226 76 3MB
English Pages 392 [386] Year 2022
The Open Past
The Open Past Subjectivity and Remembering in the Talmud
Sergey Dolgopolski
fordham university press new york 2013
Copyright © 2013 Fordham University Press All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means—electronic, mechanical, photocopy, recording, or any other— except for brief quotations in printed reviews, without the prior permission of the publisher. Fordham University Press has no responsibility for the persistence or accuracy of URLs for external or third-party Internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. Fordham University Press also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Library of Congress Cataloging-in-Publication Data is available from the publisher. Printed in the United States of America 15 14 13 5 4 3 2 1 First edition
For Lilia
contents
Preface
ix
introduction
1
part i. stakes
1. What Happens to Thinking?
21
2. Ego Cogito, Ego Meminí: I Think, Therefore I Remember
34
3. Through Talmud Criticism to the Talmud as Thought and Memory
44
part ii. who speaks? preamble: The Virtual Author
57
4. Thought and Memory in the Talmud: The Ambiguous Status of “The Author”—and Beyond
59
5. Human Existence in the Talmud: Thinking as Multiplicity and Heterogeneity
78
6. Sense in the Making: Hermeneutical Practices of the Babylonian Talmud
105 vii
viii Contents
part iii. who thinks? preamble: The Virtual Subject
131
7. Who Thinks in the Talmud?
134
8. The Hand of Augustine: Thought, Memory, and Performative Existence in the Talmud
158
part iv. who remembers? preamble: The Virtual
9. What Is the Sophist? Who Is the Rabbi?: The Virtual of Thinking
181 185
10. The Talmud as Film
212
conclusion
247
Appendix: Talmud Criticism, An Analytical Example: “Composer” versus “Redactors”: David Halivni’s and Shamma Friedman’s Competing Readings of Baba Metzi‘a 76ab
257
Notes
307
Bibliography
357
Acknowledgments
369
Index
371
preface
This book reclaims the originary power that the past exerts on the characters in the Talmud. To discern and explore that unusual power both in the late ancient text of the Talmud and in the broader context of connected disciplines of thinking and remembering, I engage competing—modern and ancient—notions of virtual agents, agencies, and of the virtual as such. That perspective helps reveal, in the final analysis, the title notion of the book, the open past. The term “virtual” is used too often, and “the open past” rarely. Therefore, before I proceed, it is important to begin with a preliminary disentanglement of the virtual from its more familiar but narrower manifestations in digital technology, of “the open past” from its homonym in analytical philosophy, and of the bordering notion of “the open future” in continental thought. I begin with the virtual. Far from being a merely technological phenomenon, the virtual characterizes certain ways of human engagement with existence in the world along with others. Shaped in medieval theology and extended through the modern to the postmodern to its most recent versions, including “digital virtual,” the dominant conceptualization of the virtual has been based on both philosophical and political urges to decide what is, what is not, and what only seems to be. In this context, the virtual has been conceived by medieval scholastics as what is not fully there in the real but nevertheless has effects in reality. Modern examples include not only computer games and hyperreality but also ideas of human existence as subject to moral responsibility rather than to natural law alone, as well as a related view that a human is a thinking subject, that is, a mind standing at the center of both thinking and volition. Ancient forms of the virtual differ from that model, however, and, by the logic of contrast, help explain the contemporary forms better. In that spirit, ix
x Preface
I approach the texts of the Babylonian Talmud. Placed in the context of rhetorical and philosophical schools of thinking in late ancient and early medieval periods, the texts of the Talmud reveal a rather different perspective on the virtual as a way of human existence, which in turn puts philosophical ideas of the virtual into a new light. By juxtaposing the Talmud, rhetoric, and philosophy as distinct and closely intertwined disciplines of thought and memory, I set out to explore these older forms of the virtual in an effort to better understand the currently dominant forms of the virtual. Despite these differences, there emerges an overarching notion of the virtual as a performative, which the book both highlights and analyzes in its particular instances. The notion of a performative, I argue, is foundational for the particular forms of the virtual, including modern ideas about transcendental subjectivity, film montage, and literary authorship and authority, as well as for ancient conceptualizations of the virtual in the Talmudic and philosophical texts. Once properly distanced from its narrower implementation in digital technology and instead understood as a performative, the virtual reveals its unique relation to time, thinking, and remembering, which I capture in the notion of the open past. Without getting ahead of myself, I can only say: Performatives in the Talmud imply the past as not only the content remembered but also as a power that puts one’s thinking and remembering into action. In that context, “the open past” differs from its homonym in analytical philosophy; it should also not be mistaken for the mirror image of the notion of the “open future” in continental thought. In the analytical tradition, the “open past”1 is a technical term referring to one’s temporary inability to tell exactly which of the descriptions of the past is accurate. That means the “open” characterizes only an extrinsic, temporary, property of the past, not its intrinsic feature. In continental philosophy, in contrast, the notion of the “open future” is far from narrowly technical. Rather, in that tradition, the notion of the “open future” is fundamental for understanding the nature of time. Life in time means imminence and thus implies having an always open future as opposed to any preorganized future, which, strictly speaking, would be only a deferred present. However, if, as that tradition has it, living in time means having an open future, what role remains for the past? If time originates from that relationship to the future, then the past can only be a fictitious beginning, a necessary phantom of a starting point, a retroactively generated chronological period of “before.” Advanced in philosophical thought of the last two centuries, this view of the past permeated the study on the Talmud
Preface xi
as well, resulting in the application of modern philosophical categories of the “thinking subject,” subjectivity, and time to the thinking displayed in the ancient texts of the Talmud. This book challenges that application, both in the specific context of studying the Talmud and in the broader context of the twentieth and twenty-first centuries’ traditions of thought.
Introduction
Zero and the “imaginary number” i in mathematics; paper money in economics; the vanishing point and the royal vista point in perspectivist painting; the free will of a person bound by the chain of causes and effects in the physical world as posited in theology; virtue as a power to practice what is known to be right (as opposed to only knowing what is right) in ethics—all these make up a short list of examples of the virtual. These examples entail agents, agencies, or instances that actively participate in the real world without being real in the same way in which other things in that world are. Historically, some of the above examples of the virtual came to life in the Middle Ages, well before computer gaming or digital consumption in general both dominated and shaped our ideas about virtual worlds, peoples, and lives. Other examples of virtual agencies hark back to even earlier times and are even more complex. Such older examples include the characters in philosophical dialogues, who are neither present on the stage nor absent, that is, present somewhere else in the world, but rather are not only discussed in the dialogue as its subject matter, but also take active, indeed leading parts in the conversation. 1
2 Introduction
The impersonal and in many cases personified placeholders in the arguments presented on the pages of the Talmud offer another example of such virtual agencies. These older types of virtual agency require study both in themselves and for the sake of better understanding new forms of the virtual: To understand digital virtuality critically is to understand and critically analyze the notion of the virtual in the first place, in its earlier historical manifestations, in general as a concept, and as a way of social and individual life, now and in the past. This broader interest in the notion of the virtual as such animates this book and defines a particular focus of my inquiry: the nature of virtual thinking and remembering beings in the Talmud in (dis)connection with and in the context of other historical disciplines of thinking that also developed and employed certain conceptions of virtual: the disciplines of philosophy and rhetoric. This informs the question of the role of the virtual in thinking and remembering that I ask in this book.
Agendas This book asks the question: How does modern thinking differ from historical methods of thinking, particularly within the Jewish tradition? Placing the Talmud, the core text of that tradition, in the context of late ancient philosophy and rhetoric allows us to see the thinking process in the Talmud quite differently from the way it is commonly viewed. In late ancient schools of philosophy and rhetoric, persons were not seen as the origin or cause of thinking; rather, they were believed only to participate in thinking as replaceable and interchangeable agents. However, philosophy and theology in the Middle Ages promoted a new point of view in which thinking was seen as originating from and being done by a person, human or divine, individual or collective. It was no longer taken to be an effect of a universal, impersonal, immobile faculty, which late ancient philosophers called “intellect.” The new, personality-centered view of thinking was epitomized in the newly coined intellectual idiom of a “thinking subject” as simultaneously the origin, cause, foundation, and center of thinking, as well as the agent who thinks.1 From medieval philosophy and theology, this intellectual idiom made its way into modern secular philosophy and political theory. Among the broader implications of this shift, one implication is that it has affected our
Introduction 3
efforts to understand what kind of text the Talmud is. Examining that question not only helps us avoid the impasses of the twentieth-century theories of the Talmud’s redaction but also offers new insight into a discussion of the relationship between thinking and remembering in general and, in particular, into the theory of both historical and contemporary forms of new media that are meant to enhance our ability to remember.2 The Talmud provides an important counterexample to how our ability to remember might be conceived and carried out in practice. Influenced by person-centered ideas about the nature of thinking, twentieth-century scholars of the Talmud attempted to explain the genesis of the text of the Talmud by a hidden activity of such a hybrid agent as the “thinking subject” (individual or collective), whom these scholars construed as either the composer3 of the Talmud’s text or as a redactor of the sayings of the earlier masters in the Talmud. In addition, the medieval philosophical notion of the “thinking subject” affected modern scholars of the Talmud in how they constructed the Talmud as an object of critical analysis. These scholars uncritically followed medieval, philosophy-driven pejorative constructions of the literary form of the Talmud as predominantly a series of rhetorical attacks and defenses, controlled by a rationally thinking subject.4 The thinking subject can dismiss these “lower” forms of rationality, such as storytelling (haggadah) and the rhetoric of the legal debate (halakhah), in favor of the philosophically and theologically “higher” forms of rationality, as Maimonides proposed and accomplished in his Mishneh Torah.5 Alternatively, the notion of the thinking subject can explain the genesis of these rhetorical forms by reference to a “redactor” who created such forms according to a rational standard of infallible arguments, as Shamma Friedman and David W. Halivni, contemporary text-critical scholars of the Talmud, have implied. However, the recent work of Alain de Libera shows the inapplicability of the modern notion of the “thinking subject” to the period of late antiquity.6 What follows both highlights and undoes the role of the medieval and modern notions of the thinking subject in understanding the late ancient text of the Talmud. I explore a new approach to thinking in the Talmud by viewing it within the context of the philosophical and rhetorical disciplines of thinking in late antiquity that would have been prevalent when the Talmud was developed. I lay bare this context by both applying and renegotiating recent developments in philosophical and rhetorical theories of thinking, subjectivity, and remembering.
