of the rabbinic study community. And that quote is not a “text” per se, but rather the mnemonic cue that might serve for the construction of a text. This same point can be made with Mishnah Sanhedrin 4:5’s second intertextual partner, ARN A:31. Let us first review the relevant intertexts, with special attention to the bold-faced material: Mishnah Sanhedrin 4:5 B. For so do we find regarding Cain who killed his brother, since it is said: “your brother’s bloods cry out.” It does not say “your brother’s blood,” but rather “bloods”—that is, his blood and the blood of his offspring … C. For this reason Adam was created as a single individual, in order to teach you that anyone who destroys a single life is acknowledged by Scripture as one who destroyed an entire world; and anyone who sustains a single life is acknowledged by Scripture as one who sustains an entire world.

ARN A:31 C. Rather, this is to teach you that one who performs a single commandment, or one who observes a single Sabbath, or one who saves a single [Jewish] life is acknowledged by Scripture as one who sustained an entire world that had been created with 10 Divine Utterances. And one who commits a single transgression, or one who desecrates a single Sabbath, or one who destroys a single [Jewish] life is acknowledged by Scripture as one who destroyed an entire world that had been created by 10 Divine Utterances.

30

HOW SHOULD RABBINIC LITERATURE BE READ? D. For thus have we found regarding Cain who murdered Abel his brother, as it is said: “the voice of your brother’s bloods cries out to me.” He spilled one blood, but many bloods are spoken of here. Rather, this teaches that the blood of his children, his grandchildren, his great grandchildren, and all his descendants until the end of all generations that would have proceeded from him, all of them arose and cried out to the Blessed Holy One.

Here there are two contiguous intertextual elements: the exegetical comment regarding Gen. 4:10 and the dichotomy of “one who sustains/destroys a human life.” But what is shared by these intertexts is hardly a completed discursive unit that is quoted by one text from another or, perhaps, might somehow be interpolated in the passage of the text through scribal tradition. The sharing, as in our earlier case, is at the level of mnemonic cue alone. In Mishnah Sanhedrin, the exegesis of Gen. 4:10 is framed as the words of judges to the witnesses, and introduces the idea of sustaining/destroying a single life that is comparable to an “entire” 26 world. This point is developed in the present context in relation to the singularity of individuals. By contrast, ARN’s deployment of these materials shares virtually nothing but the structural elements we have identified. First of all, the sequence of the intertexts found in Sanhedrin is transposed, with the exegesis of Gen. 4:10 now coming to illustrate the teaching about the preciousness of the world created with 10 Divine Utterances rather than serving as a consequence of it, as in Sanhedrin. In ARN, moreover, the dichotomy of “sustaining/destroying” is encased in a richer series of acts that link up to the distinctly rabbinic elements of performance of commandments and celebration of the Sabbath. What Sanhedrin knows as an insistence upon the sanctity of human life becomes in ARN more intensely focused upon specifically Jewish life.27 Let me now summarize the basic conclusion of the above analysis of the relationship of the mnemonic elements of our three texts to the compositional creations of their scribal performercomposers. It is likely that the textual snippets represented in the 26 ARN, B:36 agrees on the word, “entire,” (malei), thus aligning with m. Sanhedrin and ARN, A over and against m. Avot 5:1. 27See references in n. 23 for discussions of some possible explanations for this reformulation.

“ORALITY” OF RABBINIC WRITING

31

above passages by the italicized and bold-face texts circulated as core oral-compositional elements analogous to the epithets so often noticed by students of oral-epic composition or topoi employed by rhetoricians and public speakers in Greco-Roman antiquity. They were the phrases and clichés that made up the thesaurus of “words” that could be deployed in the transmission of orallymanaged tradition. In the present settings in Mishnah Avot, ARN, and Mishnah Sanhedrin, however, they have been set down in written texts. Each rendering of the topoi exemplifies one of many possible ways of deploying the orally-managed material in the purpose of creating new discourses upon the received themes. The text of ARN displays one such way. Few scholars doubt that by the late Byzantine period, during which ARN was probably reaching its extant form, written texts of the Mishnah were in circulation. Yet, the training of disciples required them to memorize the texts from orally-transmitted versions. So the scribe/composer of ARN would have surely known these written versions. But his own appropriation of those versions reveals the aural and oral mastery of the text distinctive to the oral-performative tradition, mixing and matching diverse traditional elements of 10 divine utterances and the uniqueness of Adam for new theological work.

CONCLUSION: ORALITY, PERFORMATIVITY AND INTERPRETATION In a certain sense, my approach to the orality of classical rabbinic literature has been driven by a desire to enable the oralperformative dimension of rabbinic tradition to have an impact upon the interpretive work of contemporary readers of rabbinic writings. In the academic interpretive tradition in which I was trained, much was said about rabbinic oral tradition, but the texts that purportedly emerged from this oral tradition were read as if they were composed by academic intellectuals perpetually refining their thoughts in silent studies until they finally released the version that perfectly captured every nuance of their thought. Thus we combed these texts for every subtlety that might tip the interpreter off to the inner thought concealed in the text. I still believe that rabbinic texts contain enormously subtle thoughts—but that these emerge not from the careful reconsiderations of isolated scribal intellectuals.

32

HOW SHOULD RABBINIC LITERATURE BE READ?

Despite the obvious fact that single individuals produced by their manual labor the physical manuscripts from which we know the rabbinic texts, their work is a collective project that unfolded over many centuries. Accordingly, our search for the subtleties of rabbinic textual meaning must attend as much to the interplay of meanings that extend over several textual collections as they do over the work of meaning that emerges within a single textual compilation. Thus, to stick with the single example I have offered above, the “meaning” of Mishnah Avot’s reflection upon the 10 Utterances that brought the world into being, is greater than the meaning it has within Mishnah Avot 5:1 or the way in which ARN enriches its meaning by thinking it through the exhortations of Mishnah Sanhedrin 4:5. The synergy of the textual tradition, which emerges when we seek to distinguish the various ways in which transmitted texts overlap with the scribal/redactional production of new versions of tradition is a proper focus of our own interpretive attention. In conclusion, I would like to return to the collegial question that occasions these remarks, and to which they are intended to offer a response. What difference, after all, does the recognition of the oral processes underlying the making of rabbinic manuscript traditions make for the basic interpretation of rabbinic writings? I would reply that the chief interpretive impact lies in our conception of one of the most ubiquitous aspects of rabbinic literature—its obsessive self-referentiality and internal self-quotation. It is not enough to acknowledge these traits of rabbinic writings as “sharing of common sources,” “quotation and commentary,” or other frequent characterizations that leave intact the impression that it is a simple matter to distinguish in rabbinic literature the creation of “compositions” that, in some unilinear fashion, yield “commentaries” and other sorts of appropriative literary renderings. Rather, in a basic sense, there is no strict “early and late” in rabbinic writing. When an oral-performative milieu includes not only fully-formed textual units that have benefited from written preservation and transmission, but also a continuously elaborated thesaurus of mnemonic cues that encourage ever-new textual formulations, it is precarious indeed to know the precise direction of phenomena that appear as “commentary” or “amplification.” Thus, more acute attention to the oral traits of rabbinic text-making will, I hope, render less certain—and perhaps more alive—our efforts to read rabbinic

“ORALITY” OF RABBINIC WRITING

33

texts as communications that remain incomplete and “in the air” until a illustrative range of performative versions can be reconstructed.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS: ISSUES OF LITERARY RELATIONSHIP AND REDACTION Alyssa Gray Hebrew Union College-Jewish Institute of Religion, New York ASSESSING THE RELATIONSHIP BETWEEN BAVLI AND YERUSHALMI There is a growing literature on the so-called “synoptic problem in rabbinic literature.” As Shaye J.D. Cohen has recently reminded us, medieval rabbinic scholars were well aware of the numerous parallels among the rabbinic compilations, and typically studied and commented on texts together with their parallels. The emergence of “historically minded Jewish scholarship” (his phrase) in the nineteenth century changed the synoptic problem from being an issue of using parallels to comment on texts to, inter alia, determining the literary relationship between parallel materials in rabbinic compilations.1 This paper is a contribution to the ongoing project of determining literary relationship between the Talmuds. The relationship between the Talmuds has been a contested item on the agenda of rabbinics scholarship since the nineteenth century as has, to a lesser extent, the relationship between the Yerushalmi and Yerushalmi parallels in Bereshit Rabbah. The Bavli’s relationship to Yerushalmi

1

See Shaye J.D. Cohen, The Synoptic Problem in Rabbinic Literature (Providence, RI: Brown Judaic Studies, 2000), pp. vii-viii.

35

HOW SHOULD RABBINIC LITERATURE BE READ?

36

parallels in B.R., however, has not been systematically explored.2 An analysis of a lengthy sugya at y. Ber. 7:2, 11b, par. B.R. 91:4 (Theodor-Albeck 3:1111-1118) and its parallel at b. Ber. 47b-48b addresses this lacuna.3 This paper offers the following conclusions: (1) y. Ber. and B.R. are two versions of the same Palestinian sugya explicating m. Ber. 7:2, and y. Ber. is a later reworking of B.R.4; (2) b. Ber. 47b-48b is demonstrably derivative of this edited Palestinian sugya and bears similarity to both; (3) no Babylonian amora is aware of this sugya qua sugya, rather, the Bavli’s similarity to the Palestinian 2

See Hanokh Albeck and Judah Theodor, eds., Bereschit Rabba, 3 vols. (Jerusalem: Shalem Books, 1996), 3:1-138, in which Albeck provides his “Introduction to Bereschit Rabba.” That introduction includes a section on “The Yerushalmi and Bereschit Rabba,” but only lists Bavli parallels to materials found in the midrash compilation. Recently, Hans-Jürgen Becker has revisited and rejected Albeck’s conclusion that Bereshit Rabbah drew on a version of Yerushalmi different from “our” Yerushalmi. See his Die grössen rabbinischen Sammelwerke Palästinas (Tübingen: Mohr Siebeck, 1999) and idem, “Texts and History: The Dynamic Relationship between Talmud Yerushalmi and Genesis Rabbah,” in Cohen, The Synoptic Problem in Rabbinic Literature, pp. 145-158. See also Chaim Milikowsky’s thoughtful review essay “On the Formation and Transmission of Bereshit Rabba and the Yerushalmi: Questions of Redaction, Text-Criticism, and Literary Relationships,” JQR 92:3-4 (Jan.-Apr. 2002): 521-567. 3 There is an additional parallel to a portion of the y. Ber./B.R. parallel (the story of Shimon b. Shetach’s conflict with King Yannai) at y. Nazir 5:5, 54b. This additional parallel is not treated in this paper because it is clearly not an independent tradition but was added to y. Nazir from y. Ber. (which it resembles more closely than B.R.) as we may infer from the discussion of Grace after Meals that follows the story and which is completely out of place in y. Nazir. Moreover, the Grace after Meals discussion in y. Nazir is confusing because it lacks the context provided before the story in y. Ber. Thus, it is more likely that the story was originally part of y. Ber. (as part of the lengthy sugya under study in this paper), after which it was lifted out and added to y. Nazir. This phenomenon of socalled “transferred sugyot” in the Yerushalmi has been systematically studied by Moshe Assis; see his Sugyot Maqbilot ba-Yerushalmi, Ph.D. diss., Hebrew University, 1976. 4 See Baruch Bokser, “A Minor for Zimmun (y. Ber. 7:2, 11c) and Recensions of Yerushalmi,” AJS Review 4 (1979): 1-25. Bokser came to this conclusion based on his study of the first part of this lengthy sugya, which deals with the minor and zimmun. I came to this conclusion independently based on an analysis of the entire sugya.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

37

sugya results from the Babylonian redactors’ incorporation and reworking of a redacted Palestinian sugya attached to m. Ber. 7:2; and (4) differences between the older Palestinian sugya and its reworked Babylonian version are traceable to characteristic Babylonian cultural, intellectual, and linguistic concerns, and characteristic Babylonian redactional practices. These differences do not negate conclusion (3). My goal in this paper is not simply to present argumentation in support of these discrete conclusions, but to advance scholarly inquiry about the appropriate method for assessing literary dependence within the rabbinic compilations. Determining literary relationships between synoptic parallels in rabbinic literature is difficult. The parallels are typically both similar and different, which could be due to independent reliance on a third source, or independent use of oral traditions in an oral contextȥnot literary relationship. In order to be maximally persuasive, then, a study of literary dependence must attend carefully to issues of method and spell out criteria for comparison and analysis.5 The method employed in this study may be grouped under three broad headings, each of which involves a number of careful steps. The first heading is “macro-analysis.” The point of macroanalysis, in a nutshell, is to eliminate from further consideration inter-Talmudic parallels whose similarities are clearly due to independent resort to a common mishnah pericope or to a shared stratum of baraitot and/or other tannaitic materials or amoraic materials. Macro-analysis of the parallels is a comparison of the parallel sugyot in toto. These sugyot may be found in both Talmuds (as in this 5

I am hopeful that this paper will also be seen as a contribution toward bringing rabbinics into the larger scholarly conversation about literary dependence. This conversation has been ongoing in the field of classics, as Professor Matthew Kraus has alerted me. See especially Bertil Axelson, “Lygdamus und Ovid: Zur Methodik der literarischen Prioritätsbestimmung,” Eranos 58 (1960): 92-111, which has points of contact with this study. Stephen Hinds took issue with the “Axelsonian” method of study in his “Medea in Ovid: Scenes from the Life of an Intertextual Heroine,” Materiali et discussioni per l’anlisi dei testi classici 30 (1993): 9-47. I intend to explore the question of literary dependence in a more interdisciplinary vein at another time. I thank Professor Kraus for providing me with the articles.

38

HOW SHOULD RABBINIC LITERATURE BE READ?

case), or in one or both of the Talmuds and another compilation as well (Bereshit Rabbah in this case). The key is to look for similarities in the selections and sequences of legal and non-legal materials. As a first step, if these similarities are attributable to the mishnah to which the sugyot may be attached or to other tannaitic sources cited in those materials, then it is unlikely that the later sugya (usually the Bavli version) was influenced by the earlier (usually the Palestinian version). The determination of which similarities are or are not attributable to the mishnah or to other tannaitic sources is critical to the macro-analysis. If the parallel Talmudic/B.R. sugyot discuss a topic that explicates the mishnah out of logical necessity (i.e., they define a term in the mishnah), this topic is considered “called for by the mishnah,” and may have been independently pursued in both sugyot. For example, the pertinent part of m. Ber. 7:2 rules that women, slaves, and minors are not to be included in the zimmun, or invitation to recite the Grace after Meals. Further, although the mishnah discusses the amount of food (olive’s worth or egg’s worth) that a diner must eat in order to be included in the zimmun, the mishnah is silent as to what type of food must be eatenȥmust it or need it not be grains? Since “minors” are mentioned in the mishnah and “type of food” is a clear omission from it, it is unremarkable that both sugyot devote lengthy discussions to these topics. Were the Talmudic/B.R. discussions of these topics dissimilar, it would be appropriate to conclude that the discussions were independently based on the mishnah. But, as we will see, the Talmudic/B.R. discussions of these topics are conducted using remarkably similar materials and are arranged in the same order (which is not called for by the mishnah since “type of food” is a “topic by omission”, as it were, and has no particular place). Although the presence in both sugyot of a discussion of these topics is “called for by the mishnah”, nothing in the mishnah calls for the similarities in the selection and sequence of materials employed to explore them. Similarly, the second step in macro-analysis is to find out whether any similarities between the Talmudic/B.R. sugyot are due to other tannaitic sources common to themȥsuch as a shared set of Toseftan or other baraitot, or mishnayot from other tractates. If such a shared tannaitic set is used or interpreted differently in the Talmudic/B.R. sugyot, then it is more reasonable to assume that it is present in those sugyot for reasons other than the dependence of

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

39

one on the other. But if the shared tannaitic set is embedded in sugyot which are similar overall in their selections and arrangements of amoraic and unattributed materials, then it is more reasonable to assume that the common tannaitic set is a result, rather than a cause, of the similarities between the parallel Talmudic/B.R. sugyot. The second methodological heading is “micro-analysis.” “Micro-analysis” of the sugyot aims to determine whether, when the parallels are closely studied together, they yield any evidence that the Bavli sugya demonstrates awareness of the flow of the Yerushalmi sugya. Specifically, micro-analysis leads us to ask: is the Bavli sugya “aware” of the conclusion(s) reached in the parallel Yerushalmi sugya? Does it begin its analysis at the point at which the parallel Yerushalmi sugya closed? What other points of contact are there between the sugyot that are best explained by the hypothesis that the Bavli sugya is based on the Yerushalmi sugya? The third methodological heading is the exploration and explanation of the differences that exist between the sugyot. In the case we will study, these differences can be explained as the results of intellectual and/or cultural differences between the rabbinic centers, and as conscious Bavli interventions that improve the organization of the material. Micro-analysis and study of the differences between the parallels may well go hand-in-hand, and need not be viewed as completely distinct methodological steps. Taken together, macro- and micro- analysis, and study of the differences between the sugyot, lay a foundation for our conclusion that the Bavli sugya is a conscious reworking of its Palestinian predecessor. There is a fourth methodological heading that is, strictly speaking, tangential to the work of textual analysis. Despite the fact that I have labeled this the “fourth,” this step need not come last, although the more data one assembles from macro- and microanalysis and analysis of the differences, the easier this step will be. This fourth step is that a finding of literary dependency must be able to withstand two likely competing explanations: (1) the “early talmud” hypothesis, 6 and (2) the possibility that the observable 6

The phrase “early talmud” in its Hebrew formulation (“talmud qadum”) is Shamma Friedman’s; see his Talmud ‘Arukh, 2 vols. (New York and Jerusalem: Jewish Theological Seminary, 1993 and 1996), 2:7-23. According to Friedman, Babylonian and Palestinian amoraim both drew from a pool of early rabbinic traditions consisting of amoraic memrot and anonymous materials. Due to this sharing, many Yerushalmi and Bavli

40

HOW SHOULD RABBINIC LITERATURE BE READ?

similarities between the parallel sugyot are due to oral circulation and re-use of shared halakhic or aggadic sayings, stories, interpretations, or motifs.7 As to the first hypothesis, the question must be asked: Are the observable similarities between the Talmudic/B.R. sugyot likely to have resulted from sharing between the rabbinic communities that took place during the amoraic period? Or are the similarities such that they reflect a redactional hand, and more likely are due to editorial activity after the amoraic period? As to the second, we must consider: What is the nature of the similarities and differences between the parallel sugyot? Do the parallels show such a close similarity to each other that it is reasonable to hypothesize that the Bavli appropriated the Yerushalmi sugya itself, or do the parallels show only the common and occasional use of shared motifs, expressions, or a story, such that it is more reasonable to hypothesize that the Bavli sugya is the creation resulting from an oral re-contextualization of disparate sources also found in the older Yerushalmi sugya?

Y. BER. 7:2, 11B AND B. BER. 47B-48B8 y. Ber. 7:2, 11b

b. Ber. 47b-48b

A. The Minor and the Zimmun of 10 or 3 1. R. Simon in the name of R. 1. R. Asi9 said: “A minor who is sugyot share “frameworks of memrot and stam” (p. 16). Although we cannot now know what those early shared traditions looked like, Friedman asserts that the Yerushalmi versions are more likely closer to those originals. See also Noah Aminoah, “Qit’ei Talmud mi-Siddur Qadum be-Massekhet Rosh Hashanah,” in Y.D. Gilat, et al., eds., ‘Iyunim be-Sifrut Hazal ba-Miqra u-ve-Toldot Yisrael (Ramat-Gan: Bar-Ilan University Press, 1982), pp. 185197; Yaacov Sussman, “ve-Shuv le-Yerushalmi Neziqin,” in Yaacov Sussman and David Rosenthal, eds., Mehqerei Talmud I (Jerusalem: Magnes Press, 1990), pp. 55-133. 7 See Martin Jaffee’s paper in this volume. 8 In this translation, tannaitic materials will be presented in bold, statements of the anonymous redactors in italics, and amoraic statements in regular type. The Yerushalmi translation is based on the Venice edition (1523), while the Bavli is that of the printed edition. Significant manuscript variations, as well as differences between y. Ber. and Gen. Rab. are found in the notes.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS Yehoshua b. Levi; R. Yose b. Shaul in the name of Rabbi: “A minorȥwe make him an adjunct to the 10 [for zimmun],” [meaning we make him an adjunct to a zimmun of 10, but not to one of 3, for which he will never count] 2. But wasn’t it taught: We do not carefully scrutinize a minor [t. Ber. 5:19; meaning in context that we do not scrutinize a minor to see if he really can eat an olive’s worth of food. That baraita does not distinguish between a zimmun of 10 and one of 3, and so implies that a minor can count in a zimmun of 3]. 3. R. Yose said: “R. Simon upheld 10 [the baraita]; R. Hanina, R. Simon in the name of R. Yehoshua b. Levi: “[‘We do not carefully scrutinize a minor’ is] necessary in connection with the years of minority [and not with specific connection to the minor’s ability or not to eat an olive’s worth of food]. For if he 9

41

lying in a cribȥ we include him in a zimmun.”

2. But wasn’t it taught in the mishnah: We do not include women, slaves, and minors in the zimmun?

3. [R. Asi] said like R. Yehoshua b. Levi, as R. Yehoshua b. Levi said: “Even though they said that we don’t count a minor lying in a crib in a zimmunȥwe do make him an adjunct to the 10” [and so R. Asi was talking about a zimmun of 10, not one of 3].

The printed edition reads “R. Yose.” See b. Ber. 47b (Diqduqei Soferim note quf, hereinafter “DS”), which notes that Munich 95 reads “R. Ashi,” as does Babylonian Talmud, Codex Florence, Florence National Library, Introduction by David Rosenthal, 3 vols. (Jerusalem: 1972) (hereinafter “CF”). The Ashkenazic rishonim (beginning with Rashi in his commentary to b. Ber. 47b, s.v. “senif”), Rabad, Ramban, and Rid read “R. Asi” with a samekh (which suggests that Munich 95’s reading may actually also be “R. Asi” with a sin). “R. Asi” is clearly preferable; see also Mesoret ha-Shas, ad. loc. 10 Gen. Rab. 91:4 (Theodor-Albeck, p. 1111) reads “R. Isi said: “And R. Simon upheld [the baraita]; R. Hananya and R. Simon in the name of R. Yehoshua b. Levi…”

42

HOW SHOULD RABBINIC LITERATURE BE READ?

was a minor, we consider him a doubtful case [if he had some signs of adulthood but not all], and if he is a doubtful case, we consider him a certain case [and include him in the zimmun]. 11 [Thus R. Yehoshua b. Levi can reconcile his ruling that a minor can only count in a zimmun of 10 with the Tosefta, which implied that the minor could also count in the zimmun of 3]. 4. 12 R. Yehuda b. Pazi in the name of R. Yose: “Nine which appear as tenȥdo we do a zimmun [of 10]?” 5. [No, we require ten] discrete [individuals]. 6. But even if there is a minor among them [the group constitutes a zimmun]. 7. R. Berechya said: “R. Yaqov b. Zavdi 13 asked R. Yose: ‘Just as matters are stated there that we make a minor an adjunct to the 10, let it also be said here 11 Gen. Rab. 91:4 reads: “the doubtful case is the same as the case of the minor: we make the doubtful case into a certainty and the minor into an adjunct.” 12 Gen. Rab. 91:4 puts y. §§9-10 at this point. Gen. Rab.’s version of §10 makes no mention of R. Yudan. 13 Gen. Rab. 91:4 reads “R. Berechya and R. Yaqov b. Zavdi asked R. Simon…” 14 Gen. Rab. 91:4 reads “he [R. Simon] said to them: “Thus it is [a case of] ‘shouldn’t it be all the more so’; above [in the case of a zimmun of 10] in which case the Name is mentioned we make a zimmun [including the minor], here in which the Name is not mentioned [the zimmun of 3] shouldn’t [the minor be counted] all the more so? That which you say that we make the minor an adjunct is with regard to the Grace after Meals. But as for the recitation of the Shema and the Tefillah we do not make him an adjunct until he brings forth two [pubic] hairs.”

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS that we make him an adjunct to the 3! Just as there [by a zimmun of 10] we mention [God’s] Name and make him an adjunct, shouldn’t we make him an adjunct [by a zimmun of 3] in which we don’t mention [God’s] Name all the more so?’” 8. [R. Yose] said to [R. Yaqov b. Zavdi]: “There is no ‘all the more so.’ There [in the zimmun of 10] they make [the minor] an adjunct because they say the Name; and here [in the zimmun of 3] in which they do not recite the Name, they do not make him an adjunct.”14 9. It was taught: A minor and a Torah scrollȥwe make him an adjunct. 10. R. Yudan said: “Thus do we interpret our baraitaȦA minor to a Torah scrollȦwe make him an adjunct” [that is, make the minor an adjunct to the seven who read from the Torah on the Sabbath].

43

HOW SHOULD RABBINIC LITERATURE BE READ?

44

B. The Bavli’s Digression from the Minor and Zimmun 4. 15 And R. Yehoshua b. Levi said: “Nine and a slave join together [to make a ritual unit of 10].” 5. They raised an objection: It once happened to R. Eliezer that he entered the synagogue and did not find 10, so he freed his slave and he completed the 10. If he freed [the slave], yes [he can join to make 10], if he did not free [the slave, which was the case contemplated by R. Yehoshua b. Levi], no [he cannot join to make 10]. 6. He really needed two; he freed one and used [the other] one [to make the 10]. 7. And how could he have behaved thus? For didn’t R. Yehudah say: “Anyone who frees his slave transgresses a positive commandment, as it says, ‘You shall work with them forever’” (Lev. 25:46). 8. It is different if a matter of a mitzvah is at stake [so in this case, the positive commandment can be violated in order to fulfill it]. 9. [But] this is a mitzvah that comes about through a transgression [and thus is not validly performed]! 10. But a mitzvah of the public [in order to cause the public to fulfill its obligation in some way] is 15

CF puts b. §11 at this point.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

45

different. 11. And R. Yehoshua b. Levi said: “A man should always rise early to the synagogue in order that he might merit to be counted among the first ten. For even if 100 come after him, he will receive the reward of all of them.” The reward of all of them, do you think?! Rather I would say that he receives a reward corresponding to the reward of all of them. 12. R. Huna said: “Nine and an Ark join together [to make 10].” 13. R. Nahman said to him: “And is an Ark a man?” 16 Rather, R. Huna said: “Nine who appear as 10 join together [to make 10].” 14. Some say [9 who appear as 10 are considered 10] when they come together; some say when they scatter [that is, if the 9 look like 10 as they are departing, then they are considered as 10]. 15. R. Ami said: “Two and the Sabbath join together [to make a zimmun of 3].” 16. R. Nahman said to him: “And is the Sabbath a man?” Rather, R. Ami said: “Two Torah scholars who sharpen one another in the Law join together [and are considered as 3].” 16 CF follows this question with R. Ami’s (R. Ashi in CF) tradition about two scholars and the Sabbath. Basically, whereas the printed edition orders the material as: tradition, R. Nahman’s question, and the answer to the question for each of the two traditions, CF orders the material as: traditions, R. Nahman’s questions, and then the solutions.

46

HOW SHOULD RABBINIC LITERATURE BE READ? 17. R. Hisda pointed out: “Like R. Sheshet and myself.” 18. Sheshet pointed out: “Like R. Hisda and myself.”

C. When Can a Minor Be an Adjunct? 11. From when do we make [a minor] an adjunct?17 12. R. Avina said: “R. Huna and R. Yehuda disputed [about this], both [offering interpretations] in the name of Shmuel. One said: “When he knows the nature of a blessing,” and the other one said: “He should know to Whom he blesses.”18 13. R. Nasa said: “How many times did I eat with R. Tahlifa, father, and with Ininya b. Sisi my friend, and they did not include me in the zimmun until I had brought forth two [pubic] hairs.

19. R. Yohanan said: “We include a ‘flowering minor’ in a zimmun [a minor who is showing signs of puberty although he is as yet below the legal age of majority].” 20. It is also taught thus: A minor who has brought forth two hairsȥwe include him in a zimmun. One who has not, we do not count. And we do not carefully scrutinize a minor. 21. There is a difficulty in the body [of this teaching]! You said “If he brought forth two hairs”ȥ yes, “if he did not bring forth”ȥno, and it is then taught “We do not carefully scrutinize a minor”?! [which would leave us ignorant as to whether or not he possessed two pubic hairs]. 22. [“We do not carefully scru-

17 Gen. Rab. 91:4 reads “And up to what point does he read [from the Torah]?,” which, as Theodor points out (p. 1113) is a rather confused reading found in other manuscripts of Gen. Rab. Far better is Bokser’s suggestion in “A Minor for Zimmun” that the correct reading of Gen. Rab. is found in Ms. Vat. 60: “And until when is he counted as a minor?” The difference between Ms. Vat. 60 and the printed edition is how the quf was filled in; the printed edition filled it in as “qoreh,” while Ms. Vat. 60 is more like the Yerushalmi, filling it in as “qatan.” 18 Gen. Rab. 91:4 reverses the order of these traditions.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

47

tinize a minor”] comes to include what? Doesn’t it come to include a ‘flowering minor’? [that is, if a minor has two pubic hairs, we do not carefully scrutinize to find out whether or not he has reached the age of majority]. 23. And the law is not in accordance with all of these traditions; rather, it is in accordance with that which R. Nahman said: “A minor who knows to Whom we blessȥwe include him in a zimmun.” 24. Abaye and Rava b. R. Hanan 19 sat before Rabbah. [He] said to them: “To Whom do we bless?” They said to him: “To the Merciful One.” “And where does the Merciful One reside?” Rava pointed to the ceiling. Abaye went outside and pointed toward Heaven. Rabbah said to them: “The two of you will be rabbis.” It is as people say: The young pumpkin is known by its stalk. D. Groups Consisting of Bread Eaters and Non-Bread Eaters 14. Shmuel b. Shilat asked Rav, and some say they [anonymous scholars] asked Shmuel b. Shilat: “Nine [who ate] bread and one [who ate] vegetables, [can they be counted in the zimmun]?”20

25. R. Yehudah the son of R. Shmuel b. Shilat said in the name of Rav: “Nine who ate grains and one who ate vegetablesȥ they join together [to make a zimmun of 10].”

19 I have rendered this name in accordance with b. Ber. 48a (DS note vav) and CF. 20 Gen. Rab. 91:4 presents this as a straightforward tradition of R. Shmuel b. Shilat.

HOW SHOULD RABBINIC LITERATURE BE READ?

48

15. He said to them: “We include them in the zimmun.”21 16. [They asked again:] “Eight [who ate] bread and two [who ate] vegetables?” 17. He said to them: “We include them in the zimmun.” 18. [They asked again:] “Seven [who ate] bread and three [who ate] vegetables?” 19. He said to them: “We include them in the zimmun.” 20. R. Avina asked: “What about half [who ate bread] and half who ate vegetables?” 21. R. Ze’irah said: “While I was there [in Bavel] I needed [to know the answer to this question], and now I am sorrier than you that I did not ask it.”22

21

26. R. Zera said: “I asked R. Yehudah: ‘What about 8? What about 7?,’ and he said: ‘It doesn’t make a difference.’ Six certainly 23 were not a question for me.” 27. R. Yirmiya 24 said to him: “You did well in not asking [about 6]. There [by the cases of 9, 8, and 7] what was the reason [they could join with eaters of vegetables to make 10]? Because there was a majority [of bread eaters]. Here also [in the case of 6 who ate bread] there is also a majority].” 28. 25 [R. Zera] thought that we require a recognizable majority [and since 6 is very close to half and half, he hadn’t asked because he was sure the answer would be “No”].

In Gen. Rab. 91:4 Shmuel b. Shilat is represented as asking all these questions himself, including one not mentioned in y. Ber about “six [who ate] bread and four [who ate] vegetables.” The questions in Gen. Rab. are left unanswered. 22 As Theodor points out in his notes to Gen. Rab. 91:4 (p. 1114), the text of R. Ze’irah’s statement varies widely and is difficult to interpret. As related in Gen. Rab., R. Ze’irah is saying: “While I was there [in Bavel] I wished to ask you [Shmuel b. Shilat] and was afraid to ask you.” 23 But see b. Ber. 48a (DS note lamed) which points out that this formulation (“shishah vada’i”) is problematic. Judging from his comment (s.v. “ihu savar”) Rashi likely read “voy,” as in “Woe that I didn’t ask him.” CF reads “And I did not ask him about six.” 24 CF reads “Abaye.” 25 CF reads “it is not so (ve-lo hi), we require a recognizable majority.”

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

49

E. Shimon b. Shetach, Yannai and Halakhic Discussion of the Story 22. R. Yirmiya asked: “The one who ate vegetables, can he bless [on behalf of everyone else]?”26 23. R. Yirmiya’s position is reversed [that is, from R. Yirmiya’s question in 22 it appears that he thinks the answer might be “No.” But his comment on the following story indicates that he thinks the answer might be “Yes”].27 24. It was taught: Three hundred Nazirites came up in the days of Shimon b. Shetach and he found an “opening” [to release the vows] for one hundred fifty of them, but for one hundred fifty he did not find an opening. He came to King Yannai. He said to him: “There are three hundred Nazirites here, requiring nine hundred sacrifices. You provide half from your [side], and I will provide half from my [side].” [Yannai] sent him four hundred fifty [sacrificial animals]. A bad report [about Shimon b. Shetach] went out and it was told to [Yannai]: “He did not provide anything from

29. King Yannai and the queen were rolling [and eating] bread together. And since [Yannai] had killed the Rabbis, there was no one who could bless for them. 34 He said to his wife: “Who will bring us a man who will bless for us?” She said to him: Swear to me that if I bring you a man you won’t afflict him.” 35 He swore to her. She brought her brother Shimon b. Shetach and [Yannai] seated him between himself and her. [Yannai] said to him: “Do you see how much honor I have done you?” He said to him: “It is not you who has honored me, but the Torah which has honored me, as it is written, ‘Exalt her and she will raise you up, she will honor those who embrace her’ (Prov. 4:8).” He said to him: “I see that you [still] do not accept authority.” 36 They gave him a cup [with which] to make a blessing. He said: “How should I bless? Blessed is the One from whom Yannai and his companions have eaten?” He

26 In Gen. Rab., R. Yirmiya asks whether one who ate vegetables can be included in the zimmun, not if he can bless. 27 This is missing from Gen. Rab., which nevertheless does include y. §§28-32, which relate the Shimon b. Shetach/Yannai story to R. Yirmiya’s position in y. §22.

HOW SHOULD RABBINIC LITERATURE BE READ?

50

his own [resources].” King Yannai heard and became angry. Shimon b. Shetach was afraid and fled. After some time, some great men of the Kingdom of Persia came to King Yannai. 28 When they were sitting and eating, they said to him: “We recall that there was an old man here who recited words of wisdom to us.” [Yannai] told them the story. They said to him: “Send [for him] and bring him.” He sent [for him] and conveyed to him a sign [so that Shimon b. Shetach would know that Yannai did not intend to harm 28 Gen.

drank that cup, they gave him another and he blessed. 30. R. Abba the son of R. Hiyya b. Abba said: “Shimon b. Shetach acted thusly on his own, for thus did R. Hiyya b. Abba say in the name of R. Yohanan: ‘A person cannot fulfill an obligation for the many unless he himself eats [at least] an olive’s worth of grains.’” 31. 37 They raised an objection: R. Shimon b. Gamliel says: “If he went up and reclined with them ȥeven if he only dipped a small amount and only ate a dried fig with them, he

Rab. reads “approached the table of King Yannai.” Gen. Rab. reads “He [Yannai] said to his sister: ‘Send for him and bring him…’” It is unclear whose sister is meant. 30 Gen. Rab.’s version of the exchange is slightly different. Since Shimon b. Shetach’s last action before the king speaks to him was to sit between the king and queen, Gen. Rab.’s Yannai asks him about that first. Yannai then goes on to ask him why Shimon b. Shetach made fun of him, and then adds a question not found in y. Ber.: “and why didn’t you tell me [that your half of the Nazirites’ requirements would come from your Torah]?,” to which Shimon b. Shetach replied: “Had I told you, you would not have done [your part].” Yannai then goes on to ask Shimon b. Shetach why he fled. 31 In Gen. Rab., Yannai says: “I never heard this matter [=version of the blessing] from you.” 32 In Gen. Rab., it is recounted that a second cup was poured for Shimon b. Shetach, who then blessed “for the food which we ate.” This implies that he drank the first cup as his participation in Yannai’s meal. 33 In Gen. Rab. this tradition is presented without attribution. 34 This sentence is missing from CF. 35 The queen’s statement is fuller and bolder in CF: “What can I do with you since you killed all the rabbis? But swear to me…” 36 CF reads “do not accept us.” 37 CF puts b. §33 here, and introduces §31 with “itmar nami amar R. Huna tanya nami hakhi” rather than the argumentative “metivei.” 29

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS him]. 29 And he came and sat between the king and queen. [Yannai] said to him: “Why did you make fun of me?” He said to him: “I did not make fun of you; [I meant for] you [to provide for the Nazirites] out of your money, and me from my Torah, as it is written (Qoh. 7:12): ‘for the protection of wisdom is as the protection of silver.’” He said to him: “Why did you flee?” He replied: “I heard that my master was angry with me, and I wished to fulfill this verse, ‘Hide a short while until anger passes’ (Is. 26:20). And [Shimon b. Shetach] recited concerning himself: “And the advantage of wisdom is this, in that it protects its owners” (Qoh. 7:12). [Yannai] said to him: “And why did you sit between the king and queen?” He said to him: “It is written in the Book of Ben Sira: ‘the wisdom of the humble man will lift up his head and seat him among the great’” (11:1). 30 [Yannai] said: “Give him a cup and let him bless [for us].” He took a cup and said: “Let us bless for the food that Yannai and his companions ate.” Yannai] said to him: “Are you still being stubborn?” 31 He said to him: “And what should be saidȥ ‘Let us bless] for the food we didn’t eat?’” [Yannai] said: “Give him to eat,” and they gave him, and

51

joins [with them for a zimmun].” 32. He can certainly join with themȥbut he cannot fulfill the many’s obligation for them unless he eats [at least] an olive’s worth of grains. 33. It was also said: R. Hana b. Yehuda said in the name of Rava: “Even if he only dipped a small amount with them and only ate a dried fig with them, he joins [with them to make a zimmun]. But to fulfill the many’s obligationȥhe does not fulfill it unless he eats [at least] an olive’s worth of grains.” 34. R. Hana b. Yehuda said in the name of Rava: “The law isȥ if he ate a vegetable leaf and drank a cup of wine, he can join [with others to make a zimmun]. He cannot fulfill the many’s obligation unless he eats [at least] an olive’s worth of grains.”

52

HOW SHOULD RABBINIC LITERATURE BE READ?

he ate, [whereupon] he said: “Let us bless for the food which we ate.”32 25. R. Yohanan said: “[Sages] disagreed with Shimon b. Shetach.” 26. R. Yirmiya said: “[The scholars disagreed] with the first [blessing by Shimon b. Shetach, in that he said “Let us bless… for the food that Yannai … ate.” Such a blessing violates the principle that a person cannot fulfill an obligation for others unless he too is obligated].” 27. R. Abba said: “[The scholars disagreed] with the second [blessing, because a person cannot fulfill the obligation of Grace for others unless he ate an olive’s worth of grains. Shimon b. Shetach did not, as seen by the fact that he ate without washing his hands].” 28. R. Yirmiya’s position is reversed [earlier in §22, he asked if someone who ate vegetables could bless for others, implying perhaps that the answer is “No.” But here his comment suggests that there was no problem with the sage’s having presumably eaten vegetables and then fulfilling Yannai and company’s obligation]. There [§22] he required [an answer], and here it is clear to him [that Shimon b. Shetach did right in eating vegetables and fulfilling Yannai’s obligation].

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

53

29. That he had required [an answer]ȥwas in accordance with the Rabbis; that it was clear to himȥ was in accordance with R. Shimon b. Gamliel. 30. It was taught: If he went up and reclined and ate with them, even if he did not eat an olive’s worth of grains, we include him in the zimmun (t. Ber. 5:21, in the name of R. Shimon b. Gamliel]. 31. And the words of the Sages [are conveyed by the following:] R. Yaqov b. Aha in the name of R. Yohanan: “We do not include a person in a zimmun unless he eats an olive’s worth of grains.”33 32. But wasn’t it taught: Two [who ate] bread and one [who ate] vegetablesȥwe include them in the zimmun? That baraita is in accordance with R. Shimon b. Gamliel.

THE BAVLI IN RELATION TO Y. BER. AND GENESIS RABBAH – MACRO- AND MICRO-ANALYSIS Two Versions of Yerushalmi Y. Ber. 7:2, 11b is a later version of Gen. Rab. 91:4. Let us look at the following chart:

Gen. Rab. 91:4

y. Ber. 7:2, 11b

1. R. Yehoshua b. Levi’s tradi- 1. Same tion that a minor can be an adjunct to 10. 2. Question about his tradition 2. Same

54

HOW SHOULD RABBINIC LITERATURE BE READ?

from a baraita; answer ultimately attributed to R. Yehoshua b. Levi. 3. Baraita that a minor and a To- 3. Nine who appear as 10 rah Scroll are adjunct; interpreted to mean that a minor can be an adjunct to the seven ascensions to the Torah. 4. Nine who appear as 10.

4. Question to R. Yose about whether analogy can be drawn about minor being an adjunct to 3 from the tradition that a minor can be an adjunct to 10. Analogy rejected.

5. Question to R. Simon about whether analogy can be drawn about minor being adjunct to 3 from the tradition that a minor can be an adjunct to 10. Analogy accepted.

5. Baraita that a minor and a Torah Scroll are adjunct; interpreted to mean that a minor can be an adjunct to the seven ascensions to the Torah

6. Limitation of minor’s “adjuncthood” to Grace; cannot be an adjunct to Shema or to prayer until he brings forth two hairs. R. Nisa’s recollection that he was not counted in a zimmun until he brought forth two hairs.

6. Question: from when can the minor be an adjunct? Answer in the name of Shmuel: when he knows the blessing; other answerȥwhen he knows to Whom to bless; R. Nasa’s recollection that he was not counted in a zimmun until he brought forth two hairs.

7. Question: And until when is 7. Question involving Shmuel he considered a minor? 38 An- b. Shilat: can 9 who ate bread swer in name of Shmuel: until and 1 who ate vegetables be a 38 This

formulation follows Bokser’s emendation according to Ms. Vat. 60; see “A Minor for Zimmun,” p. 11.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

55

he knows how to bless; other zimmun? Eight and 2? Seven and answerȥuntil he knows the na- 3? R. Avina asked about 5 and 5. R. Ze’irah expressed regret ture of the blessing. he hadn’t asked about that in Bavel. 8. Shmuel b. Shilat asks about 91, 8-2, 7-3, and 6-4; R. Avina asks about 5-5, and R. Ze’irah expresses regret he hadn’t asked about that in Bavel.

8. R. Yirmiya asks if one who ate vegetables can bless [on behalf of everyone else]. Y observes that R. Yirmiya’s position on this is reversed, and quotes Shimon b. Shetach story to show this.

9. R. Yirmiya asks if one who 9. R. Yohanan observes that ate vegetables can be counted in Sages disagree with Shimon b. the zimmun. Shetach. Y goes on to draw out the implications of its earlier observation that R. Yirmiya’s position is reversed. Second version of R. Yirmiya’s opinion is in accordance with R. Shimon b. Gamliel. Sugya ends with discussion of latter.

10. Shimon b. Shetach story. 11. Same as y. 9. Y. Ber. and Gen. Rab. are clearly versions of the same sugya, despite the differences between them. What we find in Gen. Rab. is a version of Yerushalmi, not a coincidentally similar sequence of materials found in both compilations. This sugya is integrally connected to m. Ber. 7:2, not to Gen. 42:4-5. It is difficult to understand why anyone would have formulated this sugya to explain the scriptural verses; on the other hand, it is understandable how this sugya may have been formulated in order to take up (in order) two issues found in (or called for by) m. Ber. 7:2: minors and the amount and the type of food that entitles one to be counted in (or

56

HOW SHOULD RABBINIC LITERATURE BE READ?

lead) a zimmun. 39 Moreover, a redactional hand likely shaped this sugya. No amora is represented as being aware of this sugya qua sugya; the selection and movement from one topic to the other was more likely an editorial construct.40 Of the various differences between the two sugyot, some differences point convincingly to the conclusion that y. Ber. is later than Gen. Rab. Baruch Bokser came to this conclusion nearly twenty-five years ago on the basis of his detailed study of the first part of this sugyaȥthe minor and zimmun. Bokser studied the redactional placement of Shmuel’s tradition (that the minor must know either the character of the blessing or to Whom the blessing is recited) in both sequences with the assistance of Ms. Vat. 60. 41 Whereas y. Ber. deploys Shmuel’s tradition as an answer to the question: “From when [do minors] count as adjuncts?,” Gen. Rab. (according to Ms. Vat. 60) deploys it to answer the question “And until when is he counted as a minor?” As Bokser noted, y. Ber. was deploying Shmuel’s tradition in furtherance of a Palestinian amoraic concern, while Gen. Rab. preserved a more original redactional placement of Shmuel’s tradition as a comment directly upon the mishnah. The mishnah mentions minors; hence it was logical for an early amora such as Shmuel to explore the question of just how long a child is considered a minor. Bokser emphasized that these two versions were the results of diachronic development rather than the versions of two contemporary study-circles: “[a]s the Gen. R. version represents a more fundamental form of the tradition and is logically prior to that of PT it is preferable. …[w]e, therefore, con39 See Albeck, introduction to Gen. Rab., 3:73. This conclusion is not undermined by Bokser’s conclusion that this sugya “forms an integral part of Gen.R. …,” see “A Minor for Zimmun,” p. 24. The sugya may very well have been introduced into Gen. Rab. at an early stage, but it was introduced; it most likely was not generated as part of the midrashic exegesis of Gen. 42:4-5. 40 Our point in stressing the redactional as opposed to amoraic role in formulating this sugya is not to suggest that the sugya was formulated and transmitted in writing, but simply to emphasize that the sugya as a constructed composition consisting of different topics and sources is more likely the product of a post-amoraic editor/redactor than of an amora, not one of whom is represented as having any awareness of this lengthy sugya qua sugya. 41 On Ms. Vat. 60, see Bokser, “A Minor for Zimmun,” p. 9 n. 22.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

57

cluded that PT adapted the version preserved in Gen. R.” (emphasis in original).42 We may extend the conclusion that Gen. Rab. is prior to y. Ber. by studying the entire sugya. In Gen. Rab., R. Yirmiya asks whether one who ate vegetables can be counted in the zimmun. But this is exactly the question just answered by Shmuel b. Shilat. Why would this question be repeated? Y. Ber., more logically, has R. Yirmiya ask whether one who ate vegetables can bless for everyone, which has not been previously asked. Second, Gen. Rab.’s Shimon b. Shetach drinks wine and eats nothing, which makes it difficult to understand how Gen. Rab. uses that story to understand R. Yirmiya, whose question was about a diner who ate vegetables.43 Y. Ber.’s Shimon b. Shetach specifically calls attention to the fact that he hasn’t eaten, whereupon Yannai orders him fed. He eats without washing his hands (thus he likely ate something other than bread) and then led Grace; his conduct thus can shed light on R. Yirmiya’s question about eating vegetables and blessing on behalf of everyone else. Y. Ber.’s version of this sugya is simply more logical than what we find in Gen. Rab. It is easy to understand how y. Ber. improves Gen. Rab.’s presentation of these two points but it is difficult to understand what anyone would have gained by moving from y. Ber.’s version to that of Gen. Rab.44 It may be objected that these differences between Gen. Rab. and y. Ber. are simply too small to be of much significance in determining chronological priority and that it is more reasonable to see these sugyot as different contemporaneous versions of the same sugya. Bokser, as noted above, was moved to respond to this objection in his 1979 essay, and the objection requires response in the 42

Bokser, “A Minor for Zimmun,” p. 22. Irit Aminof made a similar point in her fine analysis of these versions. She points out that Gen. Rab. responds to Shimon b. Shetach’s hint that he hasn’t eaten (“what do you want? Can it be said ‘Let us bless for food I did not eat?!’”) by having him drink wine as his participation in the meal, which does not fit as well as y. Ber.’s specific reference to his eating food. See Irit Aminof, “Aggadat Shimon b. Shetach ve-Yannai ha-Melekh ve-’Ibbuda’ be-Sefer ha-Aggadah,” Alei-Si’ah 24 (1986):119-120. My thanks to Dr. Hillel Newman for directing me to this article. 44 See also Irit Aminof, who agrees that Gen. Rab. is prior to y. Ber., pointing out that the “clear and complete version is the version of the Yerushalmi…,” “Aggadat Shimon ben Shetach,” p. 119. 43

58

HOW SHOULD RABBINIC LITERATURE BE READ?

present essay as well. The key to resolving this issue, as Bokser implied, is to look to the nature of the differences between the sugyot. Let us therefore generalize a bit from the analysis presented above: If the differences between the sugyot are such that (1) one sugya preserves a “more fundamental form” of a tradition than the other (such as Gen. Rab.’s preservation of a question pertaining to m. Ber. 7:2 and y. Ber.’s transformation of that question into one about an amoraic issue); (2) one sugya eliminates a redundancy found in the other by altering the redundant material so as to teach something new (such as y. Ber.’s version of R. Yirmiya’s question, according to which the issue is whether one who ate a food other than grains can bless for everyone else, unlike Gen. Rab.’s “count in the zimmun”); or (3) an apparently illogical point in one sugya appears logically formulated in the other (such as Gen. Rab.’s reference to Shimon b. Shetach drinking wine, which is altered in y. Ber. to his eating)ȥthen we are justified in hypothesizing that there was a diachronic development from one version of the sugya (Gen. Rab. in this case) to the other (y. Ber.). These three types of difference are not random and in this case they all point in the same direction: to y. Ber.’s improvement upon Gen. Rab.’s presentation, and hence to Gen. Rab.’s chronological priority to y. Ber.45 While methodological rigor is rightly called for, it seems reasonable that in this case, Gen. Rab. can be seen as chronologically prior to y. Ber.46 45 There is another noteworthy difference between y. Ber. and Gen. Rab.: in the latter Yannai tells “his sister” to summon the sage, while in y. Ber. the Persian notables tell Yannai to summon him, which he does directly. While it is tempting to suggest that y. Ber. may have eliminated the sister in order to make the story flow more smoothly (her role in summoning the sage is easily filled by the Persians and Yannai and she plays no further role in the story), this temptation must be resisted because one can just as easily argue the opposite: that Gen. Rab. “invented” the sister in order to account for the illogical fact that in y. Ber., Shimon b. Shetach fled from Yannai and yet the king was able to find him so quickly without any assistance. Therefore, since it is not possible to use this difference between the versions to decide chronological priority, the matter must be left as a “teiku,” as it were. 46 Indeed, it is precisely the concern for methodological rigor that leads me to eliminate from consideration another possible sign of Gen. Rab.’s chronological priority to y. Ber.: the absence from Gen. Rab. of a transition to the Shimon b. Shetach story, and the presence of such a transition “machlifa shitateih de-rav Yirmiyah” or “R. Yirmiya’s position is re-

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

59

Orality studies, which is gaining ground as a field valuable to rabbinics, affords another, albeit related, objection. From the perspective of orality studies, the differences between the sugyot can be assumed to stem from different oral performances of the same sugya in different contexts. What we have now in Gen. Rab. and y. Ber. are therefore two distinct “snapshots” of an ongoing series of oral performances of this sugya that are now preserved for us in these two Palestinian compilations. Relatedly, it is likely that a scholar arguing from within orality studies would object that the diachronic development model we have posited is too “literary” an account of the relationship between Gen. Rab. and y. Ber., which should be understood as more fluid, more “oral.” There is little doubt that given the current state of our knowledge it is safer to presume that these sugyot were transmitted orally, and that we are looking here at two snapshots of a sugya that may well have also circulated in other largely similar although subtly different versions now lost to us. But it is important that we not stop at these observations, but pay careful attentionȥas we did beforeȥto the nature of the differences between the sugyot. If the vagaries of oral transmission were responsible for all the differences between the sugyot, then we would reasonably expect them to be more random; in fact, we observed three key differences that suggest y. Ber. is a secondary development beyond Gen. Rab. Thus, while the changes made to y. Ber. were likely made orally, orality as such does not explain the changes, which were more likely improvements upon an earlier recitation of the sugya (Gen. Rab.). What about the Bavli? Let us examine the following outline of the Bavli sequence of sugyot, which highlights the features the Bavli has in common with the Palestinian sugya (in both its versions): 1. R. Yose’s tradition about a minor “in a crib” and R. Yehoshua b. Levi on a minor as an adjunct to ten. versed”) in y. Ber. One could argue that the absence of such a transition in Gen. Rab. is a lacuna that y. Ber. filled. Yet I am disinclined to hold that view because Gen. Rab. does after all have that line at the end of its version of the Shimon b. Shetach story, and the absence from Gen. Rab. of that line immediately preceding the story could be due to a number of causes: scribal error, the vagaries of oral transmission, etc. Moreover, and of greater importance, the presence of that line in y. Ber. does not mark a significant development in the presentation of the Shimon b. Shetach story in the two sugyot, unlike the three changes we have studied above.