4 Introduction
If placed in the context of late ancient views, the thinking processes in the Talmud emerge in a very different light. Persons in the Talmud (both the characters mentioned in the text and the composers of that text) no longer stand for either the origin or the cause of the thinking of which the Talmud is a document. Instead of being centered on the personalities of either characters or redactor(s), the process of thinking in the Talmud is decentered as a weblike and therefore apersonal and heterogeneous process. One version of that process is what the ancient philosophers described as “intellect.”7 Another version of the same process is what is documented in the text of the Talmud. In the Talmudic process, persons are no more and no less than agents and placeholders for textual traditions and for arguments about texts. However, they are not person-centered thinkers mediating their thoughts in the text. The complexity of factors involved in the thinking process in the Talmud results in the particular literary form of that document. In the Talmud, rhetorical refutations of the texts of the past, for example of the Mishnah, the third-century set of instructions for rabbinical courts, are followed by defenses of its formulations, which in turn lead to new discoveries in its content, all of which forms a complex dance of thinking in which no person stands at the center and no one is positioned as an originator of the dance or as a subject who autonomously thinks in the text. Instead, the application of all these rhetorical techniques in the Talmud is driven by the overarching task of remembering the authoritative texts of the past,8 such as the texts of the Mishnah as well as of earlier post-Mishnaic authorities, by means of first attacking and ultimately defending the formulations of these texts in as many ways as needed in order to protect the texts against any possible threats of oblivion, including any attacks or doubts concerning the reliability of memory as a medium by which these texts are delivered. This approach to memory was novel in the context of the rhetorical and philosophical schools in late ancient times, and is even more novel today. It is an apersonal, memory-oriented (anamnestic) approach to thinking.
Approaches This work marks a continuation, a rupture, and a new beginning. As a continuation, it builds on my analysis of the distinction between the Talmud as a
Introduction 5
text—either an object of critical scholarship or a traditional source of applied legalistic decision making (halakhah)—and “Talmud” (without the “the”) as an intellectual discipline in which to live a life of thinking.9 As a rupture, it captures the break in my own understanding of Talmud. Whereas in my earlier work I understood the Talmud as a text in late antiquity and “Talmud” as a discipline in the medieval/modern period, I now see that in the earlier period, (the) Talmud represented both a text and a discipline. “Rupture” also refers to the break in continuity between the Talmudic practices of late antiquity and those of the medieval/modern scholars, despite the latter’s belief that they have been directly continuing the disciplinary practices exemplified in Talmudic text. As a new beginning, this book is an attempt to show both how the Talmud as a late ancient text can be fruitfully explored by incorporating competing philosophical and rhetorical approaches, and how the philosophical and rhetorical disciplines can be illuminated by placing them in the context of the Talmud. The common ground on which these three disciplines—rhetoric, philosophy, and Talmud—meet each other is in concepts and practices of thinking and remembering, a common ground that each of these disciplines approaches in different ways. Each particular disciplinary perspective illuminates the otherwise unrevealed sides of the others. For example, in the philosophical traditions of Platonism, memory, that is, remembering ideas, is the main path to thinking, knowing the truth, and being wise. In the tradition of philosophical dialogue, in an intellectual conversation between interlocutors about a given matter, one of them helps the other recall and thereby cognize the truth of what is discussed. In this way, the interlocutors apply dialectical argumentation in order to cognize what the matter of the discussion truly is. In this view, memory is therefore a property of finite, time-bound beings by which they can cognize the eternal ideas of things. It thus helps the interlocutors transcend the finitude of their time-bound existence by taking them to the realm of eternal truths or ideas of the things. In the tradition of philosophical dialogue, memory thus serves cognition, rather than cognition serving memory. In contrast, the rhetorical traditions emphasize persuasion through speaking. In the rhetorical traditions, the art of memory, mnemonics, is only one among other technical skills that a successful orator needs, and the role of memory thus shrinks accordingly, because rhetoric differentiates the techniques of memory from the other, more argumentative rhetorical skills, such as refutation and invention. In rhetoric, memory is understood
6 Introduction
to be a technique for managing—and manipulating—data, such as circumstances, claims, reports, or any other content an orator needs to remember for speaking. It is not memory, but the practices of refuting and inventing that involve thinking and cognizing, using the techniques of remembering only to collect and manage data. Mnemonics helps to organize that data by associations and then to retrieve them from memory as needed for speaking. Unlike the traditions of philosophical dialogue, rhetoric thus weakens the connection between memory, on the one hand, and thinking—or cognizing the truth—on the other. When juxtaposed with the claims of philosophy, rhetoric’s distinction between memory, which is subject to distortion, and the other necessary skills reveals the potential vulnerability of the philosophical view that memory is the path to thinking and knowing the truth. Rhetoric helps us understand that what a philosopher thinks is authentic memory may in fact be a result of manipulation—not to mention simple forgetting. In turn, philosophy argues against the reduction of memory to a collection of data in mnemonics by showing that the data need to be understood and interpreted before they can be remembered. In rhetoric, this process of understanding and interpreting—cognition—remains unexplained. As a result, both the rhetorical and philosophical views of memory are insufficient in themselves; they complement and illuminate each other as paths of seeking knowledge and wisdom. These two disciplines can be even further illuminated by bringing in an additional discipline, one that programmatically subordinates thinking, including the rhetorical processes of refuting and inventing, to the service of memory. That is the discipline of Talmud as an intellectual project. In contrast with the disciplines of philosophy and rhetoric, Talmud links cognition to memory in a new way, using cognition in the service of memory, rather than placing memory at the service of cognition.10 That move separates Talmud from both philosophy and rhetoric, because philosophy connects cognition/thinking to memory in a different way than Talmud, and rhetoric loosens that connection altogether.
a In contrast to these other two disciplines, Talmud (without the definite article) as an intellectual discipline applies rhetorical techniques of refutation
Introduction 7
and invention for the purposes of discovering new content in the traditions of the past. Refutation puts the text11 of the tradition into question, and invention allows more refined and thus more precise understandings of that text to be developed. The practitioners of the art of Talmud discover/invent—and thereby remember—the new content in the textual tradition they receive from the past. More specifically, refuting the language of a textual tradition and defending that tradition by means of inventing its meaning anew leads, as it does in “the” Talmud, to generating new understandings of the arguments of the authorities of the past. By the same token, this new understanding amounts to remembering these traditions and their content more clearly. The practitioners of the art of Talmud discover such new content by critically analyzing the received textual traditions of the past in the series of conversations in which they refute and reinvent those traditions and thereby verify their validity—precisely by attempting to refute them and by showing that these refutations fail. In that way, the practitioners of Talmud come to remember/understand the content of the recorded traditions with greater precision. However, unlike the philosophy of Platonism, which understands truth as eternal (and thereby timeless), Talmud locates truth in the past to be recalled through refutation and invention, rather than as something located in eternity to be understood through dialogue. As I will soon explain, this difference also results in a different understanding of the past. In particular, locating truth in the past frees one’s thinking about that past from the otherwise rigid opposition between time and eternity, in contrast to the attitude toward the past, time, and eternity that philosophy has developed. While philosophy locates truth in eternity and the past in time, Talmud locates truth in the past, and thereby frees that past from the rigid opposition between what is time-bound and what is eternal. These complex relationships between the traditions of philosophical dialogue about eternal truth, the rhetorical persuasion of an opponent, and Talmudic memory of the traditions of the past have yet another important dimension. That has to do with a question of “who.” It concerns anthropology, in the strictest sense of the term: In particular, it concerns understanding the role of memory in construing the human being and his or her intellectual life. Platonic philosophy generally sees memory as an indispensable—albeit intermediate—step by which time-bound beings such as humans cognize eternal truth. These time-bound beings recall eternal ideas in the course of thinking as they dialectically argue one with another. In this construction
8 Introduction
of a philosophical dialogue, the human ability to recall is both a remedy for human finitude and for human time-boundedness and a condition of possibility by which finite humans who live in time can cognize the timeless, eternal world of ideas. In sharp contrast to that, the art of Talmud approaches memory and remembering as a goal in itself, rather than as a means to a different end. This adds yet another reason why the art of Talmud no longer associates memory with eternity, but rather with looking into the past, and in particular into the textual recorded traditions that came therefrom.12 The radical difference between the philosophical linkage of memory with eternity and the Talmudic connection of memory with the past has at least one implication for the conception of truth and its criteria in each of these disciplines. In philosophy, by definition, there can be only one eternal truth (or more precisely, one set of truths) about each subject matter. In contrast, because the art of Talmud locates truth in the past, rather than in eternity, this art affords the possibility of there being multiple ways in which a tradition of the past can be true. That also means multiple venues in which the past reveals its authenticity before the inquirer. The availability of multiple truths means that the truth about the past no longer has to be singular or exclusive. Instead, a textual tradition that comes from the past might be a true tradition precisely because it delivers several possible truths about that past while representing all of them and making each of them compensate for the potential vulnerability of the others.13 Using refutations and inventions, a practitioner of the art of Talmud can find more than one way in which a given tradition is true or more than one version of truth that a particular tradition conveys. Truth thus no longer has to be singular and, as already said, it need not be located in eternity, either. What this means, however, is that the truth also no longer has to be confined to the comprehension of the single mind of an inquirer, either individual or collective. Instead, the art of Talmud places the multiplicity of truth in the past and discovers it through intrinsically collective questioning, refuting, and reinventing the memory of that past. In this process, persons serve as placeholders for differences, rather than centers or independent agents of thinking. By applying the techniques of rhetorical refutation and invention, the art of Talmud reconfirms the validity of the memory of the textual traditions that come from the past, such as the Mishnah and other texts of earlier rabbinic authorities. In the process of remembering these traditions, the art of Talmud also expands the content of the memory of these traditions further. This is exactly why the art of
Introduction 9
Talmud disregards eternity—the ultimate location of truth for philosophers, in which there can only be one true idea or one true set of ideas involved in a given matter. In contrast, the art of Talmud concentrates on the past as the place for multiple (even if potentially mutually exclusive) truths, which collectively confirm the truth of tradition. Let’s pause for a moment to highlight this important difference between Talmud’s and philosophy’s understandings of the past. Unlike the philosophical view of memory, which links the past to eternity by making the former closed to any change, the art of Talmud takes another approach: Perhaps for the first time, it gives the past an independent value and appreciation in which the past is separate from eternity. In fact, Talmudic understanding of the past has nothing to do with the concept of time at all,14 at least as long as time is understood along the lines of Plato’s Timaeus as “the moving image of eternity.” Instead, the past is considered to be open and allowing for multiple interpretations, both mutually exclusive and complementary.
a I drew the above juxtapositions of the three arts—the philosophical art of dialogue, the rhetorical art of memory, and Talmudic art of argumentation— along three main lines: their respective conceptions of truth(s), the roles of memory in cognition, and the relationships between memory and argumentation. In the light of these linkages, Talmud is both strongly similar and strongly dissimilar to the traditions of philosophy and rhetoric. In sum, while philosophy utilizes memory to cognize eternally true ideas, thereby avoiding the problems of time and memory, Talmud considers remembering fundamental for cognizing multiple truths, and while rhetoric separates the skills of remembering from those of refuting and inventing, Talmud uses both of the latter for the tasks of the former. These differences between the philosophical, rhetorical, and Talmudic approaches to memory lead to mutual illuminations. Philosophy, with its valuation of memory, indicates the cognitive limitations of the rhetorical instrumentalist approach to remembering. Rhetoric reveals the vulnerability of philosophy’s theory of cognizing exclusivist truth in each subject matter by remembering the idea involved in it. Talmud, by using rhetorical forms of persuasion to verify the memory of traditions, highlights a third possibility, that of cognizing multiple truths in the same tradition.
10 Introduction
In addition to the above differences between these three disciplines, we can see how the variety—and complexity—of the relationships between thinking and remembering shapes both the characters and the performative existence of the Talmudic masters and students. I am using the term “character” here in two distinct ways. The first refers to those characteristics of mind and behavior that are cultivated and described in the classic texts of each discipline. The second and more important meaning refers to the human existence embedded in the text, but not described in it, just as a photographer’s gaze is not described in a photo, but nevertheless is embedded in it. Indeed, there is no photograph without such a gaze,15 and similarly, there is no text without a human existence embedded in it, even though that existence is not immediately explicit. This second, purely performative dimension of character in the Talmud is both subtle and absolutely crucial for understanding the question of thinking and remembering in this text. The first meaning of character and of its existence is conveyed through activities of the listed characters of the masters and students, named or nameless. For example, a Socratic philosopher exists as a philosopher as long as he or she performs as a philosopher in speeches and dialogues. Performing as such a philosopher requires eliciting wisdom and knowledge from another person while at the same time ironically positioning oneself as ignorant. In a similar sense, yet in contrast to the philosopher, the rhetorician performs— and thereby exists—as a rhetorician in speeches and disputes by persuading others about the ideas he or she already has in mind. In turn, a Talmudic sage performs and thereby exists by way of working from inside the arguments presented to him or her by others; the sage either refutes these arguments or defends them against a refutation presented by other sages, students, or even outsiders. Unlike either the rhetorician or the philosopher, either Socratic dialectician or Aristotelian doctrine-teacher, in interacting with the others, the performative existence of a Talmudic sage involves neither dialectical eliciting nor promoting a positive doctrine or knowledge on the part of the sage. Most important, these differences in character are not always explicitly spelled out in the texts of philosophical, rhetorical, or Talmudic classics. Rather, they are indicated on the level of the performance of the characters in the texts. This is why in comparing the disciplines, I need to ask about the different forms of personal existence performed in them, rather than exclusively about the different sets of character traits described or even implied in these texts.
Introduction 11
The second meaning of character is more complex because it implies a performed existence embedded in the text but not explicitly that of a character or characters—whether object(s) or subject(s)—described therein. Whereas the notion of human existence in the first sense is not only performative but also descriptive, the notion of human existence in the second sense is completely nondescriptive and only performative.16 As I suggested above, human existence can be seen as embedded in a text in the same way that it is embedded in a photograph in addition to what is represented in the photograph; the photographer’s gaze is captured by the frame, even though it does not appear as an object in the photograph. It is not an unrepresented, but potentially presentable, subject of the process of photography. No additional details about either the body or the mind of the photographer would be relevant for understanding the photograph as such. The gaze of the photographer is both captured in the frame and at the same time created by the framing. This gaze (call it either personal or impersonal) is thus an example of an embodied, but purely performative, existence in the sense in which I am using this term, because when the photographer, her attire, her feelings, including her presence, and more specifically, her absence in the photograph are bracketed out of consideration, what remains is a simple gaze, something without which this picture and this frame would no longer be possible. The gaze may be very simple. It may not even relate to itself or show the self- aware presence of the photographer, and it does not even have to include a realization of the being of the objects in the frame. Just as the picture might be empty, the gaze may stare nowhere, as if it was not a photographer, but a camera that took the picture. And yet, such a blind look would still be a gaze,17 even without an actual human presence being ascribed there. Neither an object in the picture nor a subject in front of it, the gaze that is embedded in the picture paradoxically has an even greater degree of existence than what is portrayed in the picture. The latter may or may not have any biographical equivalent in the reality outside of the picture, but the gaze is undeniably there. It exists as realistically as the picture. In fact, this gaze is the least noticeable and the most real of all elements in that picture: It must exist as long as the picture does, whereas all other elements of the picture may exist or may have existed but do not have to exist or to have existed either at all or in that particular posture before the picture was taken. The gaze thus performatively exists and does so much more strongly and more immediately or more “simply”
12 Introduction
than any subject or object described in the picture might exist. Its being there and its representation of being there are the same—one and the same act of performance. This example of the gaze in the photograph as a performative existence is, of course, limited to the area of the visual, but performative existence is broader than that. What the example of the photographer’s gaze illustrates is only one kind of existence: existence as an unobjectifiable presence—or, if you prefer, absence. In more general terms, the presence (or absence) of the photographer in the gaze epitomized in the frame is only one case of what can be more broadly called a personal—even if, perhaps, nameless— performative way of being there.18 Such a way of being is a more general notion of what that gaze is, and its existence is an elusive definition of the performative action that it performs. The gaze’s act is precisely its being there. It is and exists, in the active sense of the verbs “to be” and “to exist,” that is, it performs, regardless of the physicality of the photographer, and of course regardless of his or her name.19 By extension, the same applies not only to gazing, but also to thinking, speaking, and remembering. They are performative actions, and as such can be generically called “existentials,” because, unlike the more generally known category of existing subjects and objects, these existentials are epitomized, performed, or constructed in texts, while not necessary objectified in these texts as directly described subjects or objects. These actions or existentials are embodied in the text regardless of whether the agents of those actions have been present or not and regardless of whether or not these agents ever existed in biographical or historical terms—regardless of whether there can even be or have been a photograph picturing them as objects. In sum, existentials are performed in texts, but not necessarily objectified in these texts. In this respect, one difference between the Talmud and a photograph is that in the Talmud, there are many more than just one such existential. Many of the named or anonymous characters in the text of the Talmud also perform as such existentials and need to be appreciated as such. Discovering and describing such existentials in the Talmud takes the question of “who” to a different level. To use the example of the photograph one more time, bracketing either the absence or the presence of the photographer opens up a new vista on the being of his or her gaze, or, in other words, reveals the performative existence of the photographer as a gazing being, an existential, who is not as an object or subject, but rather as an action.20 By
Introduction 13
extension, bracketing the presence or absence of the agent of speaking, or of thinking, or of remembering opens up the question of the nature of the Talmudic, philosophical, and rhetorical equivalents of the gaze—opens for analysis the nature of the existentials or performative existences of speaking, thinking, or remembering beings that are embedded in texts and that at the same time make these texts possible. Like the gaze, the existence of such beings is embodied or epitomized in the classical texts of each of the disciplines and is different in each of the traditions, if not in each of its texts. The variety of existentials depends on how the relationship between thinking and remembering is configured in each text. Different configurations underline different ways of performative existence. My more general interest is thus in the relationship between thinking and remembering as a factor shaping different ways of performative existences (both individual and collective) in these traditions. This more general interest animates my particular interests in the late ancient discipline of Talmud in its relationship to the disciplines and traditions of philosophy and rhetoric. My initial question is thus hermeneutical: What can understanding Talmud as an intellectual discipline in the context of other intellectual disciplines teach us about thinking and remembering beings in the sense of existentials—either personal or impersonal—both in the Talmud as a late ancient text and, by extension from that example, in general as an interpretive principle? This question entails a new path of inquiry. I call that path the mutual hermeneutics of the philosophical, rhetorical, and Talmudic traditions and practices of thinking, remembering, and existing. The hermeneutical method by definition proceeds from what is clearer to what is more obscure. In this case, it also means moving archaeologically or backward in time from medieval/modern understandings of the role of thinking and memory in general and in Talmud in particular to late ancient philosophical, rhetorical, and Talmudic practices of thinking and remembering. However, in the case of Talmud, such a backward hermeneutical-archaeological movement runs into a rupture in discourses, a rupture that those who proceed genealogically or forward in time cannot see. For example, as de Libera’s work will help us understand, transiting to the late ancient disciplines of thinking reveals an irreparable mismatch in the conceptual frames of thinking and remembering. I take this mismatch not as an obstacle to be overcome, but rather as a symptom to read, and, more precisely, as a heuristic tool for analyzing, what is more complex, obscure, but even more important
14 Introduction
about thinking and remembering in the Talmud—the performative thinking and remembering being, that is, the existential(s) in the late ancient text of the Talmud. By way of the hermeneutics of mutual elucidation, I draw on the scholarship of thinking and memory to discern the construction of those thinking and remembering existentials by contrasting the variety of Talmudic and other practices of thinking and memory in late antiquity to their (dis)continuations through the Middle Ages and modernity. The contrast between diverse late ancient practices of thinking and remembering and their medieval and modern transformations newly illuminates contemporary, technology- driven, and, in particular, speed-driven ways of thinking and remembering. In turn, using the conceptual apparatus developed by analyzing technology (both ancient and modern) helps us understand better the Talmudic practices of thinking and remembering. This method of hermeneutical contrast also affords a better understanding of the intellectual practices in the information society, a society in which information threatens to replace thinking and data management to replace memory. In the new performative practices of existence of the digital age, thinking and remembering face new challenges and thus require a new understanding of the centuries-old practices from which they have derived their power and their importance for humans in search for a way to think, be, or live in the world.21
Structure My argument is animated by constant attention to symptomatic differences between the explicit questions that the scholars of the Talmud have asked and the implicit questions that they answered. Highlighting this difference leads to a more precise understanding of the questions and hence advances the analysis of performative existence in the Talmud further. Following that approach, the inquiry progresses by asking and answering questions that shape the three major parts of the book: The question “Who speaks?” leads to an answer about thinking; in turn, answering the question “Who thinks?” leads to the question “Who remembers?” and to memory as a mode of existence in the world, a mode of existence that the Talmud performs. The argument in this book is intrinsically interdisciplinary. It involves the studies in theory and history of rhetoric, in philosophical anthropology,
Introduction 15
and branches of the historical-literary study of the Talmud, as well as studies in both new and historically new media, all of which mutually engage and illuminate each other. At the intersection of these different and otherwise only remotely connected disciplines stand the notions of thinking and memory. I address these notions in two introductory chapters, “Ego Cogito, Ego Meminí” and “What Happens to Thinking?” which together introduce that broader context of the inquiry, in preparation for exploring the specific path of analysis that goes, as the title of the third chapter suggests, “through Talmud criticism to the Talmud as thought and memory”—a road explored in the three main parts of the book that follow. These parts address three closely connected components of Talmudic composition: speaking, thinking, and memory. Each part particularly focuses on one of the components, with two others constantly remaining in the background of analysis. Following the introductory chapters, the second part, “Who Speaks?” concentrates on speaking and speakers in the Talmud. It deals with the question of the authorship and redaction of the Talmud that text-critical scholars have asked about the text. More specifically, it focuses on the authors of the Talmud in relationship to other characters who speak therein. This part reveals an implicit question of thinking behind the explicit question of redaction: the question of who thinks in the Talmud. Scholars of the Talmud have answered that question in response to and—in a sense—instead of the explicit question of who authored or redacted the text of the Talmud. Revealing the implicit question of thinking in the Talmud, the discussion in this part both uses and reflects on the methods of literary and historical analysis. This lens of reading discerns both the process and the protocol of sense making in the text of the Talmud as the characters in the Talmud read and arrive at understandings of the Mishnah. However, even if the author(s) of the text and other characters therein collectively contribute to the process of making sense of the Mishnah, the protocol of sense making in the Talmud is impersonal and remains beyond the control of the characters. As we will see, this takes the analysis to the realm of logical and rhetorical inquiry about thinking. Part III, “Who Thinks?” focuses on thinking process in the Talmud. It directly addresses the hitherto implicit question of Talmudic scholarship of who thinks in the Talmud by reformulating the results of text-critical scholarship in the light of its own implicit question of thinking and by engaging the broader resources of the philosophical and rhetorical traditions as they analyze thinking processes. In this part of the book, I thus both highlight the