HOW SHOULD RABBINIC LITERATURE BE READ?

60

2. R. Huna’s tradition that “nine who appear as ten” join together for a zimmun. 3. R. Yohanan’s tradition about the “flowering minor,” baraita that a minor with two pubic hairs can be counted in zimmun; baraita that “ein medaqdiqin baqatan.” 4. R. Nahman’s tradition that a minor who knows to Whom we bless may be counted for a zimmun. 5. R. Yehuda b. R. Shmuel b. Shilat in name of Rav: Nine who ate bread and one who ate vegetables join together. R. Zera says that he asked R. Yehuda about eight and two, and seven and three. He says that six and four were certainly not an issue for him. R. Yirmiya approves of his omission on the ground that just as in the other cases (nine, eight, and seven) there was a majority of grain-eaters (and thus the joiningtogether was alright), so would it be in the case of six, in which there was a majority of grain-eaters. But R. Zera was concerned that perhaps a “recognizable” majority was needed (and six is not a “recognizable” majority). 6. The Shimon b. Shetach story, in which the sage drinks wine and eats nothing. 7. R. Abba b. R. Hiyya b. Abba states that Shimon b. Shetach acted for himself (i.e., no one agreed with his action), citing a tradition of R. Yohanan. 8. R. Shimon b. Gamliel quoted in baraita to support R. Abba’s objection to Shimon b. Shetach; followed by amoraic teachings to the same effect. The progression of topics clearly mirrors that of the Palestinian sugya mapped out above. Not only that, but in §§5-8 of this outline we even see a similarity in attributions: the names “Shmuel b. Shilat,” “R. Zera (=Ze’irah),” and “R. Yirmiya” appear; the latter is even the last-quoted amora cited prior to the Shimon b. Shetach story, as in the Palestinian sugya.47 “R. Abba” raises an objection to the story, just as “R. Abba” was alleged to have done in the Pales-

47

But see Codex Florence, which reads “Abaye” instead of “R. Yirmiya.”

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

61

tinian sugya, and both sugyot end with discussion involving R. Shimon b. Gamliel. What accounts for the strong similarity between the Babylonian and Palestinian sugyot? The similarity is not due to m. Ber. 7:2, which calls for discussion of minors and the type of food that diners must eat in order to be counted, but lacks the specific sequence of topics we see here. Nor is there a sequence of baraitot or other tannaitic materials that calls for it. The tannaitic materials in the Bavli and Palestinian sugyot are fully integrated into the give-andtake of the sequence, and did not generate it. “Early talmud”ȥthe hypothesis that the similarities between the sugyot are due to the transmission of the Palestinian sugya to Babylonia during the amoraic periodȥis a possible explanation that draws strength from the fact that the Bavli sugya is more similar in part to the earlier Gen. Rab. A proponent of “early talmud” may argue that the sugya was put together (in Gen. Rab.’s version), transmitted to Babylonia during the amoraic period, and ultimately reworked one way in Palestine and differently in Babylonia. But “early talmud” also fails to persuade. First, although the evidence of the nehutei (“those who went down,” i.e., “those who migrated from Palestine to Babylonia”) proves that scholars did transmit traditions and even whole sugyot between the two rabbinic centers during the amoraic period, this evidence does not permit the inference that the nehutei also transmitted edited sequences of two or more topics such as we see in this example.48 Looking at the materials attributed to R. Dimi (the nehuta who looms largest in the Bavli), we see a scholar who conveyed individual legal traditions, sometimes stories or whole sugyot, and sometimes discrete aggadic 48

See Jeffrey Rubenstein, “Some Structural Patterns of Yerushalmi Sugyot,” in Peter Schäfer, ed., The Talmud Yerushalmi in Graeco-Roman Culture III (Tübingen: Mohr Siebeck, 2003), pp. 308-309, who points out that the presence in the Bavli of complex sugyot also found in the Yerushalmi should provoke a reassessment of either the nature of the Yerushalmi’s redaction or of the types of sugyot produced and circulated by amoraim. Given the current state of our knowledge, I am more inclined to hold that the presence of complex sugyot in the Yerushalmi is a result of redactional work. As I have already noted, a sugya consisting of amoraim of various generations which also demonstrates a lack of amoraic awareness of the whole sugya qua sugya is much more likely the work of redactors than of amoraim.

HOW SHOULD RABBINIC LITERATURE BE READ?

62

traditions. 49 Transmitting edited sequences of topics such as the sugya we are studying here was not attributed to him. And if this sort of activity was not attributed to R. Dimi, the nehuta who looms largest, what can we say of the others? Second, if the transmission of this complex sugya was not done by nehutei, then how was it done? Neither y. Ber. nor Gen. Rab. nor the Bavli portray any amora as aware of the sugya qua sugya, and the sugya consists of amoraim of various generations. The most logical answer to the question of how the Palestinian sugya came about is that it is the work of redactors, and the lack of amoraic awareness of the sugya as a whole in the Bavli suggests that it was Bavli redactors who appropriated the sugya for the Bavli. To reiterate, although there is evidence that some Palestinian and Babylonian amoraim did function as editors,50 this activity does not seem to have extended so far as the creation of extended sugyot and their attachment to particular mishnayot. It is our good fortune in studying this Bavli sugya to have not one, but two parallels of the genre “Talmud Yerushalmi.” These two parallels help us see the fluidity that results from the (likely oral) transmission of the Palestinian sugya to Babylonia, as well as 49

R. Dimi’s activity in transmitting discrete traditions looms largest. I will only provide a few examples: b. Ber. 6b, 44b; b. ‘Eruv. 3b, 77a, 87a; b. Yev. 78a, 84b; b. Sanh. 7b, 57a, 69a; b. Tem. 12b-13a; b. Nid. 25a. R. Dimi is represented as recounting stories at b. Ber. 44a; b. Shab. 13b (he comments on it, which implies that he knows it), b. Shab. 50a, 74a, 125b, 147a; b. ‘Eruv. 86b; b. Suk. 16b; b. Yev. 59b; b. Qidd. 31a; b. AZ 8b, 35b. He is represented as conveying sugyot at b. Shab. 76a; b. Moȧed Qat. 10a; b. Ket. 34b-35a, 57a, 104b, 107b; b. B. Bat. 27b, 129a, 152b; b. Sanh. 70a; b. Zev. 10a; b. Men. 71b; b. Hul. 103b, 134a; b. Me’ilah 21b. Finally, R. Dimi is represented as transmitting aggadic traditions at b. Ber. 31b; b. Hag. 14a; b. Ket. 105b, 111b; b. B. Bat. 74b-75a, 79a; b. Sanh. 100a, 108a; b. AZ 8b, 35a; and b. Zev. 118b. 50 Evidence suggests that later Palestinian amoraim may have functioned as editors, see, e.g. y. Shev. 4:2, 35a (sugya on martyrdom apparently stemming from fourth Palestinian amoraic generation). See also Richard Kalmin, Sages, Stories, Authors, and Editors in Rabbinic Babylonia (Scholars Press, 1994), pp. 169-173, who demonstrates that middle generation Babylonian amoraim may have functioned as editors. But the available evidence does not go nearly so far as to support Noah Aminoah’s contention that Abaye and Rava functioned as editors in Babylonian-adapted Palestinian sugyot in which they appear; see, e.g. “Qit’ei Talmud Qadum ba-Massekhet Rosh Hashanah,” p. 188.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

63

the close similarity the Bavli sugya retains to its predecessors. When we scrutinize the Bavli closely, we see that it more closely resembles y. Ber. in two key ways and Gen. Rab. in two others. Like y. Ber., the Bavli deploys R. Nahman’s tradition (a minor who knows to Whom we bless is counted in a zimmun)ȥone of two versions of Shmuel, per the Palestinian sugyaȥ in order to establish from when a minor may be included in a zimmun.51 Also, the Bavli’s halakhic interest in the Shimon b. Shetach story is the issue of whether a diner who consumed something other than grains can bless after the meal on behalf of everyone else who did eat grains, which is y. Ber.’s halakhic interest as well. But as in Gen. Rab., the Bavli’s Shimon b. Shetach drinks wine as his participation in Yannai’s meal. Like Gen. Rab., the Bavli also includes mention of the sibling relationship between the queen and the sage, which is consistent with its representation of this relationship elsewhere (b. Sot. 47a). Summing up, the Bavli sugya is not an entirely original creation of the Bavli, but is a version of the older Palestinian sugya. More to the point, its similarities to and differences from the two versions of the Palestinian sugya show it to be an interesting hybrid of the two. We can therefore conclude that it is based on the Palestinian sugya, although it is impossible to conclude that it more closely resembles y. Ber. or Gen. Rab.

THE DIFFERENCES BETWEEN THE BAVLI AND THE PALESTINIAN SUGYOT Having demonstrated the Bavli’s similarity to the Palestinian sugya, we must now explain the obvious differences between them. To the extent that difference can be explained on linguistic, cultural, historical, or intellectual grounds, it is not a powerful argument against the theory that the Bavli relied on the Palestinian sequence of sugyot.

51

We will later explore why the Bavli deploys R. Nahman in the context of a zimmun of three rather than one of ten, as does y. Ber.

64

HOW SHOULD RABBINIC LITERATURE BE READ?

Linguistic and Terminological Differences The Yerushalmi has its own characteristic terms and expressions that are not usually found in the Bavli.52 In the lengthy passage we are studying, the Yerushalmi terms “Rabbi X ba’a’ qomei Rabbi Y” (“Rabbi X asked Rabbi Y”), “kini matnita” (“thus is the [proper interpretation of the] baraita”), “itpalgun Rabbi X and Rabbi Y” (“Rabbi X and Rabbi Y disputed”), “had amar … ve-haranah amar …” (“One said… and the other one said…”), “machlifa shitateih de Rabbi X” (“The position of Rabbi X is reversed”), are not found in the Bavli parallel. Similarly, the Bavli has its own characteristic terms and expressions that are not necessarily found in the Yerushalmi. I will divide my discussion of these terms into two groups: (1) terms that function to move along the Bavli argument; (2) other terms and expressions (such as popular sayings) found only in the Bavli. Terms That Signal Movements in the Bavli Argument The term “ve-heikhi ‘aveid hakhi?” (“and how did he do thus?”) at b. Ber. 47b is found thirty-one times in the Bavli53 and not at all in the Yerushalmi. Similarly the dialectical term “[X] salqa da’atakh? Ela eima [Y]” (“Does X occur to you? Rather I would say Y”) at 47b is also found thirty-one times exclusively in the Bavli. “Ha gufa qashya” (“there is a problem in the body [of the tradition]”) is found fifty-nine times in the Bavli alone, and the term “ve-hadar tani” (“and it was then taught”) found as part of the dialectical move beginning “ha gufa qashya” on 47b is found forty-six times only in the Bavli. The term “la-atuyei mai?” (“to include what?”) at 47b-48a is found ninety-eight times only in the Bavli, while the precise terminological sequence we find here (“la-atuyei mai? Lav la-atuyei X”) is rarer, appearing only sixteen timesȥbut all of them in the Bavli.54 52 But see b. Zev. 9a, where R. Avin b. Hiyya or R. Avin b. Kahana is represented as using the Palestinian formulation: “be-khol atar at amart ‘X’ ve-kan ‘Y.’” 53 This and all succeeding numerical counts in this section come from the Bar-Ilan Responsa Project, version 9.0 (Bar-Ilan University: 19722001). 54 In his Tequfat ha-Savoraim ve-Sifrutah (Jerusalem: Dror, 1973), Yakov Ephrathi devoted considerable space (pp. 159-277) to a discussion of the sugyot “zeh ha-klal la-atuyei mai?” (“this is the general principle [of the mishnah, which comes] to include what?”) Ephrathi’s view is that the question

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

65

The anonymous statement “ve-leit hilkheta ke-khol hanei shematta” (“and the law is not like all of these traditions”) at 48a, which announces a halakhic repudiation of traditions presented up to that point in the Talmud in favor of one about to be announced, appears six times only in the Bavli.55 Finally, “itmar nami” (“it was also said”ȥheralding the quotation of an amoraic memra) at 48a is found seventy-nine times in the Bavli alone. Thus the Bavli modified the Palestinian sugya by adding its own distinctive dialectical terminology, as it does quite often.56 The differences between the Bavli and Yerushalmi sugyot traceable to this terminology ultimately support the claim that the Bavli relied on the Palestinian sugya. Other Terms and Expressions Found Only in the Bavli At 47b the Bavli’s version of this sugya opens with R. Yose’s reference to the “qatan mutal ba-’arisa” (“the minor lying in his crib”), which is a term found only in the Bavli (twice at 47b, and at b. B. Bat. 36a and b. B. Bat. 131b). The redactors introduce the concept “mitzvah ha-ba’ah ba-’aveirah” (“a mitzvah that comes about through a transgression”) into a discussion at 47b about R. Eliezer’s freeing of his slave in order to complete a quorum of ten for prayer. I will defer until later an analysis of this “mitzvah ha-ba’ah ba-’aveirah” sugya as a whole, which I will argue is a Bavli interpolation of a Babylo“la-atuyei mai?” is late, and probably post-amoraic, and was thus not known to the Yerushalmi (pp. 162-164). But see b. B.Bat. 86a, where Ravina appears to use “la-atuyei mai?” in a question to R. Ashi. If that attribution is reliable, the phrase is not purely post-amoraic. 55 Per Otzar ha-Geonim: Berakhot (Teshuvot), siman 314, it is possible that this language is an accretion to the Bavli from a responsum of R. Natronai Gaon. If this is so, then the presence of this phrase in the sugya may not reflect redactional activity, but rather the literary “blowback” of the (likely) oral study of the Bavli together with pertinent Geonic materials in the Babylonian yeshivot. 56 For detailed analyses of the Bavli’s tendency to “Bavlicize” earlier materials, see Shamma Friedman, Talmud ‘Arukh 2:7-23; idem, “haBaraitot ba-Talmud ha-Bavli ve-Yahasan le-Maqbiloteihen she-beTosefta,” in Daniel Boyarin, et al., eds., ‘Atarah la-Hayyim: Mehqarim baSifrut ha-Talmudit ve-ha-Rabbanit le-Khvod Professor Hayyim Zalman Dimitrovsky (Jerusalem: Magnes Press, 2000), pp. 163-201; and idem, “Uncovering Literary Dependencies in the Talmudic Corpus,” in Shaye Cohen, ed., The Synoptic Problem in Rabbinic Literature, pp. 35-57.

HOW SHOULD RABBINIC LITERATURE BE READ?

66

nian sugya into the older Palestinian sequence of sugyot. My focus now is on the wording of the concept itself. Although this concept has Palestinian antecedents (y. Hal. 1:9, 58a; y. Shab. 13:3, 14a) expressed as “ein ‘aveirah mitzvah” or “ein mitzvah ‘aveirah,” the formulation “mitzvah ha-ba’ah ba-’aveirah” is uniquely Babylonian (b. B. Qam. 94a; b. Ber. 47b; and b. Suk. 30a). A related Babylonian expression of the concept (“davar ha-ba ba-’aveirah”) is attributed to Rava at b. Tem. 20b. Further down on 47b the Bavli introduces alternative understandings of R. Huna’s view that “nine who appear as ten join together [in a zimmun].” One view is that the nine appear to be as ten “ki mekhanfi,” “when they are gathered together,” and the other is that the nine appear as ten “ki mevadrei,” “when they separate.” At b. Sanh. 8a the redactors use the same verb roots in discussing the courts of three and/or twenty-three who must adjudicate a case of “motzi shem ra’.” At one point in the discussion the redactors there ask: “With what are we dealing here? With a case in which they gathered [“de-ikhnif”] as twenty-three in order to adjudicate a capital case, and separated [“ve-ibdur], [and the husband] said, ‘adjudicate a monetary case for me at least.’” The use of these verb roots in analyzing a group which must be of a certain size to fulfill a particular religious task is unique to the Bavli. Shortly after this in 47b R. Ami is represented as saying that “Two Torah scholars who sharpen one another in the law join together [to make a zimmun of three].” The expression “two Torah scholars who sharpen one another in the law” (“shnei talmidei hahamim ha-mehadidin zeh et zeh ba-halakhah”) appears in only one other place in rabbinic literature – at b. Tan. 7a. The term “hainu de-amrei inshei” (“thus do people say”) at 48a, which introduces a popular saying, is found fifty-five times, only in the Bavli.57 The saying itself is found only in the Bavli. The Bavli on the Minor and Zimmun (And Other Zimmun Issues) as Compared to the Palestinian Compilations Y. Ber. and Gen. Rab. have largely similarȥalthough not identicalȥ halakhic resolutions about zimmun. They share the notion that a minor can be an adjunct for ten, and both present R. Yudan’s view that the minor can be an adjunct for the seven weekly Torah lec57

Source is the Bar-Ilan Responsa Project.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

67

tions. Gen. Rab. adds that a minor can be an adjunct for the Shema and Tefillah if he has two pubic hairs. Y. Ber. presents two positions on when a minor can be an adjunct: two versions of Shmuel (he knows the blessing or he knows to Whom to bless), and the recollection of R. Nasa that he was not counted until he had two pubic hairs.58 Both compilations hold that mixed dining parties in which a majority ate grains (9-1, 8-2, and 7-3; Gen. Rab. adds 6-4) can constitute a zimmun. Y. Ber., as noted, improves on Gen. Rab. by having R. Yirmiya ask whether a diner who ate vegetables can bless on behalf of everyone else (as opposed to Gen. Rab.’s “count in the zimmun”). Both versions of Yerushalmi observe that R. Yirmiya’s position on a diner who ate vegetables (whether he can count in the zimmun or lead the Grace) is equivocal and root that equivocality in R. Yirmiya’s resort to either the view of the Rabbis (who hold that only diners who ate grains can count) or that of R. Shimon b. Gamliel (who was more lenient). But Yerushalmi adds its own layer of equivocality due to its ambiguous use of terminology: Yerushalmi uses the terms “mezamnin” or “mezamnin ‘alav” in its discussion of R. Yirmiya, neither of which mean “bless” (as in R. Yirmiya’s question in y. Ber.). This forces P’nei Moshe (for example) in y. Ber. to hypothesize that Yerushalmi holds that the Rabbis and R. Shimon b. Gamliel dispute both about counting in a zimmun and about blessing (leading)ȥalthough that double dispute is not explicitly mentioned in the Talmud. A similar confusion is apparent in Gen. Rab. The Bavli’s emendations both signal its halakhic differences with Yerushalmi and cut through the terminological ambiguity. First, the Bavli decisively rejects59 the Palestinian perspective that a minor can be an adjunct for ten. Thus, whereas y. Ber. had deployed Shmuel’s views about the minor (he may be an adjunct to 10 either when he knows the blessing or knows to Whom to bless) 58 Gen. Rab., it will be recalled, deployed these positions to determine how long a child is considered a minor. See Bokser, “A Minor for Zimmun,” p. 11. 59 This interpretation of the Bavli’s position makes the most sense on the basis of a straightforward reading of the text. But for other readings and their halakhic ramifications, see B.M. Lewin, ed., Otzar ha-Geonim (Berakhot), Perushim: 81, Tosafot to b. Ber. 48a, s.v. “ve-leit,” and Menahem b. Shlomo ha-Me’iri, Beit ha-Behirah (Berakhot), ed. Shmuel Daikman, p. 180.

68

HOW SHOULD RABBINIC LITERATURE BE READ?

in connection with his being an adjunct to 10, the Bavli presents the similar view of R. Nahman (the minor may count in a zimmun when he knows to Whom to bless) in connection with a zimmun of 3.60 The Bavli also rejects R. Yohanan’s notion that a “flowering minor” (a minor with pubic hairs) can be counted in a zimmun (b. §23). The Bavli instead presents R. Nahman (“a minor who knows to Whom we bless is counted in a zimmun”) as the law. We thus see clear halakhic differences between Yerushalmi and Bavli, which also explain the Bavli’s omission of Gen. Rab.’s notion that a minor with two pubic hairs can be an adjunct for Shema and Tefillah. The Bavli amplifies this halakhic difference with its story of Abaye and Rava b. R. Hanan. The deployment of Abaye and Rava as part of an “update” of Palestinian materials is seen elsewhere, 61 and the Bavli’s placement of this view last could be an indication that this is its halakhic stance.62 On 48a, the Bavli’s version of the Shmuel b. Shilat tradition differs both from y. Ber. specifically (in that R. Zera specifically mentions that he didn’t ask about 6-4, about which R. Yirmiya agrees) and from both y. Ber. and Gen. Rab. in that the Bavli does not deal with the issue of 5-5. The Bavli’s omission of 5-5 is likely due to the anonymous voice’s focus on “majority” and “we require a recognizable majority,” which renders the issue of 5-5 moot. Moreover, the Bavli avoids the terminological confusion engendered by Yerushalmi’s ubiquitous use of “mezamnin” by consistently using the distinct terms “mitstarfin” (“join together”), “b-r-kh”

60 The attribution of Shmuel’s view to R. Nahman is unsurprising for as Prof. Richard Kalmin has pointed out, R. Nahman quotes Shmuel more frequently than he quotes anyone else (personal communication). 61 See e.g. y. AZ 1:1, 39b and its reworking at b. AZ 6b (Abaye and Rava deployed in the Bavli sugya); y. AZ 5:1, 44c and its reworking at b. AZ 62a-b (same); and y. Shev. 4:2, 35a-b and its reworking at b. Sanh. 74a-75a (Abaye and Rava are deployed twice). This deployment of Abaye and Rava as the “Babylonian updates” to Palestinian materials requires further research. Noah Aminoah’s assumption that their presence in a sugya with a Palestinian parallel marks them as the Babylonian editors of that sugya is an over-reading of the evidence, since nothing in their portrayals shows any sort of editorial activity. See, e.g., Aminoah, “Qit’ei Talmud Qadum,” p. 188. 62 See e.g. Bokser, “A Minor for Zimmun,” pp. 14-15 n. 39 and sources cited. This issue also requires further research.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

69

(“bless”), and “le-hotzi… yedei hovatan” (“to fulfill others’ obligations”). Thus the Bavli makes clear that the questions about the mixed dining parties of 9-1, 8-2, and 7-3 are about whether they “mitstarfin.” A propos of this point, the Bavli eliminates R. Yirmiya’s question about the diner who ate vegetables, which would be superfluous in the Bavli in either Palestinian version. The Gen. Rab. version would be superfluous because it was answered by R. Yehuda b. R. Shmuel b. Shilat, and y. Ber.’s version is rendered superfluous because the Bavli emends Yannai’s initial statement to read “who will bring us a man who will bless for us?,” who turns out to be Shimon b. Shetach.63 The question is thus asked without the need of Yerushalmi’s R. Yirmiya. By eliminating R. Yirmiya’s question, the Bavli has eliminated y. Ber.’s framing of the Shimon b. Shetach story as a story meant to explore that question.64 Thus, after the story, the Bavli explores Shimon b. Shetach’s halakhic stance directly, unlike Yerushalmi, which discussed R. Yirmiya’s. Moreover the Bavli carefully avoids the terminological confusion in Yerushalmi’s discussion of the Rabbis and R. Shimon b. Gamliel (in relation to R. Yirmiya) by adhering to its earlier distinction between “mitstarfin” and “le-hotzi… yedei hovatan.” The Bavli even emends the text of R. Shimon b. Gamliel’s baraita to add the word “mitstaref,” which it then uses as a springboard to insist that the tanna was only talking about “joining,” not about “fulfilling the many’s obligations”; a position it maintains in the linguisticallymatched amoraic traditions that follow.

63

Although Gen. Rab.’s Yannai tells the sage to bless, while y. Ber.’s Yannai orders him to be a given a cup for that purpose (y. §24), only in the Bavli is Shimon b. Shetach summoned into the king’s presence for the express purpose of “blessing for us.” As noted, this emendation signals that the Bavli’s interest in the story is the sage’s blessing on behalf of everyone else; thus this emendation does the work that R. Yirmiya’s question about the diner who ate vegetables did in y. Ber. 64 Thus as was pointed out earlier, the fact that neither the Bavli nor Gen. Rab. have such a transition to the Shimon b. Shetach story does not mean that the Bavli is similar to Gen. Rab. in this respect; it may simply mean that the Bavli’s elimination of y. Ber.’s formulation of R. Yirmiya’s question created this lack of transition.

HOW SHOULD RABBINIC LITERATURE BE READ?

70

The Bavli’s Digression from the Minor and Zimmun (b. §§4-18) As Tosafot already recognized,65 the Bavli includes material at 47b which is not relevant to the issue of the minor and zimmun. Except for R. Huna’s tradition that “nine who appear as ten” should be considered a zimmun of ten,66 the other material in b. §§4-18 appears in neither y. Ber. nor Gen. Rab. The Bavli redactors likely added this material to the pre-existing Palestinian sugya, as the following analysis makes clear. The first point to note is the Bavli redactors’ use of association to introduce two teachings of R. Yehoshua b. Levi that are unrelated to the issue of the minor and zimmun. The first, found above at §4 of b. Ber., is that “nine and a slave join together [to make a ritual unit of ten].” It is understandable how this tradition ȥespecially in an oral contextȥwould have come to be associated with the one about the minor and zimmun, as it deals with a ritual unit of ten and a person (the slave) who may join although not on the same level as the other nine. The concept of a ritual unit (for prayer) of ten (a “minyan”) is the associative link for the second added tradition of R. Yehoshua b. Levi (b. Ber. at §11) about the desirability of being counted among the first ten each morning in 65

See Tosafot to b. Ber. 48a, s.v. “ve-leit.” R. Huna’s teaching about “nine who appear as ten” is interesting on its own terms. R. Huna is first represented as holding that “nine and an Ark (“aron”) join together [to make a zimmun of ten].” R. Nahman challenges him with the rhetorical question as to whether an Ark is a human being, whereupon the Bavli re-presents R. Huna as teaching “nine who appear (“nir’in”) as ten [join together].” In the Hebrew there is an orthographic similarity between “aron” (consisting of the root “aleph-resh-nun”) and “nir’in” (consisting in part of the letters “resh-aleph-nun”). Although the matter requires more research, it is interesting to speculate that the orthographic similarity caused a confusion to arise in R. Huna’s teaching. This confusion and its resolution are represented here as R. Huna’s having taught “aron” and being corrected to “nir’in,” which is the “correct” version of the teaching as presented in the Yerushalmi and Gen. Rab. While examining Codex Florence, I noticed that a similar orthographic confusion may have occurred in R. Ami’s teaching. In CF, the letter “nun” found in the word “shnei” (as in “shnei talmidei [hahamim]”) resembles a “bet,” so that the “shin-nun-tof appear at first glance to be “Shabbat.” If this reconstruction is correct, it may be an interesting datum relevant to the question of whether or not the Bavli or any portion of it was transmitted or redacted in writing. 66

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

71

the synagogue. The associative linking of these two additional R. Yehoshua b. Levi traditions to the first yields a unit of three R. Yehoshua b. Levi traditions. As we have learned in other contexts, the arrangement of materials in units of three is one hallmark of Bavli redactional practice.67 The Bavli redactors adapted a sugya also found in b. Git. 38a-b (R. Eliezer and “mitzvah ha-ba’ah ba-’aveirah”) for deployment at b. Ber. 47b as b. §§5-10. In Berakhot, the Bavli challenges R. Yehoshua b. Levi’s tradition that “nine and a slave join together” by citing the case of R. Eliezer, who completed a quorum of ten with a slave that he freed for that purpose. R. Eliezer’s action is subjected to further scrutiny, including the observation in the name of R. Yehudah that “whoever frees his slave has transgressed a positive commandment” and that R. Eliezer did a “mitzvah that comes about through a transgression” (a claim that is immediately rejected on the ground that a mitzvah done for the public weal is different). R. Yehudah’s tradition (in the name of Shmuel) appears without the scriptural prooftext at b. Git. 38a in the context of a lengthy discussion about the redemption of slaves from captivity. The Bavli takes up R. Yehudah’s tradition more fully at b. Git. 38b, where the discussion appears as follows: R. Yehudah said in the name of Shmuel: “Whoever frees his slave transgresses a positive commandment, as it says, ‘You shall work with them forever’” (Lev. 25:46). They raised an objection: “It once happened that R. Eliezer entered the synagogue and did not find ten. He freed his slave and used him to complete the ten.” [The case of R. Eliezer is different, because] a mitzvah is [a] different [matter; and thus the positive commandment of Lev. 25:46 could be set aside in furtherance of this other mitzvah].

A comparison of the contexts of these similar sugyot supports the conclusion that the Bavli redactors adapted the Gittin material 67

Shamma Friedman, “‘Al Derekh Heqer ha-Sugya,” in H.Z. Dimitrovsky, ed., Mehqarim u-Meqorot I (New York: Jewish Theological Seminary, 1978): 277-441. See also Friedman’s earlier findings in “Mivneh Sifruti be-Sugyot ha-Bavli,” Proceedings of the Sixth World Congress of Jewish Studies (Jerusalem, 1977), 3:389-402, esp. 391-392; see also my “The Power Conferred By Distance From Power: Redaction and Meaning in b. AZ 10a-11a,” forthcoming.

72

HOW SHOULD RABBINIC LITERATURE BE READ?

for use at b. Ber. 47b. This sugya is part of a lengthy discussion of redeeming slaves from captivity based on m. Git. 4:3. The sugya fits the Gittin context more naturally than it fits the Berakhot context, which has absolutely nothing to do with the redemption of slaves. The difference between b. Git. 38b and b. Ber. 47b (b. §§5-10) is due to the Berakhot redactors’ amplification of the Gittin sugya. Whereas R. Eliezer is used in Gittin to challenge the verse, he is deployed in Berakhot to challenge R. Yehoshua b. Levi. Having deployed R. Eliezer to challenge R. Yehoshua b. Levi, the Berakhot redactors then raise a further challenge to him in turn from Lev. 25:46, and initially give the same answer as Gittinȥa matter of mitzvah is different. But then the Berakhot redactors go a step beyond Gittin and challenge R. Eliezer again on the basis of “mitzvah haba’ah ba-’aveirah,” ultimately concluding that even this is no problem because “a mitzvah for the public weal is different.” Berakhot’s “step beyond” Gittin in the analysis of this sugya likely arose from its further reflection on the sugya in light of a concept deployed elsewhere in the Bavli. R. Ami is represented as teaching that “two and the Sabbath join [to make a zimmun],” whereupon R. Nahman challenges him with the rhetorical question of whether the Sabbath is a human being. R. Ami’s emended teaching is that two Torah scholars who sharpen one another in the Law join together to make a zimmun. This emended teaching is odd because now there are only two. How can two in any way constitute a zimmun of three? The answer may be found in the Babylonian/Palestinian halakhic difference over whether two judges can constitute a valid court for monetary cases, notwithstanding m. Sanh. 1:1’s unequivocal direction that “monetary cases [are to be adjudicated] by three.” The Palestinian amoraim R. Yohanan and Resh Laqish at y. Sanh. 1:1, 18a and R. Abbahu at b. Sanh. 2b and 6a hold that the judgment of two judges is invalid, while Shmuel is represented at both y. Sanh. 1:1, 18a and b. Sanh. 3a and 5b as validating the judgment of two judges, although they are called a “presumptuous” court. This Babylonian view that two persons can constitute a group that can legally do the work of three likely underlies the presentation at b. Ber. 47b of R. Ami’s view that two scholars can constitute a zimmun of three.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

73

THE REDACTIONAL CONTEXT AND EDITING OF THE SHIMON B. SHETACH STORY IN THE BAVLI Y. Ber. frames the Shimon b. Shetach story nicely as a means to ascertain whether R. Yirmiya thought that one who ate only vegetables could or could not bless on behalf of the other diners, who presumably ate bread. It precedes the story with the observation “R. Yirmiya’s position is reversed,” and then follows it with a repetition of that observation and a harmonization. Gen. Rab. moves abruptly from R. Yirmiya’s question “one who ate vegetables – is he included in the zimmun?” (note again the difference between this question and the Yerushalmi’s version) to the story of Shimon b. Shetach, without explicitly introducing the story as an attempt to understand his view. Moreover, Gen. Rab.’s Shimon b. Shetach drinks wine rather than eats vegetables, which makes the utility of the story for an elucidation of R. Yirmiya’s view rather slight. Nevertheless Gen. Rab. does follow the story with an attempt to use it to understand R. Yirmiya’s view, which is odd since Shimon b. Shetach’s behavior in the story cannot really shed light on that view. Earlier we saw why the Bavli eliminated R. Yirmiya’s question and how this elimination smoothes over the ill fit between Shimon b. Shetach’s wine-drinking and R. Yirmiya’s question about eating vegetables by focusing the halakhic discussion of the story on Shimon b. Shetach’s behavior rather than on R. Yirmiya’s. 68 Turning to the story itself, the Bavli omits the three hundred Nazirites. 69 The Nazirites’ narrative function is to illustrate the animosity between the sage and the king that causes Shimon b. Shetach to flee. The Bavli does not need them to illustrate the ani68

It is more reasonable to assume that the Bavli inherited the detail that Shimon b. Shetach drank wine at the meal, rather than to assume that the Bavli changed a tradition that he ate (something other than bread) to one that he drank. The reason is that (as was pointed out before) Gen. Rab.’s tradition that the sage drank wine is difficult to understand in light of Yerushalmi’s use of that story to investigate R. Yirmiya’s position about one who ate vegetables. Thus it is more logical to assume that the tradition that the sage drank wine is the “lectio difficilior,” and that y. Ber. altered this tradition in the interests of clarity. 69 See Richard Kalmin, “Portrayals of Kings in Rabbinic Literature of Late Antiquity,” JSQ 3:4 (1996): 322-323, for a discussion of the story that places it in the context of Babylonian polemic against non-rabbis who claim Hasmonean descent.

HOW SHOULD RABBINIC LITERATURE BE READ?

74

mosity of the two men because it has its own unique account of how that animosity aroseȥin Yannai’s murderous onslaught upon the Pharisees (b. Qidd. 66a; b. Sot. 47a; b. Sanh. 107b70). This tradition about Yannai is not found in Palestinian compilations.71 The Bavli directly alludes to its own tradition about Yannai’s rabbinic killings here at b. Ber. 48a in explaining why Shimon b. Shetach had to be summoned into the king’s presence: “and since he [Yannai] had killed the Rabbis, there was no one who could bless for them…” And, since the Bavli has omitted the Nazirites, it also omits Yannai’s questions to Shimon b. Shetach that pertain to that incident. After recounting Shimon b. Shetach’s flight from Yannai, the Palestinian sources mention the Persian notables’ request to Yannai for Shimon b. Shetach. The Persians serve no function in the story beyond that of being the means through which the sage is reintroduced to the royal court. The Bavli economically eliminates them and arranges for the sage’s return through the good offices of the queen, who the Bavli understands to be Shimon b. Shetach’s sister (b. Sot. 47a). By playing up the queen’s role, the Bavli has both clarified the ambiguous identity of the “sister” in Gen. Rab. and represented this familial relationship in a manner consistent with its tendency to make its characters (particularly Palestinians) relatives of one another.72 Since the Bavli holds that the queen was hiding her brother (b. Sot. 47a), it is logical that the Bavli sees her as the agent of his return. Moreover, since the queen is the agent of his return, the Bavli logically has her ask for and receive from Yannai a guarantee of his safety, rather than the sage himself, like Gen. Rab. and unlike y. Ber. There is another major difference that emerges from these two Bavli revisions of the Palestinian source. In the Palestinian story the animosity between king and sage is personalȥYannai feels that he has been duped, and Shimon b. Shetach flees. In the Bavli, by contrast, the animosity between them is not at all personal: 70

Per DS to b. Sanh. 107b, note resh. Although the version at b. Qidd. 66a is found in a baraita. The fact that the baraita is unattested in a tannaitic compilation casts some doubt on its authenticity although there is nothing prima facie suspicious about it. 72 See Shmuel Safrai, “Tales of the Sages in the Palestinian Tradition and the Babylonian Talmud,” Scripta Hierosolymitana 22 (1971), pp. 209-232. 71

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

75

Yannai’s problem is with the Rabbis as a group, and Shimon b. Shetach is simply one member of that group (who happens to be the queen’s brother). In the Bavli, then, the animosity is portrayed as an anti-rabbinic animosity directed toward one particular rabbi. Similarly, the Persians in the Palestinian story want Shimon b. Shetach to return because they crave his wisdom; they see him as a wise man generally, not as a Torah sage in particular. By contrast, the Bavli’s Yannai wants Shimon b. Shetach back in order to perform a liturgical function that falls within rabbinic expertise: leading the Grace after Meals. In other words, while Yerushalmi sees Shimon b. Shetach as a wise man and sage, the Bavli distinctly emphasizes that he is a rabbi, and a rabbi who is part of a rabbinic class that is arrayed in opposition to a murderous Yannai. Further on in the story, y. Ber. and Gen. Rab. recount that Shimon b. Shetach came back and sat between the king and queen on his own initiative, while the Bavli recounts that Yannai himself seated him there. Y. Ber. recounts three questions that Yannai asked him (“why did you make fun of me,” “why did you flee,” “why did you sit between the king and queen”), while Gen. Rab. recounts four (“why did you sit between the king and queen,” “why did you make fun of me,” “why didn’t you tell me that your half would come from your Torah,” and “why did you flee”). The Bavli logically omits the questions that flow from the men’s interaction over the Nazirite issue. The Bavli emends “why did you sit between the king and queen” to “do you see how much honor I have done you?,” because the Bavli has made the additional emendation of having Yannai seat Shimon b. Shetach in that spot. Whereas y. Ber.’s /Gen. Rab.’s Shimon b. Shetach answers Yannai’s question by quoting Ben Sira 11:1 (“the wisdom of the humble man will lift up his head and seat him among the great”), the Bavli’s sage responds with an alleged quotation of Prov. 4:8 (“Exalt her and she will raise you up, she will honor those who embrace her”).73 The Palestinian deployment of the Ben Sira verse says nothing about the Torah, while the Bavli’s deployment of Proverbs in the context of the king’s seating the sage between himself and the queen allows 73 But see b. Ber. 48a (DS note tsadi) which points out that some editions of the Bavli did cite the Ben Sira verse. Although certainty is impossible, I think Prov. 4:8 is the better reading because it connects more smoothly to Shimon b. Shetach’s emphasis on Torah, which is a clear emphasis in the Bavli overall.

76

HOW SHOULD RABBINIC LITERATURE BE READ?

the sage to point out that the honor done him is done not by the king, but by the Torah. The Bavli’s emphasis on the Torah as the bringer of honor to the sage is linked to the Bavli’s emphasis on Shimon b. Shetach’s rabbinic status and to the Bavli’s presentation of the Torah and its interpretation as a point of contention between the king and the rabbis (b. Qidd. 66a). Taken together, the Bavli’s emphasis on the rabbis’ “caste consciousness” vis-à-vis the non-rabbinic Yannai and on Shimon b. Shetach as rabbi and Torah sage are consistent with its emphasis on the importance of Torah and rabbis in other settings.74 After Shimon b. Shetach receives the second cup of wine with which to say Grace, R. Abba b. R. Hiyya b. Abba observes that he acted alone (that is, none of the Rabbis agreed with him), quoting R. Yohanan to the effect that one cannot lead Grace unless he ate at least an olive’s worth of grains. As noted earlier, this discussion continues the Bavli’s careful use of the distinct terms “mitstaref” and “le-hotzi… yedei hovatan,” the latter even finding its way into the Bavli’s version of R. Yohanan’s teaching that a person cannot fulfill the many’s obligation (“eino motzi et ha-rabim yedei hovatan”) unless he eats at least an olive’s worth of grains.

CONCLUSION Gen. Rab. 91:4 and y. Ber. 7:2, 11b are two successive versions of the same Palestinian sugya. B. Ber. 47b-48b is likely a conscious reworking of this sugya that demonstrates similarity in part to y. Ber. and in part to Gen. Rab. What larger conclusions can be drawn from this analysis? First, we cannot conclude on the basis of this one example that Bavli Berakhot knew and relied upon Yerushalmi Berakhot. The fact that the Bavli is similar to both y. Ber. and Gen. Rab. must make us cautious about reaching such a sweeping conclusion on the basis of this one example. Only a careful macro- and micro-analysis of Bavli and Yerushalmi Berakhot in toto can lay the foundation for a conclusion about their relationship. Second, we also cannot conclude that the similarities between the Bavli and Palestinian sugyot (and the Bavli’s reworking of them) are attributable to a shared stratum of amoraic “proto-Talmud”. No amora is 74 See generally, Richard Kalmin, The Sage in Jewish Society of Late Antiquity (London: Routledge, 1999); idem, “Rabbinic Traditions about Roman Persecutions of the Jews: A Reconsideration,” JJS 54:1 (2003): 21-50.

A BAVLI SUGYA AND ITS TWO YERUSHALMI PARALLELS

77

demonstrably aware of the sequence of sugyot qua sequence in any of the three compilations. The current state of scholarship does not promote confidence that amoraim joined topics and sourcematerials into lengthy sugyot and then attached them to particular mishnayot. It is more likely that Palestinian redactors assembled the sugya in a form more like the one we now see in Gen. Rab. and that this sequence underwent continued development in Palestine, resulting in the (overall) more logical sugya as we now have it in y. Ber. The sugya likely came to Babylonia75 attached to m. Ber. 7:2 (since we have no solid reason to assume it came to Babylonia any other way)76 in a form showing evidence of the continued development it underwent in Palestine, although it still retained features of the earlier version of the Palestinian sugya. This now non-extant “missing link” in the development of the Palestinian sugya is a link that we can reconstruct from the Bavli and was likely a version of the sugya in oral circulation at some point. The Bavli redactors further reworked the Palestinian sugya in keeping with their own linguistic and cultural tendencies, yielding the sugya as we now have it at b. Ber. 47b-48b.

75 For a very recent review of the literary and archaeological evidence of contacts between the Babylonian and Palestinian rabbinic communities, see Ze’ev Safrai and Aren M. Maeir, “Atta Igarta mi-Ma’arava (‘An Epistle Came From the West’): Historical and Archaeological Evidence for the Ties Between the Jewish Communities in the Land of Israel and Babylonia During the Talmudic Period,” JQR 93:3-4 (Jan.-Apr., 2003): 497-531. 76 Cf. Richard Kalmin’s suggestion that the midrash compilations were probably not known in Babylonia, in The Sage in Jewish Society of Late Antiquity, p. 113.

THE SEMIOTICS OF THE SEXED BODY IN EARLY HALAKHIC DISCOURSE Charlotte Elisheva Fonrobert Stanford University INTRODUCTION In recent years the study of gender in rabbinic literature has established itself as one of the prevalent approaches to understanding how rabbinic texts have given and continue to give shape to Jewish culture. Thinking about gender in terms of a hermeneutic category has moved far beyond the earlier approaches which analyzed the various ways in which rabbinic Judaism presented itself as a patriarchal society. It simply does not seem interesting anymore to arrive at the conclusion of patriarchy. Rather, in response to and in conversation with gender studies elsewhere in the humanities, particularly in literary studies, scholars have turned to more subtle investigations of how gender operates within rabbinic texts, both overtly and covertly. Whereas in earlier feminist studies ideology was the object of investigation, the more recent studies have turned to rhetoric and to discourse. Texts no longer merely “reflect” the ideology of a certain class of men, or “reflect” the socio-historical reality behind the text. Rather, texts now labor at constructing and shaping realities. Another significant development perhaps has been the rise of “cultural studies” with respect to rabbinic texts. In other words, rabbinic texts do not simply express “ideas” to be distilled from them to the end of writing a history of “Jewish ideas” from the Bible to modernity. Rather, texts are believed to operate within a specific discursive situatedness. The rabbis are now considered to labor in a cultural context that gives shape to their discourse, often in rather indirect, implicit ways. Within such a shift in approach to the study of rabbinic texts, gender is no longer something that rab79

80

HOW SHOULD RABBINIC LITERATURE BE READ?

binic texts encapsulate, seemingly in a cultural and historical vacuum. On the contrary, more and more the rabbis emerge as both a product and shaper of the Greco-Roman (and Sassanian) world in which they set about their labor in Torah. Rabbinic thinking about gender takes place within a complicated net of relationships to the cultural world that surrounds them. And, our thinking about rabbinic gender gains in depth when rabbinic discourse is analyzed within this complex net of relationships. Finally, by way of introduction, the relationship between sex and gender, between the body and the text, between the body and culture is considered with particular attention to the productive role that language plays. This essay takes these issues into consideration as it deals with some of the fundamental and surprisingly un-read texts dealing with the question of the body and sexual identity in rabbinic culture.