16 Introduction
impersonal nature of thinking that contributes to making sense of the Mishnah and address the question of the thinking and thinkers involved in that impersonal process. In terms of method, understanding the process of thinking requires supplementing the methods of logical and rhetorical analysis with instruments from philosophy and the phenomenology of thinking. In the light of these disciplines, when analyzing the protocols of sense making in the Talmud, I particularly focus on the relationship between memory and thinking in the exchanges of refutation and defense, as well as in larger compositional units of the text. The task of understanding the relationship between remembering and thinking involves a comparative component of research. I therefore highlight both parallels and contrasts in how the relationships between remembering, thinking, and refuting-defending are configured in the discipline of Talmud versus the rhetorical and philosophical traditions and disciplines of thinking and remembering. This part of the discussion brings to the foreground what was in the background of the questions of both authorship and thinking, which is the question of memory and remembering as the foundation of the thinking processes in the Talmud. Part IV, “Who Remembers?” changes the emphasis from speaking and thinking in the Talmud to the performance of remembering, which, as I argue, is a ground from which both speaking and thinking stem. I advance understanding of the performative of remembering in the Talmud by using comparative tools of analysis. To that end, I focus on the modes of existence that the Talmud discerns in the figure of the rabbi in comparison with how the Platonic tradition understands the existence of the sophist, and this leads to an inquiry into the impersonal foundations of remembering as a way of existence. This part heuristically juxtaposes post-Heideggerian understandings of human existence in the world, of which his notion of Dasein (as being in the world open to its finitude, as opposed to mere functioning as a part of an environment) is an example, with the structure of existence constructed in and by the discourses of the Talmud understood in the broader context of the conflicting philosophical and rhetorical disciplines of thinking and remembering. Returning to the issues raised in the introductory chapters in Part I, I first address the contemporary context of the problem of subject positioning in human thinking in relationship to technological innovation, which I now explore through a juxtaposition of the ontological foundations of redaction theory of the Talmud with the ontological foundations of Sergei Eisenstein’s theory of film montage. Using this example, I argue that
Introduction 17
understanding thinking processes in the Talmud is an intrinsic part of the broader problem of human existence in the world of technology. I consequently ask how the problem of thinking and remembering in the Talmud opens up the broader question of what text-critical, literary, historical, and rhetorical-philosophical analysis of the thinking and remembering in the Talmud can teach, by way of both comparison and contrast, about more recent—medieval and modern—views of the position of the subject in the relation between thinking and remembering. This structure represents a step-by-step broadening of the disciplinary frame of analysis from the literary and historical analysis of authors and other characters in the text of the Talmud, to the logical and rhetorical analysis of particular performatives, and procedures of making sense, to a broader inquiry into the relationship between remembering and thinking in the process of performance. Of course, to a different degree, Parts II, III, and IV of the book involve all the three elements of analysis—the literary-historical analysis of characters, the logical-rhetorical analysis of sense making, and, for lack of a better term, the rhetorical-ontological analysis of the configuration of the relationship between remembering and thinking in the Talmud, both in comparison with and in distinction from other rhetorical and philosophical disciplines and traditions. Yet the overall change of emphasis proceeds from the literary analysis of characters, to the logical analysis of sense making, to the rhetorical-ontological analysis of thinking in relation to memory and of the thinking being in relation to the performance, and to the performatives, of remembering in the Talmud.
part one
Stakes
one
What Happens to Thinking?
What does it mean “to think” in the age of technology? As unexpected as it may sound, this question—as I will show in this chapter—is highly relevant for analyzing thinking processes in the Talmud. In the first three chapters, I therefore introduce, in an inevitably general form, a broader background and the stakes of my inquiry shaped by three questions: Who speaks? Who thinks? And who remembers in the Talmud? To begin, we need to ask what thinking has to do with the “age of technology” in the first place. Did not thinking exist prior to the technological innovations of the modern age? Should not the question about thinking therefore be asked in its own terms first? Should not analyzing thinking in the Talmud be preceded by analyzing the question of what thinking is in general rather than by asking a seemingly narrower question about thinking in relation to technology? In an approach to thinking shaped by the thought of Martin Heidegger, as well as by the traditions of scholarship thereafter, the answer to these questions 21
22 Stakes
is negative. For Heidegger and for many of the other contemporary thinkers he influenced or provoked, understanding what thinking entails in general is impossible without both linking and contrasting thinking to technological innovation in particular or rather to what Heidegger calls “the essence of technology.”1 The essence of technology, he contends, is nothing technical. Instead, it has to do with a certain kind of relationship between human beings and the world around them. This relationship is informed by making things “readily available” for humans to use and consume, so that the process of consumption is given in advance. This process of consumption does not require thinking about things, much less understanding them, and even much less asking about their being: whether they really are or only seem to be. Consumption is thus not predicated on the existence or nonexistence of what is consumed.2 An early example of technology understood in this sense is alphabetical writing, which allows transmitting, repeating, and even implementing recorded knowledge without the necessity of understanding the original thinking involved. For example, when delivered in writing, military orders need to be fulfilled, but the thinking of the commander does not have to be understood. This way of responding may apply to oral command as well: It is to be fulfilled, even if the thinking of the commander is not known. In a Heideggerian perspective, this ability to implement and even to follow a thought without comprehensive understanding of that thought is precisely what the essence of technology helps achieve. Such a notion of the essence of technology associates the beginning of the technological age not with the time of either the industrial or the information revolutions but rather as early as in the Greek civic society of Plato’s time. In the philosophy of Socrates, (post-)Heideggerian philosophy finds the first critique of technology,3 as well as the first understanding that there is a related problem with human thinking. For example, in Plato’s Phaedrus, using writing as tekhnê, a technological memory aid that enables a free citizen to deliver a speech without fully understanding the original thinking behind that speech, leads to a situation in which the speaker, Phaedrus, not only fails to understand the thinking but also needs to be reminded by Socrates that he does not understand it. Otherwise, he forgets. He tacitly appropriates the speech of the sophist Lysias, whose words he forgetfully uses as if they were his own thoughts. Instead of understanding the thoughts, he substitutes words for thoughts and thinks that he does real thinking.
What Happens to Thinking? 23
Read in a Heideggerian perspective, Plato’s Phaedrus is an example of negative effects that such a simple technology as letters on paper in ink can produce. Phaedrus forgets that he did not understand the speech of the sophist Lysias. More specifically, he confuses remembering and interiorizing the words of that speech with both understanding and appropriating the thinking behind it. Remembering the words from the paper helped him mistake the memory of words for an understanding of the thoughts and thus to forget about thinking. By asking critical questions and showing young Phaedrus that he actually does not understand the claims he forgetfully considers his own, Socrates lays that forgetting bare. Socrates thus shows that Phaedrus’s ways of thinking about the speech and about himself were predicated on forgetting the thought behind remembering the words. Memorizing the words of the speech created the false effect of understanding the thoughts they convey. As a result, not only has Phaedrus forgotten that the speech is not his, but, by memorizing the words, he falsely thinks that he also has understood them. As a result, he even has forgotten his own forgetting. This example illustrates Heidegger’s idea that at the very beginning of the implementation of technology—in that case, of the transmission of speech through alphabetical writing—thinking was already closely intertwined with memory. Let’s use the same example to illustrate some important intricacies of the connection between thinking and memory in technological transmission of information and knowledge. Phaedrus forgets about the original thought of Lysias, which he perhaps never understood, and forgets about that forgetting. In this example of technological transmission, the original thinking is neither preserved nor required. Rather, it is forgotten, and that forgetfulness is forgotten, as well. Phaedrus remembers the words instead of understanding the original thinking in them. By the same token, he confuses remembering with understanding. In fact, his confusion reaches so far that even when his understanding is challenged by questions, he feels lost, but feeling lost nevertheless does not help him understand, much less remember, his own forgetting of the original thought (call it Lysias’s), the thought that he did not understand and thus did not remember. Substituting understanding of thinking by memorizing the words represents the second degree of forgetting. Having memorized a written copy of the speech of Lysias, not only had Phaedrus forgotten that the speech was not his, but, even when unable to answer questions about it, he still failed to recognize the forgetting of what he never understood—the original thinking
24 Stakes
in the speech. The second degree of forgetting achieved through substituting memory for understanding and understanding for memory is an example of the connections between technology, forgetting, and thinking. What is more, the second degree of forgetting highlights the ontological or, more broadly, as I will soon argue, the performative, rather than the purely epistemological status of thinking in its relation to a person who might think about himself or herself as an original thinker, but in fact performs as a consumer of words. What is even more, the second degree of forgetting is ontologically irreversible. Since it is no longer transparent for the understanding and knowing of the one who forgets, such as young Phaedrus, another person, such as Socrates, is required to reveal the forgetting, even if no one can recover what has been forgotten, which has never been understood. Such irreversible and nontransparent forgetting characterizes the way in which Phaedrus performs and therefore is. Forgetting himself in consuming the words of Lysias, and mistaking these words for thoughts that he, Phaedrus, thinks, is what uniquely defines his standing, and his being in the world. Even if based on forgetting, Phaedrus’s being there is nevertheless very powerful. Even Socrates cannot shatter it thoroughly enough to get to what exactly young Phaedrus had forgotten, if there even was a consistent thought of Lysias to forget. One might simplistically think that memorizing the writing makes Phae drus forgetfully mistake himself for Lysias. However, Phaedrus performs that mistake as strength rather than weakness. The power of Phaedrus’s being is that he is independent of Lysias. Phaedrus continues to perform the second degree of forgetting precisely because of and despite Socrates’ irony. Through asking questions, Socrates shows that Phaedrus must have either forgotten Lysias’s teachings, or that there was nothing consistent in these teachings in the first place. Yet which one is right remains programmatically undefined. What it means, however, is that Phaedrus performs the forgetting in the second degree, regardless of whether the words of Lysias contained a coherent thought. Socrates initially does not claim any knowledge of his own, and at the end, finds no such knowledge in the worlds of Lysias either, but Phaedrus continues to perform double forgetting of knowledge, which Socrates cannot find. Phaedrus, in other words, performs double forgetting regardless of what exactly has been forgotten, or whether or not there was something to forget in the first place. Performed in a technological setting, the double forgetting is thus not only irreversible but also independent of the being or nonbeing of what is forgotten.
What Happens to Thinking? 25
Similarly, Heidegger conducts his inquiry into the theory of being as the irreversibly forgotten foundation of both the being of the world and of one’s being therein. Dissimilarly, Heidegger assumes the forgotten being, as if there had surely been Being rather than Nonbeing to forget. Even if he highlights the impossibility of recovering from the second degree of forgetting, he has a presupposition regarding what exactly has been forgotten: that what is forgotten is there, and is therefore being rather than nonbeing. This also explains why, for Heidegger, Phaedrus is rather than performs. The explanation works best in terms of “existence,” a term that I will temporarily use for “being,” and of “existent,” which I will use for a particular being, a human being. Because there is existence to be forgotten, there should be an existent to forget it: Phaedrus, as a human being uniquely situated between Lysias, on the one hand, and Socrates, on the other. What this means, however, is that what I have more generally described as Phaedrus’s performance of double forgetting in the circumstance of technology is defined more narrowly as Phaedrus’s way of being, his unique existence, as irreducible to those of Socrates and Lysias as it is. In a more general view, Phaedrus’s double forgetting represents an undeniably strong performance; and, in Heidegger’s interpretation, that strong performance means only a strong way of Phaedrus’s being there, his Dasein. By implication, for Heidegger, since the original thinking may be occluded in the second degree of forgetting, and since that occlusion is irreversible, it has an ontological status: A performer of double forgetting is a being, who is absolutely irreducible to either Socrates or Lysias. Both memory and forgetting are no longer purely epistemological and psychological factors, but rather performatives characterizing a way to be. Phaedrus’s structure of being is such that through remembering he forgets the difference between himself and Lysias, and in that sense occludes both his own being as Phae drus and the occlusion. The consumption, and by implication, interiorization of the writing of Lysias, occludes the being of Phaedrus from Phaedrus, substituting for it a Phaedrus who is now tacitly identifying with the words of Lysias. By extension, it reveals an important characterization of the notion of human being in general: Performing in technological environments, humans can occlude their being. What is more, the conveniences of technology may occlude the first occlusion. Only the second occlusion can be undone, and human thinking can at best pass behind it. Critical thinking, similar to that of Socrates, can expose but, alas, not undo the first occlusion. Therefore
26 Stakes
no anthropological theory can describe what human being is; instead it can reveal only the occlusion, or forgetting of the being embedded in our technologically driven ways of being human. Such an anthropology is therefore ontology, Heidegger argues in his polemics with Ernst Cassirer’s, and indirectly with Hermann Cohen’s, philosophical anthropology of human being as a universal common denominator underlying all cultural differences between people. Performing double forgetting defines a character, which is irreducible to other characters. Translated into the language of Heideggerian philosophy, that performed character is a Dasein, a being there. I do not deny the validity of such a translation of a performative into a being, a Dasein, but I do not insist on its necessity either. After all, the double forgotten cannot be characterized and therefore may but does not have to be Being (Sein). What takes us to Being is a performative, but Being is not an apodictic presupposition of the performance. Instead, it is only a possibility.4 For my own argument about thinking and memory in the Talmud, an ontological rather than psychological or epistemological framework of thinking about thinking and about memory/forgetting provides a lens through which to begin to approximate the complex relationship between thinking and remembering performed in the Talmudic text. Just as in the ontological framework memory is a fundamental characteristic of one’s way of being, so also analyzing memory in Talmudic texts must go beyond a psychological and anthropological analysis of the text, to attend to the ways in which people are or, as I more broadly argue, perform in the text. As I will cumulatively argue in the following chapters, the latter term is broader because it embraces a notion of the virtual, not predicated on the ontological (and by extension political) urge of deciding on what is, what is not, and what only seems to be. A Heideggerian perspective on thinking helps regain an element of the performance of thinking in the Talmud that is otherwise forgotten by scholars. That element is the thinking performed in the Talmud, which both extends beyond and grounds what is immediately represented in the text of the Talmud by the characters or the conversations they have. Yet it is only a first approximation of a way to recover an understanding of that performance. Heidegger’s is no more and no less than one way to approach the area of thinking and remembering in which these processes go beyond what is described in the text. It points toward a broader idea of the thinking and
What Happens to Thinking? 27
remembering being performed in the text, which I am setting out to explore. It is, therefore, important for me, by way of preparation, to expand the account of the broader connections between thinking, technology, and remembering. The questions to ask Heidegger are: What does “to think” mean then? And what is thinking supposed to accomplish? For Heidegger, the task of thinking is more modest than one would expect. Nevertheless, and precisely because of that, it is a very difficult task. To think means “only” to reveal the forgetting of being in the world, as opposed to the more obvious or even more ambitious task of recovering what has been forgotten. The task of thinking must be more modest because what is forgotten due to the essence of technology cannot be regained, since the essence of technology cannot be canceled or undone. Thinking, therefore, can neither regain the forgotten being, even if that being conditions the human’s being in the world, nor undo the forgetting.5 Yet thinking still can lay bare that process of forgetting. Attending to the fact of forgetting, of which technology is both the cause and the hallmark, according to Heidegger (along with German poets, whom he reads in that way), allows the world in which humans live to bespeak both the being of the world and, as a result, forgotten being, as well. In that way, the world no longer is reduced to an environment standing ready for the humans to consume—in particular, to preserve or to destroy. In this theoretical framework, to think thus means to listen to the call of being of both the world and of the humans therein. In short, to think means to let the world bespeak its being there, to remind us about the forgetting of being that enables the specifically human ability to become either excited or bored by the environment. That also reminds us that things do not merely stand at the service of humans, but rather are in their own right. A critical reader of Heidegger might ask: Who has preferred technogenic consumption to recognizing that the world is, and thus has preferred not to think through being as the condition of possibility of the worlds’s being and hence has chosen not to lay bare the forgetting of being? Heidegger, of course, encounters this choice only after the fact, or as his interpreters often have it, in the mode of “always already.” But he seems to imply that the one who has committed the choice is the same one who lives through the consequences of that choice. Even if the choice has been already made, Heidegger thus—rather automatically—ascribes the ability to make that choice to the same (human) being who actually forgets. The one who has forgotten about being is the same one who has chosen to forget.
28 Stakes
It is, therefore, only humans (as opposed to animals) who can make a different choice. That choice is what has been made by Heidegger himself or by the German poets, who, in their poetry, lay bare the forgetting of being. One reason for granting the ability to make the choice to forget to humans at large is of course that Heidegger and the German poets are among them. By this logic, the one who lays bare the forgetting of being could, of course, also be the one who forgets. The resulting general category, “humans,” applies, by extension, to all the privileged who join Heidegger and the German poets in thinking—that is, in laying bare the forgetting of being—as well as to all the disadvantaged who merely forget who they are and naively indulge in both the excitement and the boredom that technology gives them. However, that taxonomy excludes all those who do neither of these things, but rather, as I will be arguing in the case of the Talmud, perform. I will soon turn to the political implications of that tacit exclusion. What is even more important, however, is that for Heidegger, forgetting is linked to memory. This means at least two things. First of all, that linkage implies, in line with the philosophy of Plato, that both memory and forgetting have an ontological, rather than a merely psychological significance. Second, and most important, in line with the philosophy of Descartes, it implies that thinking is independent of and stronger than either remembering or forgetting. Regardless of what is an extension of what—remembering of forgetting, or forgetting of remembering—thinking reveals forgetting and thereby reaches where memory (or forgetting) cannot reach. Both faculties require a direct object other than themselves, but thinking can focus on the working of forgetting. Thus to think means for Heidegger much more than either to forget or to remember, and thinking exceeds both memory and forgetting in power: Thinking takes control of both forgetting and memory. However, as the analysis of the thinking in the Talmud in the following chapters shows, there is a different model. Without jumping ahead of myself, for now I can say only that in that model, memory takes control over thinking, and as a result sets thinking free from the urge of the ontological (and political) choice between what is, what is not, and what only seems to be. In what remains in this chapter, let me address a practical-political aspect of Heidegger’s conception of Dasein, which, as I have argued, is a specific version of a performative existence in general. What does the structure of Dasein exclude? A seemingly very abstract framework erected around the question of thinking supports Heidegger’s practical political attitude toward
What Happens to Thinking? 29
Jews, both toward real Jews exterminated in the technocamps of World War II and to the “Jews” Heidegger theoretically implies (and politically falsifies) for the purposes of his theory of thinking. These “Jews” represent the third category, which Heidegger needs to exclude in order to find a common (even if negative) denominator between those who have chosen to indulge in technology and those who have chosen to think. On my reading, this common ground is reached by excluding the “Jews” from the new humanity.6 A common ground between those who have de facto preferred the life of technological consumption and those who undo that preference by thinking about being is that neither group is “Jewish.” The “Jews” of Heidegger are those who act unlike other groups. Unlike others, “Jews” do not de facto forget that they are finite beings in the world, which is, and which therefore might stop being and revert to nonbeing. They therefore do not forget that they are bound to die either. Rather they remember death, understand their being in the world, and understand everything that Heidegger himself understands about being in the world—but “Jews” have deliberately chosen to deny all that Heidegger chooses to affirm. Let’s take a closer look at that taxonomy of tacit exclusion. The taxonomy renders the implied “Jews” as understanding but not willing to embrace finitude of being in general and of human being in particular. Feeling, but not willing, to accept that beings are, and that therefore their being, or for that matter being in general, is finite, that is, temporal, that is, can imminently revert to nonbeing, “Jews” might forfeit death by refusing to deny afterlife. Because the “Jews” forfeit death, they, in Heidegger’s logic, by implication never have life either.7 These “Jews” thus represent the worst case of technology taking over the human. Unlike young Phaedrus, they do not de facto passively or naively forget or occlude their being in the world behind consumption, but rather both actively remember and deny it. Although other groups forget being because of technology, “Jews” do so willingly. For Heidegger, then, it is no surprise that these “Jews” fell prey to technology in the camps. The extermination of both his implied “Jews” and the real Jews, as Heidegger continued to maintain even after World War II, was thus not killing, but rather just another technological process, somewhat comparable, as Heidegger put it even after the war, to a “motorized food industry.”8 As striking and politically controversial as it is, Heidegger’s conception of thinking and forgetting is deeply rooted in one of the most tacit parts
30 Stakes
of his thought, in his theorizing of “Jews” as the negative common term for both those who have de facto forgotten/occluded being and those who, like Heidegger, have chosen to lay bare that forgetting/occlusion. However, despite its strong anti-Jewish stance, his conception was actively engaged by many Jewish intellectuals after the war. A consideration of the relevance of Heidegger’s thinking as a lens helping us understand the catastrophe of European Jewry in World War II, including its consequences for the contemporary world, provides one reason for that engagement. Another reason is in understanding the importance of Heidegger’s thought about technology for dealing with current, even more advanced, stages of technological development, which include not only the “ready availability” but also the constantly increasing speed of consumption. Collectively, these two reasons have prevented many Jewish intellectuals from boycotting Heidegger’s thought, despite his controversial relationships with National Socialism, and, instead, have stimulated their critical engagement with his thought. In the case of Emmanuel Levinas, such a critical engagement meant no less than creating an intellectual alternative to Heidegger’s framework of asking the question of thinking, including his implications about “Jews.” Levinas’s view is particularly important not only because it provides a viable alternative to Heidegger’s theory of being and thinking, but also because it stems from Levinas’s reading of the Talmud.9 A Jewish student of Heidegger’s from the years before the Nazis came to power and a prisoner of war, Levinas found intellectual resources for resisting Heidegger by studying the Talmud, which he began and continued in the postwar decades, when he was exposed to the Talmud through the Lithuanian yeshiva approach to the text. In these studies, Levinas, as he put it, attempted to “translate” the texts of the Talmud “into the language of philosophy.” With a classical philosophical sense of irony, when delivering these “translations,” he called himself “an amateur,” thereby distancing himself from those whom he called “professional Talmudists.” The reference might have been to those who studied the Talmud in and only in a traditional yeshiva. Yet it pertains even more pointedly to academic historical and philological scholars of Talmud criticism, who, according to his critical view expressed in regard to Bible criticism,10 “forgot” the nature of human beings in the texts of the Talmud, precisely in Heidegger’s meaning of “forgetting.” Levinas’s distancing from (and implied criticism of) the positivist academic scholars of the Talmud of his time becomes particularly important, both in