THE GENDER DUALITY OF JEWISH LAW Jewish law is based on a fundamental assumption of what we may identify as gender duality or dimorphism. Already the Mishnah, as the earliest text of the rabbinic movement and the foundational text for Jewish law, reflects self-consciously on the tight gender grid that serves as the framework and simultaneously the basis of rabbinic legal thinking. Hence, as diverse as the various contexts of individual commandments may appear to be, the Mishnah time and again attempts to order and categorize them in a more systematic way according to its dual gender grid. Mishnah Sotah 3:8 asks the question most directly: “What is the difference between a man and a woman?” and answers it with an eclectic list of halakhic rulings from a variety of contexts, ranging from commandments concerning leprosy to capital punishment—a list, in short, which reads like a halakhic compendium on the question of gender and law. Another, more abstract approach is taken by Mishnah Qiddushin 1:7, which famously differentiates the commandments along the line of “obligation” and “exemption.” Whether the ordering principle here is time (“time-bound” commandments), parent-child relationship (“commandments incumbent on the father with respect to the

THE SEMIOTICS OF THE SEXED BODY

81

son” and vice versa) 1 or the distinction between positive commandments and prohibitions, men are always and women sometimes obligated, whereas only women are sometimes exempt. In this mishnaic reflection on gender and law, then, exemption from religious duties is a halakhic category that can apply only to women. Similarly, the list of Mishnah Sotah 3:8 is rhetorically suspended between a positive and a negative pole: “These are the things that apply to a man, but not to a woman” (my emphasis). These fundamental mishnaic texts indicate that the gender duality of Jewish law is in fact not conceived as a symmetric duality, as least not from the perspective of the rabbis who began to build the halakhic edifice. One of the earliest critical gestures of Jewish feminists in the United States has been to point out that women’s exemption from the “time-bound commandments” (as for example, regular prayer, laying tefillin, donning a tallit) has amounted to women’s exemption from rituals that are the basis for a positive formation of religious identity as a Jew, as most prominently articulated by Rachel Adler, Judith Plaskow and Paula Hyman. Furthermore, the halakhic literature subsequent to the Mishnah develops the notion of “obligation” or “commandedness” as an integral aspect of full halakhic subjectivity, whereas exemption amounts to exclusion from the performance of religious acts on behalf of others as their representative (e.g., serving as a sheliah tsibbur), as well as ultimately to the exclusion of women from teaching and therefore reproduction of Torah.2 Finally, we may add another duality overlapping with the halakhic gender duality which is the juxtaposition of grammatical ac1

The Tosefta provides concrete examples for the abstract principles in the Mishnah: “What are the commandments of the father to the son? To circumcise him, to redeem him, to teach him Torah, to teach him a trade and to marry him off” (t. Qidd. 1:11). Elizabeth Shanks Alexander (“The Impact of Feminism on Rabbinic Studies: The Impossible Paradox of Reading Women into Rabbinic Literature,” in Jews and Gender: The Challenge to Hierarchy, ed. by Jonathan Frankel, Oxford Univeristy Press, 2000) articulates the gender dynamics inherent in this passage most succinctly: “… the passage from the Tosefta transforms the family from a site of ‘mere’ biological replication to one where central cultural values are transmitted. In so doing, however the contribution of the mother is ignored and discounted” (105). 2 See for instance b. Qidd. 29b and b. Sot. 21a.

HOW SHOULD RABBINIC LITERATURE BE READ?

82

tive and passive when defining the marital relationship between man and woman. All cognate verbs for acts of marriage (sanctify, marry, acquire and sexual relations)3 ascribe the grammatical active to the man and the passive to the woman: a man takes, but he is never taken. By way of introduction, I would suggest that the halakhic discourse on gender in its classic rabbinic form cannot be described as an ontological or teleological essentialism. That is to say, it does not posit an essential masculinity or femininity, nor is there a fixed nature to men or women,4 ascribing a given ontological superiority to men on which their halakhic privilege is to be based. Nor, as to the latter, does it attribute an essential function in the order of creation to women, such that women might be regarded as having been created solely for the purpose of reproduction.5 Such essentialism entered the halakhic discourse only later, as is the case with Maimonides’s Aristotelianism and Neoplatonic traditions, and has become a commonplace in modern day apologetic discourse. The essentialist argument is designed to justify, for instance, the exemption of women from time-bound commandments with the suggestion that, unlike men, the very nature of women’s constitution absolves them from the need to be made aware of the sanctity of time.6 Rather, rabbinic halakhic discourse institutes what I would call a functional gender duality, anchored in the need of reproduction of the Jewish collective body. Indeed, supreme value is attributed to reproduction and establishing and maintaining sexual and genealogical order. As such, the halakhic discourse aims to enforce and normativize congruence between sexed bodies and gendered identities. For example, already the biblical prohibition of cross-dressing (Deut. 22:5) presumes self-evident categories of a man’s apparel (ʸʡʢ ʩʬʫ) 3

For instance, m. Qidd. 1:1, 2:1, m. Yev. 8:6, t. Bik. 2:4. See also Daniel Boyarin, “Gender,” in Critical Terms for Religious Studies, ed. by Mark C. Taylor (Chicago: University of Chicago Press, 1998), 130. 5 As for instance later suggested by ibn Ezra, at Deut. 22:5. This is not to say that ibn Ezra espouses in any form a less authentically Jewish view. But such a view is decidedly different from earlier, i.e., late antique rabbinic articulations. 6 See Tamar Ross, “Modern Orthodoxy and the Challenge of Feminism,” in Jews and Gender: The Challenge to Hierarchy, ed. by Jonathan Frankel (Oxford University Press 2000), 10-12. 4

THE SEMIOTICS OF THE SEXED BODY

83

ȥat times interpreted in the Talmud as a man’s weapons (b. Naz. 59a)ȥand women’s clothing (ʤʹʠ ʺʬʮʹ)ȥat times interpreted in the Talmud as the habit of plucking hair (b. Naz. 59a, b. Shab. 94b, b. Mak. 20b) or as women’s jewelry.7 Some categories of appearance may be more contested, such as perfuming (b. Ber. 43b).8 In addition, sexual behavior is coded as male or female, derived from the biblical locution “a woman’s lyings” (ʤʹʠ ʩʡʫʹʮ, Lev. 18:22 and 20:13) and “a male’s lying” (ʸʫʦ ʡʫʹʮ Num. 31:17, 18.35; Judg. 21:11–12).9 In the former case, “a woman’s lyings” are not appropriate when performed by men, between men. In the latter, “a male’s lying” refers to what men do to women. In the rabbinic idiom, mishkav zakhar comes to refer to male-male anal penetration (b. Nid. 13b), 10 similar to the more exceptional term “the way of masculinity” (ʺʥʸʫʦ ʪʸʣ m. Yev. 10:2, y. Ket. 3:9, 27d).11 Another significant way of reinforcing congruence between sexed bodies and gendered identities in rabbinic literature is of course what I would call the ritualization of the sexed body. On the one hand, circumcision is the major way in which the male body is inscribed, while the laws of niddah are perhaps the major way in which the female body is circumscribed. Indeed, the question of Jewish identity often hinges in fundamental ways on this ritualization of the sexed body, and it is here that sexed bodies are most “rabbinized.”12 7 See Michael Satlow, “They Abused Him Like a Woman,” Journal of the History of Sexuality 5:1 (1994), 12. As Satlow points out, the editorial voice in b. Naz. 59a “explains this prohibition as due to the fear that a man, by appearing like a woman, will slip disguised among women, leading to a greater chance of heterosexual immorality.” 8 Satlow, “They Abused Him,” 13. 9 See Saul M.Olyan, “‘And With a Male You Shall not Lie the Lying Down of a Woman’: On the Meaning and Significance of Lev 18:22 and 20:13,” Journal of the History of Sexuality 5:2 (1994), 184. 10 See Daniel Boyarin, “Are There Any Jews in ‘The History of Sexuality’?,” in Journal of the History of Sexuality 5:3 (1995), 336. 11 See Michael Satlow, “They Abused Him Like a Woman,” 18. 12 Space does not allow here to delve into depth concerning this particular subject and the body of literature on this topic is huge, especially on circumcision and the question of Jewish identity. Indeed, the ritual act of male circumcision has been mobilized as perhaps the marker of Jewish

84

HOW SHOULD RABBINIC LITERATURE BE READ?

Such is the fundamental gender structure of rabbinic literature. However, emerging from this structure are the following questions: How, in fact, is sexual identity constituted? If, indeed, the rabbinic legal discourse by and large assumes a self-evident duality of men and women, why does it time and again discuss “abnormal” cases, such as prominently the androginos and the tumtum, but also the saris and the aylonit? What is the function of the discussions of these categories in view of the seemingly rigid gender grid that is the desired norm of rabbinic halakhic discourse? Are these, indeed, hybrids that undermine the dual gender grid? Or do they operate to reinscribe the sexual dimorphism of the rabbinic legal discourse? The following pages will be devoted to precisely these questions. In addition, the question will be raised as to how far these discussions are related to their contemporary Greco-Roman discussions.

THE SEMIOTICS OF THE SEXED BODY The rabbinic halakhic discourse grounds the gender grid imposed on the system of commandments in a semiotics of the body, in the form of a reading of body surfaces. Primarily, the genitals are “the place from which it can be recognized whether s/he is male or female” (Gen. Rab. 46:5, 46:13). Generally, halakhic discourse imagines the human body to be sexed as either male or female, in contrast to the Aristotelian model, according to which male bodies are the norm, indeed the only sex, and female bodies the aberration.13

difference from the biblical beginnings. More recently the laws of niddah have been mobilized to reflect on Jewish identity formation. In the modern context apologists regard the mikvah as a means of positive identity formation for women, as in fact reinforcing their Jewish identity, while critics regard it as oppressive of women’s embodied lives. Either way, the ritualization of menstruation serves as a major context of gender identity formation for women in rabbinic culture. See my book Menstrual Purity: Rabbinic and Christian Reconstructions of Biblical Gender (Stanford: Stanford University Press, 2000). 13 For an analysis of the Aristotelian one-sex model see Thomas Laqueur, Making Sex: Body and Gender from the Greeks to Freud (Cambridge/ Mass.: Harvard University Press, 1990), 23 and Anne Carson, “Putting Her in Her Place: Woman, Dirt, and Desire,” in Before Sexuality: The Construction of Erotic Experience in the Ancient Greek World, ed. D. Halperin, J.

THE SEMIOTICS OF THE SEXED BODY

85

The one exception in mishnaic thinking about the human body is potentially m. Ohal. 1:8 which enumerates the 248 primary parts or limbs of a human being (ʑadam), a list which does not include any sexual organ.14 The mishnaic passage itself avoids sexing the human body in the context of determining the individual parts of the human body that can convey impurity of the dead. However, talmudic discussions elsewhere (b. Bek. 45a) cite a tannaitic tradition according to which women have four body parts in addition to the normative two hundred forty-eight enumerated in the Mishnah. The four additional body-parts, two doors and two hinges, are midrashically derived.15 This talmudic passage indicates that at least as early as in the talmudic context, if not in the earlier context of the tannaitic tradition cited there, the human being (‘adam) of the mishnaic passage in m. Ohal. 1:8 could be read as referring specifically to a male body, allowing for the conclusion that the woman is other than ʑadam. This would further be corroborated by an amoraic midrashic tradition that posits a correspondence between the 248 limbs and the 248 positive commandments (b. Mak. 23b). Indeed, in medieval Jewish mysticism this reading is the one which becomes much more prominent.16 However, the predominant picture cast by early rabbinic law is still one in which the human body is imagined to be sexed as either Winkler, and F. Zeitlin (Princeton: Princeton University Press, 1990), 137 and 152. 14 Pace Jacob Neusner’s translation of this passage who renders ʤʹʩʮʧ ʥʩʡʷʰʡ [ʺʥʩʬʥʧ] as “five [vertebrae] in the genitals,” The Mishnah: A New Translation (New Haven and London: Yale University Press, 1988), 951. ʭʩʡʷʰ refers to bodily orifices in general and not merely the genitalia. See also Boyarin, Carnal Israel, 36 n.9, who notes in a different context that the word does not refer specifically to female genitalia as already assumed by David Daube. Rather, “it is well attested in rabbinic Hebrew with the meaning of all bodily orifices.” 15 See Julius Preuss, Biblical and Talmudic Medicine, tran. by F. Rosner (Northvale, N.J.: J. Aronson, 1993 [orig. German: 1911]), 46. For an extensive discussion of this passage, especially its metaphoric imagery, see my Menstrual Purity, 58. 16 See Elliot Wolfson, “WomanȥThe Feminine as Other in Theosophic Kabbalah: Some Philosophical Observations on the Divine Androgyny,” in The Other in Jewish History and Culture: Constructions of Jewish Culture and Identity, ed. by R. Cohen and L. Silberstein (New York: New York University Press, 1994).

86

HOW SHOULD RABBINIC LITERATURE BE READ?

male or female. One further context expands on this theme, namely, the discussion of the external development of the human body. Similar to the indagatio corporis in Roman law, and perhaps echoing it, early rabbinic texts devote considerable discussion to the development of secondary sexual characteristics (ʭʩʰʮʩʱ), e.g., pubic hair and breast development (m. Nid. 5:7–8, t. Nid. 6:4, b. Nid. 47a–b).17 Here, the body is discursively constructed and at the same time decoded, both to infer its capability to reproduce as well as to establish a person’s legal majority and hence responsibility with respect to halakhic observance. In what follows, we will see that the rabbinic texts strive to establish a semiotic system of reading the body. They set up a list of signs to be decoded not only as to the ability to reproduce, which is the overt purpose of the system of signs. But further, this semiotic system reinforces sexual identity as either male or female. Hence, it is designed to ultimately reinforce the dual morphology of the human body.

THE SARIS AND THE AYLONIT To the mishnaic sages, lack of pubic hair by the age of twenty (according to the School of Hillel) or eighteen (according to the School of Shammai, m. Nid. 5:9) indicates an inability to reproduce. On the basis of this evidence, such people are categorized as either saris in the male case, or aylonit in the female case: If a woman twenty years old has not grown two [pubic] hairs she must bring proof that she is twenty years old; she is considered an aylonit, and she may not perform halitzah nor may she contract levirate marriage. If a man twenty years old has not grown two [pubic] hairs he must bring proof that he is twenty years old; he is considered a saris, and he may neither submit to ʚalitzah nor may he contract levirate marriage. Thus according to the School of Hillel.

17 See Fonrobert, Menstrual Purity, 142 and Preuss, Biblical and Talmudic Medicine, 128 as well as Tirzah Meacham (LeBeit Yoreh), “Halakhic Limitations on the Use of Slaves in Physical Examinations,” in From Athens to Jerusalem: Medicine in Hellenized Jewish Lore and in Early Christian Literature (Rotterdam: Erasmus Publishing, 2000), 39-41.

THE SEMIOTICS OF THE SEXED BODY

87

The School of Shammai disagrees: In both cases [the prohibition is valid already] when they are eighteen years old. Rabbi Eliezer says: For a male the rule is according to the School of Hillel, and for a female it is according to the School of Shammai, since the growth of a woman is more speedy than that of a man. (m. Nid 5:9)

The term saris is usually translated as “eunuch,” but would better be translated as a “sterile” or “infertile” man, parallel to the woman.18 The rabbinic assumption is that the saris can in fact have sexual relations, as for instance in m. Yev. 8:5 according to which “a saris who had sex with his [widowed] sister-in-law (rather than submitting to halitzah from her), causes her to be disqualified from the priesthood.” This is not to say that in Greek the use of the concept of “eunuch” is always to be equated with castrate. Tamsyn Barton for instance notes that “Dio Chrysostom has the sage Diogenes, in conversation with Alexander the Great, explain why Sophists should be compared to eunuchs: they pretend to virility, but their unions with women are unproductive.”19 However, as we shall see further below, in the literature of the Second Sophistic eunuch is used primarily as an invective, a rhetorical force not present in the legal terminology of the rabbis, where saris exclusively connotes infertility. The Mishnah and subsequent halakhic literature distinguish further between a saris ʚamah, a man who is born with the defect,20 18 The verb ʱʸʱ is used in a variety of contexts, rendering its meaning as uprooting, destroying, or mutilating. At the time of writing this essay, I did not have access to a dissertation devoted to the figure of the saris. See now Sarra Lev, Genital Trouble: On the Innovations of Tannaitic Thought Regarding Damaged Genitals and Eunuchs (New York University Press, 2004). As far as the woman is concerned, the Babylonian Talmud (b. Ket. 11a) provides a midrashic etymology and derives the term from ʬʩʠ, to be rendered as something like “ram-like” woman. The Arukh contests this etymology, and instead derives the term from ʯʬʩʠ to be translated as “wood-like” or “tree-like” woman. Accordingly, the Arukh points out that the vocalization should be ilonit, rather than the traditional aylonit. 19 Power and Knowledge: Astrology, Physiognomics, and Medicine under the Roman Empire (Ann Arbor: The University of Michigan Press, 1994/ 2000), 116, quoting from Chrysostom’s Orations 4-35. 20 In the Tosefta, Rabbi Yishmael claims that he has heard “in the vineyard at Yavneh that a saris ʚamah is someone who has only one testi-

88

HOW SHOULD RABBINIC LITERATURE BE READ?

and a saris adam, a man who becomes incapable of reproduction by human interference (m. Yev. 8:5–6), but who was at some point capable of reproduction. M. Yev. 8:4 records a disagreement between R. Akiva and R. Eliezer regarding the different halakhic rules applying to the ʤʮʧ ʱʩʸʱ and the ʭʣʠ ʱʩʸʱ with regard to levirate marriage. According to the former, the defining moment of the ʭʣʠ ʱʩʸʱ is the fact that “there was a time in his life when he was fully functioning” (ʸʹʫʤ ʺʲʹ ʥʬ ʤʺʩʤ) as opposed to the ʤʮʧ ʱʩʸʱ.21 According to R. Eliezer, the defining moment of the ʭʣʠ ʱʩʸʱ is that “there is no cure for him” ʤʠʥʴʸ ʥʬ ʯʩʠ as opposed to the ʱʩʸʱ ʤʮʧ. Other tannaitic traditions, include additional “signs” (simanim) to infer an incapacity to reproduce: What is a saris ʚamah? Any man who turns twenty and has not produced two [pubic] hairs, and even if he produces [them] after this age limit, he still is to be considered as a saris with respect to all matters. And these are his signs: anyone who does not have a beard, and his flesh is smooth.22 Rabban Shime’on ben Gamliel in the name of Rabbi Yehudah ben Ya’ir: Anyone whose urine does not make foam.23 And there are those who say: Anyone whose urine does not turn sour. And there are those who say: Anyone whose seminal cle” (t. Yev. 10:3; see also Sifre Devarim ʠʶʺ ʩʫ, pisqa 147, y. Yev. 8:2, 9a and b. Yev. 75a). Presumably he is born this way, “from the time of seeing the sun,” in distinction from the case-story of the man who climbed up an olive-tree, fell down, was wounded in one of his testicles, and had intercourse with his wife afterwards, but died childless (t. Yev. 10:3). 21 It is perhaps noteworthy that Greek texts make a similar distinction. Thus, Dio Chrysostom’s student Polemo from Smyrna, about whom more will be said further below, accuses his rival orator Favorinus from Ephesus, both orators, of eunuchism in the following manner: “When the eye is open with a shimmer like marble and a sharp gaze, it indicates little modesty. This is the type of eye found in men who are not like the rest, like a eunuch born without testicles rather than castrated…” (quoted in Barton, Power and Knowledge, 118; my emphasis). 22 Rashi (b. Yev. 80b, ad loc.) glosses the term as “smooth as women’s flesh. The norm of men’s flesh is to be hairy, rough.” The baraita’s version in the Babylonian Talmud adds ʩʥʷʬ ʥʸʲʹʥ, which is difficult to translate, perhaps as “and his hair is soft” (thus Rashi). Perhaps this addition is a doubling from the parallel in the list for the aylonit. 23 So Rashi on b. Yev. 80b, ad loc.

THE SEMIOTICS OF THE SEXED BODY

89

ejaculation is delayed. And there are those who say: Anyone who cannot make an arch [when urinating].24 Others say: Anyone who immerses in cold water during the rainy season and vapor does not rise from his flesh. Rabbi Shim’on ben Eli’ezer says: Anyone whose voice is thin and cannot be distinguished as a man’s or woman’s voice. And what is an aylonit? Any woman who turns twenty and has not produced two [pubic] hairs, and even if she produces [them] after this age-limit, she still is to be considered as an aylonit with respect to all matters. And these are her signs: anyone who does not have breasts, and her hair is thin and who becomes stiff25 at the time of intercourse. Rabban Shime’on ben Gamli’el says: anyone who does not have a pronounced lower abdomen 26 like other women. Rabbi Shime’on ben Ele’azar says: anyone whose voice is deep and cannot be distinguished as a man’s or a woman’s. (t. Yev. 10:67; cp. b. Yev. 80b)

This Toseftan text has numerous intertexts in Greco-Roman culture, for interestingly at least some of these signs are also the received wisdom of medical theories deriving from Aristotle and

24

This description corresponds to what contemporary medical literature describes as hypospadias, in which case the urethra does not run to the tip of the penis. In mild forms, the opening is just shy of the tip of the penis, while in severe forms, it may open at the base of the penis. This disorder necessitates the person to sit while urinating. See A. FaustoSterling, Sexing the Body: Gender Politics and the Construction of Sexuality (New York: Basic Books, 2000), 52-62. I am not proposing a “realist hermeneutics” here, i.e., where the point of reading the text is to establish that the rabbis already knew what medicine knows today. Nonetheless, the potential empirical reality behind the text presents one aspect of understanding it. 25 I am uncertain about the translation here. The hifȧil verb is used also in the context of erection (b. Nid. 13b), while the piȧel and the hitpaȧel form are used to connote “severe laboring” in child-birth (m. Hul. 4:1, m. Nid. 4:1, Gen. Rab. 84). Another possibility, therefore, is that the meaning here is that she suffers pain during intercourse. 26 Rashi glosses the term ʭʩʲʮ ʩʬʥʴʹ as “a sort of roundness above ‘that place’,” meaning the pubic area.

90

HOW SHOULD RABBINIC LITERATURE BE READ?

are wide-spread in late antique culture.27 Dio Chrysostom refers to “those who violate nature’s laws in secret, but whose true character is revealed by their voice, glance, posture, hairstyle and gait,28 shifty eyes, receding neck…” (Orations 33.52).29 These are the symbola that can be decoded. His student Polemo, whom we mentioned above, provides the following description of the ideal type of woman in comparison with the man. She has “a small head and mouth, softer, black hair, a narrower face, bright, glittering eyes, a narrow neck, a weak, rather sunken chest, with weak ribs, big, very fleshy hips, narrow calves and [knock]-knees, dainty fingers and toes, the body otherwise moist and slack, with soft joints… a weak voice.”30 As Barton points out, this ideal type appears in an almost entirely negative light, with the physical description being stained by deviation from the male norm. Subsequently, Polemo draws on similar signs when polemicizing against his rival orator Favorinus: “… His voice was just like a woman’s and all the rest of his limbs and extremities were soft; and he did not walk upright, but with slack joints and limb. He took great care of his person [by nourishing] his thick hair…” 31 Similarly, Clement of Alexandria (ca. 150-215 c.e.), who offers a Christianized view of Stoic doctrine on the subject of depilation,32 remarks: For God wished women to be smooth … but has adorned man, like lions, with a beard, and endowed him, as an attribute of manhood, with shaggy breasts – a sign of his strength and rule … This, then, the mark

27 See Maud Gleason, “The Semiotics of Gender: Physiognomy and Self-Fashioning in the Second Century C.E.,” in Before Sexuality. She writes: “Although physiognomists and astrologers are unique in the clarity with which they articulated the grammar of their semiotic system, the ‘signs’ of effeminacy were a commonplace constellation in the nontechnical discourse of antiquity,” 399. For examples see Gleason’s section on “Stoic Cosmetology,” 399-402. More will have to be said about the physiognomists’ attempt to detect effeminacy further below. 28 It seems that many of the Greco-Roman texts are preoccupied with the question of gait, something which interestingly does not seem to concern the rabbis very much. 29 See Tamsyn S. Barton, Power and Knowledge, 116. 30 Barton, Power and Knowledge, 115. 31 Barton, Power and Knowledge, 118. 32 See Gleason, “The Semiotics of Gender,” 401.

THE SEMIOTICS OF THE SEXED BODY

91

of the man, the beard, by which he is seen to be a man, is older than Eve, and is the token of his superior nature. In this God deemed it right that he should excel, and dispersed hair over man’s whole body. Whatever smoothness and softness was in him He abstracted from his side when He formed the woman Eve. (Clement, Paedagogus 3.3)33

Clement here emphasizes the beard as a sign of a man’s superiority over women and emphasizes it as “the mark of the man.” The notion of the “true” man lurks in his description. The true man was assigned a certain physical nature from creation onwards, which at the same time indicates his ontological superiority over women. This text is not a far cry from the late ancient physiognomists, such as for instance Polemo, and their attempt to read the body as to the true gendered nature of a person’s character that it shrouds. Obviously, there are significant parallels between the Greek symbola and the Hebrew simanim.34 Not only do both semiotic systems include similar signs. Both develop a semiotic system of ex33

This passage from Clement is quoted by Joshua Levinson, “Cultural Androgyny in Rabbinic Literature,” in From Athens to Jerusalem: Medicine in Hellenized Jewish Lore and in Early Christian Literature (Rotterdam: Erasmus Publishing, 2000), 129. Similarly Epictetus who writes that “Nature made women smooth and men hirsute” (Discourses, 3.1.27). Levinson relates Clement’s passage to the famous and oft-interpreted talmudic narrative about Rabbi Yohanan’s and Resh Laqish’s encounter in the River Jordan, which is preceded in the Babylonian Talmud by a remark about Rabbi Yohanan’s lack of beard. Levinson refers to Clement as Rabbi Yohanan’s contemporary, without however taking into consideration that the narrative is a later Babylonian creation, rather than a third-century creation. Clement is actually a closer contemporary to the tannaitic texts cited here. Also, both Levinson and Daniel Boyarin, “Homotopia: The Feminized Jewish Male and the Lives of Women in Late Antiquity,” Differences 7 (1995), 41-81, read the narrative as a rabbinic transvaluation of the prevalent (Greco-Roman) gender-codes which value masculine men while Rabbi Yohanan, as a clearly important hero of rabbinic discourse, is represented as the effeminate man in those terms. However, it seems that the halakhic passage in t. Yev. 10:6 does not quite fit that pattern. 34 The Hebrew term is after all a Greek Lehnwort, from shmei=on. See Samuel Krauss, Griechische und Lateinische Lehnwörter im Talmud, Midrasch und Targum (Hildesheim: Georg Olms Verlagsbuchhandlung, 1964, repr.), 386.

92

HOW SHOULD RABBINIC LITERATURE BE READ?

ternal signs of the body and physiognomy that point towards a hidden reality regarding the distinction between men and women. Indeed, both semiotic systems are looking for aberrations from the unspoken rule, and both are driven by the desire to categorize. The rabbis, however, differ significantly in their approach to the semiotics of the body, while they clearly share in the discourse of establishing a semiotics of sexual identity. Or is it really reproductive identity? According to the rabbinic semiotic system, the “signs” of a saris or an aylonit or their lack would suggest gender uncertainty, since a morphologically male body should indeed produce a beard, and a female body should develop breasts, and so on. In fact, ambiguous sonority of the voice may cause confusion as to the sexual identity of a person. However, the halakhic literature nowhere suggests that the saris or the aylonit are in fact not “really” man or woman. After all, Rabbi Eliezer in m. Nid. 5:9 cited above refers to the saris as ʸʫʦʤ and the aylonit as ʤʡʷʰʤ. Nor do the tannaitic texts betray interest in the ontological superiority of “manhood” or “masculinity.” Importantly, and in contrast to late ancient physiognomy, for the rabbis the goal of decoding such signs is not to establish or to undermine a mere appearance of genuine masculinity or femininity and to detect a hidden effeminacy, but to infer a potential incapacity to reproduce before marriage is contracted. Indeed, the primary context in which these two types of potential sexual ambiguities are discussed is in the context of levirate marriage, because here the sole rationale of marriage is reproduction, reproduction that is for the deceased brother who left no children. In this context, then, our signs are signs of bodies potentially not functioning properly with respect to reproduction. Aggadic texts, on the other hand, may play with the notion of mistaken identity, as in the famous story of Rabbi Yohanan and Resh Lakish in which the latter mistakes the former for a woman because he does not have a beard (b. B. Mez. 84a).35 Ultimately, the saris and the aylonit as halakhic categories are perceived as either man or woman via their primary sexual organs, and ambiguous

35 See Daniel Boyarin, Carnal Israel, 215 and again in Unheroic Conduct: The Rise of Heterosexuality and the Invention of the Jewish Man (Berkeley: University of California Press, 1997), 127-144; also Joshua Levinson, “Cultural Androgyny,” 128-129.

THE SEMIOTICS OF THE SEXED BODY

93

gender signs are read as merely secondary indicators as to their ability to reproduce. I would argue then that the saris and the aylonit are in fact not distinct genders, but are men or women who cannot reproduce, a “defect” that is significant only in the context of the rabbinic laws of marriage, while otherwise halakhah applies to them as to men and women. Hence, if a marriage halakhically requires capability of reproduction, such as in the case of levirate marriage, both the saris ʚamah and the aylonit are excluded from it (m. Yev. 8:5). Nonetheless, even a saris is not altogether prohibited from marriage, at least not after the fact, for the Mishnah reports the case of Ben Megoset who was a married saris adam in Jerusalem (m. Yev. 8:6). Such distinctions imply that ultimately the saris and the aylonit are halakhically similar to other sub-categories of man- or womanhood, such as the qatan or qetanah (male and female minors), that affect a person’s halakhic rights and obligations in certain respect, but not their sexual identity. A tentative conclusion may be advanced in view of the wealth of parallel evidence in Greco-Roman literature concerning the semiotics of gender identities. The Stoic material is riddled with ambiguities and a scale of gender identities that allows for fluidities of a wide range. The visible body and the “true” gendered identity of a person are conceptually disconnected. Differently put, sex does not determine gender, and gender may indeed betray the body. Bodies can be read for the “true” nature that they enclose. It is perhaps because of this that in such a context masculinity and femininity need to be policed more closely and in turn become ideal canon fodder for the polemics of late ancient orators, in order to detect potential aberrations from their “true” forms. The tannaitic material then transforms this approach and integrates or absorbs the ambiguities into the rabbinic-halakhic system of gender-dimorphism. It may, in fact, be read as exerting a concerted effort to subordinate any potential ambiguities, so prominently displayed by the rhetors of the Second Sophistic as well as by Christian thinkers such as Clement, to the rigid dimorphic gender-grid that the rabbis are imposing on their evolving halakhic system. A high or low voice here is not an indicator of the person who is male according to the morphology of his body but who on an ontological scale is not genuinely masculine, or of the person who is a woman according to the morphology of her body who is

HOW SHOULD RABBINIC LITERATURE BE READ?

94

not genuinely feminine on that scale. Rather, it is merely an indicator that such a man or a woman will not be capable of reproduction and are therefore not suitable for levirate marriage.

THE ANDROGINOS AND THE TUMTUM The semiotics of body surfaces produces other yet different and seemingly more ambiguous gender possibilities, the androginos or person with both primary sexual organs, and the tumtum as one with neither, at least not recognizably.36 The latter is mostly treated as a not-yet sexed person, as somebody whose sexual organs may eventually appear or be uncovered surgically.37 Meanwhile s/he is regarded as somebody in a genderwise doubtful legal status (t. Bik. 2:7), the concern being that s/he may yet turn out to be a man, and therefore hitherto uncircumcised, or perhaps even a saris (b. Yev. 83b). The figure of the androginos, however, is understood to be clearly doubly-sexed and to remain such. Nowhere in rabbinic literature can the assumption be found that s/he will change or should be or could even be changed, that his or her hermaphroditic nature should be disambiguated. The androginos appears in both midrashic and halakhic literature. Much has been written about the midrashic androginos, particularly with respect to the midrashic reading of Genesis, its anthropomorphic theology and the origin of humanity as dually sexed.38 The androginos in halakhic discourse, on the other hand, has not yet been much analyzed in scholarly literature.39 The midrashic androginos raises the question of gender with respect to “cultural androgyny,”40 and thus has already invited stud36

The Semitic root connotes “closedness.” M. Yev. 8:6; b. Yev. 83b; b. B. Batra 126b. 38 Daniel Boyarin, Carnal Israel, chapter 1-3; J. Levinson, “Cultural Androgyny”; David H. Aaron, Imagery of the Divine and the Human: On the Mythology of Genesis Rabba 8:1,” Journal of Jewish Thought and Philosophy 5 (1995), 1-62; E. Wolfson, “Woman – The Feminine as Other in Theosophic Kabbalah.” 39 See note 18. 40 Thus the term used by Joshua Levinson, “Cultural Androgyny in Rabbinic Literature.” He does not exactly define the term, but uses it loosely to refer to the confluence of medical theories, narrative material in rabbinic literature (on Rabbi Yohanan, and the midrashic rendering of Joseph), Roman literature on gendered behavior, all of which constitute a 37

THE SEMIOTICS OF THE SEXED BODY

95

ies of the rabbinic interaction with Greco-Roman culture, particularly because the term is a transliteration from Greek. The halakhic figure of the androginos, however, is primarily what could perhaps be designated as a “genuine hermaphrodite,” 41 or a “horizontal hermaphrodite” 42 who has both primary sexual organs and as such poses a challenge to a system that is built on sexual dimorphism. In tannaitic literature the halakhic androginos is most comprehensively discussed in the Tosefta (t. Bik. 2:3–7), even though a slightly different but similar text appears in some editions of the Mishnah and is attached at the end of Tractate Bikkurim as its fourth chapter.43 Again, similar to m. Sot. 3:8, this text provides a definition of the androginos in terms of a long list of gender-specific “discourse of cultural androgyny.” The beautiful Rabbi Yohanan, for instance, and the midrashic elaboration of Joseph are androgynous figures in terms of the Greco-Roman discourse of androgyny, not just in their looks, but also in their behavior, particularly in Joseph’s case. 41 By “genuine hermaphrodite” I am referring to the person who has both sexual organs, in distinction from male figures who behave in such a way that they are categorized as “effeminate” or androginoi in Roman literature, a phenomenon that will be discussed further below. The terminology in medical literature is somewhat confusing but need not really concern us here. Briefly, the concept of the “true hermaphrodite” in distinction to the spurious or pseudo-hermaphrodite is nineteenth-century medical vocabulary and refers to the presence of both ovarian and testicular tissue in a person’s body. In contemporary medical literature formerly “true hermaphroditism” has become a subcategory of “disorders of gonadal differentiation.” Fausto-Sterling, however, argues that “the distinction between pure and pseudo-hermaphroditism should be dropped and the general term intersexuality substituted,” Sexing the Body, 277, n.15. 42 The rabbinic texts do not reflect a figure who combines a woman’s breasts and a penis, as Hermaphroditus is classically represented in Roman art. M. Bekh. 7:5 considers breasts on a kohen that are “hanging like those of a woman” to be a blemish that disqualifies the priest from serving at the altar, but not something that makes him androginos. See, for instance, the reclining hermaphrodite figure from third century Roman Egypt at the Kelsey Museum of Archaeology (www.umich.edu/~kelseydb /Exhibits/WomenandGender/other.im.html) and Andrea Raehs, Zur Ikonographie des Hermaphroditen: Begriff und Problem von Hermaphrodismus und Androgynie in der Kunst (Frankfurt am Main: P. Lang, 1990). 43 For a discussion of the textual situation and complications of this text see Saul Lieberman, Tosefta Ki-Fshutah: Order Zera’im, Part II (Jerusalem: The Jewish Theological Seminary of America) 836.

HOW SHOULD RABBINIC LITERATURE BE READ?

96

commandments that range from purity and priestly issues to marriage and inheritance laws. We may surmise that the fact that various versions of this text exist is due to its nature as a list. This list of halakhot, many of which can be found elsewhere in tannaitic literature, is ordered into four sets laid out in the introductory paragraph and repeated as titles in the following subsections: As far as the androginos is concerned: there are with regard to him ways in which he is similar to men, and there are ways with regard to him in which he is similar to women, and there are ways with regard to him in which he similar to both men and women, and there are ways in which he is dissimilar from both men and women.

Accordingly, the androginos is similar either to the halakhic category of a man, or a woman, or halakhic categories apply to him that are shared by men and women, or apply to both sexes combined. Finally, there are some laws that cannot apply to him at all. The particular way in which this list is set up then suggests that the rabbis do not even allow for seriously contemplating the androginos as a hybrid, or at least not as a hybrid that would undermine in any way the sexual dimorphism of their evolving halakhic system. In fact, the list projects that ʭʩʹʰʠ and ʭʩʹʰ are stable categories, albeit halakhic constructs, in reference to which the androginos can be integrated into the system: he is like men in some respects, and like women in others. t. Bik. 2:444 The ways in which he is similar to men: he becomes impure by a white genital discharge, like men45 he marries (a woman) but is not married (by a man), like men46 44

I am following Lieberman’s text for the Tosefta, based on the Vienna manuscript, rather than the Naples editio princeps version for the Mishnah (1492) which adds a version of the Tosefta as Chapter 4 to Tractate Bikkurim and is listed synoptically by Lieberman. 45 Cp. m. Zav. 2:1. 46 See m. Yev. 8:6.

THE SEMIOTICS OF THE SEXED BODY

97

he does not go in seclusion with womenȥlike men47 he does not receive maintenance with the daughtersȥ like men48 he does not dress and cut his hair [in a woman’s style] ȥlike men he does not (should not) get impure by the deadȥlike men49 he transgresses “you shall not round [the corners of the hair of your head]” (Lev. 19:27) and “you shall not mar [the corners of your beard]” (Lev. 19:27), like men and he is obligated with regard to all the commandments mentioned in the Torah, like men.

t. Bik. 2:5 The ways in which he is similar to women: he becomes impure by a red genital discharge, like women he does not go into seclusion with menȥlike women he is not subject to levirate marriageȥlike women he does not receive a share in the inheritance with the sonsȥlike women he does not receive a share of the holiest sacrificesȥ like women he is invalidated for all testimony required by the Torah ȥlike women and if he has transgressive sex (passiveȦʤʸʩʡʲʡ ʬʲʡʰ), he is invalidated from [acquiring the status of] the priesthood [by marriage]ȥlike women.

t. Bik. 2:6 The ways in which he is similar to both men and women: whoever causes him damage incurs guilt, whether a man or woman

47

Cp. m. Qidd. 4:12. Cp. m. Ket. 4:6; m. B. Bat. 9:2. 49 Lieberman, Tosefta Ki-Fshutah: Order Zera’im, draws a connection with m. Sot. 3:7 according to which ʭʩʺʮʬ ʤʠʮʨʮ ʺʰʤʥʫ. Accordingly, the commandment specifically refers to the priests. A female kohen can in fact become impure with death impurity, which the male priest cannot do. 48

HOW SHOULD RABBINIC LITERATURE BE READ?

98

if someone kills himȥif with intent, he receives capital punishment; if unintentionally, he is exiled to the cities of refuge his mother considers her blood of purification for him ȥlike both men and women, and she brings a sacrifice [at the end of that period]ȥlike for both men and women he inherits any inheritanceȥlike both men and women he receives a share in the sacrifices to be consumed outside of the Temple and Jerusalem (ʬʥʡʢʤ ʩʹʣʷʡ) 50 ȥlike both men and women if he says: behold, I will be a Nazir, who is both a man and a woman, [the vow is valid and] he is indeed a Nazir.

t. Bik. 2:7 The ways in which he is dissimilar from both men and women [people] do not incur guilt on account of his impurity51 [people] do not burn [the heave-offering] on account of his impurity52 he is not valuated53ȥunlike both men and women he is not sold as a Hebrew slave, unlike both men and women if he says: behold, I will be a Nazir, who is neither man nor woman, [the vow is valid and] he is indeed a Nazir.

50

Such as the heave-offering, see b. Ket. 24b. Veȧal tuma’h, rather than veȧal hata’t , following ms. Erfurt rather than Vienna, as pointed out by Lieberman, ibid., 291, meaning: others do not incure a penalty if he goes to the Temple in an impure ritual status, even if they touch his red or white genital discharge. 52 As Lieberman glosses: “if it came into contact with either his red or white genital discharge,” ibid., 291. 53 His monetary value cannot be vowed to the Temple. See m. Arak. 1:1 where the Mishnah spells out explicitly: “[The tumtum and the androginos] may vow [another’s] valuation, but their valuation cannot be vowed by others, since only the valuation of those may be vowed who are unambiguously male or unambiguously female.” Biblical law here is no less clear. The list of valuations of persons (ʺʥʹʴʰ) in Lev 27:2ff subdivides its valuations by gender, referred to as ʸʫʦ and ʤʡʷʰ , as well as age-group. 51

THE SEMIOTICS OF THE SEXED BODY

99

In terms of textual form, then, this list is set up as a parallel to the (considerably shorter) list of halakhot organized around the koi in the animal world, a creature that is a hybrid of domestic and wild animal and needs to be fitted into that respective binary system,54 which is the reason the laws concerning the androginos got assembled in the context of Tractate Bikkurim. The connection between the two lists suggests, at least at first view, that both the human “male-female” hybrid, and the animal “wild-domesticated” hybrid operate primarily as theoretical test cases of the respective binary legal systems. As theoretical text-cases their presence within the rabbinic texts could then be explained as emerging from the logic of the system itself, as a kind of “what if” scenario. At the same time, I suggest, we have to consider the contemporary GrecoRoman physiognomic material to which we will return in the conclusion. In generalȥsince space does not allow for a more detailed discussion of the individual lawsȥwe can perhaps make the following observations about the logic of the list. The third section (t. Bik. 2:6) is built on the assumption that the androginos is after all a “human” person and “human” offspring. Causing damage to him, or killing him, requires the halakhic response as to a (presumably) unambiguously male or female person. The version of the list in the earliest printed edition of the Mishnah adds a nuance to this: “one incurs guilt for inflicting him with a wound and for cursing himȥlike [the halakhah] for men and women.”55 His birth is considered the birth of a human child, for the post-partum mother needs to count of her “blood of purification,” just as after the birth of an unambiguously sexed child. At the same time he is also considered to be a valid heir (whose claims, however, need to be balanced with the claims of potential siblings, as evidenced in the preceding sections). This general approach is not invalidated by the fourth section (t. Bik. 2:7) which by its set-up may lead one to think that the fact of his exclusion from certain laws categorize him as lesser than a human person. 56 However, the laws in this section 54

M. Bik. 2:8-11; t. Bik. 2:1-2; b. Hul. 79b. See Lieberman, The Tosefta: Seder Zera’im (Part One), 290. 56 Quite differently from the opinion attributed to Rabbi Yossi (Rabbi Meir in editio princeps) that concludes our list in Tosefta Bikkurim. Rabbi Yossi insists that just like the koi “an androginos is a creature in his own right and the sages could not decide whether he is man or woman” (t. Bik. 55

100

HOW SHOULD RABBINIC LITERATURE BE READ?

simply do not apply to him. Since, for instance, the laws of valuation are subdivided according to male or female, and there are fixed valuations attributed to each gender (and age group), the sexual ambiguity of the androginos excludes him from the act of valuation. The relationship between the first two sections of the list (t. Bik. 2:4 and 5) and the underlying logic of the two creates another complexity. As a figure of legal thought the halakhic androginos may appear to be the result of adding or subtracting the laws applying to men and women, based on his or her dual sex. Or vice versa, it may function as an index of how gender operates in various halakhic areas (purity, criminal law, family law, Temple laws), 57 in which case the concern of the text would not, or at least not primarily be about the sexual identity of the hermaphrodite in and by itself. However, the texts do indicate the relative weight of both sexes. In the end, the combination of both primary sexual organs in one body does not allow for either hybridity or choice, as if he could choose to be either man or woman. Indeed, in the project of integrating the doubly-sexed body into the binary halakhic system, the presence of the male genitalia has greater signifying force than the female organ. Thus, s/he has to dress like a man (t. Bik. 2:5), but most significantly, the androginos may take a wife, but cannot be taken as a wife (t. Bik. 2:4),58 just as he is subject to the laws of levi-

2:7). This opinion, however, is entirely marginalized in the tannaitic discourse, since it is the list of laws, culled from elsewhere in the mishnaic and toseftan discussions, that constitutes the general tannaitic approach. 57 I.e., in criminal law gender is less of an issue in the act of damaging a person, than in the laws of valuation of a person with respect to Temple vows. The latter depend totally on unambiguous sexual identities, while the former emphasizes the humanity of the victim. 58 Cp. m. Yev. 8:6. On the grammatical active (male) versus passive (female) see my comments above in the first section. M. Satlow, “They Abused Him,” 18 suggests the following: “Because the androgynos possesses a penis, s/he cannot ‘be married,’ that is, be a wife to another man. Yet, although the androgynos possesses a vagina, s/he is permitted ‘to marry’ a woman. Behind this rule lurks again the rabbinic fear of making a manȥor someone who looks very much like a manȥinto a wife.” See also J. Levinson, “Cultural Androgyny,” 127 who writes more tentatively: “The prevalent attitutude seems to be that the rabbis considered the default sex of the androgyne to be male, however the extent of his ‘maleness’ is questionable. Though he is ‘man’ enough to marry a woman and to forbid a

THE SEMIOTICS OF THE SEXED BODY

101

rate marriage as a man but not as a woman. A minority opinion in the Mishnah holding that sexual relations between a man and an androginos should be punished by stoning, is glossed in the Tosefta by restricting this punishment to a sexual act ʺʥʸʫʦ ʪʸʣȥ in “the way of masculinity.” While this discussion indicates that the presence of a male organ is not necessarily entirely determinative of the maleness of the androginos, since vaginal intercourse with him can be considered to be permissible, it also reveals the anxiety driving the halakhic consideration of marriage or sexual relations with an androginos as being one about potential male penetration.59 At the same time, the presence of the vaginal opening may compromise the signifying power of the penis in some cases, such as importantly in the context of circumcision, not considered in the Tosefta’s list. M. Shab. 19:3 records a rabbinic disagreement about the case when the circumcision of a hermaphrodite child falls on the Sabbath, when for an unambiguously male child the prohibitions of the Sabbath are set aside: “… For a ‘doubtful’ child60 or an androginos child one does not profane the Sabbath [for its circumcision], but R. Yehudah permits [the Sabbath circumcision] of an androginos child.”61 Finally, the tannaitic consideration of the doubly-sexed body reaches its critical point in the halakhic context of the first-born male animal, which according to biblical law is consecrated to the Temple (Deut. 15:19–23), or can be eaten if it has a physical defect. Here, the discussion concerns the identity of the doubly sexed animal, whether it is to be treated as a first-born of doubtful gender identity or whether it should be regarded as a male animal with a blemish: “By reason of these [following] blemishes62 they may not union with another man, is his maleness sufficient to acquire the cultural and religious privileges that accompany that category?” 59 Satlow, previous note, and Boyarin, “Are There Any Jews in the ‘History of Sexuality’?,” 347. 60 The Talmudic interpretation of this mishnah holds that a baby born during the eighth month of pregnancy will not live, but a baby born in the seventh month will live. In this context this means that if there is a doubt whether the baby was born in the seventh or eighth month, he should not be circumcised on the Sabbath (b. Shab. 135a). 61 B. Shab. 136b, Tosafot ibid., ʤ”ʣ ʸʫʦʤ . 62 As opposed to blemishes listed previously in this mishnaic chapter (m. Bek. 6:1-11), for which the firstborn may be slaughtered.

102

HOW SHOULD RABBINIC LITERATURE BE READ?

be slaughtered neither in the Temple nor in the provinces: … or if it is a tumtum or an androginos. R. Ishma’el says: No blemish is greater than this. But the sages say: It does not count as a firstborn but may be shorn and used for labor” (m. Bek. 6:12).63 Or, as the medieval commentator Rashi glosses this minority opinion, “because the place of femininity (ʺʥʡʷʰ ʭʥʷʮ, i.e., the female genitalia) is like a blemish” (Rashi, ad loc., 41a). This tannaitic opinion engenders considerable disagreement in the talmudic discourse. 64 However, overall it reinforces a semiotic system in which the male organ has greater signifying force than the female organ.

CONCLUSION How do we read what appears as a close relationship between the halakhic literature on semiotics of the body and late antique physiognomy, a popular discipline of knowledge or “science” that similarly focuses on deciphering the exterior features of a person? At the same time the two systems of knowledge differ significantly as to their objective, in that late ancient physiognomy has the purpose of interpreting human character, intelligence and virtue. Based in Aristotelian physiognomical inference (Prior Analytics II 27, 70b 7–4 and the Ps-Aristotelian third century B. C. E. treatise Physiognomy), it was rooted in many disciplines, such as medicine (Galen), rhetoric (Polemo), and ethics. As a “semiotics of gender”65 the discourse is based upon the belief that exterior traits reveal a person’s inner life.66 We have seen that the tools of the techne of physiognomics are employed most explicitly in the context of detecting a saris or an aylonit, where the rabbis share the attempt of decoding certain exterior features with their contemporary orators, albeit for different purposes. One system of knowledgeȥthe halakhic system of the rabbisȥdistinguishes between reproductive aberrant identities, while the other distinguishes between aberrant gender identities.

63

For the talmudic discussion see b. Bek. 41a-43b and Tosafot on b. Bek. 42a, ʤ”ʣ ʤʮʬʠ . 64 Which I am planning to discuss elsewhere, deo volente. 65 This is Maud Gleason’s term from her earlier article, “The Semiotics of Gender.” 66 T. Barton, Power and Knowledge, 126.

THE SEMIOTICS OF THE SEXED BODY

103

The androginos, on the other hand, initially appeared more as a “figure of thought” that is the product of the inner logic of the sexual dimorphism of the rabbis’ evolving halakhic system. However, the concept of the androgyne plays also a significant role in the physiognomic literature. In this context the concept, like kinaidos, indicates effeminacy, in that androgynoi are described as people who have “something of the shape of a man, but are feminine in all other respects.”67 In fact, in Roman literature the term evokes older connotations, since “androgynoi were also to be found in the traditional catalog of ‘Monstrous Races,’” and the Romans “used to drown such androgynoi with solemn ceremony.”68 Subsequently, the androgynos turns into one of the favored invectives that orators use to denounce their rivals.69 We may surmise, then, that the halakhic androginos is not simply, or as least not exclusively to be read as a figure of halakhic thought, produced only by the inner logic of the rabbinic halakhic discourse, however much it functions this way. But the physiognomic discourse is echoed in the extensive treatment of the androginos in early rabbinic halakhic discourse, if mostly by way of recontextualizing it. For in contrast to the physiognomic concerns, in rabbinic halakhic texts the semiotics of the body is entirely dis67 Suetonius, Peri Blasphemion 61; Maud Gleason, “The Semiotics of Gender,” 396. Often the different terms, such as androgynos and kinaidos, seem to be used interchangeably. Gleason notes: “The word androgynos (which I have been translating as ‘effeminate’) in its most literal sense describes an appearance of gender-indeterminacy (qui inter virum est et feminam, Anon. Lat. 98, 2.123F). The word kinaedus, on the other hand, describes sexual deviance, in its most specific sense referring to males who prefer to play a ‘feminine’ (receptive) role in intercourse with other men. But the two terms become virtually indistinguishable when used to describe men of effeminate appearance and behavior,” 396. 68 T. Barton, Power and Knowledge, 118. 69 Thus, Polemo describes how to spot the adult androgynos: “You may recognize him by his provocatively melting glance and by the rapid movement of his intensely staring eyes. His brow is furrowed while his eyebrows and cheeks are in constant motion. His head is tilted to the side, his loins do not hold still, and his slack limbs never stay in one position. He minces along with little jumping steps; his knees knock together. He carries his hands with palms turned upward. He has a shifting gaze, and his voice is thin, weepy, shrill and drawling,” cited in Gleason, “The Semiotics of Gender,” 395.

104

HOW SHOULD RABBINIC LITERATURE BE READ?

connected from the question of character and virtue, not merely because of the difference of genre (law versus rhetorical polemic). Rather, the difference between the two cultures with respect to their conceptual use of the hermaphrodite adds another piece to corroborate the suggestion that in rabbinic literature there is actually no ontological dualism between appearance and essence. Differently put, there is no such thing as a rabbinic equivalent for the discipline of physiognomy, nor is there a rabbinic equivalent for the kinaidos.70 The halakhic androginos is merely a person, or for that matter an animal (m. Bek. 6:12, b. Bek. 41a–42a), who has both sexual organs, just as the saris is a man who cannot reproduce. When a halakhic context requires decisiveness as to the gender identity of the doubly-sexed human or animal the default sex of the androginos is the male.71 In sum, even though the rabbinic semiotics of the body open the gate towards a remarkable self-consciousness about the potential ambiguity of its signs, the same system manages to maintain its fundamental gender binarism in Jewish law.

Boyarin comes to a similar conclusion, but from a different direction. Adopting Foucault’s thesis that sexuality is an invention of the nineteenth century rather than being a meta-historical category, Boyarin argues that “it becomes … plausible to begin by assuming that Jewish culture of the biblical and talmudic periods was not organized around a system of sexual orientations defined by object choice (or for that matter in any other way) … I know of no evidence that would support the claim for a system of sexual orientations (there is not talmudic equivalent even for the cinaedus),” “Are There Any Jews in `The History of Sexuality’?,” 335. 71 Joshua Levinson, “Cultural Androgyny,” 127. 70

CLOSING THE CIRCLE: YONAH FRAENKEL, THE TALMUDIC STORY, AND RABBINIC HISTORY Hillel I. Newman University of Haifa It is only proper to begin with a story. It was told to me as a faithful account of an historical event, though it has been marred, undoubtedly, by my informant’sȦand my ownȦirrepressible literary urges and our flawed memories. Out of both tact and genuine respect, I omit the names of the two heroes, both of an earlier generation. The story is nonetheless, I contend, true in all important respects. Once, two leading scholars of the Hebrew UniversityȦthe one a prominent historian of the talmudic period, the other a specialist in the literary study of rabbinic literatureȦattended a conference together. The historian delivered his lecture, at the end of which the literary critic arose to respond. He announced ceremoniously to the audience that it was his daily practice to add a special blessing to his morning prayers: “Blessed art Thou, who hast not made me a historian of the People of Israel.” The historian stormed out in anger. I relate this anecdote to illustrate the fact that the tension between historical and literary approaches to rabbinic sources is a crux of long standing in Israeli scholarship, though the effects of this tension on the writing of history in Israel have become manifest much more slowly than in America and Europe. Everywhere, the burden of proof has shifted: if, in the past, rabbinic narrative accounts conventionally enjoyed a presumption of historical reliability unless shown to be patently fictitious, today the historian making use of an aggadic narrative is obliged to demonstrate why and how that source should be considered serviceable for the study 105

106

HOW SHOULD RABBINIC LITERATURE BE READ?

of history. 1 The pivotal role of Jacob Neusner in bringing about this paradigm shift in America and Europe needs no elaboration, and his influence on the way Jewish history is being written is increasingly recognizable in Israel as well.2 At the same time, historians’ habits and long held assumptions have faced another challenge from a different quarter, in the works of Professor Yonah Fraenkel of the Hebrew University.3 Though for the historian the implications of the arguments of these two scholars are not dissimilar, the resemblance is in part superficial, and in the final analysis Neusner For retrospective surveys of trends in scholarship as seen from both sides of the Atlantic and the Mediterranean cf.: I.Gafni, “Erets Yisrael Bitequfat Hamishnah Vehatalmud: ʗeqer Shanot Dor—Hesagim Vithiyot,” Cathedra 100 (2001), 199-226; S. Schwartz, “Historiography on the Jews in the ‘Talmudic Period’ (70-640 CE),” in: The Oxford Handbook of Jewish Studies, M. Goodman, ed. (Oxford: 2002), pp. 79-114. A broad spectrum of opinion and method may be found in J. Neusner, A.J. Avery-Peck, eds., Judaism in Late Antiquity, III,1 (Leiden: 1999). 2 A small sample of representative studies will have to suffice: J. Neusner, In Search of Talmudic Biography: The Problem of the Attributed Saying, (Chico: 1984); idem, Reading and Believing: Ancient Judaism and Contemporary Gullibility, (Atlanta: 1986); “Story and Tradition in Judaism,” in: Judaism: The Evidence of the Mishnah, (Atlanta: 1988). Note that Neusner’s thinking on the subject has evolved over time and that the studies cited here reflect its later stage. 3 Most of his separate studies of the rabbinic story, written over a period of more than thirty years, have been collected in Y. Fraenkel, Sipur Ha’agadah—’Aʚdut shel Tokhen Vetsurah (Tel Aviv: 2001). See also: ȧIyunim Beȧolamo Haruʚani shel Sipur Ha’agadah (Tel Aviv: 1981); Darkhei Ha’agadah Vehamidrash (Givatayim: 1991), esp. pp. 235-285; Y. Fraenkel, Midrash Ve’agadah (Tel Aviv: 1996), esp. pp. 329-397. For studies in English: “Bible Verses Quoted in Tales of the Sages,” Scripta Hierosolymitana, 22 (1971), 80-99; “Paronomasia in Aggadic Narrative,” Scripta Hierosolymitana, 27 (1978), 27-51.; “Chiasmus in Talmudic-Aggadic Narrative,” in Chiasmus in Antiquity, J.W. Welch, ed. (Hildesheim: 1981), pp. 183-197; “Time and Its Role in the Aggadic Story,” Binah, 2 (1989) 31-56. For reviews of Fraenkel 1991 see: M. Hirshman, “‘Al Hamidrash Keyetsirah: Yotsrav Vetsurotav,” Jewish Studies 32 (1992), 83-90; R. Kalmin, “The Modern Study of Ancient Rabbinic Literature: Yonah Fraenkel’s Darkhei ha’aggadah vehamidrash,” Prooftexts, 14 (1994), 189-204. J.L. Rubenstein, Talmudic Stories: Narrative Art, Composition, and Culture (Baltimore: 1999), pp. 8-12 provides an insightful assessment of Fraenkel’s method against the backdrop of various schools of twentieth-century literary theory. 1

YONAH FRAENKEL

107

and Fraenkel are distinguished by profound differences in method and point of view. In what follows, I attempt a critical evaluation of the implications of Fraenkel’s work for the historian, with an eye towards marking out his position in the larger landscape of current scholarship.

GENRE Over a period of more than thirty years, Fraenkel has proven himself to be this generation’s most sensitive, insightful and eloquent interpreter of the rabbinic story. Even if we do not find all his readings equally compelling, at his best he is simply unmatched. In the process, he has virtually created a new discipline, adapting the theory and tools of general literary criticism to the specific case of the rabbinic narrative. His writing has often shown a polemical edge, directed not least of all against historians of rabbinic Judaism and the talmudic period (a scurvy lot, to which the present writer must confess that he belongs). The challenge he addresses to the historian is fundamentally proprietary: to what genre should we assign the rabbinic narrativesȥhistoriography or literature? His answer, needless to say, is the latter. In this he is not unique. What sets him apart is his rigorous demonstration of the genre by means of the painstaking analysis to which he subjects his sources, thereby uncovering, time and time again, sophisticated consistency of structure (most notably chiasmus) and literary embellishments (e.g., various forms of paronomasia). His readings are founded on a thorough philological examination of the texts, taking into account the full range of available manuscript evidence, and display an acute appreciation of the finest linguistic, halachic and exegetical nuances. Yet all this is secondary to Fraenkel’s ultimate purpose as an interpreter: to expose the resolution of the psychological and dramatic tensions in each story, which holds the key to the didactic intent of the storyteller, whose goal is to bring the listener or reader to some new level of insight and appreciation of reality.4 The rabbinic story 4

Fraenkel 2001, pp. 131-132. Note also his acknowledgement of the usefulness of structuralist criticism as an analytical tool on the one hand, together with his critique of its limitations for the purpose of uncovering the unique inner meaning of rabbinic stories on the other (p. 130). For a similar assessment (in the context, however, of a very different approach to uncovering the purpose of the storyteller) cf. Neusner 1988, pp. 324-

108

HOW SHOULD RABBINIC LITERATURE BE READ?

is thus shown to be fraught with design and rife with artifice, far removed from the methods of even ancient historiography, and contrived for purposes other than portraying real events of the past or present. True, the storyteller may incorporate elements of reality not only in the props and scenery of the narrative, but even in its plot. But it is the burden of the historian to demonstrate why this or that part of the story should be taken as witness to a real event.5 One obvious difference between Fraenkel’s historical skepticism and that of the Neusnerian school is how extremely genrespecific his position is. For Fraenkel, rabbinic texts are not wholly unreliable as repositories of conventional historical knowledge: rabbinic stories are literature, as demonstrated by the tools of literary analysis, but the sources defining and elaborating halacha are not. Nor, for that matter, are non-narrative aggadic materials (such as midrashic exegesis or gnomic sayings) subject to the same canons as the rabbinic story, though they are susceptible to literary analysis. Fraenkel further discriminates between “artistic” aggadic stories and “realistic” halachic stories (not to be confused with aggadic stories dealing with halachic themes). Halachic stories, which by his count number in the hundreds, containȥaccording to his definitionȥonly a description of the case or circumstances in question, an exchange of legal opinions and arguments, and a halachic resolution. They are generally devoid of the artistic contrivances that characterize the aggadic story and hence are “realistic.”6 What complicates matters is that some halachic stories, says Fraenkel, do indeed show signs of literary form and design, yet he insists that these too “are without a doubt realistic and not fictional.”7 This is not an obvious conclusion, for it is not clear why the literary devices that contribute to undermining the historical realism of the aggadic story should not do the same elsewhere. As Fraenkel’s 325 (the article has appeared several times under different titles, with minor variations). 5 Against the manner in which historians use rabbinic stories: Fraenkel 1991, pp. 235-238, 550-552; Fraenkel 1996, pp. 345-361; Fraenkel 2001, pp. 15-19, 373. For other specific examples: Fraenkel 1991, p. 358; Fraenkel 1996, p. 655; Fraenkel 2001, pp. 104 n. 53, 215-216 nn. 59 and 63. Against historicizing interpretations of rabbinic parables: Fraenkel 1991, pp. 379-383. 6 Fraenkel 2001, pp. 220-235, 366-368; cf. Fraenkel 1991, p. 633 n. 111. 7 Fraenkel 2001, p. 366 n. 3.

YONAH FRAENKEL

109

studies deal overwhelmingly with the aggadic story alone, his characterization of the halachic story awaits further investigation. One of Fraenkel’s contributions to our appreciation of rabbinic stories and the circles which produced them is his emphasis on their self-critical quality.8 These storiesȥby, for, and about Sages (a point to which we will return below)ȥare often shockingly frank in their depiction of the dramatic tensions and conflicts among Sages belonging ostensibly to a single fraternity. Strikingly absent is the hagiographic deference we might be predisposed to expect in literature devoted to the exploits of the audience’s religious paragons. The more overt dramatic confrontations in these stories have not, of course, escaped the notice of other readers. The tendency among some scholars (not unjustifiable, at least in some cases) is to read these biographical tales along the dividing lines of the factionalism of rabbinic society and see them as taking polemical positions with regard to the party politics of the day: disciple circle versus disciple circle, Patriarch versus Sages, and so forth. Read in this light, the issues at stake in these stories are power and authority.9 Fraenkel suggests another path. For him, the real battleground of rabbinic stories (including ones in which the latent criticism is by no means obvious) is the psyche of the individual, forced to confront his or her 10 own human failings. 11 The more we recognize edifying self-criticism to stand at or near the very heart of the rabbinic storytelling agenda, the more we are led to mute the political overtones of any one story.

PROVENANCE Let us turn to Fraenkel’s views on three aspects of the provenance of rabbinic stories: chronology, geography, and attribution.

8 See especially Fraenkel 2001, pp. 11-50; J. Fraenkel, “Teshuvah,” Tarbiz 49 (1980) 429. 9 See, for example, Neusner 1988; A.I. Baumgarten, “Rabbi Judah I and his Opponents,” JSJ, 12 (1982) 135-172. This list could be extended greatly. 10 Actually “his” alone, since Fraenkel conceives of the intended audience as consisting exclusively of Sages of the beit midrash (see below). 11 For this distinction: D. Stern, Parables in Midrash: Narrative and Exegesis in Rabbinic Literature (Cambridge, MA: 1991), p. 242, with n. 6.

110

HOW SHOULD RABBINIC LITERATURE BE READ?

Chronology It is important, when reading Fraenkel’s writings, to understand the somewhat idiosyncratic manner in which he categorizes the sources of rabbinic stories as classical and post-classical. The earliest evidence of rabbinic storytelling emerges, in his opinion, roughly at the end of the first century CE.12 From its first appearance in the tannaitic period, the rabbinic story shows all the distinctive traits of its full-blown classical form; in literary terms, the stylistic differences between tannaitic and amoraic stories are negligible. The classical story is to be found in all tannaitic sources (Mishna; Tosefta; the halachic midrashim), in both talmudim, and in Genesis Rabba. Leviticus Rabba and Pesikta De’rav Kahana are liminal works, containing classical stories, but also showing signs of the literary deterioration and contamination characterizing, by Fraenkel’s standards, the post-classical story prevalent in all other aggadic collections, including all those that would appear to precede the Muslim conquest and are often described by others as “classical.”13 Fraenkel’s post-classical story lacks not only the stylistic perfection of its classical predecessor, but also its ethical austerity. The postclassical story breaches the walls of the beit midrash and enters the realm of the synagogue, where it is tainted by a vulgar folk ethos.14 These categories are not just a matter of nomenclature or literary taste, for they have chronological significance as well. In absolute terms, the watershed between the classical and post-classical in both Palestine and Babylonia lies for Fraenkel roughly around 12 He refers to the story of Simon the Just and the Nazirite, according to t. Nez. 4:7 (ed. Lieberman, pp. 138-139), rejecting the opinion of those who would see this as an exceedingly early text of the Second Temple period (Fraenkel 1991, pp. 274-275; cf. pp. 497-498). The story is framed in the first person, told by Simon the Just himself. Assuming this is Simon of Avot 1:2 (cf., however, J. Schwartz, Hayishuv Hayehudi Biyehudah, [Jerusalem: 1986], p. 36), we are led to understand that Fraenkel considers this story to be pseudepigraphic. 13 Fraenkel 1991, pp. 3-10, 463; cf. Fraenkel 1996, pp. 761-846. Note that Fraenkel 2001, p. 244 n. 40, concedes that occasional post-classical stories may be found even in the Bavli. 14 Fraenkel 1991, pp. 274-280; Fraenkel 2001, pp. 51-74, 223. I make use of loaded terms advisedly. Fraenkel makes no effort to hide his identification with the classical story and his distaste for the post-classical (see, for example, Fraenkel 2001, p. 244 n. 39).

YONAH FRAENKEL

111

the year 500 CE. The relative chronology that Fraenkel gives for the Palestinian sources is in fact fairly conventional, though generally derived from arguments of literary dependence and not literary style. There is, however, something artificial and unconvincing about the alleged correlation of Palestinian and Babylonian chronologies. The schematic date for the end of classical rabbinic storytelling is derived from Rav Sherira Gaon’s date for the death of Ravina b. R. Huna in Babylonia.15 Yet it is not clear that this is a terribly meaningful date for the history of the talmudic text, let alone for the talmudic story.16 Probably the most popular topic in current academic study of the Bavli is the question of the extent and nature of the “stammaitic” contribution to its formation. Rubenstein has gone so far as to argue that the rabbinic story in the Bavli is primarily a product of the stammaitic redactors.17 Yet even if we do not subscribe to that position, it is unwarranted to deny the “stammaim” a priori any role in shaping the stories of the Bavli. What is more, our knowledge of the precise dating of the redaction of Genesis Rabba and Leviticus Rabba is so limited, that it is ultimately misleading to present an idealized picture of a unified period of classical rabbinic storytelling culture in the two rabbinic centers, coming to an end simultaneously at the end of the fifth century. We must allow for a more flexible reconstruction, in which contemporaries in Babylonia and Palestine are not necessarily doing precisely the same thing at the same time.18

15

Fraenkel 1996, p. 764; cf. Iggeret Rav Sherira Gaon, ed. Lewin, p. 95. On Ravina b. R. Huna see A. Cohen, Rabina Veʚakhamei Doro (Ramat Gan: 2001), pp. 54-55. 17 Rubenstein 1999 (esp. pp. 15-22); for a critique see H. Fox, “Biography, Stories, Tall Tales: Fishing for Gullibility,” Jewish Studies, 41 (2002), 105-141. For an example of the potential historical implications of such an approach see J.L. Rubenstein, “The Rise of the Babylonian Rabbinic Academy: A Reexamination of the Talmudic Evidence,” Jewish Studies Internet Journal, 1 (2002), 55-68 (http://www.biu.ac.il/JS/JSIJ/12002/Rubenstein.pdf). 18 For that matter, we would not be surprised to find such diversity in Palestine itself. Much depends on the vexed question of the dating of Tanʘuma-Yelammedenu materials, which are decidedly post-classical in nature. For an early dating: Marc Bregman, “Masorot Umekorot Qedumim Besifrut Tanʘuma-Yelamdenu’ Tarbiz, 60 (1991), 269-274. 16

112

HOW SHOULD RABBINIC LITERATURE BE READ?

Geography The most striking thing about Fraenkel’s view of the difference between the rabbinic story in Palestine and Babylonia is that there is hardly any difference worth mentioning. We have already seen how he attempts to unify the chronology of the two geographical centers of rabbinic Judaism. This unity is consistent with his evaluation of the aggada of the Bavli in general, for, in contrast to the prevailing opinion, he does not find Babylonia any less preoccupied with aggada than Palestine or inferior in its aggadic creativity. If anything, it is in the Bavli that the rabbinic story finds its richest artistic expression, hence Fraenkel’s predilection for the stories of the Bavli.19 While Fraenkel nowhere explicitly objects to the trend of current scholarship to seek out and underscore the social and cultural distinctions between the Jews of Babylonia and Palestine as reflected in the stories of the two talmudim (often in parallel versions of the same story),20 neither does he express much interest in such inquiry. This is probably a result not so much of his overall unifying perspective on the aggada of the two centers, as of his methodological insistence on the “closure” of the rabbinic story (see below), a conviction which tends to dismiss the importance of social context as well as suppress the kind of synoptic reading which highlights these differences.21 Let us look at one example which illustrates how attention to regional distinctions may supplement a strictly “closed” reading of a rabbinic story, and perhaps even call into question some of the conclusions of that reading. In his seminal article on the herme19

Fraenkel 1991, p. 6; Fraenkel 1996, pp. 772-779. Cf. the observations of Hirshman 1992, pp. 87, 89. 20 D. Rosenthal, “Masorot ’Erets Yisre’eliyot Vedarkan Levavel,” Cathedra 92 (1999), 7-48 provides a useful overview of this sort of investigation, with numerous examples. 21 The exceptional cases in which Fraenkel emphasizes the Babylonian social context of the stories of the Bavli are those portraying encounters between a pious Jew, a wicked Jew, and a gentile (usually a king). He points to several examples of such a pattern and suggests that they reflect the galut anxieties of the Babylonian Jews: Fraenkel 1991, p. 721 n. 26 (cf. p. 642 n. 42); Fraenkel 2001, pp. 128 n. 97, 169 n.72, 210 n. 44, 215, 271 n. 32. For a possible example of the projection of a Babylonian practice onto the landscape of Palestine: Fraenkel 1991, p. 576 n. 163.

YONAH FRAENKEL

113

neutical questions governing the interpretation of the rabbinic story, Fraenkel provides us with an interpretation of three different versions of the famous story of the appointment of Hillel as nasi.22 In each version, Hillel is rewarded for demonstrating scholarly acumen with regard to the halachic question of the permissibility of offering a paschal sacrifice on the Sabbath. Fraenkel skillfully analyzes each story, pointing out the differences among them and arguing for the artistic integrity of each version unto itself. He demonstrates that while the Tosefta version is unequivocally laudatory of Hillel’s scholarly performance, the two talmudim critically present him as being punished with temporary forgetfulness due to a display of rabbinic hubris. Yet the talmudim also differ. In the Yerushalmi, Hillel displays his prowess in legal argumentation and exegesis but is only rewarded with the appointment to the Patriarchate when he succumbs to the insistence of his audience and concedes that he has received a tradition to the same effect from his masters, Shemaya and Avtalion, regarding the question at hand. In the Bavli he is rewarded for his intellectual fireworks alone, and we do not hear explicitly of a tradition received from his teachers. The differences in the remainder of the story are less substantial. Let us look at the conclusion of the tale according to the Bavli: Immediately they seated him at the head and appointed him patriarch over them, and he expounded the laws of the Passover throughout the day. He began to rebuke them with words. He said to them, “What caused it that I should come up from Babylonia and that I should be the patriarch over you? Your laziness, that you did not serve the two great men of the generation, Shemaya and Avtalion.” They said to him, “Master: If one forgot and did not bring a knife on the eve of the Sabbath, what [is the law]?” He said to them, “This law I heard and forgot. But leave Israel be. If they are not prophets, they are the sons of prophets.” The next day, he whose Passover was a lamb stuck it [the knife] in its wool. He whose Passover was a kid [goat], stuck it between its horns. He [Hillel] saw the act and remem-

22

Fraenkel 2001, pp. 22-32. A convenient translation of the three may be found in J.L. Rubenstein, Rabbinic Stories (New York: 2002), pp. 71-79.

114

HOW SHOULD RABBINIC LITERATURE BE READ? bered the law and said, “Thus I received the tradition from the mouth of Shemaya and Avtalion.”23

Fraenkel finds in the Bavli, as in the Yerushalmi, implied criticism of Hillel, who prefers to highlight his own intellect instead of the received tradition. Yet there remains an outstanding question: if, as Fraenkel argues, Hillel’s audience prefers and expects an answer to its initial question based on received tradition (as in the Yerushalmi), why do his listeners pounce and respond “immediately” to his arrogant display by appointing him nasi? If, on the other hand, there is no such expectation on their part, how are we to explain the prominent place of received tradition in the concluding part of the story, which we have quoted? Fraenkel is conscious of this dissonance, which he resolves by seeking to expose implied tension between Hillel and his listeners in the first part of the Bavli story.24 According to Fraenkel’s hermeneutical ground rules, this is an exegetical imperative, predicated on the assumption of the integrity of the story and the principle that it must contain within itself all the answers to the questions it generates.25 Nevertheless, the fact remains that the audience in the Bavli readily accepts that which the audience in the Yerushalmi finds totally unacceptable. The problem may be attacked differently, however, as has been shown by David Rosenthal (in the name of his father, the late E.S. Rosenthal).26 His approach may be described as historicizing, but it has little to do with the sort of historical reading to which Fraenkel objected in the first place. Rosenthal distinguishes between two diametrically opposed academic tendencies in the development of halacha: the devotion to the careful collection and transmission of received tradition versus the creative process of 23

Rubenstein, Rabbinic Stories, pp. 75-76. I am not convinced that the question posed by the Bnei Betera to Hillel in the first part of the story – Minain lakh? (i.e., whence do you know this?)ȥis an appeal for a received tradition (thus Fraenkel 2001, p. 30). Some passages in the Bavli could be taken to suggest precisely the opposite. Cf. for example b. Zev. 57a (Yose Hagelili to R. Tarfon). 25 The potential hazards of this principle when taken to extremes are apparent. In the worst-case scenario we enter a vicious circle: we imagine our way to solutions that take us through the rough spots of an artistically flawed story, on the assumption that the story cannot be artistically flawed and hence there can be no genuine rough spots. 26 Rosenthal 1999, pp. 30-36. 24

YONAH FRAENKEL

115

inductive reasoning and dialecticȥ“Sinai” versus “the uprooter of mountains” in the rabbinic metaphor. He continues to demonstrate that the former is particularly characteristic of the Palestinian academy, the latter of the Babylonian. In this light, he says, the difference between the two versions of Hillel’s appointment takes on new meaning. The Yerushalmi portrays Hillel’s audience in its own imageȥrewarding him for the tradition he transmits. 27 The Bavli does the sameȥbut in reverse, rewarding skills of argumentation and induction. And what of the conclusion of the Bavli version? Here we encounter a more substantial deviation from Fraenkel’s method. For Rosenthal, it is precisely the contradiction between the implicit values of the first and second parts of the story which proves the priority of the more consistent version in the Yerushalmi, as well as the genetic dependence of the Bavli upon it. The latter is not a perfectly crafted whole, but a patchy and not entirely successful reworking of the former, in the spirit of Babylonian norms and values.28 Here, attention to the social and historical 27 Perhaps we may even identify an oblique polemic against a certain type of Babylonian scholar who finds his way to the beit midrash in Palestine and seeks to impress with his sophistry. 28 This kind of analysis has obvious implications for another hotly debated topic in current talmudic scholarship: the precise nature of the “synoptic” relationship among various rabbinic compositions. When do we assume dependence of one text on another, and when do we assume that two parallel texts borrow independently from a third? The implications of this problem for the historian should not be underestimated. Where we can prove dependence, it is necessarily the earliest version of the story which may make the strongest claim to historical authenticity, relatively speaking. For example: in the case of Hillel, I have no idea what, if any, connection there is between our stories and the life of the historical figure. I do know, however, that if the Bavli version is dependent on that of the Yerushalmi, and the Yerushalmi on the Tosefta, then it is with the Tosefta that I should begin my questȥ quixotic, perhapsȥin search of the historical Hillel, and that the deviations in the talmudim should be taken as reflections of no more (and no less) than the agenda of the later storytellers. For a model of genetic dependence together with a case study of its implications for the historian see S. Friedman, “La’agadah Hahistorit Batalmud Habavli,” in Saul Lieberman Memorial Volume, S. Friedman, ed. (New York: 1993), pp. 119-164. On the problem in general: S.J.D. Cohen (ed.), The Synoptic Problem in Rabbinic Literature (Providence: 2000); cf. the remarks of S. Schwartz, Imperialism and Jewish Society, 200 B.C.E. to 640 C.E.

116

HOW SHOULD RABBINIC LITERATURE BE READ?

context of the storyteller himself provides us with an interpretive solution, while a closed literary approach proves inadequate. Attribution Did the Sages really say what rabbinic texts attribute to them? A thorough discussion of the problem of attributed sayings in the aggada is impossible within the scope of this paper, yet a few words are in order, even though in general this is not an issue of special concern in Fraenkel’s writings.29 For the historian, the question of attributions has implications both for establishing the chronology of the sayings and for the possibility of arriving at some kind of portrait of the named Sage. Inasmuch as for Fraenkel the rabbinic story is an artistic creation, the dialogue placed in the mouths of the rabbis in these stories is naturally understood to be an invention of the storyteller and not a protocol of authentic statements. What of the attributions of the stories themselves to particular Sages as storytellers? Here the issue is less clear-cut. For example, Fraenkel occasionally expresses his admiration for the artistic merits of stories brought by the Bavli in the names of Rav and R. Yehuda, and he finds these attributions credible;30 this often seems to be the case in his comprehensive discussion of gnomic sayings as well.31 On the other hand, he suspects that the stories attributed by the Bavli to the Palestinian R. Joshua b. Levi have undergone Baby-

(Princeton 2001), pp. 8-9. I do not mean to imply that Fraenkel takes a Neusnerian view of the self-containment of individual rabbinic “documents” and their contents. The superficial similarity of approach derives mainly from Fraenkel’s notion of the literary closure of the individual rabbinic story. In broader perspective, however, there is indeed a common thread connecting the dogmatic excesses of both literary and “documentary” solipsism, each of which threatens to render us conceptually hamstrung. 29 For recent statements of the problem, with references to earlier literature: M. Bregman, “Pseudepigraphy in Rabbinic Literature,” in: Pseudepigraphic Perspectives: The Apocrypha and Pseudepigrapha in Light of the Dead Sea Scrolls, E.G. Chazon, M. Stone, eds. (Leiden 1999), pp. 27-41; Gafni 2001, pp. 218-219. On the significance of attributed sayings in general see Fraenkel 1996, p. 165; Fraenkel 2001, pp. 346-347. 30 Fraenkel 1991, pp. 275, 281; Fraenkel 2001, p. 240. 31 Fraenkel 1991, pp. 395-434.

YONAH FRAENKEL

117

lonian reworking,32 and in fact he often addresses the possibility of redactorial involvement in transmitted sayings. Earlier we encountered a story told in the first person by Simon the Just which Fraenkel seems to consider pseudepigraphic.33 We may suspect that this is also his opinion concerning “folk tales” brought in Leviticus Rabba and Pesikta De’rav Kahana in the names of R. Aha (fourth century CE) and R. Levi (late third century CE).34 In short, the attribution of stories remains by and large an open question in Fraenkel’s work.35

CLOSURE Without a doubt, the single most significant literary concept informing Fraenkel’s analyses of rabbinic stories is closure (segirut),36 which we have already encountered several times in the course of our discussion. In Rubenstein’s words: A dogmatic principle of Fraenkel’s is that stories display “internal and external closure.” By “internal closure” Fraenkel means that stories are self-contained units, complete in and of themselves. The story supplies all the information necessary for its interpretations. The end refers back to the beginning, completing the circle and sealing the story in a world of its own. Indeed, this self-contained quality is part of the literary artistry of the rabbinic story. By “external closure” Fraenkel means that each story is independent of a wider narrative framework. One cannot ask what happened before or after the events related in the story, what caused the situation, or what consequences ensued. Nor do stories relate in any way to other stories, even if they portray the same character or address a common theme. Thus stories about a given sage may contradict since each story was created for its own di-

32

Fraenkel 1991, p. 194; Fraenkel 2001, p. 362 n. 66. See above, n. 12. Cf. also his remarks on the first-person story attributed to R. Yose in b. Ber. 3a (Fraenkel 2001, p. 154). 34 Fraenkel 1996, pp. 794, 802. 35 Kalmin 1994, pp. 202-203, is critical of Fraenkel’s lack of consistency in this regard. 36 Fraenkel 2001, pp. 32-34. 33

118

HOW SHOULD RABBINIC LITERATURE BE READ? dactic purpose. A story exists in a secluded textual space with no allusion or reference to other texts.37

Rubenstein emphasizes that Fraenkel’s critical orientation must be understood as deriving from the New Criticism of the mid-twentieth century, which postulated the self-containment of the text and dismissed the importance of context, thereby effectively snatching the text out of the hands of historical positivists. As more recent movements in general literary criticism (collectively and somewhat vaguely labeled as “New Historicism”) have rediscovered context, so have literary critics of rabbinic stories.38 Needless to say, this suggests new possibilities for the historian, who may once again be permitted to pose historical questionsȥalbeit not exactly the same ones as beforeȥwithout being considered a complete Neanderthal by his or her colleagues. Reading a story both in and out of isolation is a necessary exercise, which may just as easily yield complementary results as contradictory ones, all of which must be weighed against each other. Before turning to several examples to examine the limits of closure in the rabbinic story, I would like to discuss one more type of closure which is characteristic of Fraenkel’s view of rabbinic culture and rabbinic literature in general. With unremitting consistency, Fraenkel argues that the intended audience of rabbinic sources consists of rabbis alone. Here again, he is not unique,39 but there are few who make the case so ardently. Fraenkel’s social and literary closure are not entirely unrelated. Comparison of a rabbinic story with folk motifs is not a regular part of his agenda, not only because a story should be able to explain itself, but because the rabbis are not “folk,” but rather an intellectual elite living in a rarified atmosphere of shared values setting them apart from other Jews. More importantly, Fraenkel objects to reading midrash as homiletical literature of the synagogueȥan approach commonly 37

Rubenstein 1999, p. 10. Rubenstein 1999, pp. 10-15; cf. Gafni 2001, p. 220. As Rubenstein notes, D. Boyarin, Carnal Israel: Reading Sex in Talmudic Culture (Berkeley: 1993), pp. 11-15, was the first to introduce the language of New Historicism explicitly into the study of the rabbinic story. 39 See for example R.S. Sarason, “Toward a New Agendum for the Study of Rabbinic Midrashic Literature,” in: Studies in Aggadah, Targum, and Jewish Liturgy in Memory of Joseph Heinemann, J.J. Petuchowski, E. Fleischer, eds. (Jerusalem 1981), pp. 55-73. 38

YONAH FRAENKEL

119

associated with the name of Joseph Heinemann. Midrash, including the midrashic proem (petiʚta), belongs properly in the beit midrash in the company of Sages, not in the synagogue with its popular audience. If Fraenkel is correct, then all we know of synagogue literature comes mainly from the statutory liturgy and the piyyut.40 Incidentally, his reservations find indirect but weighty support in the recent work of Naeh on the Palestinian lectionary cycle. Naeh rejects the fluid model of almost countless local cyclesȥbased in large measure on assumptions concerning the lectionary significance of midrashic loci in the petiʚtot and targumimȥin favor of a more limited model of shared and normative sedarim. According to this point of view, the abundance of such loci tells us nothing of lectionary practice in the synagogue, but is rather a reflection of the wide-ranging curriculum of the beit midrash.41 Fraenkel is undoubtedly correct in the fundamentals of his argument, which comes as an important corrective to the excesses of a “synagogal” reading of the texts. Some have objected, however, to the extent to which he denies the preservation in classical midrash of the contents of homiletical materials from the synagogue (or from sermons in the beit midrash directed to a popular audience).42 The debate revolves in part around assumptions concerning the degree of sophistication of the putative audience and its ability to appreciate the artistry of finely crafted aggada, in part around the precise meaning of hints in various rabbinic stories to the location of sermons and the nature of their audience. I will not 40 Fraenkel 1991, pp. 16-23; Fraenkel 1996, pp. 87-98. On the petiʚta: Fraenkel 1991, pp. 445-448; Fraenkel 1996, pp. 206-216 (note that the apparently late petiʚtot of the Tanhuma type are relegated in part to the synagogue). In favor of seeking the origin of the proem in the beit midrash cf.: R.S. Sarason, “The Petiʘtot in Leviticus Rabba: ‘Oral Homilies’ or Redactional Constructions?” JJS, 33 (1982), 564; M.S. Jaffee, “The ‘Midrashic’ Proem: Towards the Description of Rabbinic Exegesis,” Approaches to Ancient Judaism, 4 (1983), 107. In Fraenkel’s opinion, the targumim are also literature of the beit midrash, with the exception of the later Ps.-Jonathan, which belongs in the synagogue (Fraenkel 1991, p. 7; Fraenkel 2001, p. 340). 41 S. Naeh, “Seder Qeri’at Hatorah Be’erets-Yisra’el: Iyun Meʘudash,” Tarbiz 67 (1998), 167-187 (esp. p. 168 n. 10); cf. Sarason 1982, p. 564. 42 Hirshman 1992, pp. 84-88; Kalmin 1994, pp. 199-202. Fraenkel, incidentally, does not deny that rabbis gave sermons in synagogues; he just denies that rabbinic sources preserve them.

120

HOW SHOULD RABBINIC LITERATURE BE READ?

repeat the argument in all its details, but will make do with brief remarks on a few passages that properly belong in the dossier of this discussion, but are not mentioned by Fraenkel or his reviewers:43 a) A sermon in the name of Yose of Maon, set in a synagogue of Tiberias, is brought in y. Sanhedrin 2:6, 20c-d (cf. Gen. Rab. 80:1, ed. Theodor-Albeck, pp. 950-953). b) In Leviticus Rabbah 35:12 (ed. Margulies, pp. 830-831) we find a description of a sermon of R. Samuel b. Nahman given in a synagogue in Lod. c) The late Midrash Shemuel 7:6 (ed. Buber, p. 68; cf. Midrash Hagadol to Exodus 20:20 [ed. Margulies, p. 443]) describes a sermon of Jacob of Kfar Nevoraiah in a synagogue of Caesarea. This is probably a secondary interpolation. Cf. y. Bikkurim 3:3, 65d, which is silent about the venue of the sermon, but refers to the synagogue of Caesarea in a different context. d) An appendix to Pesikta Rabbati (ed. Ish-Shalom, p. 196b) describes a sermon of R. Abahu in a synagogue of Caesarea, but this source would certainly qualify as “postclassical.” e) In Ecclesiastes Rabbah 6:2 (cf. Ecclesiastes Zuta 5:17 [ed. Buber, p. 130]) we find an account of a synagogue sermon of R. Levi, but in the parallel passage in the earlier y. Horayot 3:7, 48c, it is located in “a certain place” (ʚad atar). While we could dismiss this too as a post-classical reworking of an earlier source, there is some evidence that the plain meaning of Aramaic atar in this context is in fact “synagogue.”44 Of course, in the overwhelming majority of homiletical passages there is no indication whatsoever of venue. Thus, we are left to our own devices to extrapolate our conclusions from the relatively meager evidence of the more explicit sources. We may agree with Fraenkel’s preference for reading our sources in the context of

43 Most of the relevant sources are collected in L.I. Levine, The Ancient Synagogue: The First Thousand Years (New Haven: 2000), pp. 461-462. 44 See M. Sokoloff, A Dictionary of Jewish Palestinian Aramaic of the Byzantine Period (Ramat-Gan: 1990), p. 82, and cf. Steven Fine’s contribution to this volume. Fraenkel 1991, pp. 23-24, does not deal with this question.

YONAH FRAENKEL

121

the beit midrash, but we should not dogmatically rule out the incorporation of some synagogue material as well. How truly closed is the rabbinic story? Since there are so many aspects to this closure, it is helpful to break down the question into smaller parts. I begin with Fraenkel’s method of divorcing rabbinic stories from one another. We have already seen some of the theoretical underpinnings of this approach, whichȥ though sometimes overextendedȥis not fundamentally unsound. In the view of Fraenkel and others, the linkage of one story to another, or of a story to an attributed saying, belongs to the naïve genre of rabbinic biography of days gone by.45 Yet the question should first be framed in literary terms: can we ever speak of the coherence of different stories, from which emerge the pieces of a literary persona, whose historical worth must then be critically evaluated? 46 Most scholars would concede that there is some body of “common knowledge” of the Bible, halacha, and aggada that the rabbinic storyteller rightfully expects of his reader; to deny this altogether leads to absurdity. It is at this point that we may begin to haggle over what should qualify as “common knowledge.” In fact, when we examine Fraenkel’s work closely, we find a case where he himself stretches the limits of closureȥquite rightlyȥto introduce external “biographical” information which he presumes was shared by the storyteller and the audience. Thus, when reading the deathbed narrative of R. Eliezer, Fraenkel assumes knowledge of the tale of R. Eliezer’s excommunication over the dispute concerning the oven of Akhnai.47 In this instance the connection is so reasonable, that we must insist on the legitimacy of the search for such resonance elsewhere. 45

Fraenkel 2001, pp. 273-274. Cf. two recent examples of extended “literary” biographies: O. Meir, Rabi Yehudah Hanasi’—Deyokno shel Manhig Bemesoret ’Erets Yisrael Uvavel (Tel Aviv: 1999); A. Goshen-Gottstein, The Sinner and the Amnesiac: The Rabbinic Invention of Elisha ben Abuya and Eleazar ben Arach (Stanford: 2000). 47 Fraenkel 2001, p. 161. Cf. Fraenkel 2001, p. 361 n. 59, for a comparison of the excommunication story and the story of R. Eliezer in m. Yad. 4:3. Among other cases where Fraenkel regards external biographical information as “common knowledge”: Fraenkel 1991, p. 278 (R. Hanina b. Dosa is poor); Fraenkel 2001, p. 208 n. 36 (R. Sheshet is blind); pp. 208209 (R. Eleazar b. Azariah is rich). See further on the problem of closure and the presumption of halachic knowledge: Fraenkel 2001, pp. 371-372. 46

122

HOW SHOULD RABBINIC LITERATURE BE READ?

What knowledge does the storyteller have of literary sources or motifs from outside the realm of Judaism, and how important is it for us to identify them in order to appreciate the story? Fraenkel objects strenuously to the efforts of some folklorists to read rabbinic stories (that is to say, classical rabbinic stories) as folk literature. Superficial similarities between the rabbinic story and the folk tale result in part from the fact that both are essentially oral literature. Certain motifs are common to folk tales and rabbinic stories, but this has no diagnostic significance for the latter: so-called folk motifs are not restricted to folk literature. Even when we identify a genuine folk motif in a rabbinic story, on closer examination it will become apparent that the storyteller has not taken over the material mechanically, but has rather used it as grist for the rabbinic mill, adapting it to the rabbinic value system in the process.48 These are powerful arguments, and Fraenkel’s approach is surely correctȥwith one reservation. While he does not deny the existence of borrowed motifs in rabbinic stories, Fraenkel does not in practice assign much importance to their identification, let alone to an analysis of their adaptation. When he does acknowledge such an element, he does so dismissingly. Thus, in his discussion of the story of Simon the Just and the Nazirite (who was taken with his own reflection in the river), Fraenkel mentions the apparent affinity to the myth of Narcissus, though he does not rule out the possibility of the independent origin of the motif in Palestine. “One way or another,” he continues, “the use of this motif in our story proves indisputably how the mere fact of the use of a popular or foreign motif still says nothing about the ideological contents of the rabbinic story.”49 Perhaps the use of the motif does not explain the story, but it does tell us some interesting things about the storyteller, the cultural environment by which he is influenced, and the manner in which he works. It is the pre-history of the motif as much as anything else which highlights that which is distinctive in the rabbinic story. These facts are not devoid of historical or artistic significance. We could identify King David as such in the Dura fresco or the Gaza synagogue mosaic without looking for foreign

48

Fraenkel 1991, pp. 280-285; Fraenkel 2001, pp. 236-260. Fraenkel 1991, p. 498. Other folk motifs are rarely mentioned. See Fraenkel 2001, pp. 163 n. 60 (Hadrian and the Old Man); p. 183 n. 24 (ʙoni Hameȧagel). 49

YONAH FRAENKEL

123

influence, but when we recognize Orphic motifs in these depictions we enjoy a new and far richer appreciation of what the artists have done and the world in which they live. The “Jewishness” of the art is not diminished by the comparison; rather, we learn something new about the nature of that Jewishness. The same, by way of analogy, is true of the rabbinic story. The identification of parallels may have further implications for the historian. Uncovering dependence on a foreign narrative can be the most elegant way to undermine the historical foundation of a rabbinic story. We may, for example, waffle all we like about the historical kernel of the tale of Judah b. Tabai and his disciple in Alexandria (y. Hag. 2:2, 77d; y. San. 6:9, 23c), but we may spare ourselves the trouble after learning from Stephen Gero that similar stories are told of Christian holy men in sources unlikely to have been influenced by the Jewish narrative.50 In some cases, the reworking of sources contrary to their original intent may reveal polemical motives.51 The search for comparative materials is an ongoing process. By way of example, let us examine one of the stories which Fraenkel has analyzed to try and shed new light on it with the aid of a non-rabbinic parallel. In b. Shabbat 119a we find the following famous story:

50

S. Gero, “The Stern Master and His Wayward Disciple: A ‘Jesus’ Story in the Talmud and in Christian Hagiography,” JSJ, 25 (1994), 287311. Cf. I. Lévi , “Légendes judéo-chrétiennes,” Revue des études juives, 8 (1884), 200-202 and “Encore un mot sur la légende de Bartalmion,” Revue des études juives, 10 (1885), 66-73 on the legend of R. Shimon b. Yoʘai and the demon in b. Meȧila 17b. 51 I have argued elsewhere that one rabbinic storyteller has deliberately used apocalyptic materials with ironic anti-apocalyptic intent (“Leidat Hamashiaʘ Beyom HaʘurbanȥHe‘arot Historiot Ve’anti-historiot,” in For Uriel: Studies in the History of Israel in Antiquity Presented to Professor Uriel Rappaport, M. Mor et al., eds. [Jerusalem: 2005]. Pp. 85-110). For a noncomparative reading of the same story see Fraenkel “‘Iyunim,” pp. 158-163. On the deliberate inversion of genres see further Jeffrey Rubenstein’s paper in this volume. Kalmin 1994, pp. 196-197, criticizes Fraenkel for paying insufficient attention to polemical factors. In a small number of cases Fraenkel does suggest an anti-Christian reading of homiletical passages: Fraenkel 1991, pp. 293, 599 n. 18.