What Happens to Thinking? 31
terms of its achievement and in terms of its limitations. He has explicitly criticized Bible criticism for making the Bible into an object of scholarship artificially alienated from the consequent tradition of its commentary.11 By extension, that critique may be applied to Talmud criticism. From that point of view, the text-critical scholars of the Talmud limited themselves to the comparative-historical and empirical-philological study of the text of the Talmud, as if the complex nature of human beings involved in the text as characters, creators, or readers was already clear for them. The “professional Talmudists,” as Levinas envisions them, axiomatically used a universalistic notion of human being as a common denominator that underlies and thereby bridges any cultural and historical differences between people, allowing them to draw these comparisons. However, such a universalistic notion of human being necessarily relied on a reductive understanding of a human being as either an object or, by extension, a subject of scientific knowledge, which means, as he argued, the occlusion of the being of the human being. That “scientific” (including philosophical) anthropology was precisely the target of Heidegger’s critique when he was developing his much more nuanced theory of thinking as laying bare the forgetting of being. Distancing himself from the anthropological approach of the “professional Talmudists” afforded Levinas the opportunity to ask questions about thinking and about the ways of being human (or more precisely, of being the sage) in the Talmud. Levinas’s alternative to Heidegger’s theory of Dasein as a thinking being was his theory of “ethics as first philosophy.” In that theory, either assuming or refusing ethical responsibility toward the other predefines human thinking about being in the world. Unlike the more common understanding of ethical responsibility in terms of either the reciprocation of gifts or the rational calculation of universal principles behind each action, Levinas12 locates responsibility where neither reciprocity nor universality can apply. These are the areas where a person truly encounters the otherness of the other. Levinas raises the issue of ethical responsibility toward the other precisely in such areas where a person cannot truly control the powers of reality, or, as Levinas says, “is no longer able to be able.” Some examples are death, love, and parenting. How does a person respond to a situation he or she cannot truly control? Does she attempt to take control over it or to escape, or to withdraw, or, alternatively, to respond adequately by being both responsive and responsible to what is beyond control? For Levinas, in choosing a response, the
32 Stakes
person predefines his or her vision of the world and of being in the world. In technical terms, this means that the ethics of responsibility precedes what Heidegger would consider fundamental—the theory of being in the world. Levinas’s theory also implies that either assuming or denying ethical responsibility toward the other both precedes and predefines the ways in which humans “think” about their being in the world. Ethical responsibility thus comes before “thinking” about being in the world. To put it in the terms I proposed in the Introduction, the performance of an ethics of responsibility is prior, as performance, to the performance of thinking about being. If applied to criticizing Heidegger, this theory suggests that in asking what it means to think, Heidegger already has made an ethical choice: In view of the otherness of Jews, he irresponsibly attempted to control them by grounding his theory of “thinking” on the exclusivist construct of a “Jew.” Where does the outcome of this critique leave the Talmud? I am asking now not about the Talmud the way it was (and is) studied in Lithuanian yeshivas of the last two centuries, informing, as they did, Levinas’s perception of the Talmud, but rather about the text shaped in the late ancient schools of the rabbis, not of Lithuania, but of Babylon. By all accounts, something must be lost in the translation of the Talmud into the language of philosophy. However, what difference does the sense of that potential loss make for a scholar who, instead of dealing with the text of the Talmud in the way in which it was studied in nineteenth- and twentieth-century Lithuania, and instead of translating it into the language of philosophy, asks the question “What does it mean to think in the Talmud?” A full answer to this question will require analyzing the relation of thinking and memory in the text of the Talmud, which, however comparable it may be in subtlety and nuance with the relation of these concepts in the philosophy of Heidegger or in Levinas, is configured in a rather different way. Certainly, understanding this possible other perspective of relating thinking to memory remains outside the theoretical tasks undertaken by Levinas in responding to Heidegger’s theory of thinking in the age of technology. As I began by noting, Heidegger’s notion of the age of technology applies as much to the time of Plato as it does to the modern era. Yet in just a few decades after Heidegger’s and Levinas’s theories took shape, the question of thinking in the age of technology became even more complex. Modern conveniences of technology provided for, or, in his idiom, made readily available, more support for knowing, decision making, learning, problem solving, and
What Happens to Thinking? 33
for other aspects of thinking and remembering. Many acts that humans have previously had to do by themselves can now be done for them by technology. If so, what happens to thinking? Did the advance of technological services put thinking in danger of being forgotten? If so, what, if at all, does technological innovation leave to the realm of the purely human? And if there is a “purely human,” does thinking still belong to it? This book contributes to an understanding of one aspect of these crucial, but broad, questions by placing the relationship between thinking and memory in a perspective arising from a close analysis of Jewish practices of thinking and remembering in what became the core recorded text, and the living archive of Jewish tradition— the Talmud.
t wo
Ego Cogito, Ego Meminí: I Think, Therefore I Remember
How does modern thinking about thinking differ from historically known methods of thinking, particularly within the Jewish tradition? Late ancient rabbinic schools of thought differ from modern intellectual habits in their understanding of the role of memory in thinking. Modern practices of thinking limit the role of memory to providing data. In contrast, the rabbis viewed thinking as an essential instrument of memory. This contrast in the relationship between thinking and memory allows rabbinic—specifically Talmudic— practices of thinking to provide valuable insight into the complexity of human thinking in the modern age. Through a critical assessment of the process of thinking inherent in the the Talmud, the following chapters contribute to the fields of Jewish studies and philosophy. In addition to offering new insights into the Talmud by analyzing it in terms of the philosophical theory of subjectivity and memory, which 34
Ego Cogito, Ego Meminí 35
has not yet been used to address the texts of the Talmud, they more broadly allow scholars to assess the utility of the philosophical theory of subjectivity beyond the realm of either philosophy or its usual application to the contemporary understandings of human thinking. However, the question of the nature of thinking and remembering in the Talmud arises not only from the context of philosophical and rhetorical exploration of thinking and remembering but also from the context of academic scholarship on the Talmud in the last two centuries. In that context, the question of thinking and remembering in the Talmud comes to light subtly and implicitly, rather than openly and explicitly. And yet, there, the emphasis in this question remains on thinking, as opposed to remembering, and in particular on the question of who allegedly thinks in the Talmud— whose thinking is expressed in it? I do not author this question. Rather, it arose because Talmudic scholars of the twentieth century asked the question “Who redacted the Talmud?” Redaction theory emphasizes the historical, chronological production of a text, rather than the intellectual practices that the text demonstrates. According to these scholars, the teachings of the earlier masters, or amor’aim, underwent successive revisions by anonymous redactors, resulting in a text that became the Talmud. However, a crucial gap remains open between the question these scholars asked and the question they answered. Their question focused on the process of redaction, but their answer posited a thinking agency that they personified as the anonymous redactors in the Talmud. These scholars tried to learn who redacted or rather reinvented the teachings of the earlier masters or amor’aim, but by positing a thinking subject in the process of redaction, they revealed the actual question underlying their inquiry, that is, “Who thinks in the Talmud?” My project here interrogates precisely what is left out when thinking is thus uncritically attributed to an agent. The concept of a redactor implies a finished process that ended with the publication of the work. Liberating the process of thinking from the conception of a thinking and writing agent allows us to reopen the Talmud as an archive that is both the protocol for and the actual record of live thinking that invites ongoing engagement, but that was artificially interrupted by extrinsic forces of history. The existence of these anonymous redactors as thinking and designing subjects was theoretically conceived in the process of analyzing the texts of the Talmud and empirically verified exclusively by the same texts. The scholars
36 Stakes
of text-critical approaches of Talmud criticism have neither initially reconstructed nor consequently succeeded in supporting the existence of these rationally thinking subjects by any external historical evidence outside the Talmud. Needless to say, their identification of the activity of redactors or designers has depended on how these scholars have understood the intellectual norms of the Talmud’s composition and consequently on how they have interpreted the literary and intellectual design of the Talmud. In turn, their understanding of the intellectual norms of the Talmud was informed by the traditions of Talmudic interpretation in the Middle Ages as well as in early modernity. These post-Talmudic traditions of interpretation (paradigmatic examples include Maimonides in the twelfth century, Canpanton in the fifteenth century, and Lutsatto in the eighteenth century) either highlighted or downplayed (but never doubted) the role of refutations and counterrefutations or defenses as the core form of legalistic discussion in the text of the Talmud. To take one prominent example of the result of focusing on Talmudic refutations and defenses as the main form of the Talmud, both traditional interpreters and contemporary text-critical scholars invariably have rigidly separated the legalistic and homiletic parts of Talmudic text. They called these parts by the terms borrowed from the Talmud, halakhah and haggadah, respectively. They have both broadened these two terms in meaning and privileged the former over the latter. However, an option that both the traditional interpretive and text-critical approaches excluded was to relate refutations in the Talmud to a larger intellectual paradigm—that of memory as the main thinking activity performed in the Talmudic text. In what follows, I discern and explore that option and examine the text of (the) Talmud as a literary and intellectual form of memory. To that end, I propose to reconnect the scholarly interest in thinking that is implicit in contemporary text-critical analyses of the Talmud and explicit in studies of both the philosophy and rhetoric of thinking processes. That is, I propose to show how the question “Who thinks in the Talmud?” needs to be rephrased as “Who remembers in the Talmud?” or even more precisely, “Who performs Talmud as an intellectual discipline in the text of the late ancient Talmud?” Advancing this inquiry presents four interconnected demands: direct analysis of the process of thinking in the Talmud; differentiating between the Talmud as a text and Talmud as an intellectual discipline; historicizing Talmud, that is, that intellectual discipline; and mapping Talmud’s relationship
Ego Cogito, Ego Meminí 37
to other disciplines of thinking, both philosophical and rhetorical.1 Answering these demands will form the woof and warp of the arguments out of which the texture of the book that follows will be woven. Contemporary text-critical understandings of Talmudic discussions have resources that can contribute to meeting these four demands. In particular, text-critical discussions dating the text of the Talmud illuminate the nature of the intelligent design of the text. The text-critical scholars addressed the question of the dating of the text as the question of identifying the speakers in and/or contributors to the text. To answer the question “Who speaks in the Talmud?” the text-critical scholars tried to determine whether it was the late ancient Talmudic masters, as it had seemed to traditional readers, or historically later anonymous redactors of these masters’ teachings, as critical analysis of the texts seemed to suggest. Another way in which these scholars approached the question of speakers in the Talmud, was by asking whether the conversations in the Talmud are the direct records of live discussions in rabbinic academies in late antiquity, as they had traditionally been assumed to be, or carefully and beautifully constructed literary designs made at a later time, as critical analysis suggested. The text-critical problem of the chronological dating of the text not only posed difficult questions for scholars to solve, but it also highlighted the nature of the text of the Talmud. The Talmud indeed presents—and, as I will show, performs—the conversation of rabbis staged in an academy in late antiquity, yet that presentation, unlike, say, a classic novel, bears neither what might be called the date stamp of its creation nor the signature of composers, and it does not start with any explicit indication of the time, place, or characters involved. Still, like some modern literary forms of the novel, instead of describing actions for the reader from the point of view of a narrator, this text directly performs these actions—attacks, defenses, inferences, and so on—right in front of its readers or even involves these readers in that performance by putting them in the footsteps left by the arguments (and identities) of the characters. An analogy between the texts of the Talmud and the sample performative utterance “I will always love you” can highlight the performative nature of the text of the Talmud. As John Austin’s theory of performative actions suggests, in saying “I will always love you,” the action coincides with the expression of that action, which makes it a performative utterance, relating, as it does, to “loving now” and to “loving in the future” at the same time.