124

HOW SHOULD RABBINIC LITERATURE BE READ? There was a gentile in the neighborhood of Joseph Who Honors the Sabbath whose property was very great. The astrologers said to him: “Joseph Who Honors the Sabbath will consume all your property.” He went and sold all his property and bought with it a pearl, which he put in his cap. While the ferry was crossing, the wind blew it off, cast it into the water, and a fish swallowed it. They caught it [the fish] and brought it on the verge of the Sabbath eve. They said: “Who buys at a time like this?” They told them: “Go and take it to Joseph Who Honors the Sabbath, who is accustomed to buy.” They brought it to him, and he bought it. He cut it open, found the pearl inside, and sold it for thirteen measures of golden dinars. A certain old man met him and said: “Whoever lends to the Sabbath, the Sabbath repays him.”52

Fraenkel expounds this story at length, pointing out the underlying tension between the determinism of the astrologers at the outset and the reward for Joseph at the conclusion. Joseph lives in a deterministic universe, but he lives indifferent to fate, even ignorant of it, making his choices freely and therefore truly entitled to the reward he receives. I will approach this tale from a different direction. First, we note that the story has not escaped the attention of folklorists, who look at it, of course, as a folk tale.53 To the best of my knowledge, though, they have not compared it to what is probably the parallel closest to it in provenance, both chronologically and geographically. I refer to a chapter in the Pratum spirituale, a collection of anecdotes recorded by the peripatetic John Moschos, a monk of Palestine (d. 619). Moschos relates a story he heard from a woman on the Greek island of Samos, who told him a tale of a Christian woman of Nisibis in Babylonia who was married to a pagan: 52 See Fraenkel “‘Iyunim,” pp. 16-18. For a different fish story in the Palestinian tradition cf. Gen. Rab. 11:4 (ed. Theodor-Albeck, pp. 91-92); Pesikta Rabbati 23 (ed. Ish-Shalom, pp. 119a-b). The story is also discussed in Fraenkel “Chiasmus,” pp. 184-185; compare the similar conclusions of N.J. Cohen, “Structural Analysis of a Talmudic Story: JosephWho-Honors-The-Sabbath,” JQR, 72 (1982), 161-177. 53 H. Schwarzbaum, Studies in Jewish and World Folklore (Berlin 1968), pp. 269-270; E. Yassif, The Hebrew Folktale: History, Genre, Meaning, trans. J.S. Teitelbaum (Bloomington: 1999), pp. 124-125.

YONAH FRAENKEL

125

They possessed fifty miliarisia. One day, the husband said to his wife: ‘Let us lend out that money and get some advantage from it, for in drawing on it a little at a time, we are going to spend it all’. The wife answered: ‘If you insist on lending this money, come: lend it to the God of the Christians’. He said to her: ‘Well, where is this God of the Christians, so we can lend it to him’? She said: ‘I will show you. Not only shall you not lose your money, but it shall even earn interest for you and the capital shall be doubled’. He said to her: ‘Come on then; show me and we will lend to him’.54

The woman takes her husband to the church of Nisibis, where he gives his money to the poor. Three months later, unable to meet his own expenses, the husband returns to the church, where he finds a coin on the floor. His wife sends him off to buy food. He buys bread, wine, and fish, and when his wife opens the fish, she finds a precious gem inside. He sells the gem for three hundred miliarisia. Filled with wonder at the goodness of God, she said to him: ‘Oh husband, see how good and generous and affluent is the God of the Christians! Look how he has not merely returned to you the fifty miliarisia you lent him together with interest, but in only a few days has given you the capital multiplied by six! Know therefore that there is no other God, neither on earth nor in heaven, but him alone’. Convinced by this miracle and learning the truth by experience, he immediately became a Christian and glorified our God and Saviour, Jesus Christ, with the Father and the Holy Spirit…

The importance of this parallel lies not only in the shared motif of the fabulous gem in the fish, but also in the fact that both Jewish and Christian stories explicitly cast the appearance of the gem as repayment of a religious loan, thereby validating the extraordinary stature of the borrower. Joseph lends to the Jewish 54

Pratum spirituale, c. 185 (PG 87, pp. 3057-3061). All quotations are from the translation of J. Wortley (trans.), The Spiritual Meadow (Kalamazoo: 1992), pp. 155-158. A similar tale is found in the Apophthegmata patrum (cf. H. Chadwick, “John Moschus and His Friend Sophronius the Sophist,” Journal of Theological Studies, 25 [1974], p. 54; Wortley 1992, p. 256-257).

126

HOW SHOULD RABBINIC LITERATURE BE READ?

Sabbath; the pagan lends to the god of the Christians. Joseph’s belief is vindicated at the expense of the gentile; the faith of the pagan’s wife is vindicated and he becomes a Christian. Without belittling all that is unique to each story, I suggest that the similarities are not merely coincidental, especially because both would appear to be Babylonian,55 and they are roughly contemporary. Chadwick has speculated, without being aware of the Jewish story in the Bavli, that the Christian story is an adaptation of a Persian folk tale.56 I am not familiar with such a Persian source and leave the search for it to specialists. The suggestion, however, is reasonable, and is made more so by the existence of the Babylonian Jewish parallel. Nevertheless, we cannot rule out the possibility of Jewish influence on the Christian storyteller, or the opposite. However we look at it, the comparison suggests intriguing interactions among neighboring religious and ethnic communities of Babylonia. Recognizing such processes is part of the job of the historian, and reading the story within the context of cultural and religious dialogue may expose within it another level of meaning. We come finally to what is the most fascinating and at the same time the most elusive question of closure. How “closed” is the rabbinic story to the reality of historical events and persons? I refer specifically to the events and persons in the story and not its contribution to the recovery of the social and intellectual world of the storytellerȥa separate though equally important realm of historical inquiry.57 Besides recognizing in principle that stories may begin their lives in the memory of some real event, Fraenkel dis55 In the Palestinian fish story of n. 52, there is originally no gem and no repayment of a loan. In Pesikta Rabbati these elements appear secondarily as a Hebrew addition to the Aramaic story, probably under the influence of our passage in the Bavli, and from there have contaminated the poorer textual traditions of Genesis Rabbah. For shades of the Babylonian milieu of the Bavli tale see also Fraenkel “Chiasmus,” p. 196 n. 5. 56 Chadwick 1974, p. 54 (following I. Opelt, “Der Edelstein im Bauch des Fisches,” in: Mullus: Festschrift Theodor Klauser, A. Stuiber, A. Hermann, eds. [Münster Westfalen: 1964], pp. 268-272). He also compares it to the story of Polycrates of Samos in Herodotus III, 42. 57 See D. Goodblatt, “Towards the Rehabilitation of Talmudic History,” in: History of Judaism: The Next Ten Years, B.M. Bokser, ed. (Chico: 1980), pp. 31-44, on the continuing importance of the latter, despite, as he calls it, “the debiographization of rabbinic literature.”

YONAH FRAENKEL

127

misses the possibility of giving meaningful answers to the question and focuses all his attention elsewhere.58 But having acknowledged the theoretical significance of these sources for the historian, is it possible for us to translate this recognition into anything resembling critical method? We have already addressed some of the difficulties of trying to recover an eventȥnot to mention a biographyȥfrom the rabbinic story. Even where we can readily identify the hand of the storyteller in contrivances of form and substance, we are generally not in a position to perform an act of literary dissection, as if discrete components of the invented and the real (assuming it to be there) could be peeled apart mechanically. When attempting to evaluate the historical content of stories about the lives of the Sages and the inner world of the beit midrash, we lack the most important of all controls: independent comparative materials from outside the rabbinic frame of reference. The only familiar figures of the rabbinic world and its literature of whom Greek and Latin authors, whether pagan or Christian, provide any information unmediated by Jewish sources are members of the patriarchal dynasty. Only a small fraction of the stories in the entire corpus of rabbinic literature deals even remotely with matters of Jewish political history, that is to say with events of the sort that might rightfully be expected to leave their mark on non-rabbinic sources as well. The largest body of such comparative material is found, not surprisingly, in the writings of Josephus, so our comparisons tend to be weighted towards the Second Temple period. They show us that while it is clear that the rabbis do not give us scientific accounts of the Second Temple period, it is equally clear that these rabbinic stories do sometimes preserve authentic historical information. Fraenkel himself discusses the case of the civil war between the Hasmonean brothers, Hyrcanus and Aristobulus, which came to a head around 65 BCE. Josephus (Antiquities 14:1928) tells that the forces of Hyrcanus besieged Aristobulus and his party in the Temple, but after receiving payment from those within the walls to provide animals for the paschal sacrifice, they reneged on their agreement, bringing upon the entire people divine retribution in the form of a storm which ravaged crops throughout the country. A similar account is found in three parallel passages in the

58

Fraenkel 1996, pp. 357-361, 683-684; Fraenkel 2001, p. 347.

128

HOW SHOULD RABBINIC LITERATURE BE READ?

Bavli (b. Sot. 49b; b. B. Qam. 82b; b. Men. 64b),59 albeit with some significant differences, not the least of which being that for the Bavli the story stands apart from any sequential historical narrative and that the storyteller takes obvious liberties with the material at hand. For Fraenkelȥquite correctlyȥthis is a parade example of the fundamental differences between the approaches of traditional historiography (specifically Josephus) and rabbinic storytelling with respect to a given event. He also exploits the differences to highlight the distance that separates the contemporary historian from the literary critic.60 For the historian, comparisons with parallel or similar accounts of an event in a non-rabbinic source are helpful not simply because they may corroborate the historicity of certain elements in a particular aggadic narrative, but also because they highlight what is unique to the rabbinic perception of the event.61 Yet no less important for our purposes, they teach that, despite the objections of some, the search for the impression of historical events and persons in rabbinic stories should not be dismissed as unwarranted in and of itself.62 Frustratingly, our external sources are far less helpful when we come to the time and circles of the storytellers themselves. The 59 Why so many of the closest parallels to Josephus are foundȥ counterintuitivelyȥin the Bavli alone remains one of the great unanswered questions of talmudic historiography. 60 Fraenkel 1991, pp. 236-238; Fraenkel 1996, pp. 346-361, 544-545. Characteristically, most of his discussion deals not so much with truly “contemporary” historians as with some of the outstanding figures of Wissenschaft des Judentums, in this case Grätz and Dubnow. 61 For recent discussion of such comparisons see inter alia: S.J.D. Cohen, “Parallel Historical Tradition in Josephus and rabbinic Literature,” in: Proceedings of the Ninth World Congress of Jewish Studies, B/1 (Jerusalem: 1986), pp. 7-14; A.I. Baumgarten, “Rabbinic Literature as a Source for the History of Jewish Sectarianism in the Second Temple Period,” Dead Sea Discoveries, 2 (1995), 14-57; A. Taran, “Yosefus Flavius Vete’orei Haʘurban Besifrut Hatalmudit,” Zion, 61 (1996), 141-157. 62 Asking potentially unanswerable questions is not an intrinsically misguided endeavor. I do not mean to suggest that Fraenkel himself levels this charge. One does, however, encounter it occasionally, and at its most extreme it takes the form of dogmatic obscurantism. No less harmful to scholarship than traditional hypo-criticism is the methodological fallacy of a neo-fundamentalism which delegitimizes a priori even qualified speculation about the traces of real events in the rabbinic story.

YONAH FRAENKEL

129

paradigm of the Second Temple tales, among other things, suggests that rabbinic stories should continue to contain a mixture of historical memory and literary invention. Yet even for the later period we may occasionally benefit from comparisons with external materials. How might this affect the way we read a story? In y. Shabbat 16:1, 15c we read: Rabbi,63 R. Hiyya the Great and R. Yishmael b. R. Yose sat and interpreted the Book of Lamentations on the eve of the Ninth of Av which fell on the Sabbath,64 and they left over one chapter. They said: “Tomorrow we will come and complete it.” When Rabbi departed for his house, he was injured in his finger and applied the following Scripture to himself: “Many are the pains of the wicked” (Psalms 32:10). R. Hiyya the Great said to him: “On our account this happened to you, as it is written: ‘Our breath, the anointed (meshiaʚ) of the Lord, was caught in their traps” (Lamentations 4:20). R. Yishmael b. R. Yose said to him: “Even if we had not been dealing with that matter65 we would have had to say this; all the more so since we were dealing with that matter.”

In his interpretation of the story, Fraenkel mentions the opinion of Urbach, among others, that the implied identification of R. Judah Hanasi with the Lord’s “anointed” should be read in light of other passages suggesting an air of messianic expectation surrounding that Patriarch.66 Fraenkel argues, not altogether convincingly,67 that the three protagonists describe three increasingly broad circles of responsibility for Rabbi’s injury in a finely crafted narrative.

63

= R. Judah Hanasi. Fraenkel 1991, p. 250, notes that study of Lamentations before the Ninth of Av is not otherwise mentioned in rabbinic literature. For the public reading of Lamentations on the Ninth of Av in a third-century sourceȥapparently the earliest reference to this practiceȥsee Didascalia apostolorum, c. 21 (trans. Vööbus, p. 201). 65 The verse in Lamentations. 66 Fraenkel 1991, pp. 250-251. Cf. E.E. Urbach, The Sages: Their Concepts and Beliefs (Jerusalem: 1975), p. 678. 67 It is doubtful, for example, whether Rabbi and R. Hiyya really seek the cause of the injury in the omission of the last chapter of Lamentations from the study session. 64

130

HOW SHOULD RABBINIC LITERATURE BE READ?

Even if we were to concede the craftsmanship, it is not clear why we should agree to the following: When we see how the exegesis and interpretations are integrated into the literary creation, we sense how much lovely playfulness there is in such interpretations and how doubtful it is whether trulyȥin the mind of the storytellerȥeach of the interpreters of the verses in the story has in mind the depth of the significance of the interpretation. Therefore a certain doubt arises regarding the opinion … that we may learn from this story that Rabbi was really considered the Messiah.68

Leaving aside the overstated presentation of Urbach’s position, 69 note that playful craftsmanship and deep exegetical intent are not mutually exclusive; if anything, the better the craftsman, the more thoroughgoing the exegetical construction should be. Fraenkel’s “non-messianic” reading is in fact problematic according to the very literary criteria he invokes. If we do not see ourselves bound by the canons of hermeneutical closure, we will also notice, with Urbach and others, that the “messianic” reading is consistent with a coherent pattern comprising other rabbinic passages referring to R. Judah Hanasi and also with roughly contemporary patristic evidence, specifically, the writings of Origen.70 This is not to say that the story in its entirety necessarily describes a genuine event. It does mean that we are entitled to regard it as further historical evidence for the quasi-messianic public persona of R. Judah Hanasi.71 68

Fraenkel 1991, p. 625 n. 94. Cf. Meir 1999, p. 357 n. 20. Urbach is careful not to say that “Rabbi was really considered the Messiah.” He does portray him as the focus of still unfulfilled messianic hopes. 70 D. Goodblatt, The Monarchic Principle: Studies in Jewish Self-Government in Antiquity (Tübingen 1994), pp. 146-175. 71 For rejection of another historicizing reading of a midrashic passage attributed to a Sage within a story see Fraenkel 2001, p. 216 n. 63, on M. Kister, “Bei’urim Be’agadot Haʘurban Be’avot Derabi Natan,” Tarbiz, 58 (1998), pp. 512-517. The midrash in question is that brought in the name of R. Yohanan b. Zakai on Isaiah 10:34 (as told to Vespasian), which, Kister argues, reflects a genuine exegetical tradition with polemical intent from the generation of the Great Revolt, though he hesitates to endorse the historicity of the narrative itself. He relates the midrash to passages in writings from Qumran and in Josephus. Rejecting this attempt to identify 69

YONAH FRAENKEL

131

What of the many stories for which no comparative nonrabbinic material presents itself? One familiar technique of historicizing readers is first to eliminate the supernatural, the miraculous, and the fabulous. The limitations of such an approach are obvious and have frequently been noted. For one thing, plausibility is a necessary but insufficient criterion for measuring the historicity of literary narratives. For another, when we ignore the fantastic elements of a story, we often willfully undermine precisely that which is most meaningful to the storyteller. Nevertheless, this technique should not be sold short, not because its results are guaranteed, but because sometimes it works, making it a legitimate starting point for intellectually honest speculation. Take the case of the Nicanor Gates, mentioned in various rabbinic accounts of the structure of the Second Temple in Jerusalem. The aggada of the Tosefta (with parallels in both the Yerushalmi and the Bavli)72 tells of one Nicanor, who brought the doors by sea from Alexandria to the Land of Israel. En route, a wave threatened to sink the ship. To save it, one of the doorsȥwhich were made of Corinthian bronzeȥwas cast into the sea; the second was spared, as Nicanor insisted on being thrown overboard along with it. When the ship sailed into the port of Jaffa (or Acco), the lost door floated up miraculously to the surface (or was spat up by a giant sea creature). The influence of the biblical model of Jonah is obvious here, and there would seem to be every indication that the story is a fantastic etiological invention, devoid of any biographical value regarding the historical

in the midrash an embedded tradition originating in a discrete historical context, Fraenkel insists that the exegesis of R. Yohanan b. Zakai is a thoroughly typical piece of rabbinic interpretation and should, with the story containing it, be read in literary isolation, without recourse to the sources of the period in which the story is set. Could we explain the emergence of the midrash in question in this way, without benefit of historical comparison? Undoubtedly, but that is insufficient reason to exclude comparative materials by force from the discussion when they present themselves. Kister argues for the appearance of literary-historical patterning among the sources. The rejection of the observation as illusory and unmeaningful is only as compelling as one’s conviction that such linkage is improbable a priori. Here Fraenkel has overstated his case, and the ideology of literary self-sufficiency threatens to close off a reasonable path of inquiry. 72 T. Kip. 2:4 (ed. Lieberman, p. 231); y. Yoma 3:8, 41a; b. Yoma 38a.

132

HOW SHOULD RABBINIC LITERATURE BE READ?

Nicanor for whom the gates were named. After all, the Mishnah merely states that an unspecified miracle happened to the doors of Nicanor, apparently counting him among prominent donors to the Second Temple, including members of the royal house of Adiabene, but suggesting nothing else of his identity or place of origin (m. Yoma 3:10; cf. m. Mid. 2:3). Our aggada, however, is quite emphatic about his bringing the doors from Alexandria. A rationalizing positivist (the present writer, for example) looking at this story together with the Mishnah and seeking the Holy Grail of the “historical kernel” might say: I don’t buy that business about the floating door, but Nicanor may very well have been a wealthy Jewish Alexandrian who donated the bronze gates to the Temple in Jerusalem.73 In this case, our positivist would probably not be far off the mark. This much can be learned from the famous ossuary inscription found in a burial cave on Mt. Scopus in 1902, which reads (in a combination of Greek and Aramaic): “Bones of the sons of Nicanor the Alexandrian who made the doors. Nicanor the Alexandrian.” 74 But in the final analysis we must admit that this approach too is no more than a limited palliative for an acute problem. The fissures in the closure of rabbinic storytelling and the incidental demonstrations of the intersection of story and history are sufficient reason for the historian not to abandon altogether the old pursuit of person and event. Yet seeking license to do so with73

Fraenkel “‘Iyunim,” pp. 25-28, argues that in the Bavli Nicanor is understood to be a Temple representative sent on a mission to Alexandria to deliver the doors to Jerusalem. Even if we take this to be the intention of the Bavli, it is not implied in the two Palestinian versions of the storyȥboth arguably earlier than the Bavli. 74 On Nicanor and his gate see J. Schwartz, “Once More on the Nicanor Gate,” HUCA, 62 (1991), 245-283; for the inscription see pp. 248249. Schwartz (following S. Safrai) in fact argues that the gate was named to commemorate the punishment visited upon Nicanor the Seleucid general and that the Mt. Scopus inscription has no bearing on our subject. He relies on the testimony of late witnesses to an alleged passage in the Yerushalmi to that effect. But, as others have already observed, the fact that different witnesses assign the passage to two different tractates and that it is absent from both in all extant manuscripts suggests that it is actually a late invention and interpolation (probably under the influence of Josippon), not a genuine “lost” Yerushalmi tradition. See the reservations of V. Noam, Megilat Taȧanit (Jerusalem: 2003), p. 299 n. 18.

YONAH FRAENKEL

133

out being charged with fundamentalist recidivism is one thing; making an honest living of it, given the inadequacy of our tools, is another. Thus, in practice Fraenkel’s challenge still stands, for the most part. But while he argues that stories are not the best point of departure for an appreciation of what distinguishes the individual Sage, we recall that at the same time he points in the direction of other materialsȥhalacha, non-narrative aggadaȥwhich are not beset by the same problems with which we are grappling here.75 This suggests a more hopeful view of the possibility of addressing at least some of the conventional issues of “rabbinic biography,” a program that rightly fell into disrepute in its old form. 76 At the same time, any new program of rabbinic biography must also include the study of rabbinic narrativeȥeven the most patently fictitious, so that proceeding beyond literary closure, we may understand the manner in which each literary persona develops over time.

CONCLUSIONS The classical form of the rabbinic story, as Fraenkel teaches us, obliges me to close where I began. Our two heroes, the historian and the literary critic, had difficulty finding a common language. Yet increasingly the methodological boundaries separating the two disciplines are becoming less distinct. Historians are conceding the significance of the literary components of rabbinic literature; literary criticism is rediscovering the importance of context, including historical context. Strictly speaking, there is indeed no such genre

75 For the hypothesis that brief stories are historically more reliable than lengthier, more elaborate ones see Kalmin 1994, p. 197; Rubenstein 1999, p. 297 n. 3. 76 See especially W.S. Green, “What’s in a NameȥThe Problematic of ‘Rabbinic Biography,’” Approaches to Ancient Judaism, 1 (1978), pp. 77-96. I am well aware that I am implicitly taking a position here in the ongoing debate over the reliability of attribution in rabbinic literature, being of the opinion that the rumors of the death of attribution have been greatly exaggerated (cf. Gafni 2001, pp. 218-219, and see above, n. 29). I do not mean to trivialize the problem (most pronounced in the Bavli), but the textual and redactional complexities of attribution within and among rabbinic works are too often misused as evidence of the allegedly arbitrary nature of rabbinic prosopography in general. In any event, these questions fall outside the scope of this paper.

134

HOW SHOULD RABBINIC LITERATURE BE READ?

as rabbinic historiography. 77 Precisely because there is no more conventional narrative repository, we are forced to seek out historical memory, such as it is, in more unwieldy forms and contexts. In their defense, historians may argue that the boundaries between genres are overly rigid in the first place and that even that which goes by the name of historiography is formally just one more kind of literature.78 Yet the genuine distinctions between rabbinic storytelling and anything that may rightfully be called historiography still remain. Various scholars (including Fraenkel) have drawn an analogy between rabbinic stories of the lives of the Sages and midrash: the stories are to real events as midrash is to Scripture.79 What are the implications of this analogy? One could argue that this is reason enough to abandon any attempt to speculate about connections between what the stories tell and the lives of the Sages who feature in them. But one may approach the problem differently. Suppose, for the sake of argument, that the Hebrew Bible and its translations were lost to us and all we knew of them were impressions derived indirectly from rabbinic midrash and Second Temple Jewish literature (strike out all direct quotations to make the exercise more interesting). In such a situation one could argue: I will ignore the influence of the lost text and pretend it never existed, reading each surviving composition in literary isolation. Champions of this path could invoke the need for methodological restraint, but in doing so would in practice introduce a patently false principle into their hermeneuticsȥi.e., the absence of a foundational textȥand by definition their readings would be skewed at least in part. Another apSee M. D. Herr, “Tefisat Hahistoriyah ’etsel ʗaZa"L,” Proceedings of the Sixth World Congress of Jewish Studies, vol. 3 (Jerusalem: 1977), pp. 129142. 78 This point was suggested to me in conversation by Jeffrey Rubenstein. The literary dimensions of historiography and its rhetorical tropes feature most prominently in the studies of Hayden White (see most recently H. White, Figural Realism: Studies in the Mimesis Effect, [Baltimore: 1999]). A. Momigliano, “The Rhetoric of History and the History of Rhetoric: On Hayden White’s Tropes,” Comparative Criticism, 3 (1981), 259-268, criticizes White for failing to give an account of that which distinguishes historical writing from all other sorts of writing. 79 Herr 1977, pp. 136-142; Fraenkel 1991, p. 238; Goshen-Gottstein 2000, p. 263. 77

YONAH FRAENKEL

135

proachȥthe one in fact generally adopted by textual scholars confronting analogous situations in other literary contextsȥwould be to try to restore fragments of the lost text, or at the very least speculate about their content. Such speculation would proceed haltingly, making false starts, but at its best would be subjected constantly to criticism and correction. I recommend the latter path. Historicizing readings are not all equally valid; some are just claptrap. There are occasions, however, when a historicist interpretation is plausible, though not demonstrable with anything approaching mathematical certainty. We need not fear such deliberation, but we are compelled in the circumstances to accustom ourselves to more elastic locutions which reflect in each case the range of conviction and contestability.80 Burdening ourselves with such qualifiers can be tiresome, but for the moment I know of no more intellectually honest solution.

80

Fox 2002, p. 135, speaks of a “minimax model.”

CONTEXT AND GENRE: ELEMENTS OF A LITERARY APPROACH TO THE RABBINIC NARRATIVE Jeffrey L. Rubenstein, New York University That the study of rabbinic literature entails a “literary approach” is almost a tautology. Even the most legally oriented rabbinic traditions have been transmitted to us as literary texts that require readers to interpret their literary forms. The same obviously holds true for aggadic midrashim and narratives, which are more routinely apprehended as “literature.” To a certain extent therefore the interpretation of rabbinic texts will involve the same general problems and difficulties as the interpretation of all other literature. But because rabbinic literature exhibits some distinctive features, its interpretation will involve particular challenges and issues, and even those general problems will be governed by specific considerations. Here I would like to discuss context and genre as two components of a literary approach to the interpretation of sage-narratives, with a constant eye towards the issue of the redaction of the text. I will focus on the Bavli, though much of my discussion applies to the other rabbinic documents once the different processes of their redaction are factored in. Context and genre are considerations that apply to all literature; redaction is less universal a concern, and the specific process of redaction of the Bavli is unique.

137

138

HOW SHOULD RABBINIC LITERATURE BE READ?

CONTEXT All texts exist in multiple contexts: historical, social, political, literary, cultural, institutional, situational, biographical, and so forth. 1 Few scholars, I assume, would argue that a text could be correctly interpreted outside of its general historical and cultural context. We would never interpret a talmudic story in terms of the social or political climate of the Renaissance for obvious reasons. The degree to which other contexts are relevant or necessary for interpretation is more open to debate. At all events, in many cases of ancient literature, including that of the Bavli, most of these contexts cannot be recovered. Bavli stories (and other traditions) appear in texts redacted many years after their original expression, the contexts of which are, for the most part, unknown. The main context—or contexts—to which we have access are the literary contexts. I say “contexts” because the parameters of the literary context can be defined in various ways. Do we mean the immediate literary context, the texts directly preceding and following the story? A slightly more extended literary context, say the sugya in which the story is found? The series of sugyot in the section of Talmud commenting on the proximate Mishnah paragraph? The entire chapter or Tractate of Talmud? The entire Talmud, including all relevant intertexts? A concrete example will illustrate the degree to which the identification of the boundaries of the literary context impacts interpretation. Yonah Fraenkel, as is well known, is a minimalist when it comes to relating the immediate literary context to his interpretation. In the manner of the New Critics, Fraenkel isolates a story and analyzes it on the basis of its structure, content and poetics. A striking, if characteristic, illustration of his technique is his analysis of the following story of Elisha b. Abuya or Aʘer (b. Hag. 15a): Our sages taught: It once happened that Aʘer was riding his horse on the Sabbath going on his way and R. Meir was walking after him and learning Torah from his mouth. When they reached the Sabbath limit, he said, “Meir, return (ʚazor) back since I measured by the footsteps of my horse that the Sabbath boundary is un-

1

Dan Ben-Amos, “‘Context’ in Context” Western Folklore 52 (1993), 209-26.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 139 til here.” He said to him, “Then you too repent (ʚazor).” He said to him, “I have already heard from behind the curtain, Return, rebellious children (Jer. 3:22)— except Aʘer.”

Fraenkel opens his interpretation by commenting: “Riding a horse on the Sabbath constitutes the profanation of the Sabbath in a public domain, and is intended to provoke or ridicule, and therefore it is no wonder at all that there can be no atonement or repentance for this prohibition. This implies that we have here the closed ‘circle’ which in and of itself tells us that Aʘer certainly is not riding the horse on the Sabbath for the first time, and therefore he knows that he cannot repent.”2 Even granting Fraenkel’s interpretive assumptions (i.e. ignoring the larger literary context), this analysis seems extremely problematic to me. There is no internal evidence that riding a horse on the Sabbath is intended “to provoke and ridicule.” Fraenkel argues, in a footnote, that although riding a horse technically does not amount to a serious violation of the Sabbath (it falls in the category of shevut), it is nonetheless perceived as a serious offense: in a story found at b.Yev. 90a the court stoned someone who rode on the Sabbath “because the [exigencies of the] times required it.” Whether this intertext is relevant I will consider below; but even granting its relevance it still falls short of demonstrating the intentions of the rider as provocation or insult. The source does not explain what circumstances of the time required such strong action; the court may have stoned the rider for any number of reasons. Moreover, there is no compelling reason to assume that Meir and Elisha are in public. Though technically in a “public domain” where they approach the Sabbath limit, they may well be traveling in isolated areas or on the outskirts of town where no others are present. Fraenkel has assumed the public nature of the act to make it seem more provocative. A more straightforward explanation is that Aʘer is simply not concerned about violating the Sabbath; he rides because it is easier and more comfortable than walking, just as nonobservant Jews today drive on the Sabbath because it is faster and easier and less tiring. Fraenkel’s interpretation depends in part on his assumption of “closure,” that the end of the story must relate to the beginning, 2

Yonah Fraenkel, Darkhei Ha’aggada Vehamidrash (Masada: Yad Letalmud, 1991), 264.

140

HOW SHOULD RABBINIC LITERATURE BE READ?

sealing the story in a circle of its own. Why should riding on the Sabbath (the beginning) precipitate such a harsh punishment as precluding repentance (the end)? It must have been habitual, hardened behavior, which in turn indicates an attitude of provocation and ridicule (the beginning). Yet here too the beginning and end seem to me to be related in a more straightforward way. From his disclosure that he cannot repent due to the voice “from behind the curtain” he has no possibility of reward in the next world, hence no motivation to observe the Sabbath or other commandments, consequently no reason not to ride. He rides out of convenience, not contempt. Fraenkel notes that the middle of the story, the fact that Aʘer counts the paces measuring the distance to the Sabbath limit, calls for explanation. After discerning a chiastic structure, he suggests that Aʘer counts because he is eager to sin and wants to know precisely when he will attain his goal of exiting the limit: “Leaving the boundary is a complete abandonment, a distancing for which there is no return, and this is the intention of ‘Aʘer’ (the Other).” R. Meir, on the other hand, misunderstands Aʘer’s reason for counting. He interprets Aʘer’s warning to him to go no further as a sign of Aʘer’s concern for the law, hence indicating his potential to repent, in turn prompting the exhortation to do so. Indeed, this was Meir’s true purpose in following Aʘer. Although the story tells us that he was “walking after him and learning [velomed] Torah from his mouth” (and in the printings and in other manuscripts: “in order to learn Torah” [lilmod]3), he was in truth hoping for an opportunity to move Aʘer to repentance. After all, “Does R. Meir not have a superior and more accessible source to learn Torah than from the mouth of one who violates the Sabbath in public?”4 The story, alas, is tragic and dramatic, for the true reality, as Aʘer knows, is that there can be no repentance. Here too there is much that can be criticized on its own terms, including the attribution of purposes and motives to each character that are nowhere disclosed in the text, and especially the privileging, in the case of R. Meir, of this putative motivation over 3

So in manuscripts London 5508, Munich 6, Vatican 134, Goettingen 3. The reading “and learning Torah” is found in Munich 95 and Vatican 171. 4 Ibid., 265.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 141 the explicit reason, that he follows Aʘer to learn Torah. That Aʘer would be so eager to commit the sin of leaving the Sabbath boundary makes little sense if he has been a habitual violator of the Sabbath, as Fraenkel claims. What is so special about this sin now, even granting the symbolic value of going beyond the boundary? Moreover, if Aʘer is such a dedicated sinner, why bother warning Meir not to sin? But my real purpose is to show how much must be “read in” and simply conjectured when the larger literary context is ignored. Fraenkel has no choice but to try to fill the enormous gaps in the narrative through speculation and clever inferences. And if that larger context is taken into account, Fraenkel’s interpretation appears not merely implausible or speculative but flat out wrong. This story actually appears as a scene in the middle of a lengthy biographical narrative of Elisha b. Abuya. The exposition relates that Elisha saw the angel Metatron sitting down and wondered whether there were “two powers in heaven.” Metatron is then punished with sixty lashes of fire, apparently to demonstrate to Elisha that he is in fact not a deity, and subsequently receives permission to burn out the merits of Elisha, presumably revenge for his suffering. At this point Elisha hears the voice from behind the curtain precluding his repentance. He understandably reasons, “Since that man (= I) has been banished from that world (= the next world), I will go and enjoy myself in this world.” Because he has lost his merits and can never repent, he cannot atone for sin and has no chance of entry to the world to come. So why worry about observing the commandments? He therefore propositions a prostitute, violating the Sabbath to prove that he is not the famous sage she believes him to be. Given this context, it is abundantly clear that the reason Aʘer cannot repent has to do with the Metatron incident, not because he habitually rode his horse on the Sabbath. Indeed, the phrase “I have already heard from behind the curtain” refers directly to the Metatron scene. Riding on the Sabbath is not the cause of the voice, as Fraenkel would have it, but the consequence. Second, Aʘer rides not to “provoke and ridicule,” but rather, as he explicitly explains, because he might as well enjoy himself in this world. Riding, as noted above, is more comfortable and “enjoyable” than walking.

142

HOW SHOULD RABBINIC LITERATURE BE READ?

Third, Aʘer does not count the horse’s paces out of his zeal to sin.5 Pleasure or comfort, not sin, is his goal. Sin is simply incidental and irrelevant to him. After consorting with the prostitute the narrative continues with two scenes in which Aʘer asks R. Meir the meaning of biblical verses. In both, R. Meir offers rather straightforward explanations which Aʘer rejects, supplying Meir with Aqiba’s midrashic interpretations. These scenes establish Aʘer as the superior master of Torah, essentially as R. Meir’s teacher. He knows Aqiba’s traditions, which Meir apparently does not, and teaches them to Meir. At this point the story that Fraenkel analyzes appears. This context belies Fraenkel’s claim that R. Meir’s real motivation in following after Elisha is to invite him to repent and not to learn Torah. As the two preceding scenes make clear, Elisha has a great deal of Torah to teach R. Meir. So we should take the text at face value: R. Meir follows Aʘer in order to learn Torah, though he surely would like his master to repent, too. At all events, Fraenkel’s rhetorical question, “Does R. Meir not have a superior and more accessible source to learn Torah than from the mouth of one who violates the Sabbath in public?” is greatly weakened. R. Meir may very well not have a superior source of Torah, and he has much to learn from Aʘer. That Elisha has counted the paces of the horse and knows exactly where the border lies fits well with this reading. Not only is he Meir’s master in Torah, but his knowledge and awareness are so profound that he is able to calculate such a distance while in the midst of a discussion. What then is Fraenkel’s justification for reading the scene as a self-contained story independent of its wider literary context? No doubt Fraenkel would argue that the scene is introduced by the term “tanu rabbamam” (“Our sages taught”), indicating a baraita, and it appears in a tannaitic Hebrew that contrasts with the Aramaic of

5

His reason for counting the paces, though not provided in the narrative—hence a challenging gap to fill—in my opinion is almost the opposite of Fraenkel’s explanation. Before his “fall” he was a meticulous, brilliant sage (see below) who seems to have been conscious always of the distances traveled on the Sabbath so as not to sin. Though in his present condition, after the voice from the curtain, sin is irrelevant, he nevertheless continues to count out of habit.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 143 the preceding and following scenes.6 He would claim that the Bavli redactors have recontextualized this baraita in a narrative of their own making, but the baraita should be read on its own terms as the independent story it originally was. And he presumably would claim that the redactors transmitted their sources faithfully without reworking them, else there is no guarantee that the source as currently found in the Bavli is the same as the (putative) original tradition that Fraenkel means to analyze. To this we can respond, first, that this may well be a pseudobaraita: It does not appear in any tannaitic document. Moreover, it bears an uncanny resemblance to a scene in the Yerushalmi’s version of this story that appears in Aramaic. Such recasting of an Aramaic—presumably amoraic—narrative of the Yerushalmi into a (pseudo-)tannaitic Hebrew version in the Bavli is attested in other cases.7 So there is in fact no solid evidence that this scene ever circulated independently. And the differences from the Yerushalmi version suggest that even if the tradition circulated independently at one point, that tradition was reworked before its inclusion in the respective Talmuds. These arguments, of course, are not absolutely conclusive. It is still possible that the source is an authentic baraita, and this would lend more justification to a detached reading a la Fraenkel. The main justification for reading the story as one scene of a more lengthy narrative is that such is the context provided by the Bavli. The boundaries of the narrative can be determined based on plausible and well-defined criteria, so we have an empirical—albeit limited—literary context.8 The Bavli redactors have either created this narrative or transmitted a narrative created by earlier authorities, perhaps by combining antecedent sources (which may have included the scene [=putative baraita] analyzed by Fraenkel). The important point here is that the question of literary context, at least for the Bavli, cannot be approached without addressing source6

Actually the immediately preceding scenes are predominantly Hebrew, though with a smattering of Aramaic. 7 See Jeffrey L. Rubenstein, Talmudic Stories: Narrative Art, Composition, and Culture (Baltimore: Johns Hopkins University Press, 1999), 261-62. 8 The narrative begins with the first mention of Aʘer and ends with final fate of Meir, the other main character in the narrative. But see my discussion below on the uncertainty here.

144

HOW SHOULD RABBINIC LITERATURE BE READ?

critical issues and without a theory of redaction. The more we see the redactors as authors as opposed to transmitters, the more active a redactional process, and the more weight should be placed on the wider literary context. Even if we grant the scene the status of a baraita, it is crucial to understand that Fraenkel’s reading is not really as independent of context as one might assume at first glance. In place of the wider literary context Fraenkel has supplied his own context, or contexts, based on other rabbinic traditions—what we might call a cultural context. Thus Fraenkel “knows” that according to rabbinic theology and law one isolated instance of riding a horse on the Sabbath would not be punished with a heavenly voice precluding repentance, hence Aʘer must have been a hardened, continual sinner bent on provocation (lehakhis). The source he footnotes in b. Yev. 90b supports the claim that horse-riding can be seriously punished in certain situations (though I have observed above that it need not be analogous.) He “knows” that Torah scholars do not seek out sinners from whom to learn tradition, hence Meir must be following Aʘer not to learn, but to persuade him to repent. And so forth. The main difficulty is that the story is so brief, the narrative so gapped, that a great deal of information must be supplied from this general cultural context. But there is no way to know what data from the vast rabbinic tradition are most relevant, nor how those gaps should be filled. That is why Fraenkel’s reading, or really any such reading, will be subject to criticisms such as those I provide above. There is no need to rehearse all the objections against the New Criticism to make this argument. From this point of view, the advantage of taking into account the wider literary context is that it provides general information relevant to the interpretation of the source, which results in fewer and smaller narrative gaps. Thus we know why Aʘer cannot repent (because of Metatron), and need not attempt to guess, based on our familiarity with other rabbinic traditions, into what category of “those who cannot repent” he falls.9 9

Fraenkel claims that a story provides all the information necessary for its interpretation, hence we should understand that Aʘer cannot repent because of the horse-riding on the Sabbath. But this principle is obviously false. The story does not even tell us that riding is forbidden on the Sabbath, does not explain what a Sabbath limit is, does not identify R. Meir and Aʘer, etc. All this knowledge is assumed. The problem is, we do

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 145 Let me now shift to another reading of this story, that of Yehudah Liebes, to assess these issues from another direction. Liebes takes an approach almost completely opposed to that of Fraenkel. In the manner of the Tosafists, Liebes reads all Bavli traditions about Elisha b. Abuya—really all rabbinic traditions—in light of each other in order to produce a general, synthetic reading. 10 In particular the tradition of Elisha b. Abuya found in b. Qidd. 39b plays a significant role in his interpretation. There Elisha’s turn to sin is attributed to a crisis of faith caused by the problem of theodicy. He saw either (1) a son fall from a ladder and die despite climbing up at the behest of his father, and despite shooing the mother bird away before taking the eggs—two commandments for which the Torah promises long life, or (2), he saw the tongue of a great man dragged along by a swine, apparently following his martyrdom at the hands of the Romans.11 Liebes assumes that all this happened before Elisha even encountered Metatron, so that the sage was incensed before seeing the angel. Moreover, Elisha held Metatron responsible for the martyr’s death (and the death of other martyrs), because Metatron is in fact responsible for recording the sins of Israel. When the text in b. Hag. 15a says that Elisha saw Metatron to whom “was given permission to write the merits of Israel,” it means “the sins of Israel.” Based on a doubtful reading of a single attestation elsewhere in the Bavli, Liebes claims “merits” can be a euphemism for “sins.” So Metatron is partially to blame for causing the suffering of Jewish martyrs in that he recorded their (minor) sins for which they were cruelly punished.12 This is all part of the theme of rivalry between humans and angels culled from other passages, which Liebes claims is at work here too. Is not God nevertheless responsible for the punishment of the martyrs? To be not know what knowledge about Aʘer’s act that precipitated the voice is presupposed. See Rubenstein, Talmudic Stories, 254-55. 10 Yehuda Liebes, Het’o shel Elisha (2nd ed., Jerusalem: Academon, 1990), 27: “In the following analysis we will exploit extensively all the traditions of the sages that pertain to Elisha (without ignoring the problems of different tendencies and later explanations), and especially the Babylonian Talmud, and with the aid of them all we will paint the picture of Elisha.” 11 The printed version reads “Huzpith the Meturgeman.” But all mss read “a great man.” 12 Liebes, ibid., 61.

146

HOW SHOULD RABBINIC LITERATURE BE READ?

sure, but Elisha’s hubris causes him to get angry at the functionary, not the authority (he blames the messenger, as it were). Ironically, Liebes’s broad contextual approach leads to a reading of the scene of Meir and the horse that bears some affinities to Fraenkel’s. He claims that riding a horse on the Sabbath is “the ultimate heresy and rebellion.” This is based on the assertion that riding a horse is an unusual activity for a sage and therefore constitutes a “gesture of rebellion, pride and provocation.” So for both Liebes, the maximalist, and Fraenkel, the minimalist, Elisha is a rebel and arch-sinner. They differ, however, in their assessment of the emotional bond between Meir and Elisha. In contrast to Fraenkel, Liebes sees no bond of warmth between the two sages, hence no tragic aspect in the scene. Indeed, he claims that Aher’s exchange with Meir expresses “contempt for the commandments, contempt for Torah itself and contempt for the sages.” I must confess that this reading seems particularly forced to me (and in this respect I side more with Fraenkel). Rather than contempt, Aʘer’s warning that Meir not violate the boundary seems to express respect and concern. Here Liebes has been influenced by the personality of Elisha that he constructs based on other sources, a pitfall clearly avoided by Fraenkel’s method. The weaknesses of Liebes’s readings are straightforward and need not detain us here.13 What interests me is the theory of redaction and source-criticism that underpins the reading. It seems to me that Liebes must be assuming one of two things. First, he could claim that all the sources are historically true, that all these events really happened; hence they must amount to a coherent, synthetic picture, whatever they seem to say. Second, he could claim that the same tradents or redactors transmitted the different sources and would not have contradicted themselves. His thinking seems to combine both possibilities. Although he denies that all traditions are historically accurate, he nevertheless understands his research as pursuit of the “true sin” of Elisha, and sees the Bavli as the earliest and most authentic interpretation of the Toseftan source of the “four who entered the pardes.” So he is delineating the rabbinic portrait of Elisha, which is anchored in historical fact. Essentially Lie13

See Rubenstein, “Elisha ben Abuyah: Torah and the Sinful Sage,” The Journal of Jewish Thought and Philosophy 7 (1998), 211-25, for detailed criticisms.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 147 bes privileges b. ʗag. 15a-15b as the most reliable “historical kernel” but then harmonizes b. Hag. with other traditions and other traditions with b. ʗag. In this way Liebes’s reading rests on a theory of genre (that the stories are history, not fiction) and redaction (that Bavli redactors transmitted sources faithfully and would not consciously include contradictory sources). His approach therefore entails a maximal literary context, essentially extending through the entire Bavli (if not all of rabbinic literature). I will deal with the issue of genre in more detail below, though it should be noted that, whatever the genre of rabbinic narratives, they are not best approached as history. The Tosafistic approach to context is no more plausible, as there are countless instances where the Bavli contradicts itself in both halakhic and aggadic materials. The redactors either could not or chose not to harmonize all the sources that they included in the Bavli. I would make the case that a more satisfactory context is the sugya or literary unit, which seems to be the basic building-block of Bavli text (granted the problem of how to define the sugya.). The commentaries to various chapters of the Bavli published by Shamma Friedman and his students demonstrate the utility of dividing the text into sugyot as the basis for analysis.14 That approach has proven itself able to explicate a great many issues and resolve numerous problems in a consistent manner. If we apply this standard academic theory to our case, we would include the entire narrative in our purview, granting that it comprises different scenes and possibly incorporates earlier sources (contra Fraenkel). But we would not include the entire Talmud or other narratives found in different sugyot, as nothing indicates the redactors relentlessly harmonized everything (contra Liebes). The important point again is that the choice of literary context is related to theories of redaction, source-criticism and genre. But the issue is in fact far more complicated. Immediately following the narrative appears the following passage.

14

Shamma Friedman, “Pereq ha’isha rabba babavli,” Mehqarim Umeqorot, ed. H. Dimitrovksi (New York: Jewish Theological Seminary, 1977), 277-441; idem, Talmud ‘arukh: Perek ha-sokher et ha-omanin (Jerusalem: Jewish Theological Seminary, 1990-96); Stephen Wald, BT Peshaim III: Critical Edition with Comprehensive Commentary (New York: Jewish Theological Seminary, 2000).

148

HOW SHOULD RABBINIC LITERATURE BE READ? Shmuel came upon Rav Yehudah leaning his hands and standing against the door-bolt and weeping. He said to him, “Keen scholar—why are you weeping?” He replied, “See what is written about the sages, Where is one who could count? Where is one who could weigh? Where is one who could count [all these] towers? (Isa 33:18). ‘Where is one who could count?’—for they would count all the letters in the Torah. ‘Where is one who could weigh?’—for they weighed the light and heavy in the Torah. ‘Where is one who could count [all these] towers?’—for they would teach three hundred laws about a tower that flies in the air. (And R. Asi said: ‘Doeg and Aʘitophel asked four hundred questions about a tower that flies in the air.’) Yet it is taught: Three kings and four commoners have no share in the world to come [… Bilaam, Doeg, Ahitophel, and Gehazi (m. San. 10:2)]. As for us—what will become of us?” He [Shmuel] said to him, “Keen scholar—there was filth in their hearts.” What about Aʘer (aʚer mai; or ‘What is Aʘer’?)?— He said to him, “Greek song never ceased from his house.” They said about Aʘer: “When he would stand up in the schoolhouse, many books of the heretics fell from his lap.”

The discussion between Samuel and Rav Yehudah is thematically related to the story in that it deals with a similar question: how is it that great sages go astray, sin and lose their share in the world to come, as did Doeg and Aʘitofel, who were masters of Torah in the rabbinic imagination. The answer: they had “filth in their hearts.” But now the discussion turns back to Aʘer. The question can be read as “What is Aʘer,” i.e., what is the meaning of the name “The Other.” The answer is that his alterity is due to his immersion in the culture of the other and alien beliefs.15 But the question can be understood in relation to the discussion of Doeg and Ahitofel— why did Aʘer sin (since he apparently did not have filth in his heart)? And here we have a rather different answer, attributing his fall not to the Metatron incident but to Hellenistic influence and heresy.

15

See Alon Goshen-Gottstein, The Sinner and the Amnesiac (Stanford: Stanford University Press, 2000), 78-80.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 149 Now I would argue that we have here some independent, in fact, contradictory traditions about Aʘer that were added because of the associative connection. 16 The redactors juxtaposed other traditions about Aʘer after the lengthy narrative that features him. We know from other passages such as the traditions at b. Qidd. 39b that there was a great deal of speculation as to the cause of Aʘer’s sin reported in the cryptic Toseftan passage. The discussion of Doeg and Ahitofel begins a separate sugya; these characters have not been mentioned previously. Based on this theory of redaction—that redactors juxtapose related, though contradictory traditions based on associative links—I would defend my limitation of the literary context to the boundaries of the narrative.17 But I think the potential weaknesses of this claim are apparent. In some ways I am opening myself to the same criticism that I raised against Fraenkel. I choose a limited literary context rather than a wider literary context, ignoring material that immediately follows the narrative. And while the traditions appear to conflict with the narrative in identifying the cause of Aʘer’s sin, they can be reconciled with enough ingenuity. We could explain that Aʘer mistakenly believed Metatron to be a God because he read too many Greek (pagan?) poems and read too many heretical books. This would move us toward a more synthetic reading such as that of Liebes, though more defensible in this case because we draw upon traditions in the proximate talmudic discussion and not from a different tractate.18 I am therefore positing two levels (or two processes) of redaction. One set of redactors composed independent, self-contained sugyot out of amoraic traditions; a second set juxtaposed other traditions, often from distant sugyot, and either added to the original sugyot or reworked them in complex ways. In defining the context as the entire narrative about Aʘer, but not the appended traditions, I am interpreting the text from the point of view of the authors, those initial redactors. There is some evidence attesting to these

16

See Rubenstein, Talmudic Stories, 94-99. For what it is worth, Alon Goshen-Gottstein independently came to similar conclusions about this passage in The Sinner and the Amnesiac, 78-80. 18 Ironically, in this case Liebes, ibid., 11-12, takes the tradition as an independent and separate source not related to the main narrative. 17

150

HOW SHOULD RABBINIC LITERATURE BE READ?

two types of redaction.19 But obviously the theory of redaction becomes somewhat cumbersome at this point. One has to question whether my contextual boundaries are more justified than others, either narrower or broader. Another way of broaching this question of the proper boundaries of the literary context is to focus on reception rather than composition, on the intended or actual audience of the story. How was the narrative of Elisha studied or recited or “performed” in the talmudic academy? Was it studied as a self-contained narrative, albeit predicated on the Toseftan baraita? Or was it studied in connection with the following traditions? Is the answer different if we focus on the Geonic academy when the text of the Talmud was more or less fixed? I do not know exactly how to answer these questions, but they are potentially relevant to the definition of the context, and hence to interpretation. Before concluding this section let me touch on the Yerushalmi version of the story, found in the corresponding location in the Talmud, y. ʗag. 2:1, 77b-c. Here we have two completely independent sources juxtaposed with one another. The first is a type of midrashic exegesis of t. Hag 2:3-4 which offers three interpretations of Elisha’s sin based on the information supplied by the Tosefta. Thus Elisha “cutting the shoots” is interpreted first as killing young students of Torah, then as drawing them away from Torah to other professions.20 These interpretations portray Elisha as an arch-sinner, a murderer, collaborator and informer. There follows a lengthy narrative, roughly parallel to the Bavli narrative, which portrays Elisha more sympathetically. The main sin recounted here is riding a horse on the Sabbath/Yom Kippur; there is no violence or antagonism to others. Moreover, Elisha’s ultimate fall is attributed in part to factors beyond his control (see below). After Elisha’s death R. Meir (who does not appear in the first source, but is a central character in the second) persuades God to 19

See Jay Rovner, “Rhetorical Stragecy and Dialectical Necessity in the Babylonian Talmud: The Case of Kiddushin 34a-35a,” HUCA 65 (1994), 177-231; idem, “Pseudepigraphic Invention and Diachronic Stratification in the Stammaitic Component of the Bavli; The Case of Sukka 28,” HUCA 68 (1997) 11-62. 20 See Rubenstein, “Elisha ben Abuyah: Torah and the Sinful Sage,” 151-55, for detailed explanations.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 151 rehabilitate or forgive Elisha, and explains to his students that Elisha is ultimately saved “for the merit of his Torah.” One is very hard pressed to believe that R. Meir (or God) would rehabilitate a murderer and collaborator, which would be the case if the two sources are read together. So it seems reasonable to read each source independently, despite the fact that they follow one another in the current text. 21 Limiting the context in this way entails the assumption that the Yerushalmi redactors juxtaposed the traditions based on mere association. They did not hesitate to place two distinct, even contradictory, traditions about Elisha one after the other. I believe that this is a reasonable theory regarding the nature of the redaction of the Yerushalmi. But optimally it is a theory that should be articulated and defended before venturing to interpret the narratives.

GENRE Identification of the genre of any text is crucial to its interpretation. If satire is not recognized as satire, parody as parody, fiction as fiction, then the interpreter cannot even begin to interpret a text correctly. Indeed, it was a basic question about the genre of the sagenarrative that led Fraenkel to his “Kuhnian paradigm shift” in the study of these texts.22 Fraenkel argued, “Every text must be understood according to its genre (sugo), and with respect to most aggadic stories, we must ask whether they should be understood as historical texts or literary texts.” 23 Having argued compellingly for the latter, Fraenkel further specified the genre as “dramatic” (as opposed to epic or lyric).24 To read rabbinic narratives as historical sources upon which to construct the history of the rabbis in late 21

Goshen-Gottstein, The Sinner and the Amnesiac, 81-88, comes to the same conclusion. 22 See Joshua Levinson, “From Parable to Fiction: The Rise of Fiction as a Cultural Category,” Sefer hayovel leYonah Fraenkel (forthcoming). 23 Yonah Fraenkel, “Hermeneutic Problems in the Study of the Aggadic Narrative,” Tarbiz 47 (1978), 142 (Hebrew). 24 In “Hermeneutic Problems,” Fraenkel simply categorizes rabbinic stories as “literature” (sifrut). In Darkhei ha’aggada vehamidrash, 240-244, he argues that dramatic, epic and lyric are subgenres of “literature” and that rabbinic stories are dramatic.

152

HOW SHOULD RABBINIC LITERATURE BE READ?

antiquity, as had generations of scholars since the early days of Wissenschaft, was a mistake. Fraenkel’s classification of rabbinic narratives as “literature” or even “drama” is rather crude. Recent scholarship, especially that of folklorists, has provided more complex and sophisticated discussions of genre with salutory results, as I will discuss below. Yet the interest in genre, I will argue, sometimes comes at the expense of an interest in context. While these interests need not be mutually exclusive, the focus on one has often led to less focus on the other. Before turning to the folklorists, I will discuss some of the work of Henry Fischel, who actually anticipated Fraenkel in questioning the genre of the rabbinic narrative and rejecting much of its historicity. Fischel brought to the study of rabbinics a comprehensive knowledge of the classical tradition. He classified sage-narratives within the general category of “rhetoric” and saw the storytellers as “rhetoricians” who constructed stories with many of the same techniques and methods as did Hellenistic authors. This raised the question of the genre of the material and problematized its usage for historical purposes: “Before any effort is made to utilize materials of rhetorical coloration, whether Greco-Roman or Near Eastern, for historiography or biography, the question of the literary genre of the material must be clarified.”25 In the process of clarifying the question, Fischel considered an impressive array of genres and literary forms: anecdote, chria, parody, diatribic-rhetorical tract, letter, gnomology, doxography, popular bioi or vita, 26 satire, comedy, mime, epigram,27 sententiae, stoic paradoxa,28 oration, symposium, epistle.29 While Fischel does not state matters as strongly as Fraenkel, he clearly rejects using the sources for historical purposes in any straightforward manner. He implies, for example, that the stories of the patient Hillel and intolerant Shammai have about as much historical worth as Hellenistic fables of good and bad ani25

Henry A. Fischel, “Story and History: Observations on GrecoRoman Rhetoric and Pharisaism,” Essays in Greco-Roman and Related Talmudic Literature, ed. Henry A. Fischel (New York: Ktav, 1977), 449. 26 Henry A. Fischel, Rabbinic Literature and Greco-Roman Philosophy (Leiden: Brill, 1973), 3-4. 27 Ibid., 25. 28 Ibid., 69-70. 29 Essays, 444-45.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 153 mals.30 In a footnote he observes that the Yerushalmi’s portrayal of Elisha b. Abuya as an arch-sinner collaborating with the persecutors “has strong legendary features: the anonymity of the event, the ingenious evasions and betrayals, the artificiality of the plot.”31 Yet for all of Fischel’s erudition and his presentation of stunning parallels between rabbinic and Greco-Roman material, there remains something unsatisfying about his conclusions. Take, for example, the analysis of the baraita of the four who enter the pardes. 32 Fischel claims that the tradition was originally an antiEpicurean polemic. Later Tannaim and Amoraim no longer understood the philosophical background or the rhetorical form, so they reinterpreted the tradition in terms of mystical praxis and added prooftexts or testimonia. But originally the “literary form” consisted of “two popular typologies: the first referring to four types of Epicureans and the second to four types of fate destined for Epicureans. The typologies are separated from each other by a parody of Epicurean pseudapocalypses uttered in the form of an admonition (paraenesis).” 33 (This “admonition” is the cryptic warning of R. Aqiba, “when you reach the stones of pure marble do not say ‘water, water.’”) To substantiate this claim Fischel marshals parallels between Epicurean traditions and the depictions and dicta of the four sages mentioned in rabbinic sources, Ben Azzai, Ben Zoma, Aʘer and R. Aqiba. Thus rabbinic traditions portray Ben Azzai as celibate or refusing to marry, an anomaly in rabbinic culture, but paralleled by traditions about Epicurus, who disdained marriage and procreation. The celibate Epicurean therefore became a “stereotype in rhetorical culture” and apparently the model for depictions of Ben Azzai.34 Fischel draws attention to a great deal of “chriic-rhetorical and biographic anti-Epicurean items” in the Aʘer traditions, especially in Aʘer’s reputed denial of divine retribution. That the esteemed R. Aqiba should be the subject of an anti-Epircurean polemic Fischel concedes is surprising. Nevertheless, he suggests that the traditions 30

Essays, 449. Rabbinic Literature, 113. See too p. 10 where he considers the traditions about Aher’s heresy to be of “uncertain historical value.” 32 Rabbinic Literature, 1-24. 33 Ibid., 4. 34 Ibid., 7. 31

154

HOW SHOULD RABBINIC LITERATURE BE READ?

concerning Aqiba’s early years as an am-haarets or ignoramus is paralleled by unflattering epithets such as “vulgaris” and “indoctus” directed at Epicureans. Hence “the classification of Aqiba as an Epicurean must have been on the basis of his ignorance.”35 The pardes, Fischel notes, parallels the “Garden,” a sobriquet for the Epicurean philosophical school. 36 Fischel’s conclusion is that the passage should be understood as follows: The statement “Four entered the pardes” means “Four entered into undue trafficking with the Epicureans.”37 The first typology (the names of the four sages) boils down to: “There are four types of Epicureans: the celibate, the wild speculator, the denier of divine retribution, and the intentional ignoramus.” And the second typology, the statement of the fates of the fours sages/Epicureans, amounts to “death – insanity – nihilism – conversion.” The brief summary above cannot do justice to Fischel’s discussion, which can only be appreciated by the full presentation of the striking classical parallels. The depth of knowledge, breadth of scholarship and complete mastery of both rabbinic and classical sources can be discerned in each and every paragraph. Classical literary genres—chria, typologies, rhetorical stereotypes, pseudapocalypses, admonitions, testimony, parodies—are all brought into impressive comparisons with rabbinic sources. The generic considerations successfully shed light on a great many puzzling and obscure elements of the pardes and related traditions. Personally I am not persuaded by his final interpretation of the passage, which strikes me as being rather reductionistic. But the connections he draws between rabbinic traditions and many of the classical genres and forms are convincing. What is most important for my purposes here is how the generic considerations serve to remove the passage from its literary context. The genres which Fischel discerns emerge from a specific social and institutional setting, the competing philosophical academies of the classical world, in which intellectuals satirized their rivals. This is apparently not the case of talmudic Babylonia, so Fischel focuses on an earlier period, that of the tradition’s genesis, rather than its reception or integration in the Talmud. The literary 35

Ibid., 15. Ibid., 22-23. 37 Ibid., 23. 36

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 155 context within the Bavli (or Tosefta) is therefore irrelevent. Second, Fischel explicitly states that the received version of the tradition has been changed from its original form. Because the reconstructed, original tradition (from the early tannaitic period) differed from the version available to the Amoraim or Talmud redactors, there is little value in considering their interpretations, another reason the literary context can be ignored. In place of the literary context he postulates a type of social context analogous to the academic rivalries of the classical world: contending rabbinic schools of the early tannaitic period, in which sages satirized their opponents. The question then becomes whether this social context is plausible, and whether the reconstruction is accurate. By no means am I rejecting this type of scholarship that seeks to reconstruct the original form of traditions preserved in later sources and to understand them on their own terms, or to derive information about earlier historical periods. But I wish to underline how the emphasis on genre functions in this case to isolate the source from its literary context. In this respect one should note the similarity to Fraenkel’s method, also heavily based on considerations of genre (that the rabbinic story is not history but literature/dramatic fiction), that decontextualizes stories from their literary contexts. Where Fraenkel postulates an amorphous, atemporal rabbinic school (bet midrash) as his social context, Fischel assumes early, competing rabbinic schools. A similar tendency can be seen in some folkloristic approaches to rabbinic stories. In recent years scholars of folklore have made some of the most significant contributions to the understanding of the genre of rabbinic stories and consequently to their interpretation. Eli Yassif, for example, in his magnum opus, The Hebrew Folktale, devotes his longest chapter to the rabbinic period (pp. 70-244). His primary generic classifications include: narrative traditions from the Second Temple period, the biographical legend, the exemplum, the historical legend, tales of magic and demonology, the comic tale, parables and fables, and the story cycle, though his discussion ranges over many other genres and subcategories of these primary genres. This taxonomy includes both genres attested in other cultures (e.g. the exemplum) and genres defined by a sort of induction on the basis of the rabbinic sources themselves (e.g. narrative traditions from the Second Temple period, the story cycle).

156

HOW SHOULD RABBINIC LITERATURE BE READ?

In his discussion of Elisha b. Abuya’s sin, Yassif comments on the strong folkloric motifs in the explanations for his fall. From the Yerushalmi’s account Yassif quotes Elisha’s explanation to R. Aqiba that his father, Avuya, dedicated Elisha to Torah after witnessing the great power of Torah when R. Eliezer and R. Joshua studied together at Elisha’s circumcision celebration. Because Avuyah’s original intention “was not for the sake of heaven,” he did not achieve his goal, that Elisha become a Torah scholar, and Elisha ended up a sinner. Yassif also quotes “another version of his birth” from Ecclesiastes Rabbah 7:8: when Elisha’s mother was pregnant she passed by idolatrous temples and smelled the offerings. She even ate from them “and it burned in her stomach like the venom of a serpent [and infected him].” 38 Yassif then comments: In these legends, as in other birth legends, the biography begins with events in the lives of the hero’s parents, which set the tone for the hero’s destiny and deeds. But while in all other birth legends, the parents perform exceptional deeds, such as withstanding temptation or giving charity, Elisha ben Avuyah’s parents sin against society’s norms (in this case, those of the sages), and this sin is an omen (or cause) of the birth of the anti-saint. Here too, a “learned” legend made brilliant use of structures and motifs of the biographical folk legend in order to create an anti-legend. Its power indeed stems from the traditional associations of the tale—folk motifs familiar to the audience of listeners from other traditions, but its moral significance and psychological effect are based on the breaking of these traditional narrative norms. These legends were intended to set Elisha ben Avuyah apart from other holy men and present him in all his negativity by means of reverse signs of the same narrative-traditional structures particular to the “true” saints.39

This analysis, in my opinion, is very illuminating. The characterization of Elisha as an “anti-saint” or anti-hero is apt, and the inversion of the common signs typically related of the hero’s parents fits nicely. The attribution of his fall to his parents’ actions or 38 39

Yassif, Folktale, 120. Ibid., 120.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 157 sins, though not unprecedented, stands in some tension with rabbinic theology, which tends to emphasize free will and individual responsibility. Understanding this anomaly as a reflex of typical folkloristic motifs supplies a useful explanation. Note, however, how Yassif’s analysis ignores the rest of the Yerushalmi’s narrative. True, his goal is to identify folk motifs in rabbinic narratives, not to provide comprehensive analysis of an entire narrative, and perhaps one should not criticize him for what he is not trying to do. Nevertheless, one can see how the focus on the folkloristic genre leads him away from analyzing these scenes within their literary context within the extended narrative. First, it should be noted that the Yerushalmi’s narrative actually includes a close parallel to the tradition Yassif cites from Ecclesiastes Rabbah about Elisha’s mother. The Yerushalmi’s version is slightly less detailed, noting that she “would pass by houses of idol worship and the aroma seeped into his body like the venom of a snake.” I assume Yassif quotes Ecclesiastes Rabbah because it mentions that she actually ate the sacrifices, which is not only a more substantive manner of ingestion than smell, but indicates the intention to sin. That “she would pass by” the idolatrous temples, as the Yerushalmi puts it, renders her intentions more ambiguous. Still, that the Yerushalmi mentions both the sins or contributing actions of Elisha’s mother and father is significant. Do these stand in tension within the Yerushalmi’s narrative, one tradition blaming the father, the other blaming the mother? Are they complementary? Why mention both? More importantly, Yassif neglects the actual setting of the recounting of Avuya’s sin within the narrative discourse. The narrator does not relate this event, but rather Elisha tells it to Meir as a flashback in the course of their dialogue. Moreover, Elisha uses the account of his father to illustrate the meaning of a verse: “The end of a thing is better than its beginning” (Eccl. 7:8). When Meir offers several parables that present unexceptional, straightforward applications of this verse (e.g. “[By comparing it] to a man who had children in his youth who died, and in his old age who lived.”), Elisha tells him that Aqiba gave a different interpretation, “when it is good from its beginning.” This midrash renders meireishito (ʥʺʩʹʠʸʮ) not as “than its beginning” but “from its beginning” by reading the ʮ against its contextual meaning. Elisha then presents the account of Avuya as a real-life example of the verse: his father’s intention

158

HOW SHOULD RABBINIC LITERATURE BE READ?

was not good “from its beginning” hence Elisha ultimately sinned. Thus Elisha (1) knows the interpretation of Aqiba which Meir has forgotten or never learned, (2) offers a complex midrash rather than a straightforward paraphrase, and (3) presents a true manifestation of the verse rather than hypothetical parables. All this serves to establish that Elisha’s knowledge of Torah is far superior to Meir’s. This is crucial to the narrative dynamic and its meaning, as I have argued at length elsewhere.40 Yassif not only ignores this entire point, but effaces it completely: his quotation of the passage eliminates the words that create the setting: “[Elisha] said… ‘The end of a thing is better than its beginning’ so long as it is good from its beginning. And so it happened to me.” The ellipsis skips over Meir’s interpretation and Elisha mentioning Aqiba’s midrashic interpretation. The function of the Avuya episode and the report of the mother’s inhalation of idolatrous substances within the larger narrative context is also important. The narrative essentially poses the question of whether the merit accrued by Torah study is inviolable or can be negated by sin. To answer the question the narrative presents the figure of a sinning sage, a master of Torah who goes astray, and depicts his fate, which culminates in salvation in the next world. But it is no easy task to construct the figure of a sinning sage: how can sin and knowledge of Torah stably coexist? Either the power of Torah should influence the sage to repent and cease sinning (and perhaps protect him from error in the first place), or the love of sin should predominate to cause the neglect and loss of Torah. The narrative solves the problem by several strategies. It attributes Elisha’s sin to multiple factors including the problem of theodicy (a notoriously difficult issue) and the actions of his father and mother, which were clearly beyond his control: despite his Torah he was predestined to come out bad. The voice from the temple is critical, as it precludes Elisha from repenting, thereby excluding the obvious solution to the problem (which Meir indeed mentions several times). Elisha may wish to repent, but cannot. Hence sin and Torah coexist until death when rewards and punishments materialize, and the answer to the question is given.

40

Rubenstein, “Elisha ben Abuya: Torah and the Sinful Sage,” 155-58.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 159 As Meir says, “They save Elisha-Aʘer for the merit of his Torah.”41 The (inverted) folkloristic motifs thus serve a specific and critical function within the larger narrative, contributing to the plausible construction of an anti-hero, namely a sinning sage. These traditions may have originated independently of the Yerushalmi’s narrative, as Yassif assumes, and they can be appreciated on their own terms. (Here again we encounter the issue of the nature of the redaction of the Yerushalmi, and some sort of theory of redaction should be articulated). But a great deal is lost by neglecting to assess their function within their larger literary context. Focus on the generic context can lead to the analysis of the passage in terms of the typical characteristics of the genre rather than the function in the extended narrative. Let me conclude with one final example. A significant contribution of folklorists has been to offer new perspectives on some of the most unusual talmudic stories such as the exaggerated travel tales of Rabbah bar bar ʗana in b. B. Bat. 73b-74a.42 These tales of humongous birds and eggs, giant fish and frogs, and other such wonders, were an embarrassment to the Geonim and medieval exegetes, who were troubled that the sages would makes such “flagrant lies” by claiming to have witnessed impossibilities. They either allegorized the stories or explained them as dreams and visions. Dan Ben-Amos argued that these stories belong to the genre of “tall tales,” found in many cultures, including the tales of Paul Bunyan in 19th century America. He refers to the account of R. Joshua b. ʗanina and the wise men of Athens, which mentions milei debdiei (“words of lying”; b. Bek. 8b), attesting to the genre. According to Ben-Amos: “The talmudic rabbis themselves understood these narratives for what they were: tall tales. They constituted a distinct genre within the oral tradition of talmudic society,

41

The name “Aʘer” here may be a later addition. See Leib Moscovitz, “Leʘeqer hagufim hazarim ha’agadiim birushalmi,” Tarbiz 66 (1997). 42 Dan Ben-Amos, “Talmudic Tall Tales,” Folklore Today: A Festschrift for Richard M. Dorson, ed. Linda Degh, Henry Glassie, and Felix Oinas (Bloomington, 1976), 25-43; Dina Stein, “Devarim shero’im misham lo ro’im mikan: iyyun bevava batra 73a-75b,” Jerusalem Studies in Hebrew Literature 17 (1999), 9-32; Yassif, Jewish Folktale, 182-90; Galit Hasan-Rokem, Web of Life: Folklore and Midrash in Rabbinic Literature, trans. Batya Stein (Stanford: Stanford University Press, 2000), 159-60.

160

HOW SHOULD RABBINIC LITERATURE BE READ?

marked by distinctive features and their rhetorical significance.”43 He provides an erudite discussion delineating the characteristics of the tales, their function, themes and social context. In my opinion, this understanding of the genre goes a long way to explaining why they are in the Talmud and how they should be understood. Yet here too the generic focus can lead away from important considerations of the wider literary context. Yassif, for example, takes up Ben-Amos’s characterization of these stories as tall tales. He classifies tall tales as a subgenre of “comic tales” and identifies the mechanisms that create the humor, especially the incongruity between realistic and fantastic elements.44 He also notes that rabbinic tall tales include distinctive features beyond those found in international tall tales, including references to “Jewish national mythology,” such as Leviathan, the desert dead and the fertility of the Land of Israel. This folk genre “was recruited to substantiate and fortify the national and religious consciousness of the period.” 45 Yassif then proceeds to discuss a distinct type of tall tale, the “agricultural tall tale,” comprised of accounts of the “amazing fertility of the Land of Israel.” He observes that agricultural societies, like that of Palestine, naturally told tall tales of an agricultural nature. This type of tale: displays the two principal elements of the tall tale in general: the first is the comic-entertainment aspect. Experienced farmers can only laugh upon hearing fantastic tales of giant produce. But there is a secret desire for great success which, in the case of farmers whose agricultural output is their livelihood, would be fruits of extraordinary size. As the tall tale of travelers to distant parts oversteps the boundaries of imagination to vistas beyond the reality familiar to the audience of listeners, so the agricultural tall tales involve a leap of the imagination, in this case inward, into familiar, day-to-day reality of the narrating society. The tales are at once a bit of comic relief in a life filled with hard work, and a faint hope that even some small part of the fantasy could come true.

43

Ben-Amos, “Talmudic Tall Tales,” 28. Yassif, “Hebrew Folktale,” 184. 45 Ibid., 189. 44

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 161 All this I consider very insightful, especially the attention to the distinctive Jewish or rabbinic dimension over and above the standard features common to “international” tall tales. However, in his description of the agricultural tall tales, Yassif makes a telling error. He quotes verbatim a series of such tales from b. Ket. 111b-112a, accounts of a grape vine that yielded hundreds of clusters of grapes each producing a keg of wine, of a three-square mile area filled with fig honey, and of a sixteen square mile flow of milk and honey. He then continues with a semiquotation, semi-paraphrase of the next story as follows: The text goes on to describe the sages’ discovery of a peach as large as a village cooking pot with a five-seah capacity—“One third [of the fruit] they ate, one third they declared free of all, and one third they put before their beasts.” A year later R Eleazar found himself in the same place, and saw it as well (pp. 189-90).

This is not really how the story goes. Here is the full text of the Bavli together with the parallel found in the Yerushalmi, to which I will refer below:

y. Peah 7:3, 20a

b. Ket. 112a

Once R. Abbahu, R. Yose b. ʗanina and R. Shimon b. Laqish passed by a certain vineyard in Doron. The farmer brought them a peach. They and their ass-drivers ate, and there was some left over. They measured its size as equal to a pot of Kefar ʗananiah that holds a seah of lentils

R. ʗelbo, R. Avira and R. Yose bar ʗanina visited a certain place (in the Land of Israel). They brought them a peach as big as a pot of Kefar ʗino. And how big is a pot of Kefar ʗino? Five se’ah. They ate one-third, renounced ownership of onethird and gave one-third to their beasts.

A few days later they passed by. He brought them two or three [peaches] in the palm of his hand. They said, “We want from that same tree.” He said to them, “I brought you from that tree.” They applied the verse,

The following year R. Eleazar visited there and they brought him [a peach]. He took it in one hand and said, [God turns] fruitful land into a salty marsh because of the wickedness of its inhabitants (Ps 107:34).

162

HOW SHOULD RABBINIC LITERATURE BE READ?

[God turns] fruitful land into a salty marsh because of the wickedness of its inhabitants (Ps 107:34). So it is not at all the case that R. Eleazar “saw it as well.” The opposite—he saw that the amazing fertility experienced by his colleagues had disappeared due to the sins of the residents of the land, as explained by the verse he quotes. In place of the mammoth peach of the preceding year he holds the puny peach in one hand. Yassif paraphrases the story, instead of quoting it in its entirety, in order to pass it off as structurally similar to the other tales which he quotes verbatim. But this is a distortion. In fact, to the extent that the story of Elisha can be called an anti-legend, this story should be designated an anti-tall tale. Or perhaps it should be seen as a parody of a tall tale. At least the second half plays off the generic expectations by shifting from the typical tall tale mode of the first half.46 The tale suddenly loses its “fantastic” dimension and returns to realistic and sinful daily life. Yassif has been so seduced by the typical characteristics of the tall tale genre that he fails to recognize its inversion or parody. More significant is the lack of attention to the literary context. Yassif does not consider why these stories appear towards the end of Ketubot in the Talmud’s commentary to m. Ket. 13:11. But this is extremely important, and sheds considerable light on the reason for the anomalous tall tale. The sugya that includes these tales begins with a number of Palestinian traditions that celebrate the advantages of living in the Land of Israel and detail the disadvantages of living in the diaspora. (These traditions appear here because m. Ket. 13:11 gives certain advantages to a spouse who wants to move to the Land of Israel, which implies that living in the Land of Israel is a meritorious act, if not a full-blown mitzvah.) These traditions include, “He who dwells outside of the Land, it is as if he worships idols,” “He who walks four cubits in Israel is assured a place in the world to come,” and “He who is buried in the Land of Israel, it is as if he is buried under the altar.” The agricultural tall tales illustrating the amazing fertility of the Land fit perfectly here in that they 46

For a brilliant analysis of a rabbinic tale that employs the genre of the late antique novel in an inverted way to create precisely the opposite effect, see Joshua Levinson, “The Tragedy of Romance: A Case of Literary Exile,” HTR 89 (1996), 227-44.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 163 express another aspect of the superiority of the Land and benefits of living therein. At the same time, a stratum that runs throughout the sugya attempts to attenuate the exaggerated praise of the Land and to neutralize the traditions that denigrate diaspora life—an obvious interest of Babylonian sages living in the diaspora.47 Here we find the famous midrash of the “three oaths” forbidding mass immigration to the Land of Israel, frequently cited by medieval rabbis, and to this day the basis for the anti-Zionism of the Satmar Hasidim. The anti-tall tale is part of this stratum. It neutralizes the traditions of the Land’s fertility by claiming that it is contingent, ephemeral and elusive. Perhaps the point is even that such fertility no longer is to be found. R. Eleazar already discovered that the fruit had returned to its typical size. At all events, the two halves of the story beautifully exemplify the two warring tendencies within the larger sugya. To fully appreciate the tradition it is not enough to recognize it as a tall tale (or parody of a tall tale), but to view it within this broader context. And yet there is more. One can observe an additional peculiarity in the source. The protagonists shift from R. ʗelbo, R. Avira and R. Yose bar ʗanina in the first half of the story to R. Eleazar in the second. This is something of a non-sequitur. How did R. Eleazar know of the place’s fertility? We would have to say that the three rabbis told him about the peach they experienced, but that datum is hardly self-evident, and should be given in the story. Note how much more smoothly the parallel in the Yerushalmi reads. The same three rabbis return and they want the same sort of giant fruit which they had eaten a little while before. There is no shift in protagonist. The Bavli must accordingly dispense with the explicit request to eat from the “same tree” since R. Eleazar had not been there before. The citation of the verse also makes less sense. It does not explain the astonishing decrease in size, as in the Yerushalmi, since R. Eleazar had not seen the giant peaches previously. Rather it simply explains the small size of the fruit he receives. The Yerushalmi’s version is also an anti-tall tale or a parody of a tall tale, but it is a much better one than that of the Bavli. 47

See Jeffrey L. Rubenstein, “Coping with the Virtues of the Land of Israel: An Analysis of Bavli Ketubot 110b-112a,” Israel-Diapora Relations in the Second Temple and Talmudic Periods, ed. I Gafni (Jerusalem: Shazar Institute, 2005), 155-188 (Hebrew).

164

HOW SHOULD RABBINIC LITERATURE BE READ?

I would suggest, in light of the larger context, that the Bavli redactors replaced the three rabbis in the second half of the story with R. Eleazar. (Here of course I am assuming an active, interventionist mode of redaction, a position that I have argued elsewhere.48) They did so because several of the most pro-Israel and anti-diaspora traditions quoted earlier in the sugya are attributed to R. Eleazar, such as “He who dwells in the Land of Israel lives without sin” and “The dead outside of the Land will not live [again]” (b. Ket. 111a). The change to R. Eleazar is thus very effective and ironic. The most ardent pro-Israel advocate sees with his own eyes that the yield of crops in the Land of Israel is reduced by sin, and acknowledges that truth with his own mouth, thus neutralizing his own claim that those in the land of Israel live without sin. This version of the anti-tall tale reads less neatly than that of the Yerushalmi, but makes a more powerful statement within the overall sugya. It contributes much more effectively to the neutralization of the anti-diasporic traditions. Only by taking into account genre, context and redactional process can this tradition be appreciated in all its complexity. In sum, an adequate literary theory of the rabbinic narrative must address issues of context and genre while articulating a theory of redaction. These considerations are by no means mutually exclusive: a focus on context need not lead to neglect of genre, nor vice versa. In practice, however, concentrating on the one has tended to lead to less emphasis on the other. In both cases a theory of redaction is crucial. Proper definition of the literary context, as well as the decision to ignore that context completely, will depend on the conception of the process of the redaction. Likewise, the redactional process may have caused distortions or modifications of typical genres. While there are clearly other elements that must be 48

Talmudic Stories, 255-67. And see Rubenstein, “Criteria of Stammaitic Intervention in Aggada,” Creation and Contribution: The Contribution of the Bavli Redactors (Stammaim) to the Aggada, ed. J.L. Rubenstein (Tübingen: Mohr-Siebeck, 2005), 417-440.

A LITERARY APPROACH TO THE RABBINIC NARRATIVE 165 assessed, these three are necessary components of a successful literary approach to rabbinic narratives.

EPHRAIM E. URBACH AND THE STUDY OF JUDEO-CHRISTIAN DIALOGUE IN LATE ANTIQUITY ȥ SOME PRELIMINARY OBSERVATIONS Oded Irshai Hebrew University URBACH’S RESEARCH AND THE MODERN SCHOLARLY SETTING: METHOD AND CONTENT1 In one of the most comprehensive and penetrating appraisals of an intellectual biography compiled in recent years, Jacob Sussmann described the life and letters of the late Ephraim Urbach, no doubt one of the most prominent scholars in Jewish Studies in the world in the second half of the 20th century.2 In one sense at least, this was a tribute to a scholar who himself resorted to this very same

1

I would like to thank deeply Mrs. Hannah Urbach, the late Prof. Urbach’s wife, for sharing with me her private recollections and memories concerning several aspects of her husband’s research, interests and encounters with some of his colleagues. 2 Y. Sussmann, “The Scholarly Oeuvre of Professor Ephraim Elimelech Urbach,” Jewish Studies (Supplement 1), 1993, pp. 7-116 (Hebrew), (hereafter, Sussmann). Though Urbach’s scholarly work and career are at the center of Sussmann’s study, it reflects essentially a much wider prism of Rabbinic Studies spanning 60 years, and as such deserves to be translated into English for the benefit of a wider readership. A short account of Urbach’s life and work was written by Moshe D. Herr in the introduction to the English volume of Urbach’s Collected Writings in Jewish Studies, R. Brody & M. D. Herr (eds.), Jerusalem, 1999, pp. xi-xviii.

167

168

HOW SHOULD RABBINIC LITERATURE BE READ?

genre of perceptive evaluation of his own colleagues’ or mentors’ contributions to the field of Jewish Studies.3 A close reading of Sussman’s most detailed and fascinating account of Urbach’s research throughout his entire career, spanning over 55 years, reveals that rather little attention has been paid to this scholar’s contribution to the study of Judeo-Christian polemics in general and during late antiquity in particular.4 It would seem rather difficult to ignore the presence of the Christian factor in Urbach’s overall portrayal of the rabbinic world. Urbach’s own presentation of his collected articles on rabbinic thought (1988), where he allotted a section to his research on rabbinic polemics, in particular polemics with Christianity, speaks for itself. 5 However, none of this rather large body of material has received the attention it deserves in Sussmann’s overall assessment of Urbach’s research.6 Indeed we have come a long way since S. Schechter’s memorable and rather dismaying conclusion in his 1899 lecture at University College London: I have often heard the wish expressed that a history of the rise of Christianity might be written by a Jew who could bring rabbinic learning to bear upon the subject. I do not think that the time is as yet ripe for such an experiment. The best thing to be done is that Christians devote themselves to the study of rabbinic literature. 3

I am referring to Urbach’s assessments of I. Baer, Y. Yadin and J. Bernays, the great late 19th century classical philologist of the Breslau Rabbinical Seminary where Urbach himself spent his formative scholarly years 1930-1938 (Sussman, ibid. pp. 13-14). Urbach’s admiration for the Breslau Seminary’s founding fathers and their philological critical method of talmudic studies was immense, as can be gleaned from his appraisal of their work published in the centenary volume edited by G. Kisch, Das Breslauer Seminar (Tübingen: 1963), pp. 175-185 (see more below). 4 This is not meant to diminish in any way my great indebtedness to Sussmann’s masterly study which is by all accounts well-documented, tightly argued and neatly presented. 5 The World of the Sages: Collected Articles (Jerusalem: 1988, Hebrew). 6 Sussmann did succeed however (ibid. pp. 66-71), in setting Urbach’s work on the rabbis in its rightful position within the modern context of Religionsgeschichtliche studies spanning the world of the Wissenschaft des Judentums of the 19th and 20th centuries and the world of Catholic and Protestant scholars seeking the antecedents of and parallels to early Christianity in the thought-world of the rabbis.

EPHRAIM E. URBACH

169

The history which would be written after such a study would certainly be more scientific and more critical, though perhaps less edifying.7

7

Later published as “Some Rabbinical Parallels to the New Testament,” JQR o.s. 12 (1900), pp. 415-433, at p. 433 (later published with notes in Studies in Judaism, 2nd ser. Philadelphia, 1908, pp. 102-125). Schechter opened his lecture with Ernest Renan’s question: “Has Jewish tradition anything to teach us concerning Jesus?,”( meaning the historical Jesus), to which he replied in the negative, but went on to assert the importance of the rabbinic texts to the understanding of the “times in which Jesus lived”. He proceeded to demonstrate this premise by drawing parallels between terms used in the New Testament discourse and those found in the rabbinic texts (parallels he labelled as “according to the letter”) to which he appended parallels in the field of ideas and miracles (“parallels according to the spirit”). He followed with the rather strange conclusion, quoted above, to which he might have been led by his own observation earlier (see further in this note) of the poor quality of rabbinic studies in his day (ibid. pp. 416-417). Perhaps, armed with this conviction, Schechter himself expressed his disdain for the line of research utilizing the rabbinical sources as mines for the study of the New Testament (Sussmann, p. 69, note 134). As seen from his 1909 survey, Aspects of Rabbinic Theology, in which the terms Christianity or for that matter Jesus are absent, he continued to abide by this sentiment. Schechter also expressed general reservations concerning comprehensive studies of rabbinic thought and lore, because he had extreme methodological qualms about ever coming to grips with the talmudic sea of rabbinic philosophical, theological and historical sayings. He describes these sayings as obiter dicta and hardly resembling any systematic body of thought. Schecter goes on to lament how the state of rabbinic texts lacks the philologically established authority of the Greco-Roman classics. All this led him to the conclusion that no scientific analysis could begin prior to the completion of a full-scale critical philological work on the entire rabbinic corpus (see among others his essay “On the Study of the Talmud,” Westminster Review,1885, later published in his collection Studies in Judaism 3rd ser., Philadelphia 1924, pp. 143-193, at 150156, 158-160). This situation continues to hamper current students of rabbinic texts, and serves as a barrier to progress in the more sublime topics such as history and thought. For a more detailed perspective of the latter issue, see Y. Sussman, “Schechter the Scholar,” Jewish Studies 38 (1998), pp. 213-229, at 219-220, 229 (Hebrew). Concerning Schechter’s wish that “Christians devote themselves to the study of rabbinic literature,” one should note that at the time it was unrealistic given the poor achievements in rabbinic studies of the late 19th century Christian scholars.

170

HOW SHOULD RABBINIC LITERATURE BE READ?

Schechter’s dream was realized to a great extent in the running commentary on the New Testament (4 vols.) heavily based on rabbinic material produced by Hermann Strack and Paul Billerbeck in 1924 and nearly half a century later in a much more comprehensive manner in the studies by W. D. Davies (1958 and 1974) followed by E. D. Sanders’ groundbreaking work (1977, 1987) revolving around the Judaic (historical, social and theological) setting of the New Testament era and its heroes. Given, however, Schechter’s deep disappointment with the poor state of rabbinic studies and students at the time, he was reluctant to contemplate the possibility of including the study of Christianity in rabbinic traditions as an approach to understanding the rabbinic world. Urbach was definitely not the first8 to tackle this challenge, but he was by far the most prominent scholar in the field for half a century before and after the Second World War, and as such, merits special scrutiny. Sussmann who focuses much of his analysis on Urbach’s magnum opus in Hebrew on the rabbinic world of thought, The Sages (first published in 1969),9 opens with the complaint that since its publication no comprehensive critical, or for that matter, valuable review of the work has yet appeared.10 He endeavors to fulfill the long awaited task himself by initially presenting an overview of Urbach’s intentions in this study, and later by discussing its lasting influence on rabbinic studies. Though appearing somewhat late in Urbach’s publications, its inception was very early. 11 Similar to some of the leading scholars in Jewish Studies such as Baer and

See George Foot Moore’s rather critical evaluation, “Christian Writers on Judaism,” HTR 14 (1921), pp. 197-254, particularly pp. 228-254. 8 Three years after Schechter voiced his views , R. Travers Herford (a British Protestant scholar) published his renowned Christianity in Talmud and Midrash, London 1903, a study that was well received by rabbinic scholars such as Willhelm Bacher (see his extensive review in JQR o.s. 17 [1905], pp. 171-183). 9 The Sages: Their Concepts and Beliefs, translated by Israel Abrahams, 2 vols. (Jerusalem: The Magnes Press, 1975). A paperback edition of this book was published by Harvard University Press in 1987. 10 Sussman, p.66. 11 The lack of a study on the rabbinic history of ideas bothered Urbach already in the early forties; it was probably there and then that the notion of this magnum opus was conceived (Mrs. Urbach in a private communication).

EPHRAIM E. URBACH

171

Epstein, Urbach made his way from the study of the Jewish medieval world to that of the late antique rabbis, realizing that later phenomena had their formative roots in the ancient world.12 Within this major shift in Urbach’s intellectual biography one may discern a distinctive element, namely, Urbach’s deep interest in JudeoChristian polemics. Thus, his first major publication early in 1935 (based on his dissertation) was in the field of medieval polemics. His subsequent and most important publications (in 1946/47) on the place of prophecy in the rabbinic mind, were actually meant also to portray the boundaries between the rabbinic and Christian conceptual worlds.13 However, a close reading of Sussmann’s assessment of Urbach’s work, especially in regard to the rabbinic period, reveals an implicit agenda predominantly revolving around questions of philology and methodology rather than on Urbach’s understanding and evaluation of the historical setting in which the rabbinic mind was formed.14 One cannot escape the thought that, in Sussman’s eyes, Urbach’s attempt to draw a detailed, if not comprehensive portrait of the rabbinic world of thought, with all his critical and philological pedantry, was still premature.15 12

Sussmann, ibid. p. 11. “Études sur la literature polémique au moyen-âge,” REJ 100 (1935), pp. 49-77; 11-1 'ʮʲ (ʥʿʹʺ) ʦʩ ʵʩʡʸʺ, 'ʤʠʥʡʰʤ ʤʷʱʴ ʩʺʮ'; ʵʩʡʸʺ, 'ʤʠʥʡʰʥ ʤʫʬʤ' 27-1 'ʮʲ (ʦʿʹʺ) ʧʩ . On the latter two articles see, infra p. 177. 14 Sussmann’s agenda (meticulously presented on pp. 79-87) reflects an ongoing genuine reservation concerning our ability at this current stage of the critical–philological study of rabbinic texts to make clear historical judgments about the historical development of rabbinic thought. 15 See his current and subtle perspective on Urbach’s achievements in The Sages, ibid. pp. 82-87. It might be added that in that sense Urbach the scholar, trained at Breslau under the strictest historical philological regimen (introduced into the Hebrew University Judaic studies curriculum by Jacob N. Epstein), had in a sense deviated from his meticulous training. Several decades ago, Jacob Neusner published an extensive, but unsubstantiated review that was rather disparaging. Neusner criticized Urbach for what he saw as the misuse of talmudic materials for historical knowledge, advocating instead, the critical decipherment, separation and analysis of the historical layers of the different rabbinical sayings, so as to get a handle on the real historical setting and the developments in the rabbinic mindset (see his review, “The Teaching of the Rabbis: Approaches Old and New,” JJS 27 [1976], pp. 23-35). 13

172

HOW SHOULD RABBINIC LITERATURE BE READ?

Hence, the following is a modest attempt to fill in the gap and, notwithstanding the above impressions, to draw fresh attention to Urbach’s contributions in the field of polemics by reviewing his predisposition towards the subject and his methodological assumptions when treating the rabbis’ attitude toward Christianity. However, before we immerse ourselves more deeply into the topic at hand a quick sketch of Urbach’s intellectual pedigree is needed. Urbach’s study of the rabbinic world was influenced by his strong rabbinical Eastern European background fused with, but not in anyway compromised by the new secular and critical atmosphere of Jewish Studies advocated by the rabbinical School in Breslau (Germany). Its leading talmudic and midrashic scholars, Israel Levi and Isaak Heinemann, held great sway over Urbach’s intellectual and spiritual life. Urbach, the confident Beit Midrash scholar, soon adopted their historical methodology coupled with their deep philological acumen and expertise in evaluating rabbinic sources in the best traditions of classical philology. He created a brand of scholarship that assessed the rabbis for what he believed they really were within their own internal world and the uniqueness of the Jewish historical context. In his understanding, one of the most vital forces that shaped the rabbinic mind was the world of Halakhah which was structured by a self-contained a priori rigid set of rules. Conceived thusly, this composite world afforded very little room for the play and influence of external forces. One cannot claim for a moment that Urbach dismissed any sort of contact or cultural dialogue between the rabbis and the surrounding cultures, particularly Greco-Roman society, but he regarded their impact on the rabbinic mind as marginal. As will become apparent later on, in his eyes those cultures neither shaped the religious world and beliefs of the rabbis, nor did they inspire any form or sort of acculturation by the rabbinic world to the surrounding cultures. One may infer that Urbach did not exceed the guidelines laid down by his mentor Israel Levi and realized by his close friend and colleague, Saul Lieberman, author of the now classic works on the presence of Greek language and culture in Jewish Palestine.16 16 See one of Urbach’s last public lectures on past attempts to study rabbinic thought in his opening address to the First Conference on Rabbinic Thought (Mahshevet ʙazal), University of Haifa, December 1987, edited by Marc Hirshman and Tzvi Gruner and published by Haifa U. Press, 1990, pp. 13-19, especially pp. 15-16.

EPHRAIM E. URBACH

173

Therefore, in order to appreciate Urbach’s contributions to the understanding of the rabbinic world in the context of contemporary cultures in general and in the field of early Judeo-Christian polemics in particular, a quick perusal of Urbach’s scholarly path into this world is necessary. In tracing this aspect of Urbach’s intellectual biography, I will situate relevant features of its evolution within the wider framework of his studies. A glance through a list of his publications quickly makes it apparent that when he began his studies on the world of the Second Temple and of the rabbinic period he was no newcomer to the field of polemics. As indicated above, already in his initial years of research back in 1935 he showed keen interest in the exploration of Jewish - Christian relations, albeit during the Middle Ages.17 During the following years Urbach immersed himself in the study of the history of medieval Jewish literary world, where he traced anti-Christian sentiments among the Ashkenazi poets of the 12th and 13th centuries and described in detail the parallel intellectual enterprises of the Tosaphists’ exegetical work and that of the medieval Christian glossators.18 It was in those years, that, dedicated to medieval Judaism, he noticed more and more the great void in critical research into the formative role played by rabbinic thought of the tannaitic and talmudic periods. He decided to take it upon himself to fill this void. Indeed, this was a formidable undertaking and it resulted years later, apart from a substantial number of groundbreaking articles, in two monumental studies: The Sages, Their Concepts and Beliefs and later The Halakhah, Its Sources and Development.19 It was also then in the late thirties that Urbach began to formulate his guidelines on the ways to approach and understand the medieval and early rabbinic worlds, though it must be stressed that much of the following became apparent in his later work. To Urbach, the leading methodological principle was the rejection of external, especially Christian influence on Jewish religious or cultural thought and custom, although more often than 17

Sussmann, pp. 14, 117 (no.2). The dissertation was written in Rome under the supervision of M.D. Cassutto. 18 This has been recently judged as one of the book’s greatest merits. See the appraisal offered of Urbach’s study by I. Ta-Shma, “Prof. E.E. Urbach’s ‘Baalei Ha-Tosafot’—50 Years after Its Appearance,” Jewish Studies 41 (2002), p. 38 (Hebrew). 19 The Sages, see (supra, note 9); The Halakhah, ET, Jerusalem 1986.

174

HOW SHOULD RABBINIC LITERATURE BE READ?

not he readily entertained the possibility of an opposite influence.20 As a result of his adamant attitude concerning external influence,21 he was ultimately led to dismiss the mutual influence of parallel intellectual and cultural phenomena and to come up with the following claim: “similar conditions bring about similar results.”22 From our current perspective it would seem that absent in Urbach’s thought (as in that of some of his contemporaries), was the notion of a dialogue or encounter in the modern sense that leads to the creation of a common cultural ground. Hence, in Ur20

On medieval Jewish Piyyut see his prolegomena to R. Abraham son of R. Azriel’s treatise Arugat Ha’bosem, vol. IV, Jerusalem 1963, pp. 47-48, 183-185; On the parallels between the Tosaphists and the Christian Glossators, see his The Tosaphists: Their History, Writings and Methods, (4th ed.), vol. 2, Jerusalem 1980, pp. 744-752 (Hebrew). On Urbach’s deep reservations concerning a possible Christian influence on the Jewish medieval mind as suggested by his friend and colleague, the historian Yitzhak Baer, see his introduction to the first volume of Arugat Ha’bosem, Jerusalem 1939, pp. 12-13 note 1 (omitted from later copies of the same edition). In this note Urbach strongly advocates the view that rather than seeking external parallels and giving them great significance, one ought to at least initially judge the cultural phenomena as fruits of internal developments rooted in the past Jewish world of thought. In the case of the Tosaphists for instance, Urbach claimed one should not forget their deeply rooted foundations in the talmudic world. For more details consult Sussmann, ibid. pp. 29-30, 60-61. 21 See the previous note. In his rejection of Christianity as a formative and influential force to be reckoned with when assessing developments within Judaism, Urbach was not alone. It seems that Gershom Scholem too had the same views as can be gleaned from his magisterial portrayal of Jewish mysticism, Major Trends in Jewish Mysticism, NY 1941. On Scholem’s stance see recently, A. Raz – Krakotzkin, “‘Without Regard for External Considerations’ – The Question of Christianity in Scholem and Baer’s Writings,” Jewish Studies 38 (1998), pp. 73-96 (Hebrew). 22 Urbach, The Tosaphists, 1st ed., Jerusalem 1955, 27; see further in Sussmann, pp. 42-43. In Urbach’s understanding the Tosaphists’ world was first and foremost a product of the talmudic world. They were its continuators, and thus the fact that they had contemporary parallels was secondary. It seems that Urbach waved this banner even more enthusiastically when he explored the rabbinic world. See for instance his rejection of Baer’s suggestion that inheritance laws during the Second Temple and the Tannaitic periods contained concepts adopted from the Hellenistic world (see more below).