38 Stakes
Expressing it is the only way of doing it. By the same token, repeating it is the only way of verifying it is true.2 Properly, if the action is not expressed as the promise of love, it is not a promise of love. What is more, when a person says to another “I will always love you,” the details of the context, of time, place, or scenery, may vary, but the performative action as simultaneously the action and the expression of that action remains the same. In that sense, as indispensable as they are, the details of “scenery” or actual context still have only peripheral importance and thereby ultimately remain extrinsic. In short, the details change, but the expression of love as the promise remains. “I will always love you” is simultaneously an expression and a performative assertion of what is thereby expressed. Its truth or falsity, “the reality check,” can be verified only by repeating that performed utterance or falsified by not repeating it. Where and when it is performed doesn’t matter. The text of the Talmud can also be seen as a performative utterance, as long, complex, and multifaceted as it is. The action it expresses (make it refer not to “love,” but to “Talmud”)3 and the expression of that action coincide. The action of “Talmud” too is performed only by the performative promise to remember, which is achieved through following—that is, repeating—the protocols of memory in the text of the Talmud. In understanding the nature of the text of the Talmud, a second analogy with the performative utterance “I will always love you” can help. If presented as direct speech, the utterance “I will always love you” may or may not require the explicit presence of either a narrator or even the voice of the narrator. The latter can remain totally mute. Not only does a performer perform the performance, but the performance performs the performer. Therefore, if a narrator is needed to introduce the performer of an action, but if the performer is always already given in the performance, the text may have what could be called a zero degree of narration and be only direct speech, although in this speech, the speakers may of course serve as narrators if they quote or otherwise deliver speeches they attribute to others. In fact, to make the phrase “I will always love you” work, it is unnecessary and perhaps even excessive to introduce the point of view of a narrator, who might say, for example, “Tom said to Becky, ‘I will always love you.’ ” The utterance sufficiently defines the performer as a lover. And again, verification would depend only on the reiteration of the performative utterance. Therefore, a narrator does not need to say, “And Tom meant to keep this promise.” In sum, the positioning of the performer as a lover who promises
Ego Cogito, Ego Meminí 39
love does not necessarily require any further help from a narrator, not even an introduction, much less a confirmation of what is uttered. Furthermore, the uttered phrase automatically tells that besides the performer, there is some context: someone else to respond, or perhaps to deny, or perhaps even simply to hear. In short, when uttered, the phrase automatically defines the lover, even before and regardless of his or her name. What is more, that phrase already implies a place, a context, or even a language, without either defining them in any specific way or depending on such a definition. “I will always love you” is thus a performatively self-sufficient and self-referential utterance. By way of that second analogy, the text of the Talmud can also be approached as a performative utterance in the mode of a direct speech that always implies a “now” and a context in which the speakers perform and are performed. Of course, the Talmud as direct speech is much longer and much more multifaceted in terms of the performers and implied contexts. What is uttered and performed in it is obviously not love, at least not love par excellence, but rather the intellectual life of a rabbinic academy “in Babylon” and “in late antiquity.” Just as in “I will always love you,” the context is implied, but the details remain extrinsic, so also “Babylon” and “late antiquity” are external elements of the context for doing “Talmud” as performed in the text of the Talmud. Of course, these external elements could and should be inferred and “reconstructed” by an attentive reader.4 Yet however correctly they are inferred by modern readers and researchers, these literary circumstances remain of such extrinsic importance to the main performative act of the text of the Talmud that in the text they do not deserve direct introduction but get only indirect mentions or else appear as a part of direct speeches. What ultimately matters in the Talmud is not the implied, yet not described, scenery and characters, but the performative act, the doing of what is done in the rabbinic academy. But what exactly is that doing in the Talmud? If I had to give a brief answer to that question, I would say, as I suggested above, that the doing of the academy is remembering the teachings of the sages of the Mishnah,5 a third-century code of instructions for rabbinical courts. And if I indeed ventured to give such a simple, if not simplistic answer, this chapter would be a long (and therefore unnecessary) version of that simple answer. However, my concern about any answer in such a simple form has to do with the profound complexity of the notion of memory in terms of who
40 Stakes
remembers, or rather what it means to remember. The complexity of that question defies any brief, much less any simple, answer. And at this point, the hitherto useful analogy with the performative utterance “I will always love you” helps no longer, because simply uttering “I (will) remember” does not quite perform anything: The object of love, the “you,” is already given or performed in the utterance “I will always love you,” but in “I (will) remember,” the objects of memory are not yet performed and are yet to be performed to make “I remember” a performative utterance. An even greater complication is that, as I have already mentioned in the Introduction and as I will be arguing in more detail in the following chapters, in the Talmud, in contrast to the rhetorical practices of memory in the art of mnemonics, the protocols involved in memory also involve another faculty—that of analytical thinking—and in contrast to philosophical theories of analytical thinking as a vehicle of memory, Talmud performatively connects memory with the rhetorical technique of heuristic refutation and defense. Finally, even more complexity emerges from the fact that understanding memory in the Talmud involves the question of personal being as such, of a performed existence, and, more precisely here, of a being that is performed as a person in the Talmud: the remembering being who is performed in the Talmud. Brief mention of the terms “memory” or “remembering” or even “thinking” cannot address the question of the nature and status of that personal being, because asking the question “Who speaks?” or “thinks,” or “remembers” seems automatically to assume that the answer must involve the thinking subject or cogito of modern philosophy—precisely the assumption I wish to interrogate here. Grammatically, the pronoun “who,” in the subject position of a sentence, instantiates the proposition I wish to bracket and question in terms of the subject position in general. Therefore, without jumping ahead of myself, I can say only that what follows throughout this book will pay close attention to the performer(s) who are performed both in and by the performative utterances that make for the text of Talmud, and make the Talmud not only a closed set of texts, but also an open archive. A more complicated, but more accurate way of putting the question that I wish to ask here, therefore, would have to be something like “What sort of way of being in the world is involved in performing that long utterance for which ‘the Talmud’ is a retroactively given name?” That is, what way of being in the world is involved in thinking and remembering in both the closed text
Ego Cogito, Ego Meminí 41
and the open archive called the Talmud? It is fundamentally that question that I ask when I ask, “What does it mean to remember?”: What way of being in the world is involved in performing the intellectual discipline that is Talmud? How, in other words, are “remembering” and “being” related? To answer this question, I begin by employing the resources of contemporary Talmud criticism and, based on what Talmudic criticism has both succeeded and failed to achieve, proceed from what is clearer to what is more obscure about a remembering being, or, in the final analysis, the performance of remembering, in the Talmud. I initially highlight how that remembering being was first discovered (and occluded) as a thinking being in the Talmud by contemporary text-critical Talmud scholars as they constructed the Talmud as an object of scholarship by following in the footsteps of their early modern and medieval predecessors. In the light of, and, most importantly, in the darkness revealed by both the discovery and occlusion of who thinks in the Talmud, I will examine the complex relationship between memory and thinking as performed in that late ancient text. This task is intrinsically comparative and hermeneutical in nature. It requires placing the Talmud as the performance of memory as thinking and of thinking as memory in a broader context and on a larger map—both synchronic and diachronic—of other intellectual disciplines, such as rhetoric and philosophy, including, in particular, the role of refutation and invention in each, including different versions of what it means to think and remember in the Greek, Latin, Syriac, and other schools and practices, both in late antiquity and in later texts that help illuminate earlier ones, either directly or by way of heuristic interpretations that I will consequently put into question by applying them to reading the Talmudic texts. Approached as a performance of intellectual activity, the late ancient Talmud reemerges as a protocol of thinking and remembering and thereby shows what exactly is involved in thinking, remembering, and existing—simultaneously in the literary, historical, and ontological senses of the terms—as a performed thinking and remembering being in the Talmud. A first, hypothetical approximation of a comparative hermeneutical framework in which to ask the question of thinking, remembering, and personal existence in the Talmud, a simplistic and heuristic first approximation of the comparative framework to be adjusted and renegotiated as the inquiry goes on, can begin with the difference between cogito, ergo sum in modernity and meminí, ergo sum in antiquity: “I think, therefore I am” versus “I remember,
42 Stakes
therefore I am.” In modern times, in Descartes’s principle of cogito, ergo sum, existing as a person requires, first and foremost, thinking, or cognition in general. In contrast, in the Talmud, it requires remembering. The difference between these two constitutive principles of individual existence does not form a linear opposition. Rather, their relationship escapes direct comparison. Instead, a more fine-tuned picture of the relationship between thinking and remembering must be drawn. For Descartes, cogito does not exclude remembering, yet remembering is only an extrinsic part of personal existence. Remembering is comparable to other accidental activities, such as eating, walking, or reading a text. In order to exist, Descartes requires only that we think (cogitare), which, again, may, but does not have to include remembering. (Perhaps this is why some interpreters and critics of Descartes have suggested reading his cogito ergo sum as cogitat ergo est [“It (he, she) thinks, therefore it exists”],6 for, not unlike a computer or, to take an older example of the “it,” the “intellect” in the Neoplatonic theory of Amelius, a thinking “it” does not have to remember itself.) In contrast, in the Talmud, an individual—call it a rabbi, a student, or talmid h.akham, “one who learns from the wise”—exists first of all as one who remembers the traditions and teachings of the sages of the past (whether this past has either existed or is claimed to have been). His or her remembering may include and in many cases even requires the critical analysis of the data of memory, but thinking remains an added feature, and remembering stands at the core of what it takes to exist as a talmid h.akham. This gives a better approximation, but it is not nuanced or complicated enough, either. The following chapters build, step by step, a theoretical framework discerning the thinking and remembering performed in the Talmud. I structure the argument there around three focal points of the performative utterance that is the Talmud: speaking, thinking, and remembering. Although each of the main parts of what follows emphasizes one of these foci, each also deals with its relationships with the other two. What follows thus comprises three parts. The part on “speaking” puts at the center of analysis the question of who speaks in the Talmud. I address this question by highlighting both the achievement and the limitations of Talmud criticism’s conceptions of speaking beings in the Talmud as anonymous speakers (“redactors”) either of the text as a whole, or of the dicta of earlier, named teachers. The next part, on “thinking,” helps establish a theoretical framework in which to ask about the role of thinking in the activity of the anonymous speaking subjects whom the text-critical
Ego Cogito, Ego Meminí 43
scholars have identified in the Talmud. This theoretical framework provides what is necessary (but not what is sufficient) for asking about thinking in the Talmud directly and explicitly without occluding it by the notion of anonymous speaking subjects. The last part, on “remembering,” both employs and adjusts that framework of analysis in order to get even closer to the elusive target of my inquiry, the question of what sort of way of being in the world is involved in remembering in the Talmud. In focusing on remembering, I emphasize the role of refutation and invention in its relationship to memory and thinking in the Talmud in juxtaposition with the rhetorical and philosophical compartmentalization of remembering and refuting. This structure forms the larger tridimensional system of coordinates that defines the territory I now explore and map.