EPHRAIM E. URBACH

175

bach’s world the approach by which the rabbis confronted external challenges was that of response or reaction but less so in the mode of adaptation or acculturation, in contrast to the current belief that external influences regularly mold cultural self-identity. 23 Judaism was and remained a self-contained entity and its transmutations throughout history were first and foremost to be attributed to internal developments.24 The latter principle became deeply ingrained 23 This is not meant at all to deny the fact that the Greek language and traces of Greco-Roman lore and literature, as well as Greek concepts of logic and thinking (for instance, exegetical methods used in the Midrash and the like), did enter the rabbinic world. What is meant by us here concerns a deep adaptation of social, religious and legal principles. A possible glimpse into the inner world of scholars from Urbach’s generation regarding their qualms and the issues they grappled with when they conducted their research on the Judeo-Christian interaction has been offered by S. Sandmel, “The Jewish Scholar and Christianity,” The SeventyFifth Anniversary Volume of the Jewish Quarterly Review, Abraham A. Neuman and S. Zeitlin (eds.), Philadelphia 1967, pp. 473-481. A more complex reading of the cultural matrix in which late antique Judaism operated has been recently presented in a set of papers edited by D. Biale, Cultures of the Jews: A New History, New York, 2002, pp. 135-265. 24 See too Raz-Krakotzkin, ibid. A typical example of this approach in Urbach’s work is the following case in which he discusses in detail the notion of martyrdom in rabbinic thought. Urbach sets out to explore this ideal as a corollary to the concept of asceticism. Urbach was determined to refute Y. Baer’s theory that asceticism was a central trait in the world of the Second Temple pietists (ʗasidim ) and Pharisees and subsequently left its deep mark on the later rabbinic world. Urbach rejected this view by claiming that the characteristics of the latter asceticism were products of the special historical circumstances of the rabbinic age and were not to be linked to the formative age of the Second Temple era. Martyrdom was part of that scenery and was thus drawn by Urbach into the equation. However, in the process of his examination he tended to ignore its wider cultural connotations. Thus, though well aware of the comparisons drawn in contemporary scholarship between Jewish and Christian or pagan martyrology, he nonetheless refrained from commenting on the possible parallel historical setting. While acknowledging the possibility that Jewish martyrdom had an influence on Christianity, he proceeded to demonstrate the uniqueness of the historical circumstances (the destruction of the Temple and the failure of the Bar Kokhba Revolt) that enhanced the ideal of martyrdom among the Jews of the period, as well as the significance attributed to it by the rabbis, thus, effectively denying martyrdom its wider

176

HOW SHOULD RABBINIC LITERATURE BE READ?

into Urbach’s Weltanschauung. One gets the impression, (though I am quite hesitant to express it) that this outlook of Urbach tended to limit his use of comparative insights to be gained from the study of parallels, be they Greco-Roman or Christian religious and cultural phenomena.25 Urbach, so it seems, when it came to the study of the early rabbinic world was less prepared to use the same yardstick he used in his explorations of rabbinic intellectual and cultural life during the Middle Ages.

URBACH’S EXCURSIONS IN THE WORLD OF EARLY RABBINIC POLEMICS It was only after his immigration to Eretz-Israel in 1939 and a further partial lull in his scholarly work due to his service in the British Army during the Second World War, that Urbach began to realize his dream. 26 During the years 1946-7 he published his first articultural import; see, “Ascesis and Suffering in Talmudic and Midrashic Sources,” Yitzhak F. Baer Jubilee Volume, S.W. Baron et al. (eds.), Jerusalem, 1960, pp. 48-68, at 56-61 esp. note 44 (Hebrew). Urbach seems to have continued to ignore the plausible premise that later 2nd-3rd century accounts of Jewish martyrdom might have at least in a way mirrored some of the characteristics or ideological notions of the nascent Christian martyrology. He treated this extreme form of Jewish devotion as a religious phenomenon to be studied entirely within the inner constraints of the Jewish world of thought. See his later exposition on the subject in his: “Kidush Ha’Shem” (a concise historical survey, published posthumously) in: M.D. Herr & J. Fraenkel (eds.), Ephraim E. Urbach: Studies in Judaica, Vol. I, Jerusalem 1998, pp. 510-519 (Hebrew). However, his reluctance on this point could be attributed to the reservations he held concerning Y. Baer’s opinions on the matter. See more on that below. 25 Without being judgmental, I would suggest that this approach manifests itself in some of Urbach’s seminal works. Two examples will suffice. In his studies on the Sages and their authority, “Class–Status and Leadership in the World of the Palestinian Sages,” Proceedings of the Israel Academy of Sciences and Humanities 2, 4 (1966), pp. 1-37 and “The Talmudic Sage: Character and Authority,” Journal of World History 11 (1968), pp. 116-147 (both reprinted in the collective studies in English [supra note 2] pp. 436472; 404-435 [respectively]), there is no attempt (apart from a fleeting remark in the study on the Talmudic Sage pp. 135-136) to compare the rabbinic elite with any of its contemporary elites in the Greco-Roman or Christian worlds. 26 Although see supra note 11.

EPHRAIM E. URBACH

177

cles ?ʤʠʥʡʰʤ ʤʷʱʴ ʩʺʮ and ʤʠʥʡʰʥ ʤʫʬʤ which, when combined, served, according to Sussmann, as Urbach’s manifesto on the way one should approach the rabbinic world: as a society characterized by a set of rational principles, not dominated by the guidance of revelation and prophecy typical of the biblical world.27 While only implicit in Sussmann’s review, however, the corollary to Urbach’s main argument in these initial studies (especially in the former) was to draw attention to the dividing line between the Rabbis and the revelatory world of Christianity.28 By laying down this fundamental premise, Urbach decisively maintained that the two worlds were miles apart, but at the same time he re-established (following a growing line of scholars) the relevance, albeit relative, of early Christian thought to the understanding of the rabbinic world. 29 11-1 'ʮʲ (ʥʿʹʺ) ʦʩ ʵʩʡʸʺ, 'ʤʠʥʡʰʤ ʤʷʱʴ ʩʺʮ'; ʧʩ ʵʩʡʸʺ, 'ʤʠʥʡʰʥ ʤʫʬʤ' 27-1 'ʮʲ (ʦʿʹʺ). cf. Sussmann, ibid. pp. 63-64. 28 ‘ʩʺʮ’ pp. 9-11 29 Urbach further enhanced the position advocated earlier (with some hesitancy) by Marmorstein that rabbinic literature carries traces of antiChristian polemics, contra the sweeping view promoted by M. Freiman, that the rabbis kept aloof and were entirely indifferent towards Christianity (compare Schechter’s views, supra pp. 170-172). On this protracted scholarly debate see W. Horbury’s dissertation, A Critical Examination of the Toledoth Jeshu (Cambridge 1971), pp. 311-353. Having said that, however, I would note that Urbach’s stance was also restricted. The role he was prepared to grant Christianity was limited, as seen in the following passage (taken from The Sages, as are the following examples): The Jews saw no need to engage in a serious controversy with the New Testament, and were content to underscore the hesitancy and inconsistency of those who professed it (the example used by Urbach to demonstrate his point is the famous anecdote about the encounter between the philosopher [a Christian] and Imma Shalom and R. Gamaliel her brother, reported in Babylonian Talmud, Shabbat, 116a-116b). Hence, it is not surprising that we find no signs of struggle against Paulinian ideas, and certainly not of the influence of his views before the end of the Tannaitic era and the beginning of the Amoraic period… (The Sages, p. 303). Thus, in the eyes of Urbach, Christianity in its formative period, with or without its Pauline garb did not pose a major challenge to the JewishRabbinic world apart from a few restricted cases, like the significance and effect of miracles and magic (The Sages, pp. 115-117) or (though to a lesser 27

178

HOW SHOULD RABBINIC LITERATURE BE READ?

This outlook, as we shall try to demonstrate, determined to a great extent his later work on the subject. In the following fourteen years, while bringing to completion his monumental work on the Tosaphists, Urbach devoted more and more time to the conceptual and spiritual world of the mishnaic and talmudic sages. During this period, in a growing number of studies, he focused his attention more and more on the rabbinic polemic against the New Testament and Patristic thought.30 In fact, it is quite astounding to notice that during the years between 1945– 1963 the lion’s share of Urbach’s publications in the two leading scholarly periodicals in Israel, Zion and Tarbiz, touched in some way or another on the rabbinic refutation of or attitude towards Christian views, thus, no doubt amplifying the centrality of this topic to rabbinic studies. 31 It might be plausible to add that Urbach was carrying here yet another banner, namely, the extension of the quite rigid limits of contemporary research on the relationship of rabbinic society with the surrounding cultures. By then, towards the end of the 50’s, and under the spell of Saul Lieberman’s classic extent) the doctrine of Adam and the primordial sin as interpreted by Paul in Romans 5:12-21 (The Sages, p. 426 ff.). Even in the case of the Election of Israel and the issue of conversion, Urbach did not attribute any prominence to the Judeo–Christian conflict; the pagan world too had its fair share in the debate (The Sages, pp. 524-526; 541-550). The atmosphere was to change during the later 3rd century and more so following the Christianization of the empire, (ibid.). Thus, what emerges from this handful of examples is that Urbach, faithful to his critical, philological historical method, judged every saying or anecdote on its own merit aspiring to contextualize it as much as he could and in all events refraining from simplistic generalizations. 30 ; 27-1 'ʮʲ (ʥ”ʨʹʺ) ʦʨ ʯʥʩʶ ,'ʬʿʦʧ ʬʹ ʤʷʣʶʤ ʺʸʥʺʡ ʺʥʩʺʸʡʧʥ ʺʥʩʺʣ ʺʥʮʢʮ' ,ʤʰʹ ʭʩʲʡʹ ʥʬ ʺʥʠʬʮʡ ʸʲʡ ʷʧʶʩʬ ʬʡʥʩʤ ʸʴʱ ,'ʬʿʦʧ ʺʸʥʺʡ ʭʩʸʥʱʩʥ ʭʩʦʷʱʠ' 68-48 'ʮʲ ,ʠʿʫʹʺ ,ʭʩʬʹʥʸʩ. “The Rabbinical Laws of Idolatry in the Second and Third Centuries in the Light of Archaeological and Historical Facts,” originally only partially published in IEJ 9 (1959), pp. 149-165, 229-245, and recently fully translated from the Hebrew and published in Urbach’s Collected Writings (supra, note 2), pp. 151-193; “The Laws Regarding Slavery as a Source for Social History of the Period of the Second Temple, Mishnah and Talmud,” Annual of Jewish Studies 1 (1964), pp. 1-94. 31 The only other scholar to make major contributions to the field in the same periodicals was Yitzhak Baer, Urbach’s main target of criticism as well as a close and admired colleague, about whom see more infra.

EPHRAIM E. URBACH

179

studies, the predominant cultural world with which it was assumed the rabbis had most contacts, was that of pagan Greece and Rome.32 In these studies Urbach introduced another religious and cultural force to be reckoned with, that of Christianity, only to reject its impact (as will be later demonstrated). And though he dedicated three full studies to the rabbinic–patristic controversy, all of which were to attain an almost canonical status, it would seem that he did not find it rewarding in any sense to write yet another study on the topic of Jesus in rabbinic sources as the one produced, for instance, in those very same years by Jacob Lauterbach.33 Thereaf32

It should be stressed, however, that our contention is not in any way intended to blur the fact that Lieberman himself made some very important contributions in retrieving lost rabbinic (or other Jewish) traditions from Christian and other sources as well as to the study of the rabbinic world within the Judeo–Christian context. See his famous study “Palestine During the Third and Fourth Centuries” JQR, 36 (1946), pp. 329-370; ibid. 37 (1946), pp. 31-54, or “The Martyrs of Caesarea,” Annuaire de ĻInstitute de Philologie et đHistoire Orientales et Slaves 7 (1939-1944), pp. 395-445. For Lieberman’s refutation of Y. Baer’s excessive interpretations in the service of proving the presence of dominant features like asceticism or martyrdom that determined Jewish life and culture from the Hasmonean period onwards see, ʯʥʸʠʡ ʭʥʬʹʬ ʬʡʥʩʤ ʸʴʱ, ‘(ʭʩʠʮʥʸʤ ʯʥʨʬʹ ʩʮʩʡ) ʬʠʸʹʩ ʺʣ ʺʥʴʩʣʸ’ ʤʮʸ-ʢʩʸ 'ʮʲ ,ʤ''ʹʺ ʭʩʬʹʥʸʩ. Notwithstanding, Lieberman’s main thrust was to interpret the rabbinic sources and to seek their contextualization within the realia of the Greco-Roman world between the first and fourth centuries. See for instance the following treatment of a limited topic in his “Roman Legal Institutions in Early Rabbinics and in the Acta Martyrorum,” JQR 35 (1944), pp. 1-57, and on a much wider scale in his formidable studies, Greek in Jewish Palestine, New York, 1942; Hellenism in Jewish Palestine, New York, 1950. The fine traits of Lieberman’s work and his role in rabbinic studies as well as his unique achievements in exposing the contemporary realia in which the rabbis operated have been meticulously described in Eliezer Shimshon Rosenthal’s appraisal, ‘ʤʸʥʮʤ’ published in PAAJR 31 (1963). In this context it is interesting to note that what we have designated as the “Lieberman spirit” still holds sway among current scholars. See for instance the passing remark of Sussmann, Urbach, p. 79, as well as in the recent broadly ranged project launched by Peter Schaefer on the Talmud Yerushalmi and Greco-Roman culture (so far comprising three volumes of varied studies), in which Christianity or Christian culture have been accorded a rather minor role. 33 “Apologetics” EJ Vol. 3, pp. 188-201; “The Homiletical Interpretations of the Sages and the Expositions of Origen on Canticles and the

180

HOW SHOULD RABBINIC LITERATURE BE READ?

ter (beginning in 1962) he spent less and less time on specific issues in rabbinic thought, and became absorbed in the compilation of his magnum opus The Sages, where, as far as the topic at hand is concerned, he incorporatedȦthough not verbatim and with some minor modifications and additionsȦhis previous work.34 After that publication, Urbach’s work on the rabbinicChristian controversy was mainly aimed at the rejection of allegations concerning the impact Christianity might have had on the rabbinic world.35 In fact, his aforementioned outright rejection or

Jewish-Christian Disputation” in: Urbach’s Collected Writings, (supra, note 2), pp. 318-346; ʧʥʫʩʥʤ ʸʥʠʬ ʭʲʬʡ ʺʹʸʴ ʬʲʥ ʭʬʥʲʤ ʺʥʮʥʠ ʩʠʩʡʰ ʬʲ ʬʿʦʧ ʺʥʹʸʣ' ʩʣʥʤʩʤ ʧʥʫʩʥʤʥ ʤʥʥʰʩʰ ʩʹʰʠ ʺʡʥʹʺ'; 289-272 'ʮʲ (ʦʿʨʹʺ) ʤʫ ʵʩʡʸʺ 'ʩʸʶʥʰ ʩʣʥʤʩʤ 152-148 ,ʩ''ʹʺ ʭʩʬʹʥʸʩ [ʫ ʵʩʡʸʺ=] ʯʩʩʨʹʴʠ ʰʿʩ ʣʥʡʫʬ ʬʡʥʩʤ ʸʴʱ ,'ʩʸʶʥʰ. While all three studies have been cited time and again in the scholarly work in the field, the foundations and guidelines articulated in the article on Origen and the Song of Songs were later subjected to new scrutiny by R. Kimelman, “Rabbi Yohanan and Origen on the Song of Songs: A Third Century Jewish-Christian Disputation,” HTR 73 (1980), pp. 567-595. For an example of a study focusing on Jesus in the rabbinic sources, see Jacob Z. Lauterbach, “Jesus in the Talmud,” Rabbinical Essays, Cincinnati 1951, pp. 473-570. 34 It seems quite reasonable to assert that one of Urbach’s aims in this magnum opus (which bears some affinity in form with Solomon Schechter’s, Aspects of Rabbinic Theology, 1909 ) was to revise Louis Ginzberg’s manner of presentation of the rabbinic world of thought in his now classical compilation, The Legends of the Jews. Cf. Urbach’s critique (published posthumously) in: M.D. Herr and J. Fraenkel (supra, note 24) Vol. II, pp. 872877 (Hebrew). 35 (ʪʸʥʲ) ʸʥʥʠʸʴ 'ʩ, ʤʩʺʥʸʥʣʬ ʭʩʬʹʥʸʩ, 'ʤʬʲʮ ʬʹ ʭʩʬʹʥʸʩʥ ʤʨʮ ʬʹ ʭʩʬʹʥʸʩ' 171-156 'ʮʲ ,ʨ”ʫʹʺ ʭʩʬʹʥʸʩ. In that paper Urbach dismissed Avigdor Aptowitzer’s longstanding assertion (voiced back in 1931) that the rabbis refrained from discussing the notion of the Heavenly Jerusalem because it was adopted by Patristic circles. Yet on another occasion Urbach did not simply contest a possible assertion as in the former case but dismissed outright a long propounded notion based on what seemed to be an explicit rabbinic saying attributing a change in a custom sanctioned by the rabbis to the challenge posed by the Minim (lit. heretics, but in many cases Judeo-Christians or Christians). See his ʤʣʥʡʲʡ ʺʥʸʡʣʤ ʺʸʹʲ ʬʹ ʭʣʮʲʮ' 'ʮʲ ʭʹʥ 145-127 'ʮʲ ,ʥʿʮʹʺ ʭʩʬʹʥʸʩ ,ʺʥʸʥʣʤ ʩʠʸʡ ʺʥʸʡʣʤ ʺʸʹʲ ,'ʤʬʩʴʺʡʥ 138 ,133-132. It must be stated that in the latter case, Urbach, because of his careful evaluation of the sources was not prepared to accept them at

EPHRAIM E. URBACH

181

in some cases extreme skepticism of any sort of external influence on rabbinic thought or custom, made him approach more cautiously the task of tracing the marks of anti-Christian sentiments in rabbinic sources. 36 Hence, apart from dismissing the widely accepted view that Minim in the rabbinic sources (especially in those of the early and late tannaitic period) designated in general JudeoChristians or Christians, he also displayed hesitancy when it came to pointing out early rabbinic (prior to the mid-3rd century) polemics with Pauline views.37 Nonetheless, certain aspects of religious face value, doubting the accepted identification of those Minim as Christians; see more, infra note 37. 36 This trait in his research was brought into special prominence in his critical evaluation of the great historian Yitzhak Baer’s work, on the occasion of his 90th birthday: ʩʸʡʣ .'ʸʲʡ ʷʧʶʩ ʩʰʩʲʡ ʤʰʹʮʤ ʺʴʥʷʺʥ ʩʰʹʤ ʺʩʡʤ ʩʮʩ' 82-59 'ʮʲ ,ʭʿʹʺ ʭʩʬʹʥʸʩ ,(4) ʥ ʭʩʲʣʮʬ ʺʩʮʥʠʬʤ ʤʩʮʣʷʠʤ. Only in a few cases was Urbach prepared to concede the possibility that a custom was changed because of an external challenge. Such was the case concerning the “direction of prayer” (either toward Jerusalem or toward the East or West), where the rabbinic change of opinion was seen as a reaction to the more recent Christian custom (in this case there seems to have been no other plausible reason), see, The Sages, pp. 59-63. However, ibid. p. 59, Urbach voiced a more sweeping note of endorsement: “The forms of cultic ritual and worship are not only decided by inner reasons of faith; frequently they are influenced by the rites of other religions either through adoption accompanied by the advancement of new areas, or by negation and rejection.” In any case Urbach here seems to point openly to the possibility of a halakhic dialectical encounter with different religious and cultural forces. 37 Concerning the identification of the Minim one may quote his explicit words: “The texts in which minim are mentioned belong to the second century, mostly after 135, or the third century. In these the term is used to designate various groups. Sometimes they represent dualists, sometimes Gnostics of various sects, sometimes Gentile Christians, sometimes Judeo-Christians, and often pagans” (my emphasis). See, “Self– Isolation or Self-Affirmation in Judaism in the First Three Centuries: Theory and Practice,” Jewish and Christian Self-Definition II, Aspects of Judaism in the Graeco-Roman Period, Philadelphia 1981, pp. 290-291 (see his examples ibid., where he points out some misinterpretations of the term). As recent scholarship has demonstrated, Urbach’s cautious method might have been justified. A similar methodological caution yielding similar results has been recently demonstrated by Christine E. Hayes, “Displaced Self – Perceptions: The Deployment of Minim and Romans in b. Sanhed-

182

HOW SHOULD RABBINIC LITERATURE BE READ?

life essential to both religions did trigger, as Urbach saw it, a rather considerable shift in perspective. Thus, when dealing with the question of the presence of prophetic inspiration and revelation during the Second Temple period, Urbach presents the rabbinic view that claimed the cessation of prophecy during that period by arguing that their stance on the matter was determined by the Christian claim concerning Christ’s prophetic powers.38 This seems to have been, however, one of the rare occasions in which he considered a rabbinic doctrine to have been mediated through a Christian idea. Years later when dealing with a very similar topic, the presence of prophecy among the nations ( i.e. the biblical Balaam), Urbach, well aware of the possible parallels between Balaam and Jesus, tended to dismiss such a connection, but not solely on account of the flimsiness of the evidence.39 In that study we encounter for the first time yet another of Urbach’s explicit methodological guidelines, cautioning against what he conceived as a growing tendency to trace rabbinic anti-Christian sentiments by an instinctive and superficial application of similarities and contextual markers. In this type of research, he advocated, rather than falling into the pitfalls of far-fetched speculations, one ought to exercise discretion and stringently criticize the sources even at the

rin 90b-91a,” in: H. Lappin (ed.), Religious and Ethnic Communities in Later Roman Palestine, Bethesda MD, 1998, pp. 249-289, where she points out the distinct possibility that the term Minim is sometimes used in Amoraic sources to denote discarded rabbinic views. As to Urbach’s inconsistent evaluations on rabbinic attitudes to Pauline Christianity, see his blatant remark (The Sages, p.303) cited above note 29, a view he contradicts on several occasions by pointing out examples to the contrary, for instance on pp. 426-427. 38 See Urbach’s study: 'ʤʠʥʡʰʤ ʤʷʱʴ ʩʺʮ' cited above (note 27), pp. 8-10. The rabbinic position went against the grain of the popular and sectarian belief that continued to hold that the powers of prophecy did not diminish during the Second Temple period, ( ibid. pp. 3-7 ). 39 See his article, 'ʭʲʬʡ ʺʹʸʴ ʬʲʥ ʭʬʥʲʤ ʺʥʮʥʠ ʩʠʩʡʰ ʬʲ ʬ”ʦʧ ʺʥʹʸʣ' cited above (note 33), pp. 281-287, where he refutes a long chain of scholarly opinions (the last being, Lauterbach, supra, ibid., pp. 545ff.) drawing a parallel between Balaam and Jesus. However, Urbach tended to accept the anti-Christian sentiments in various rabbinic interpretations of the Balaam episode (ibid.). See too my short note on the matter in Zion 47 (1982), pp. 173-177 (Hebrew).

EPHRAIM E. URBACH

183

expense of missing something of value in the process. 40 Harsh words indeed. What triggered Urbach to utter them? Implicit in Urbach’s aforementioned cautionary statement was a methodological principle, namely, that it is not worthwhile to study rabbinical thought through external sources. In our case, that is, using Christian exegesis and homiletics as tools to examine rabbinic thinking could easily lead us astray by ignoring the probable explanation (the rabbis’ own agenda) and opting for a possible explanation (a vaguely established polemical context). Thus faithful to his convictions, Urbach tended to limit his research into rabbinic anti-Christian polemics to those instances in which he could demonstrate or at least assume their possible presence. Hence in the above mentioned rather obvious topics for contention such as that of the presence and powers of prophecy or the tension between prophet and sage (i.e., revealed versus received tradition), Urbach was inclined to attribute the rabbinic treatment of these issues to the challenge posed by Christianity. His discussion of the possible Christian impact on the rabbinic adaptations and changes of the laws pertaining to idolatry provides an interesting example. Here we encounter yet another major component in Urbach’s studies of the rabbinic world, the world of Halakhah, which in his thought was fully integrated with rabbinic theology. At this stage in our inquiry it is important to stress some of Urbach’s principles concerning the foundations of Halakhah and its historical setting. Urbach regarded the Halakhah as the most fundamental force shaping Jewish life and culture throughout the ages, a fact that caused astonishment among some of the most ardent enemies of the Jews.41 The study of Halakhah, its evolution and its role in history should be carried out with the utmost caution. On the one hand, Urbach dismissed the view of many “who tend to speak about Halakhah as existing in a vacuum unconnected with existing ideas or realities which have been conditioned by circumstances.” On the other hand, he admitted that “historians have not always succeeded in differentiating between the hypothetical possibilities and well-established findings, an inclination that resulted in a num40

Ibid. p. 272. Urbach, The Halakhah: Its Sources and Development, Jerusalem 1986, p.5. Among those who were astounded was Augustine, Contra Faustum, Migne PL, Vol. 42, col. 216 (cited by Urbach, ibid.). 41

184

HOW SHOULD RABBINIC LITERATURE BE READ?

ber of historical constructions based upon flimsy foundations”.42 Thus, he did not deny the role of historical circumstances in the shaping of the Halakhah, and especially when it concerned the boundaries between Judaism and other religions, he took into account the transformations taking place in the surrounding religious atmosphere.43 But he remained very reserved when seemingly parallel phenomena in polemical contexts were being overstated in the service of far-fetched theories about possible interfaith links and ideological proximity because such theories ignored the unique innate considerations and notions that molded the halakhic world. Hence, when Urbach introduced the Christian factor and its importance to the understanding of the rabbinic laws on idolatry he also engaged in refuting overstated similarities between Christian and rabbinic hostility to idolatry. Thus the faulty approach (so, Urbach) held by Y. Baer emanated from what Baer deemed to be a striking similarity in concept and custom between Tertullian’s De idolatria and the Mishnah tractate Avodah Zarah, a treatise which, he advocated, was introduced by the Jews of North Africa to their local Christian contemporaries. Urbach while refuting the historical plausibility of this link went on to dismiss Baer’s historical ideological premises on which he based his far- reaching conclusions.44 The field in which Urbach was most prepared to confirm the existence of polemics was also the most obvious, namely, the sphere of biblical exegesis, and that only when he believed that the textual setting and the historical circumstances were pointing in that direction.45 In fact, biblical exegesis was perhaps the only area 42 “Halakha and History,” Jews, Greeks and ChristiansȥReligious Cultures in Late Antiquity: Essays in Honour of W.D. Davies, Leiden 1976, p. 39. 43 “The Rabbinical Laws of Idolatry in the Second and Third Centuries in the Light of Archaeological and Historical Facts,” See Urbach’s Collected Studies, (supra, note 2), pp. 151-193, at pp. 158, 175. 44 See his lengthy discussion, ibid. pp. 183-193. 45 Even when the historical circumstances were quite vague and the Christian behavior might have set a possible challenge to the rabbinic world, Urbach tends to dismiss this possibility or for that matter any Christian impact on rabbinic thought. An example of Urbach’s reluctance can be gleaned from his criticism of some of Daniel Goldshmidt’s interpretations in his annotated edition of the Passover Haggadah, cf. Kiriat Sepher, 36 (1961), pp. 143-150, at note, 16 (Hebrew). Note there his slighting remark about David Daube’s interpretation of one of the strangest midrashic statements in the Haggadah “smacking of a Christian twist.”

EPHRAIM E. URBACH

185

in which Urbach would contemplate evaluating the rabbinic points of view through the prism of Christian exegetical tradition, thus endorsing the possibility of an implicit or hidden controversy between the rabbis and the Church Fathers.46 This was the methodological setting for his most famous article on rabbinical polemics with the Church Fathers: ‘The Homiletical Interpretations of the Sages and the Expositions of Origen on Canticles and JewishChristian Disputation’.47 This study was to generate nearly twenty years later an attempt by Reuven Kimelman to interpret anew the very same texts.48 On the one hand, the rabbinic tradition of exegesis on The Song of Songs was particularly problematic as was their overall attitude to this rather unusual biblical book.49 On the other hand, Origen’s commentary and homilies on The Song of Songs composed in Caesarea during the second quarter of the third century could well prove to be a backdrop against which one could analyze the rabbinic sayings. Urbach did just that. Capitalizing on the similarities between Origen and the rabbis on central issues (e.g. the election and the history of Israel, the status of the Law, etc.) as well as on the exegetical method used by both parties (allegory), Urbach created what appears to be a synoptic reading of the contemporaneous (= 3rd century, of R. Yohanan’s age) rabbinic commentary with that of Origen. Urbach, unfortunately, proceeded one step too far in what seems to be the very direction he was cautioning against. He added to the already existing synoptic-like commentary an assemblage of sayings perceived by him to be a set of rabbinic 46 This mode of research into the rabbinic-Christian exegetical disputations was later to be expanded and elaborated on by a few scholars, for instance by Marc Hirshman in his, A Rivalry of Genius: Jewish and Christian Biblical Interpretation in Late Antiquity, Albany NY, 1996. In the last decade or so this particular field has witnessed a slight shift whereby a greater attention is given to the comparative research of patristic writings and their contemporaneous rabbinic traditions, most notably in the case of Jerome exhibited in the studies of Adam Kamesar, and in Hillel Newman’s as well as in Matthew Kraus’ dissertations and ensuing studies. 47 Urbach, Collected Studies, (supra, note 2), pp. 318-346. 48 See supra n. 33. 49 Rabbinic commentary on Song of Songs is located in three different midrashic compendia. Their textual redaction poses many questions regarding their order, segmentation, and not in the least their overwhelmingly anonymous nature. On the general rabbinic attitude to the book, see Urbach, ibid. pp. 148-151.

186

HOW SHOULD RABBINIC LITERATURE BE READ?

anti-Christian reflections and reactions from the 4th century, thereby forming a rabbinic anti-Christian catena-like commentary. Urbach thus creates the impression that the later rabbis were reacting to Origen’s polemical comments.50 Another case study of hidden polemics treated by Urbach sought to clarify what was perceived to be an inexplicable inconsistency in rabbinic biblical exegesis. This inconsistency, however, gave Urbach the opportunity to demonstrate yet another one of his premises: the difference between Babylonian and Palestinian sages’ attitudes toward Christianity. The rabbinic pericope concerned the attitude towards the repentance of the people of Nineveh described in the Book of Jonah. While on the one hand the Mishnah (m. Taan. 2:1) as well as the Babylonian Amoraim hailed the Ninevites as exemplary figures teaching the ways of repentance (b. Taan. 16a), the Palestinian sages denigrated them by describing them as conniving deceivers (y. Taan. 2:1 65b). Urbach resolved this dis50 Recently, forty years after, in a lengthy article, Alon GoshenGotstein, informed by the novel trends and methods in rabbinic philology, criticizes Urbach’s conclusions as well as his methodological presuppositions about the process of comparative reading of patristic and rabbinic exegesis. After nicknaming the entire trend of seeking similarities as well as differences between both literatures “parallelomania” (taken from Samuel Sandmel’s famous paper published in the JBL 81 [1962], pp. 1-11), Goshen-Gotstein focuses on a case study of Origen’s Song of Songs commentary vis à vis rabbinic exegesis and raises the following doubts: (1) Is it enough to establish a possible historical context in order to claim the existence of a polemical exchange? (2) How much of Origen’s commentary could pass as an internal Christian set of ideas without specifically addressing Jewish views? (3) What do we really know about the rabbinic exegesis of the Song of Songs, its methodology and its setting? Does not Urbach’s creation of a catena-like reading of the midrashic traditions parallel to Origen’s exegeses artificially form a polemical exchange? Does it really reflect the mode in which the midrashic commentary was formed? See A. Goshen-Gotstein, “Polemo-maniaȥMethodological Reflections on the Study of the Judeo-Christian Controversy Between the Talmudic Sages and Origen over the Interpretation of the Song of Songs”, Jewish Studies 42 (2003/4), pp. 119-190 (Hebrew). Though some of GoshenGotstein’s more general premises could be refuted, nonetheless, nowadays when the study of Judeo-Christian dialogue has reached a peak, one ought to pose and grapple with the important doubts he has raised (ibid., pp. 138-145).

EPHRAIM E. URBACH

187

crepancy by taking it to be a predominantly Palestinian rabbinic reaction to Christian (Ephrem the Syrian and others) appropriation of this case of gentile repentance over against the failure of the Israelites to repent. 51 In his view, the Babylonian rabbis were not concerned at all with the Christian point of view. The Babylonian Aggadah did not see fit to combat the newly emerging religion, because Christianity in Babylonia was an unimportant segment of the local society and thus did not pose any threat to the vibrant Jewish minority. 52 In hindsight, Urbach’s sweeping conclusion is quite difficult to accept. How would he account for all the antiChristian texts in the Babylonian Talmud? In fact, one could say that the Babylonian Talmud contained much more explicit antiChristian material than the Palestinian Talmud, and much of that material did not originate in the Palestinian traditions integrated into the Babylonian Talmud by its later redactors. Be that as it may, Urbach’s merit lay in his demonstration of caution when tackling the intricacies of rabbinic-Christian relations, since according to him, only clear exegetical hints, textual signs, and a plausible historical background could serve as indications of a polemical situation or utterance. This was as far as it went, however. Urbach did not launch a deeper and all-encompassing investigation into the relations between the rabbis and Christianity. Thus, in places where one would have expected a somewhat more thorough comparison between the rabbis and the world of Christianity, as in the cases of apocalypticism, redemption or scenarios for the End of Days, we encounter total silence.53 It becomes quite apparent that with all his 122-118, 'ʩʸʶʥʰ ʩʣʥʤʩʤ ʧʥʫʩʥʤʥ ʤʥʥʰʩʰ ʩʹʰʠ ʺʡʥʹʺ'. Ibid. p. 121. It might be that Urbach’s opinion was enhanced by Avigdor Aptowitzer’s extensive studies on the deep impact of Jewish law on the early Christian Armenian and Syrian law codes. Aptowitzer asserted that Babylonia served as the arena in which the intense contacts between Jews and Christians forging this deep influence took place. Cf. “The Influence of Jewish Law on the Development of the Law in Eastern Christendom,” Studies in Memory of A. Gulack and S. Klein, Jerusalem 1942, p. 225 (Heb.). 53 Thus when introducing the rabbinic concepts of geulah, Urbach (The Sages, p. 651) writes: “When we seek to explain the doctrines of the Sages concerning redemption and to distinguish between the different elements of which they were composed, we are compelled to have recourse … to 51 52

188

HOW SHOULD RABBINIC LITERATURE BE READ?

scholarly effort to introduce Christianity as an important factor in the study of the rabbinic world, Urbach remained skeptical and reserved concerning the scope of rabbinic disputationes cum Christianis. Therefore, the question to ask is what lay behind this reticent attitude towards the assumption of a closer, intimate and at the same time tense relationship between Jews and Christians in late antiquity?

SELF-AFFIRMATION VERSUS APOLOGETICS: EPHRAIM URBACH AND YITZHAK BAER Some possible clues are to be found in what Sussmann saw as Urbach’s intentions in The Sages, a bold attempt to describe rabbinic thought within its own internal constraints.54 This assessment, to which we shall return later, sounds rather vague. Regardless, given this set of priorities, it is not surprising that the rabbinic disputations with Christianity became secondary. One ought to add yet another important trait in Urbach’s personality, apparent in his re-

the writings of the early Christians….They are all important both for determining the dates of very many ideas…and for differentiating between concepts that are a heritage from ancient generations and views that represent reactions to certain events and actions…” to which he adds in his notes the following qualification: “Although the polemical aims are not explicitly stated in Rabbinic dicta, as for example in the remarks of Nachmanides….” Though quite right in this observation, he does not go on to explore the implicit background of rabbinic anti-Christian insinuations apart from a few more observations on some concepts. For instance, we find a discussion on the idea of the suffering Messiah (= The Messiah Son of Joseph, ibid. pp. 686-687) a figure whose most attractive parallel was Jesus, a parallel Urbach tends very much to tone down. Apart from those short observations, this is about all that Urbach has in his evaluation of the rabbinic doctrine of redemption. However, the similarities and, some would advocate, even links between the apocalyptic scenarios and redemption schemes of early Christianity and those found in Jewish and particularly rabbinic traditions are in no need of proof. For a limited study which points towards a wealth of possibilities, see my “Dating the Eschaton: Jewish Apocalyptic Calculations in Late Antiquity,” in: A. Baumgarten (ed.), Apocalyptic Time, Leiden 2000, pp. 113-154. See too supra, note 25. 54 Sussmann, ibid. pp. 71-72.

EPHRAIM E. URBACH

189

search, an antipathy to contemporary apologetics.55 In his introduction to The Sages he even declares that “the apologetic orientation compels the author to explain away everything, or, more correctly, to ‘sweeten’ everything and make it plausible.” He further criticized the type of apologetic “which was customary in the ‘Wissenschaft des Judenthums’ in Europe. The aim of the latter was to demonstrate that ‘the teachings of Judaism’ are in no way inferior to Christianity, and even transcend it.”56 An extreme instance of this tendency, in the eyes of Urbach, was Louis Finkelstein’s study: The Pharisees,57 in which Christianity was portrayed as “a mere facet of Pharisaism” and Paul as a “propagator of Pharisaism.” 58 Finkelstein’s and other past appraisals of the rabbinic mind were among the factors that triggered Urbach’s differing narrative in his Sages.59 I doubt, however, whether there is enough here to explain what we have described as his rather limited treatment of the rabbinic Adversus Christianos theme. In an attempt to explain Urbach’s reservations, I would like to offer the following modest speculation. As described above in some detail, Urbach’s scholarly premises (based on his disposition fashioned by his yeshivah studies and at the Breslau Seminary) and understanding of the rabbinic world (early and medieval) stood in stark contrast to Yitzhak Baer’s evaluation of Jewish culture and history. Baer’s work, I would contend, stood in the background of Urbach’s above described reticence. What Urbach seems to have utterly rejected was a particular brand of the “historicization” of the rabbinic world, whereby modern historians consciously or unconsciously projected their own personal agenda shaped by their contemporary historical context 55

Sussmann, ibid. p. 75. Urbach, The Sages, p.14. 57 The Pharisees: The Sociological Background of their Faith, 3rd ed. N.Y. 1962. 58 Urbach, ibid. 59 There are some notable omissions in Urbach’s introduction to his magnum opus, where he skips from Finkelstein’s work to Yitzhak Baer’s studies, namely, Max Kadushin’s work and also that of Jacob Neusner, though in the latter’s case one might advocate either of the following two explanations: (1) Neusner’s serious work on the rabbinic world post-dated the publication of The Sages (at least the Hebrew version), or (2) on a more personal level, the developing sour relations between the two scholars during the 70’s and 80’s might have caused Urbach to ignore entirely Neusner’s contributions, however controversial, to the field. 56

190

HOW SHOULD RABBINIC LITERATURE BE READ?

and cultural makeup into worlds long gone by.60 Yitzhak Baer was one such “historicizing” scholar who was also the sole “player” in the field of Late Antique Jewish History in Israel during Urbach’s prime academic career (between the forties and sixties). Baer was also the scholar who encountered some of the most severe criticism from Urbach’s pen.61 In order to understand the differences between the two scholars it is only appropriate to review Baer’s work briefly. Just like Urbach, Baer began his scholarly work by researching the history of medieval Jewry (especially in Spain), and similar to Urbach he shifted his academic and cultural interests to the world of Second Temple and rabbinic Jewry, though as will be revealed later on, in a different mode. Unlike his younger colleague Urbach, however, with his seemingly critical objective approach, Baer carried some personal convictions and an ideological agenda. Baer’s perception of history was voiced already back in 1931 in his first lecture at the Hebrew University: “Judaism is one of the forces of general History, it is also influenced by them and influences them. Today there is no more room for the Wissenschaft des Judenthums without continual attention to the relations between Israel and the nations…” This remained his conviction ever after as apparent from the title of his most influential essay Yisrael Ba-Amim (Israel among the Nations) published in Jerusalem in 1955, which marked the transition of his research interest from the world of medieval Jewry to that of the Second Temple. It seems that Baer’s historical convictions were molded first and foremost by his own experience as a Jew born and raised in 60

It is doubtful whether Urbach’s definition of this method falls in line with any of the accepted understandings of historicism which in itself is open to many interpretations. In all events it is important to stress that Baer was quite far from the modern concept(s) of historicism, about which see recently, P. Hamilton, Historicism, London 1996. For a recent concise and illuminating review of the development and shortcomings of the historicizing method see Hans U. Gumbrecht, The Powers of Philology: Dynamics of Textual Scholarship, Urbana IL 2003, pp. 54-67. 61 As related to me by Urbach’s widow, Hannah, the relations between Baer and his younger colleague Urbach were of deep appreciation and affection. They used to have a continuous exchange of views in consecutive discussions and written exchanges, which as yet have not been studied and analyzed.

EPHRAIM E. URBACH

191

Germany where he was enveloped by Christian culture and surroundings, to which he credited in his scholarly work a great deal of influence on the formation of medieval Jewish life and culture. While endorsing the fact that “the Christian influence did not affect the innermost layers of the immanent qualities of Judaism,” “he complained that the historical study of medieval Jewry was being prejudiced by the ghetto, and the deep-seated hatred between Jews and Christians, and the enormous abyss between the Torah and the Christian religion….”62 His criticism of such views in face of what he thought were the bare facts of history, such as the everyday contact between Jews and Christians, their sharing of concepts and systems of communal life and above all their common roots of religious and historical heritage, led to what was recently succinctly portrayed in a most apt manner by Israel Yuval: “The picture seen from his (=Baer’s) broad overview…was one of a Jewry that, in some senses, could be considered a miniature duplication of the Christian environment. With bold brush-strokes, he painted the Jewish community in the likeness of the Christian city; he compared the piety of R. Judah he-Hassid to that of St. Francis of Assisi and Jewish martyrdom to Crusader martyrology, and he showed that Kabbalistic literature had absorbed and internalized the Christian religious symbolism of the circle of Joachim da Fiore.”63 The catch word and modus operandi of Baer the historian was to form a constant and, as much as possible, a comprehensive comparison of both cultures within the evolution of their historical context. Even in the most extreme setting of medieval European Jewry, it is very doubtful whether Urbach would have endorsed the foregoing presentation of Jewish thought and culture; 64 however, this is beyond the scope of our preliminary study. 62 Here and in what follows, I am greatly indebted to my friend Israel Yuval’s article, “Yitzhak Baer and the Search for Authentic Judaism,” in: David N. Myers & David B. Ruderman (eds.), The Jewish Past Revisited: Reflections on Modern Jewish Historians, New Haven CT, 1998, pp. 77-87, at pp. 78-79. 63 Yuval, ibid. p.78. 64 An example of his reservations about what he seemed to perceive as modern apologetic interpretations of halakhic rulings attributed to medieval rabbis can be found in his criticism of Jacob Katz’s evaluation of the halakhic attitude of R. Menahem Ha-Meiri towards the gentiles, cf. ʺʨʩʹ' ʺʩʣʥʤʩʤ ʤʸʡʧʤ ʺʥʣʬʥʺʡ ʭʩʷʸʴ ,'ʤʩʺʥʬʡʢʮʥ ʤʸʥʷʮ -ʩʸʩʠʮʤ ʭʧʰʮ 'ʸ ʬʹ ʺʥʰʬʡʥʱʤ

192

HOW SHOULD RABBINIC LITERATURE BE READ?

Baer was to change his tune entirely during the 1940s, most probably in the wake of the Holocaust. Despite all that he had written in previous years, he remained deeply apprehensive about Christianity’s role in medieval Jewish history. This induced him to cast his net more widely into the sea of past events. The results of his research were far-reaching. Contrary to his former views, Baer began to attribute the similarities between Judaism and Christianity to the receptivity of the latter to the views emanating from the former. His goal was now to find the roots of the rabbinic preChristian Judaism, which he discovered in the world of the Hasidic, pietistic philosopher peasants of the 3rd century BCE, who fused the biblical ideals with the best cultural traits of the classical, preHellenistic Greek world. Baer advocated a monolithic view of Second Temple Judaism, a religion thriving on pietism, humble existence and a deep adherence to the principles shaping the Torah as code of law. When fused with ideals emanating from the Platonic philosophical school, these principles formed the basis of an ideal society, that of the Hasidim. In Baer’s view, during the period before the destruction of the Temple there was no room for divisiveness or sectarianism. The Essenes, Pharisees and Sadducees were all figments retrojected during a later period. Thus, when portraying the Jewish community, its ideals and its governing system, Baer contended that there were no manifestation of exilic institutions (products of the medieval Christian city), but institutions rooted in the free world of early Judaism and the conceptual world of the Greek polis.65 In his eyes, again contrary to the accepted scholarly opinion, the foundations of the rabbinic period were to be located ca. 300200 BCE, rather than during the Mishnaic period which began to crystallize two or more centuries later. The real authentic Jewish spirit and culture was formed, according to him, in a Christian-free cultural environment. Baer believed that by discovering what he had, he finally turned the tables on an old misinformed conception, which he anticipated would lead to the understanding of “the ways of those first pious men who laid the foundations for the talmudic and rabbinic Judaism, for Christianity and for the European cul,ʤʰʹ ʹʮʧʥ ʭʩʲʡʹ ʥʬ ʺʠʬʮʡ ʵʫ ʡʷʲʩʬ ʭʩʹʣʷʥʮ ,ʤʹʣʧʤ ʺʲʡʥ ʭʩʩʰʩʡʤ ʩʮʩʡ ʣʮ -ʣʬ 'ʮʲ ,ʭʿʹʺ ,ʭʩʬʹʥʸʩ. 65 Yuval, ibid. p. 81.

EPHRAIM E. URBACH

193

ture, and, in the final analysis, for our entire history in the Diaspora – foundations on which we might have built our lives today as well.” Was this not a neo-Finkelsteinian depiction? Here as Yuval rightly and ironically acclaims, “Baer the historian has almost become Baer the theologian.”66 It is only fair to surmise that from that point on Baer’s pursuit of early rabbinic cultural traits assumed a rather obsessive aura,67 in the process of which his attitude to the presence of Christianity in the formation of the rabbinic mind was entirely transformed. Faithful to his conviction that Second Temple Judaism was monolithic, he attributed the outbreak of sectarianism to the period following the destruction of the Temple, and the major force that wrestled with the world of early rabbinic Jewry was the rising power of Christianity. He dedicated a series of articles to the exposure of what he saw as the early manifestations of JudeoChristian ideology struggling with the rabbinic world. Relying on the newly discovered Qumran documents, Baer tried to revise the order of ancient historical developments. Thus, his reformed disposition towards Christianity caused him to promulgate the strangest theories. Dismissing the scholarly tendency to date the Gospels to the latter decades of the first century, he began post-dating to the post-Bar Kochba era not only the Gospels, but also key documents from Qumran like the Manual of Discipline and Pesher Habbakuk. He dated the two Qumran texts to the second and fourth centuries CE respectively, and both were seen to be creations of the Judeo-Christian circles.68 It becomes apparent that Baer’s concerted effort was actually aimed at distancing those works from their Second Temple milieu. Or, in the case of the Qumran treatises, he sought to discredit their affinity with primitive Christianity in a period that was, according to his judgment, the era of an early rabbinic utopia. Establishing the centrality of Christianity in the religious and cultural world of post-destruction Judaism, Baer advocated a double-edged view: on the one hand Christianity was molded by the ideals and customs of an early idealized Judaism, manifested for instance, in the straight line to be drawn between the martyrs of the Hasmonean period and the Christian martyrs 66

Yuval, ibid. p. 83. Urbach’s subtle reaction to Baer’s perceptions was as follows: “The historian is also permitted to be a mentor and preacher, but he must distinguish between these two functions,” The Sages, p. 16. 68 Urbach, (supra n. 36), pp. 71-74. 67

194

HOW SHOULD RABBINIC LITERATURE BE READ?

persecuted by the Roman emperors Decius and Diocletian. On the other hand, Christianity fiercely struggled against Judaism until the final rift took place in the fourth century when Christianity went astray by casting its lot with the Roman Empire.69 Christianity thus became an oppressing power and it no longer played an important role in Baer’s monolithic view of late antique Jewish history. An astonishing example of the latter was noticed by Yuval in his most penetrating study on the rabbinic legends of the destruction of the Temple (b. Git. 55a-57a). According to him, Baer, who, some would contend, convincingly dismisses the authenticity of these fables 70 and attributes their inception to the fifth century CE by demonstrating their affinity to a set of Christian traditions, fails to follow up his own findings and to explore their later Sitz im Leben, thereby showing his lack of interest in their possible role in the framework of the later Judeo-Christian dialogue.71 Returning to Urbach, Urbach’s reaction to all this was decisive: “Baer’s attitude to the sources caused whole chapters to be

69

Baer’s view that the final rift between Christianity and Judaism occurred during the fourth century CE has recently gained some adherents in more than one study, albeit depicted in a different manner. This view carries some merit though it could and should be evaluated against a broader backdrop. 70 According to Urbach, what Baer was utterly unable to perceive was the possibility that a national and spiritual leader of the caliber of Rabban Yohanan Ben Zakkai was capable of what he interpreted as defecting to the enemy’s (Roman) camp. He therefore dismissed all the traditions regarding this episode as historically unreliable and as later figments of the rabbinic imagination, cf. Urbach, (supra n. 15), pp.77-79. It should be noted that this very intriguing chain of legends has again recently attracted new fascinating interpretations: see for instance in Galit Hasan Rokem’s study Web of Life: Folklore and Midrash in Rabbinic Literature, Stanford CA 2000, pp. 171-190, and in Jeffery L. Rubinstein’s Talmudic Stories: Narrative Art, Composition, and Culture, Baltimore MD, 1999, pp. 139-175. It goes without saying that these fine legendary narratives of fundamental value for Jewish self-perception in late antiquity are bound to continue to attract scholarly attention. 71 Yuval, ibid. p. 81. Yuval himself did however explore this confluence of rival traditions: see his “‘The Lord Will Take Vengeance, Vengeance for his Temple’- Historia sine ira et studio,” Zion 29 (1994), pp. 362373 (Hebrew).

EPHRAIM E. URBACH

195

erased from historical reality.” 72 Behind the aura of homage and politeness towards Baer’s immense imprint on the study of Jewish history in the period under discussion (which has yet to be fully assessed), Urbach apparently remained dismayed by what he saw as Baer’s abuse of the sources and the historical account.73 Baer’s obsessive tendency to portray the early and formative Judaism of the Second Temple period as a uniform organism of religion and culture was in Urbach’s view a most significant handicap. Urbach, who was ready at the time to concede that the case for Second Temple sectarianism might have been overstated, was not ready to follow Baer’s purism when it was extended to the rabbinic world. To him the diversity of opinion and beliefs among the rabbis, even within the sphere of HalakhahȦa result of an extended period of internal transmutations and ever changing historical conditionsȦwas an established historical principle as well as a leading working hypothesis. 74 It is important to reiterate what has been stressed above time and again that for Urbach the rabbinic world of thought was more a product of the deliberations on, and the insights into, an internal Jewish tradition than of the disputations as well as the dialogues with external religious and cultural forces.75 72

Urbach, p. 80 (supra, n. 36). This seems to be the overall message in his magisterial appraisal of Baer’s work (supra, n. 36). 74 Urbach, pp. 80-81 (supra n. 36). 75 I believe one can discern Urbach’s manifesto on Jewish existence in the manner portrayed here in his article, cited above, note 35. In that study, which deals primarily with the halakhic sphere, Urbach presents a picture of a self-affirming Judaism and yet not an isolated organism. He does that by exposing the constant play between rabbinic universalism and ethno-religious constraints. However, the method by which this conclusion is reached underscores the point we are making. For, after dismissing the older theories about a mood of isolation that dawned on Jews following the Bar Kochba revolt, and cautioning us from treating the subject which “is highly charged with emotional and practical implications” lest “we confuse the past with the present”! (p. 271), he goes on to declare that: “historical and sociological considerations should not lead us into forgetting that the primary factors in operation here were religious presuppositions” (p. 274). For the use of race and ethnicity to make, paradoxically, universalizing claims about Christianity being a “race” open to all and the implications of race and ethnicity for early Christian selfidentity as well as for its relations with other peoples as well as the Jews, 73

196

HOW SHOULD RABBINIC LITERATURE BE READ?

Thus, commenting on the famous story in b. Shabbat 116b, the mockingly narrated encounter between a (Christian) philosopher and R. Gamaliel and his sister Imma Shalom, Urbach states that: “The Jews saw no need to engage in serious controversy with the N.T., and were content to underscore the hesitancy and the inconsistency of those who professed it.” 76 It is not enough that one might be able to spot here some sort of an autobiographical reflection, as well as an implicit refutation of Baer’s fundamental concept about the deeply embedded role of Christianity in the shaping of Jewish life and culture. Put bluntly, it seems to me that in Urbach’s historical portrayal (primarily during the Mishnaic and early to middle amoraic periods), 77 Christianity was not afforded any significant impact on Judaism, as seems to be to a growing extent the case today. 78 In concluding these preliminary observations, I would like to think that Urbach’s own careful79 and calculated output on rabbinic anti-Christianity was at least partially formed in the shadow of Baer’s widely-canvassed presentation of late antique Judaism, and it might have served also as a somber statement of a rabbinic scholar

see Denise Buell’s fascinating article, “Race and Universalism in Early Christianity,” JECS 10 (2002), pp. 429-468. 76 Urbach, The Sages, p. 303. 77 Urbach, ibid. (see supra, notes 29, 43, 44). 78 A possible follow-up to the current study should examine research on Judeo-Christian dialogue and polemic since the publication of Urbach’s The Sages, as well as methodological presuppositions and implications (cf. supra, n. 50). 79 As we have stated above (n. 50) some recent critics of Urbach’s work in the field suggest that he too overstated the case.

EPHRAIM E. URBACH

197

disenchanted by what seemed to him to be an abuse of methodology and historical material in the service of ideology.

ARCHAEOLOGY AND THE INTERPRETATION OF RABBINIC LITERATURE: SOME THOUGHTS1 Steven Fine Yeshiva University Archaeological discoveries in Palestine and throughout the Roman world have provided new ways of looking at the ancient Rabbis. The impact of this material has been characterized as “revolutionary,” as indeed it has been.2 Who now could imagine the world of the Rabbis without considering ceramics, numismatics, epigraphic discoveries, transportation systems, domestic architecture, the Dura Europos synagogue paintings, the Beth Alpha mosaic, the monumental façade of the Baram synagogue, the Beth She‘arim catacombs, and on and on? Throughout the twentieth century academic talmudists demonstrated important relationships between the archaeological record and rabbinic sources. Samuel Krauss called this approach “talmudische Archäologie.” 3 The effects of archaeology are still wider, however. Archaeology has provided new “glasses” through which to view the Rabbis and their place in Jewish culture during the late Roman and Byzantine periods, a time frame that is also known as “Late Antiquity” and “the period of the Mishnah and Talmud.” While at times social-historical issues have 1 This essay is dedicated in memory of Joseph Heinemann ʬʿʦ, on the twenty-fifth anniversary of his death, 26 Tevet 5738. Many thanks to Lieve Teugels for her comments on my manuscript. 2 L. I. Levine, “The Revolutionary Effects of Archaeology on the Study of Jewish History: The Case of the Ancient Synagogue,” The Archaeology of Israel: Constructing the Past, Interpreting the Present, eds. N. A. Silberman and D. Small (Sheffield: Sheffield Academic Press, 1997), 166-189. 3 S. Krauss, Talmudische Archäologie (Leipzig: G. Fock, 1910-12).

199

200

HOW SHOULD RABBINIC LITERATURE BE READ?

been overwrought by historians of this period (and by text scholars too positivistic in their interpretation of archaeological remains),4 there is no doubt that archaeology has provided a new and exciting perspective on the Rabbis, their texts, and general Jewish and Roman/Byzantine culture at this formative period in Western civilization. The creative interaction of rabbinic texts with archaeology is fraught with both promise and danger. Building bridges between silent artifacts and the “Oral Torah” requires great care that neither type of evidence dominates the other. On the one side, the urge to find rabbinic parallels to archaeological sources can lead to a kind of “parallelomania”. On the other, a scholarly nihilism has developed that minimalizes the significance of rabbinics for understanding the archaeological record—and vice versa. I will explore just a few of the ways that archaeology can be used to better understand rabbinic literature. My focus will be on non-legal material. I begin by illustrating ways that archaeology can inform the interpretation of rabbinic texts. I then turn to the use of archaeology in the discovery of previously unknown midrashim, describing some of the more significant discoveries of midrash “in stone.”

ARCHAEOLOGY AS COMMENTARY ON RABBINIC SOURCES Twentieth-century scholars made extensive use of archaeology as a kind of commentary on rabbinic sources. This practice has roots that go back as far as the Gaonic period, particularly regarding numismatics. It is still of interest even in “ultra-Orthodox” “Haredi” contexts (where modern archaeological practice is usually suspect).5 4 See my discussion in Art and Judaism in the Greco-Roman Period: Toward a New Jewish Archaeology (Cambridge: Cambridge University Press, 2005), 552. 5 Some of these sources are discussed in S. Z. Reich, Mesoras Hashekel: Tracing the Tradition of the Torah Money/Weight System from the Ancient Era as it Affects Contemporary Halachic Issues (Toronto: Kolel Publications; Spring Valley and Jerusalem: Feldheim, 1989), Hebrew. See also, Moses son of Isaac Alashqar, She’elot Utshuvot Maharam Al Ashkar (Jerusalem: Stizberg and Sons, 1959), no. 74, pp. 230-231; Azariah de Rossi, Meor Enayim (Mantua, 1574), ch. 56. A thorough study of medieval rabbinic discussion of Jewish archaeological remains is a desideratum. Haredi interest in archaeology has increased in recent years, as is evidenced by the number of

ARCHEOLOGY AND RABBINIC LITERATURE

201

The interpretation of archaeology entered the Wissenschaft des Judentums thanks to the visually-conscious Budapest scholar David Kaufmann and two of his students.6 Krauss’s famous first attempt at Talmudic Archaeology is still a classic. His more historiographically attuned colleague, Ludwig Blau, 7 as well as “Palestinian” E. L. Sukenik referred to this study more broadly as “Jewish Archaeology.”8 The “talmudic archaeology” approach is today called “talmudic realia,” and is practiced mainly in the Talmud departments of Israeli universities. 9 The “talmudic realia” approach made impressive accomplishments over the last century. Perhaps the finest product of this approach is Jacob Sussman’s monograph-length study of the twenty-line inscription from the synagogue narthex at Reʘov. This inscription cites known rabbinic sources, and applies them to the local context.10 Archaeological artifacts often present concrete illustrations (“realia”) of literary traditions. This is true of both specifically Jewish content, such as symbols and language, and of artifacts that were not specifically “Jewish,” but were part of the general material culture of late antiquity. For more than a century, scholars have recognized a close relationship between the languages of Jewish inscriptions from Eretz Israel and the Hebrew and Aramaic of Palestinian rabbinic literature. Indicative of this proximity is the fact that Michael Sokoloff integrated Palestinian epigraphic materials seamlessly into his A

articles on archaeological themes that have appeared in the American haredi press. The Living Torah Museum has even opened in the Borough Park section of Brooklyn, which already contains an impressive archaeological collection. See: http://www.torahmuseum.com. 6 On Kaufmann, see: M. Olin, The Nation Without Art: Examining Modern Discourses on Jewish Art (Omaha: University of Nebraska Press), 73-98. 7 L. Blau, “Early Christian Archaeology from a Jewish Point of View,” Hebrew Union College Annual 3 (1926). 8 See my discussion of Sukenik’s method: Art and Judaism, 27-34. 9 The most recent statement of this approach is: D. Sperber, Material Culture in Eretz–Israel during the Talmudic Period (Jerusalem: Ben Zvi Institute, 1993), Hebrew. 10 J. Sussman, “A Halakhic Inscription from the Beth-Shean Valley” Tarbiz 43 (1973-1974): 88-158, 44 (1974-1975):193-195, Hebrew.

202

HOW SHOULD RABBINIC LITERATURE BE READ?

Dictionary of Palestinian Jewish Aramaic.11 The vocabulary and syntax of inscriptions are regularly used to interpret difficult words in rabbinic literature and particularly to resolve scribal difficulties in our received manuscripts. This relationship holds despite the fact that inscriptions are often two or more centuries later than received traditions and texts. Inscriptions show an amazing stability over time. It is often impossible to judge whether a dedicatory inscription dates from, say, the fourth or the eighth centuries based upon paleographic or literary criteria. I will suggest just one example that illustrates ways that inscriptions are used by philologists in the interpretation of ancient texts. A fine example of the interrelationship between epigraphic evidence and rabbinic sources was presented by Sokoloff to resolve a difficulty in y. Berakhot 5:3, 9c, 12 ʣʧʡ ʯʩʡʺʩ ʩʦʴ ʯʡ ʠʣʥʩ ʩʡʸʥ ʠʧʠ ʩʡʸ ʠʺʥʡʩʺ ʩʮʥʷ ʣʧ ʸʡʲ ʩʺʠ .ʠʺʹʩʰʫ. Through the study of manuscripts and inscriptions, Sokoloff came to the conclusion that this text contains a scribal error and a gloss. The original text described two rabbis who “were sitting in a place (ʸʺʠ).” ʩʺʠ was originally ʸʺʠ, a simple orthographic error. Sokoloff postulates that a later scribe added a gloss, ʠʺʹʩʰʫ. Support for this interpretation is widespread in extant inscriptions, where ʤʸʺʠ, ʤʹʩʣʷ ʤʸʺʠ, and ʠʺʹʩʰʫ (ʤʺʹʩʰʫ) are well attested.13 To complicate matters, I would add that in an inscription from the ʗammat Gader synagogue, as throughout rabbinic literature, ʸʺʠ refers to either the synagogue or the general place. A similar problem appears at Alma (with a parallel at Baram), 11

Ramat Gan: Bar Ilan University Press, 1991. On Hebrew: M. BarAsher, “Mishnaic Hebrew: An Introductory Survey,” Hebrew Studies 40 (1999) 115-151. 12 M. Sokoloff, “Epigraphical Notes on the Palestinian Talmud” Bar Ilan 18-19 (1980): 218-219 (Hebrew); idem, A Dictionary of Jewish Palestinian Aramaic (Ramat Gan: Bar Ilan University Press, 1990), 82. 13 Knisha: appears in inscriptions from ʗammath Gader and Beth Guvrin (Naveh, On Stone and Mosaic, 60-61, 109-111); atar: ʗammath Gader, probably in reference to the synagogue (Naveh, op. cit., 154, 156); atra: Naaran (100-101); maqom: Baram (small synagogue) and Alma (19-20, 2223). It is unclear, however, whether these inscriptions refer to the synagogue upon which it is inscribed, to the village that it faced out toward, or to both. Atra qadisha: Kefar ʗananyah (Naveh, op. cit., 34-36); ʗammath Tiberias B (pp. 48-49), Beth Shean B (pp. 77-78), Naaran (4 times, pp. 9596, 99-102).

ARCHEOLOGY AND RABBINIC LITERATURE

203

where we find a blessing for peace ʬʠʸʹʩ ʺʥʮʥʷʮ ʬʫʡʥ ʤʦʤ ʭʥʷʮʡ. Finally, it is quite likely that the term ʤʹʩʣʷ ʤʸʺʠ in a second inscription from the Beth Shean “study house” may refer to a study house (which, along with study, were always places of prayer), and not to a synagogue.14 Conversely, in a footnote to the same article, Sokoloff uses scribal method in order to interpret an inscription at ʗorvat Ammudim. 15 This very lightly incised inscription had originally been interpreted by archaeologist N. Avigad and by philologist Y. Kutscher to read ʠʩʮʥʹ ʩʸʮʣ ʠʸʺ, “gate of the master of heaven.”16 In the tumult of 1948 the inscription was misplaced, and only recently located by the staff of the Hebrew University’s Institute of Archaeology.17 Based upon photographs, epigrapher Joseph Naveh reinterpreted the resh as a qof, though he was aware that this letter is unclear. According to Naveh, in consultation with Jonas Greenfield, the inscription reads ʠʩʮʥʹ ʩʸʮʣ ʠʷʺ, “receptacle of the Master of Heaven,” referring to the Torah shrine.18 ʷʩʺ is a common word in rabbinic Hebrew and ʤʷʩʺ appears in Targumic literature. 19 Both are loan words from the Greek théke. An inscription from Dalton supports this reading. There the Torah shrine is called ʤʰʮʧʸ ʤʷʩʺ “Receptacle of the Merciful (One).”20 In 1978, F. G. Hüttenmeister tentatively argued, based upon Avigad’s 14

Naveh, On Stone and Mosaic, 77-78; Fine, This Holy Place, 67-72, 100-

101. 15

Sokoloff, “Epigraphical Notes,” 218, n. 2. Reading ʠʸʺ as a variant of ʲʸʺ. N. Avigad, “An Aramaic Inscription from the Synagogue at Umm El-’Amed in Galilee,” Louis M. Rabbinowitz Fund for the Exploration of Ancient Synagogues Bulletin 3 (1960), 62-64; E. Y. Kutscher, “Jewish Palestinian Aramaic,” in F. Rosenthal, An Aramaic Handbook (Wiesbaden: Harrassowitz, 1967), I/1: 70. 17 See Naveh, On Stone and Mosaic, 41, "ʯʥʫʮʤ ʳʱʥʠʡ ʺʥʶʮʩʤʬ ʤʫʩʸʶ ʯʡʠʤ", idem, “The Aramaic and Hebrew Inscriptions from Ancient Synagogues,” Eretz-Israel 20 (1989), 307. It was examined by the author in 2001. 18 Naveh, On Stone and Mosaic, 40-41. 19 Naveh, ad. loc., Sokoloff, Dictionary of Jewish Palestinian Aramaic, 581. 20 Naveh, On Stone and Mosaic, 144-145; S. Krauss, Griechische und lateinische Lehnwörter im Talmud, Midrasch und Targum (Berlin: S. Calvary and Co., 1898-1899), 588. Cf. Sokoloff, Dictionary of Jewish Palestinian Aramaic, 581, who translates “sheath.” Rahmana: “The Merciful One,” following Sokoloff, op. cit., 522. See also Naveh, op. cit, nos. 70-72, 106-109. 16

204

HOW SHOULD RABBINIC LITERATURE BE READ?

Hüttenmeister tentatively argued, based upon Avigad’s reading, that ʠʸʺ is a variant of ʠʸʺʠ.21 The inscription would then translate “place of the master of Heaven.” In 1980 Sokoloff strengthened Hüttenmeister’s position, based upon orthographic parallels in Palestinian Jewish Aramaic. In a marginal gloss to Targum Neofiti to Num. 21:8, for example, ʸʺʠ is written ʸʺ.22 Naveh revisited the inscription in an article published in 1989, some time after it was rediscovered. In this article, he accepted Avigad’s original reading, ʠʸʺ, as well as the Hüttenmeister-Sokoloff interpretation. I cite these examples in order to exemplify the amazingly close linguistic relationship between extant inscriptions and rabbinic literature. This relationship is not to be taken lightly. The proximity of these two bodies of related texts is a necessary precondition for the kinds of interpretation suggested in the example cited, and sets a firm foundation for higher level interpretation of the interconnectedness of Palestinian rabbinic texts and archaeological sources. Ever since the Middle Ages, scholars have been aware that archaeology can be useful for the interpretation of artifacts described in rabbinic sources. Individual artifacts, and even cityscapes, provide important information for interpreting specific rabbinic texts. Numerous parallels have been adduced, and exhibitions highlighting the relationships have been organized (two curated by the present author).23 This approach has been popularized in Adin Steinsaltz’s Hebrew edition of the Babylonian Talmud, where archaeological artifacts are used as illustrations of the talmudic discussion, and even the title page is modeled upon late antique synagogue discoveries.24 I will cite just two examples. The first is drawn from the realm of Roman period jewelry, the second from the lighting of late antique Jewish public spaces. 21 F. Hüttenmeister, “The Aramaic Inscription from the Synagogue at H. ‘Ammudim,” IEJ 18 (1978), 111-112. 22 See Sokoloff, Dictionary, 81-82. 23 U. Zevulun and Y. Olenik. Form and Function in the Talmudic Period (Tel Aviv: Haaretz Museum, 1979); Fine, The Tangible Talmud: Text and Artifact in the Greco-Roman Period (Los Angeles, U. S. C. Archaeological Research Collection and Fullerton, California State University, 1987); Fine, ed., Sacred Realm: The Emergence of the Synagogue in the Ancient World (Oxford University Press and Yeshiva University Museum, 1996). 24 Talmud Bavli, punctuated, interpreted and tr. A. Steinsaltz (Jerusalem: ha-Makhon ha-Yisraeli le-Firsumim Talmudiyim, 1980-).

ARCHEOLOGY AND RABBINIC LITERATURE

205

Mishnah Shabbat 6:1 instructs that “A woman may not go out (on the Sabbath … wearing) a city of gold.” Similarly, Mishnah Kelim 11:8 explains that “All women’s ornaments are susceptible to defilement, for example, a city of gold ….” Nowhere does Tannaitic literature attempt to define or identify this artifact, which apparently was obvious to all. Y. Shabbat 6:1, 7d, however, goes into some detail:25 “and not (wearing) a city of gold” (m. Shabbat 6:1). Rav Judah said: For example, a Jerusalem of gold (ʡʤʣʣ ʭʩʬʹʥʸʩ). Our Rabbis of Caesarea say: ʯʩʬʷʨ ʷʥʨ ʱʥʸʴ

Queen Esther wearing a turreted crown from the Dura Europos Synagogue (Photograph by Fred Anderegg, after Goodenough, Jewish Symbols 11, pl. vi).

The Caesarean rabbis explain the “city of gold” through a Greek loan word, with Samuel Krauss identifies as: chruskastellon, a

25

Cf. y. Sot. 9:16, 24c, b. Shab. 59b.

206

HOW SHOULD RABBINIC LITERATURE BE READ?

golden castle. 26 This correspondence translation suggests that an artifact known from the general context stands behind the “city of gold,” which is little more than a translation of the Greek. In 1967 Shalom Paul correctly associated the “city of gold” with tiaras in the shape of cities that appear in Roman art, particularly in sculptures of Tyche (Fortuna). 27 Significantly Paul shows that this type of tiara is also worn by the goddess Tyche and by Esther in the Dura Europos synagogue paintings. The comment of Rav Judah, a second generation Babylonian Amora cited in our Yerushalmi text finds important parallels in b. Shabbat 59a and b. Sotah 49b. These texts ask: “‘What is a city of gold?’: Said Rabba bar bar Hanna said Rabbi Johanan (a Palestinian amora of the second generation): a Jerusalem of Gold” (ʠʡʤʣʣ ʭʩʬʹʥʸʩ). In this way a typical Roman artifact (which, to the best of my knowledge was unknown in Sassanian Persia) was judaized in amoraic literature through association with the holy city of Jerusalem. Our second example sets rabbinic sources within the context of rabbinic study houses and within synagogues.28 Happily, a large

26 S. Krauss, Griechische und lateinische Lehnwörter im Talmud, Midrasch und Targum (Berlin: S. Calvary and Co., 1898-1899). See also S. Lieberman, Hayerushalmi Kiphshuto (New York and Jerusalem, 1995), pt. 1, vol. 1, 102. 27 See S. M. Paul, “Jerusalem—City of Gold.” Israel Exploration Journal 17 (1967): 257-263; idem, “Jerusalem of Gold—A Song and an Ancient Crown.” Biblical Archaeology Review 3, no. 4 (1977): 38-40. On the use of Roman jewelry fashions in rabbinic culture, see U. Zevulun and Y. Olenik. Form and Function in the Talmudic Period. 28 On the synagogue setting, see: J. Heinemann and J. J. Petuchowski, The Literature of the Synagogue, (New York: Behrman House, 1975). On the study house setting: S. D. Fraade, “Interpreting Midrash 1: Midrash and the History of Judaism,” Prooftexts 7, no. 2 (1987), 179-194; idem, “Sifre Deuteronomy 26 (ad Deuteronomy 3:23): How Conscious the Composition?” Hebrew Union College Annual 54 (1988), 245-301; idem, From Tradition to Commentary: Torah and Its Interpretation in the Midrash Sifre to Deuteronomy (Albany: State University of New York Press 1991); Yonah Fraenkel is extreme in his virtual negation of the significance of the synagogue in the construction of rabbinic homilies. For Fraenkel, the Sitz-im-Leben of these homilies is essentially the study house. See: Y. Fraenkel, The Ways of Aggadah and Midrash (Israel: Yad la-Talmud, 1991), 17-43, Hebrew. More recently, see Fine, “‘Their Faces Shine with the Brightness of the Firmament:’ Study Houses and Synagogues in the Targumim to the Pentateuch,”

ARCHEOLOGY AND RABBINIC LITERATURE

207

number of synagogues (and one or more study houses)29 have been uncovered in the Land of Israel that can help in the process of imagining the physical context in which the images painted on walls, set on floors and carved on stones interacted with communities that were thoroughly infused with midrashic themes and ways of thinking. What of these “sets” upon which rabbinic literature was performed? Rabbinic texts describe synagogues as having podia (bimot), Torah shrines generally on the Jerusalem-oriented walls, and other appurtenances. Archaeology gives us a sense of the sizes, proportions, colors and textures of some of the places where midrash thrived during the late Roman and Byzantine periods. It is my sense that the floors, lamps, painted walls and human actors functioned together to define and give meaning to the synagogue or study house space. This three-dimensional way of reading both texts and artifacts, and most importantly, the flesh and blood communities who “lived” the texts and artifacts, often facilitates new ways of looking at both sorts of evidence.30 To cite just one example: Genesis Rabbah 4:231 describes a Rabbi using the lamps of a synagogue or a study house to explain a complicated issue of Biblical physics: Said Rabbi Tanʘuma: I will explain. If it said (in Genesis 1:7): “And God made the firmament and He separated between the wain Biblical Translation in Context, ed. F. W. Knobloch (Bethesda, Md.: University Press of Maryland, 2002), 63-92. 29 On the Dabbura inscription, “This is the study house of Rabbi Eliezer ha-Qappar,” see Naveh, On Stone and Mosaic, 25-26. On a possible study house in Beth Shean: D. Bahat, “A Synagogue at Beth-Shean,” in Ancient Synagogues Revealed, ed. L. I. Levine (Jerusalem: Israel Exploration Society, 1981), 82-85; Fine, This Holy Place, 100-101. See also D. Urman, “The House of Assembly and the House of Study: Are They One and the Same?” in Ancient Synagogues: Historical Analysis and Archaeological Discovery, ed. D. Urman and P. V. M. Flesher (Leiden: E. J. Brill, 1995), 1: 232-255. 30 See my extensive discussions of this phenomenon in Art and Judaism, 172-204, and “A Liturgical Interpretation of Synagogue Remains in Late Antique Palestine,” in Continuity and Renewal: Jews and Judaism in ByzantineChristian Palestine, ed. L. I. Levine, Jerusalem: Dinur Center and Ben Zvi Institute, 2004, 402-429. 31 J. Theodor and Ch. Albeck, eds., Midrasch Bereschit Rabba, 3 vols. (Jerusalem: Shalem Books, 1996) p. 27.

208

HOW SHOULD RABBINIC LITERATURE BE READ? ter” which is on (ʬʲ) the firmament, I would say that the water is placed on the body of the firmament (ʲʩʷʸ ʬʹ ʥʴʥʢ ʬʲ). Since it says (in Genesis 1:7): “[And God made the firmament and He separated] between the water which is above (ʬʲʮ) the firmament,” I say that the upper waters are held in place by [divine] command. Said Rabbi Aʘa: “Like this lamp” (ʠʬʩʣʰʷ ʯʩʣʤʫ).

This midrash is part of an extended discussion by Amoraim regarding the nature of the biblical firmament. It is set as a conversation between Rabbis, presumably within a study house setting. Responding to Rabbi Tanʘuma, Rabbi Aʘa apparently pointed toward a lamp in the room where they were studying to explain Rabbi Tanʘuma’s position. His comment, “this lamp,” is meant to explain the biblical verse. Late antique synagogues, churches and study houses were illuminated with both clay and translucent glass lamps. Clay lamps were uncovered in the Gush ʗalav synagogue, and fragments of glass lamps and their holders were found at a number of sites.32 Glass lamps are illustrated in numerous synagogue mosaics. While seven-branched menorahs may have had all kinds of associations, their primary purpose was to illuminate an otherwise dark building. The lights of a synagogue bema, like those of churches, served as spot lights, highlighting the more holy section of the synagogue. In synagogues, lamps facilitated the public reading of Scripture. The way that such glass lamps were (and are) used provides ample context for Rabbi Aʘa’s analogy. Krauss and Theodor both had considerable difficulty imagining the realia be32

Gush ʗalav, see E. M. Meyers, C. Meyers and J. Strange, Excavations at the Ancient Synagogue of Gush Halav (Winona Lake, Ind.: Eisenbrauns, 1990), 128-129, 158-165. The authors suggest that oil lamps were discovered in this synagogue that are “unique to Gush ʗalav and found nowhere outside the synagogue site.” They argue that “it seems reasonable to suggest that they were manufactured solely for use within this building.” On Meroth, see Z. Ilan and E. Damati, Meroth: The Ancient Jewish Village (Tel Aviv: Society for the Preservation of Nature in Israel, 1987), 144; At ʗammath Tiberias B, level 2, M. Dothan, Hammath Tiberias: Early Synagogues (Jerusalem: Israel Exploration Society, 1983), 62, notes, “Almost all of the lamp fragments were found on the floor of Locus 52, which was probably the treasury of the synagogue.”

ARCHEOLOGY AND RABBINIC LITERATURE

209

hind this description, though the Maharzu, Ze’ev Wolf Einhorn, got it right in his commentary on Midrash Rabbah.33 The text imagines a glass lamp containing water and oil. In fact, glass lamps are always filled partially with water, so that the glass does not overheat and shatter. Owing to the differing relative gravities of water and oil, the oil floats above the water. The wick floats in the oil. Rabbi Aʘa refers to the apparent miracle of the separation of the water from the oil. By analogy, the heavenly firmament also separates “the waters above from the waters below.”34 This midrashic interpretation, set in the material culture of late antique Jewish life, also teaches much about modes of learning and the ways that analogies were drawn by the Rabbis from their material environment.35

ARCHAEOLOGY AND THE DISCOVERY OF LOST MIDRASHIM Through the entire twentieth century scholars of rabbinic literature were engaged in a sustained search for “lost midrashim.” Roman sources, the writings of the Church Fathers, early Christian art, Karaite and Islamic literatures were carefully combed in an attempt to “redeem” lost Jewish traditions. Louis Ginzberg was the master 33

Krauss, Talmudische Archäologie, 522. See the commentary of Theodor-Albeck, and parallels cited there. Z. W. Einhorn, Commentary of the Maharzu, published in the Vilna Romm edition of Midrash Rabba. Rpt. Jerusalem, ad. loc. See also S. D. Goitein, A Mediterranean Society, 2 (Berkeley and Los Angeles: University of California Press, 1971), 2: 150 who notes that “In 1075 the synagogue of the Palestinians (in Fustat) was illuminated by fifty-one large and small chandeliers, called buqandalt i.e., abu qandalat), a term not found thus far elsewhere…”; idem, “The Synagogue Building and Its Furnishings According to the Record of the Cairo Geniza,” in Religion in a Secular Age, ed. S. D. Goitein (Cambridge, Mass: Harvard University Press, 1964), 170*. M. L. Trowbridge, Philological Studies in Ancient Glass (Urbana, Ill., 1930), 190, notes that, “The first reference I find to glass lamps is in the fourth century where many large candelae are described as hanging in a church.” See L. Bouras, “Byzantine Lighting Devices,” XVI Internationaler Byzantinstenkongress Akten (= Jahrbuch der Österreichischen Byzantinistik 32, no. 3) 2, no. 3 (1982), 479; Sokoloff, Dictionary of Jewish Palestinian Aramaic, 1991: 496. 34 For other examples, see M. Bregman, “The Darshan: Preacher and Teacher of Talmudic Times,” The Melton Journal 14 (1982), 3, 19, 26. 35 Ibid, 47-48.

210

HOW SHOULD RABBINIC LITERATURE BE READ?

of this process, beginning with his doctoral dissertation, Die Haggada bei den Kirchenvätern und in der apokryphischen Litteratur (1900) 36 and continuing through Legends of the Jews (1909) 37 and the Ginzei Schechter series (1928/29).38 With the frequent discovery of Jewish artifacts after World War I, archaeology too came to be seen as a source of midrashim. E. L. Sukenik is a good example. His first survey article on synagogue archaeology, published in 1923 in Rimon: A Hebrew Magazine of Art and Letters, reflects a real attempt to attract generally image-shy Wissenschaft scholars. He describes the recentlydiscovered ʗammath Tiberias stone menorah. Sukenik notes that the branches of the menorah are decorated in pomegranate pattern, reminiscent of the ʧʸʴʥ ʸʥʺʴʫ of Exodus 25 (and parallels). He writes with obvious pleading that: “Every interpreter of Holy Scripture will need to take into account the image of this menorah that was buried in the past, certainly during a period of danger, and revealed to us in the first Hebrew excavations in the Land of Israel.”39

Limestone Menorah from the Synagogue of Hammath Tiberias A (Courtesy of the Israel Exploration Society).

The discovery of the Beth Alpha synagogue in 1928/29 and particularly of the Dura Europos synagogue in 1932, with their signifi36

L. Ginzberg, Die Haggada bei den Kirchenvätern und in der apokryphischen Litteratur, (Berlin: S. Calvary), 1900. 37 L. Ginzberg, Legends of the Jews (Philadelphia: Jewish Publication Society, 1909). 38 L. Ginzberg, Genizah Studies in Memory of Solomon Schechter (New York: Jewish Theological Seminary, 1928/29), 3 vols, Hebrew. Reprinted with an introduction by Burton Visotzky by Gorgias Press, Piscataway, 2003. 39 E. L. Sukenik, “Ancient Synagogues in Palestine,” Rimon: A Hebrew Magazine of Art and Letters, 5 (1923), 19 (Hebrew), my translation.

ARCHEOLOGY AND RABBINIC LITERATURE

211

cant arrays of biblical imagery, solidified the significance of archaeology in the search for lost midrashim. It is not insignificant that a major advisor to Carl Kraeling in his publication of the final report of the Dura synagogue was none other than Louis Ginzberg. Inscriptions too yield texts that closely parallel the rabbinic corpus. The most spectacular example is the halakhic inscription from Reʘov, though other examples exist that bear liturgical, aggadic and midrashic import. Scraps of biblical interpretation, “midrash,” are imbedded in a number of inscriptions. 40 In addition, images discovered within ancient Jewish contexts provide important parallels to midrashic texts, and on occasion even reveal otherwise unknown midrashim. While numerous visual representations of biblical themes have been uncovered, no sustained discussion of midrashic and other rabbinic themes in ancient Jewish art has been prepared. We are still in the “show and tell” mode. In a 1983 article entitled “The Illustrated Midrash in the Dura Synagogue Paintings: A New Dimension for the Study of Judaism,” art historian Joseph Gutmann attempted to summarize and convince the audience of the Proceedings of the American Academy of Jewish Research to take the relationship between ancient art and midrash seriously. 41 Gutmann describes the most important examples from Dura, and sets them within the contexts of rabbinic midrashim. I refer the reader to Gutmann’s discussion. Taken together with C. Kraeling’s final report of the Dura synagogue and Sukenik’s 1947 monograph Beit ha-Keneset shel DuraEuropos ve-Tsiyurav,42 this article highlights in a sophisticated manner the many midrashic parallels discovered there. Archaeology from the Land of Israel provides few examples as stirring as those at Dura. Still, numerous parallels between archaeology and midrash 40 These texts have yet to be studied as a group. See in the meantime G. Foerster, “Synagogue Inscriptions and Their Relation to Liturgical Versions.” Cathedra 17 (1981), 12–40, Hebrew; A. Shinan, The World of the Aggadah, tr. J. Glucker (Tel Aviv: MOD Books, 1990), 81-91; J. Yahalom, “Synagogue Inscriptions in Palestine-A Stylistic Classification,” Immanuel 10 (1980), 47-57. 41 J. Gutmann, “The Illustrated Midrash in the Dura Synagogue Paintings: A New Dimension for the Study of Judaism,” PAAJR 50 (1983): 91104. 42 E. L. Sukenik, Beit ha-Keneset shel Dura-Europos ve-Tsiyurav (Jerusalem: Mosad Bialik), 1947.

212

HOW SHOULD RABBINIC LITERATURE BE READ?

are presented visually—and occasionally new discoveries are possible. I will suggest just a few, focusing upon details that have generally been overlooked. Sukenik was right that the ʗammath Tiberias synagogue mosaic would be important for biblical interpretation. The artisan who carved this lampstand interpreted ʧʸʴʥ ʸʥʺʴʫ as recurring pomegranates. This interpretation appears in numerous depictions found more recently, most prominently in the mosaic of the synagogue of Severos, also at ʗammath Tiberias.43 I have not found any rabbinic text that makes this connection. This connection was made, however, by Josephus. He writes that the menorah was “… was made up of globules and lilies, along with pomegranates and little bowls.”44 Our menorah depictions thus stand in a long tradition of interpretations of the menorah. Only through archaeology do we now know of the presence of this interpretive tradition among Jews in late antique Palestine. The mosaic depictions of menorahs at ʗammath Tiberias provide other “midrashic” details. As in many images, the bases of the menorahs are in the form of tripods. This too is a kind of midrash, as the biblical text provides no information on the existence of a base. Targumic tradition resolves this textual lacuna, translating Exodus 25:31 ʤʫʸʩ as ʤʣʩʣ ʱʩʱʡ.45 The choice of a threelegged base was far from a foregone conclusion. One could imagine a base, for example, like that of the Arch of Titus menorah. It is my sense that the tripod bases of this and many other depictions relate to a practical consideration: bronze lampstands in this period generally had three-legged bases.46 It is likely that genuine bronze menorahs actually stood in synagogues during this period. Once constructed and then depicted, the three-legged base was de facto integrated into the ways that Jews imagined the menorah. In the mosaic depictions of menorahs in the “Synagogue of Severos” at ʗammath Tiberias the flames of each lamp are inclined

43 R. Hachlili, Ancient Jewish Art and Archaeology in the Land of Israel (Leiden: E. J. Brill. 1988), 241-249. 44 Ant. 3: 144-146. 45 Neofiti, Fragment Targum, Pseudo-Jonathan. 46 J. Brand, Ceramics in Talmudic Literature (Jerusalem: Mossad Harav Kook, 1953), 296-314 (Hebrew).

ARCHEOLOGY AND RABBINIC LITERATURE

213

toward the central stalk of the menorah. This image nicely parallels a tradition in Sifre Zuta, Beha’alotekha to Numbers 8:2:47 … And whence do I know that each lamp was pointed toward the middle lamp? Scripture says: “toward the lampstand (menorah)” (Num. 8:2). And thus it says: “and he dwells turned toward me” (memuli, Num. 22:5). Said Rabbi Simeon: When I went to Rome there I saw the menorah. All of the lamps were pointed toward the middle lamp.

In fact, among the numerous images of menorahs discovered in the Roman catacombs, quite a few show lamps inclined toward the center.48 Is this based upon a Palestinian or even local interpretation of Numbers 8:2, or is this imagery drawn from actual observation of the Temple menorah? In addition to this midrash, Josephus relates that the menorah of the Second Temple was displayed in Vespasian’s Temple of Peace on the Roman Forum during this period,49 and the once colorful Arch of Titus bas relief was in far better condition for viewing than it is now. Interestingly, the shape of most menorah depictions from Rome is different from Palestinian depictions. In these depictions, the ratio of the height of the branches to the width of the branches approximates the depiction on the arch of Titus. Significantly, the base of the Arch of Titus menorah is nowhere to be found, and the menorah is depicted as a tripod (as in Palestinian synagogue images). Interpretations of biblical themes from these communities are exceedingly rare, and give voice to ways that Jews in the Roman diaspora inter47

Siphre ad Numeros adjecto Siphre Zutta. Ed. H. S. Horovitz (Jerusalem: Wahrman, 1976); Midrash ha-Gadol, ed. Z. M. Rabinowitz (Jerusalem: Rav Kook Institute, 1967), ad. loc. 48 The illustrations are most conveniently arranged by Goodenough, Jewish Symbols During the Greco-Roman Period (Princeton: Princeton UP, 1953), 3: nos. 769, 808, 810, 817, 973. Images of menorahs from Rome and other diaspora centers interpret the bulbs and calyxes as floral in form, though not as pomegranates. 49 War 7, 158. For a full discussion, see my “‘When I went to Rome, I Saw the Menorah…’: The Jerusalem Temple Implements between 70 C.E. and the Fall of Rome,” Festschrift for Eric Meyers, ed. D. Edwards, forthcoming.

214

HOW SHOULD RABBINIC LITERATURE BE READ?

preted visually the biblical menorah. The great similarity of these depictions, and particularly the midrashic elements, to art from the Land of Israel is worthy of note. It is suggestive of a “Jewish koine” that stretched beyond the frontiers of Jewish Palestine.50 Interpretations of biblical narrative scenes have also been discovered in the Land of Israel. The visit of the angels to Abraham was found at Sepphoris; the Binding of Isaac was found at Beth Alpha and Sepphoris; Daniel in the Lion’s Den was found at Naaran (near Jericho), Ein Samsam (in the Golan) and in a very poor state of preservation at Susiya (in the Hebron Hills). David playing his harp was discovered in Gaza; Noah’s Ark at Gerasa; and fragments of Aaron before the Tabernacle in Sepphoris. 51 Scholars have found numerous parallels to rabbinic literature in these artifacts. More importantly, parallels in Byzantine period liturgical texts and Targumim, which are roughly equivalent in date to the artifacts, have been very important. This form of interpretation invites caution. Biblical iconography and themes were generally drawn directly from Christian models. I would nevertheless suggest that once they crossed the threshold into the synagogue, interpretations that might be obvious to art historians and even perhaps to ancient mosaic craftsmen (who worked on commission for Jews, Christians and Samaritans) would not necessarily have been common knowledge to local Jewish viewers. The Binding of Isaac at Beth Alpha is a fascinating example of a specifically Jewish presentation of a theme that is ubiquitous in Christian art.52 This panel is unusual in late antique art specifically because Abraham “the father of faith” in Christian contexts, is not 50

Fine, This Holy Place, 125-157. See Hachlili, Ancient Jewish Art and Archaeology in the Land of Israel, 287-300; Weiss and Netzer, Promise and Redemption. 52 On the binding of Isaac in Christian art, see A. M. Smith, “The Iconography of the Sacrifice of Isaac in Early Christian Art,” American Journal of Archaeology 26 (1922); I. S. Van Woerden, “The Iconography of the Sacrifice of Isaac,” Vigiliae Christianae 15 (1961); R. Jensen, “The Offering of Isaac in Jewish and Christian Tradition: Image and Text,” Biblical Interpretation 11, no. 1 (1994); J. Gutmann, “The Sacrifice of Isaac: Variations on a Theme in Early Jewish and Christian Art,” in Sacred Images: Studies in Jewish Art from Antiquity to the Middle Ages (Northampton, Variorum Reprints 1989); “Revisiting the Binding of Isaac Mosaic at Beth Alpha,” Bulletin of the Asia Institute 6 (1992). 51

ARCHEOLOGY AND RABBINIC LITERATURE

215

the focal point of this composition. Rather, at the focal center of the panel is the Hand of God reaching down from the heavens, the ram caught in the thicket immediately below. The focal point is the redemptive moment, when God cries out, “don’t do it!” (in biblical verse) and the ram is revealed ready to serve as Isaac’s substitute. This focus fits well with Jewish reflection on the Binding of Isaac, where Abraham’s faith is subsumed to God’s eternal pledge to redeem the children of Israel.53 The horn of the ram is much more bright than the rest of the creature, and draws attention. My sense is that this is quite intentional. The ram’s horn is emphasized, I would suggest, specifically because of its enduring liturgical significance. Its blowing on Rosh ha-Shanah was considered to be a reminder of the Covenant, the Binding of Isaac being the fullest statement of zekhut avot, the protective and enduring “merit of the fathers.” 54 The Beth Alpha mosaic does not present a “new” midrash. It does present, I would suggest, a Jewish transformation of a theme that is well-known in Christian art and its “Judaization.”55 “Archaeological midrashim” have indeed added to the corpus of rabbinic sources discovered from “unconventional” sources during the twentieth century. These are both textual and visual. Having suggested what I see as examples of archaeological midrashim, and my interpretation of these images, I hasten to add the proviso that visual sources are not so specific in their associations as written texts, even when the imagery is labeled. Scholars have spent inordinate energy asserting that specific images represent one biblical theme to the exclusion of every other possibility. Gutmann has catalogued multiple interpretations of every detail of the Dura

53

A. Shinan has surveyed the ways that the Binding of Isaac (Genesis 22) is presented in midrash, liturgy and targum. See his “Synagogues in the Land of Israel: The Literature of the Ancient Synagogue and Synagogue Archaeology,” in Fine, Sacred Realm, 130-152. 54 S. Schechter, Aspects of Rabbinic Theology (New York: Macmillan, 1909), 170-198; G. F. Moore, Judaism in the First Centuries of the Common Era (Cambridge, Mass.: Harvard University Press 1927-30), 1: 538-546. 55 On this process in later Jewish art, see S. S. Kayser, “Defining Jewish Art,” Mordecai M. Kaplan Jubilee Volume (New York: Jewish Theological Seminary of America, 1953), 457-467.

216

HOW SHOULD RABBINIC LITERATURE BE READ?

paintings that have been suggested by scholars.56 Some are based upon biblical texts, other rabbinic texts, and others on spurious assertions of the influences of neo-platonism, a Jewish mystical religion, structuralism, class struggle, and overstated notions of Christian influence. Scholars have asserted strongly deterministic programmatic narratives of ancient Jewish art, first at Dura and more recently regarding the Sepphoris synagogue mosaic.57 Unfortunately, the images cannot speak back for themselves in the ways that literary sources can. My own sense is that contemporaneous Jewish texts, used judiciously can set the general context for archaeological discoveries. Such an interpretation must not be monistic or exclusive of other possibilities, but should learn from the Rabbis (as well as from contemporary literary studies) to accept and even celebrate a multiplicity of meanings. ʭʩʸʮʥʠ ʹʩ is the order of the day—interpretations depending upon the liturgical, pedagogic or aesthetic context in which the art “lived.”

CONCLUSION The impact of archaeology upon the study of rabbinic literature, including midrash, is indeed huge. Interpretation of specific texts in terms of archaeological sources, philological use of epigraphic sources to interpret the languages of rabbinic literature, and the discovery (or, “recovery”) of otherwise unknown midrashim and other rabbinic traditions were important factors in the development of rabbinics as an academic discipline during the twentieth century. I have provided a number of specific examples of ways that archeology has been integrated in order to illustrate specific aspects of this approach. I have also suggested that archaeology provides the “set” upon which many forms of rabbinic literature were performed in antiquity, both the text and the context being essentially linked. More than any particular literary hermeneutic, archaeology has “revolutionized” the study of rabbinic literature and of the world in which it was composed. For the first time since 56 “Early Synagogue and Jewish Catacomb Art and Its Relation to Christian Art.” In Aufstieg und Niedergang der Römische Welt. Berlin and New York: Walther de Gruyter. 2.21.2:1313-1342. For a similar analysis, see A. J. Wharton, Refiguring the Post Classical City: Dura Europos, Jerash, Jerusalem, and Ravenna (Cambridge: Cambridge University Press, 1995), 43-45. 57 Eg. Weiss and Netzer, Promise and Redemption.

ARCHEOLOGY AND RABBINIC LITERATURE

217

talmudic times it is possible not only to “read” rabbinic sources, but to “see” and “touch” them as well. The essential project, then, is building the hermeneutical bridges between archaeology and rabbinic texts that allow for their creative interaction without allowing either type of source to dominate the other. This requires depth knowledge of rabbinic literature and archaeology, of the problems inherent in the study of each separately, and of the even greater issues involved in discussing the two together.