three
Through Talmud Criticism to the Talmud as Thought and Memory
In analyzing the tridimensional territory of speaking, thinking, and remembering displayed in the Talmud’s text, I proceed through contemporary Talmud text-critical scholarship to the Talmud as a thought form, and in particular a memory form. For brevity I dub the hitherto predominant text-critical scholarship on the Talmud “Talmud criticism.”1 By that I mean a set of theories and empirical textual analyses in contemporary Talmud study, which are comparable, in a sense, to biblical criticism, if the latter is understood as using philological and historical tools of analysis to see the Bible as a document produced by historically distinct groups of authors rather than by any single, or dominant, author. Talmud criticism too is asking about authors and groups of authors of the Talmud, calling these authors either anonymous “redactors” or “composers/designers” of the Talmud. The necessity by which exploring that tridimensional territory of speaking, thinking, and remembering in the Talmud requires moving through Talmud 44
Through Talmud Criticism to the Talmud as Thought and Memory 45
criticism to the Talmud as a thought and memory form will be explained in the following chapters, as well as my claim that, both more specifically and more fundamentally, the Talmudic thought form would be that of memory and remembering. My choice of the strategy of proceeding through Talmud criticism answers, albeit initially all too briefly, the question of where my approach stands with respect to major approaches and achievements of Talmud criticism. To begin answering this question in proper detail, I engage with Talmud criticism scholarship on more specific points in the chapters below. In this chapter, in advance of such local engagements, I generally outline and interpret two hitherto prevalent and competing approaches in the theory of Talmud criticism. I use Shamma Friedman and David W. Halivni as paradigmatic representatives of both achievements and impasses of these schools. Readers who are already familiar with both theoretical foundations and practical examples of the analyses in Talmud criticism may skip this chapter. Others, if they find the exposition too technical or too difficult, may also skip it and then come back later as, or if, needed for reading other chapters, where I engage with, interrogate, and build on more specific concepts and approaches. Both schools in Talmud criticism treat the text of the Talmud as a literary text rather than a direct record of real conversation in a rabbinic academy; yet these scholars differ in the degree of conflation between historical reality of the Talmud’s production and the reality literarily represented or, if one prefers, constructed in the Talmud. As I will explain, Halivni must allow a degree of conflation between historical reality of Talmud’s production and reality represented in the Talmud as a literary text, whereas Friedman limits his analysis to discerning purely formal acts of literary composition. To highlight this difference, I dub Friedman’s approach literary-formal and Halivni’s literary-realist. Literary-formalist and literary-realist approaches to the Talmud’s genesis both compete with each other and collectively dominate the field. By way of comparison, I could call the approach I am taking in this book “literary- philosophical.” It explicitly and directly engages philosophical, rhetorical, and logical theory in understanding and analyzing a literary text, in particular, the ways in which the text generates its sense for its readers; it simultaneously uses that engagement as a way of renegotiating the theory. The former two approaches also foster philosophical strategies, albeit implicitly and indirectly. As I will momentarily show, these approaches adopt (post-) Cartesian ideas about language and disembodied thinking subject, as de-personified
46 Stakes
and anonymous as it is. Explicit engagement and implicit application seem to create a sharp contrast: Whereas the first two approaches show no explicit engagement with a philosophical theory of sense-production or of “making sense” in the text, the third one explicitly engages and renegotiates it. Yet, divergence between these three approaches is less than it may seem. None of the differences between them are more definitive than they are complementary. All three deal with practices of making sense in the Talmud or, in more general terms, with the processes of thinking observed in the text. Literary-formalist and literary-realist approaches ascribe these thinking processes in the Talmud to an agent (either individual or collective), hypothecating that the agent both constructs and manipulates discussions in rabbinic academy yet does not necessarily appear on stage. As a result, the hypothecated agent of the thinking process assumes the metaphysical role of the subject, substance, center, foundation, beginning, origin, moving force, and root of the process of thinking. That role is to compose Talmudic discussions, as we currently know them. In contrast, the literary-philosophical approach takes the Talmudic practice of thinking in stride: It approaches the thinking processes in the Talmud on their own terms, that is, as thinking; it gives no agent of thinking the role of the subject or of the foundation of that thinking. In analyzing thinking processes in the Talmud, the literary-philosophical approach allows agents but vetoes subjects.2 Yet again, this difference is not strong enough to disallow engaging all three approaches in a dialogue that complements limitations embedded in each of them. This strategy replaces the model of competition between approaches by that of their complementary character. The new model of complementary approaches recognizes that literary-formalist and literary-realist approaches to the Talmud’s becoming are not sufficient, because, if read critically, they highlight problems with the historical genesis-oriented approach. Isolating these problems in historical approaches to the Talmud’s becoming opens up room for a dialogue between historically oriented and philosophically oriented approaches to the Talmud. The problems of historical approaches to the Talmud have to do with assuming a historically empirically unverified (and perhaps unverifiable) agency responsible for the Talmud’s genesis, while claiming to produce an account of the empirically verifiable history of the Talmud’s production. Some researchers construed that agency as the master of a given Talmudic discourse, that is, the designer of that discourse, while others contemplated that agency as the hidden or anonymous composers or “redactors” of or in the
Through Talmud Criticism to the Talmud as Thought and Memory 47
Talmud—a hypothetical distinct historical group, who acted in the Talmud behind the constructed scenes of the discourses of the rabbis. As a result, a historical-empirical account of the Talmud’s genesis had to start from and rely on a theoretical notion of agency rather than directly on empirical data about the agent. The problems arose not only because historical sources other than the texts of the Talmud provide no empirical-historical evidence of the existence of any agents behind that agency but also because different scholars took mutually exclusive theoretical venues of situating the agency and the agent in the realities of literature versus the realities of history. For some scholars, that agency was entirely literary, such that the historical existence of the agent came only by way of application of the results of a literary-theoretical analysis of the literary construction of the discourses of the rabbis. For the scholars of this approach the resulting historical account of the Talmud’s becoming was only an illustration or application of literary analysis, while literary analysis remained programmatically distinct from any historical application of its results. For other scholars, the historically existing agent who acted behind the scenes of Talmudic compositions was a conditio sine qua non of the literary analysis of the Talmudic text per se, which made history and literature inextricably intertwined one with the other. I will juxtapose these two approaches in the names of their most prominent contemporary representatives, Shamma Friedman and David Halivni. Shamma Friedman and the researchers of his literary-formalist school see Talmudic discourses as designed by a hypothetical “Master of the Sugya” (baal hasugya)—a composer or designer of a given Talmudic discourse. Friedman disproves earlier theories, which somewhat naively considered the Talmud a nearly stenographic record of live conversations in the rabbinic academies in Babylon. To that end, he highlights the presence of a well- balanced design in the composition of any given Talmudic discussion that he analyzes. Thereby he proves that the Talmudic discussion was not a record of a live conversation. Having discovered an undeniably well-balanced literary design in the Talmud, Friedman then more problematically concludes that there must have also been a designer or the Master (M) of the discourse. Friedman illuminates the work of M by reconstructing the initial or “earlier” materials that M might have had available for processing. This approach allows Friedman to trace what M or the designer has done on the way to the composition of a Talmudic discourse as we currently know it.
48 Stakes
Friedman’s concluding from the design to the designer or M is not logically apodictic. The fact that there is a design by logical necessity means that there is an eye of the researcher noticing that design, but it only by logical probability means that there has been a designer, M or the master of the discourse. There, of course, always is a master, that is, the one who masters the discourse, or understands it well enough to discern a design in it, as Friedman finely does. But this does not by any logical necessity mean that there was a master who put the composition together. Friedman’s M can be sufficiently construed as a literary agency—indeed a literary fiction, an alleged author, namely, a literary counterpart of the composition or design that the researcher discovers. Similarly to Homer, or to the J, E, D, and P of biblical criticism, M might but does not have to have any empirical-historical equivalency; nor is Friedman intrinsically crucially interested in finding that equivalency. Instead, the results of his textual analysis successfully stand even without any specific historical-empirical group acting as M. In that sense, his analysis of the genesis of Talmudic discourse remains in the area of purely literary analysis of the text. This of course does not mean that M cannot exist historically; it just does not have to. M can be equally successfully placed either inside of the Talmud as a literary figure of its genesis, or imagined outside of it, as a historical group of people acting as M. In other words, M. comfortably resides in literary reality or in historical reality, or in both; however M’s primary placement is and will always be literary. The relationship between literary and historical realities is even more complex in another, literary-realist approach to the Talmud’s genesis. That other direction of Talmudic scholarship has much stronger interest in the genesis of Talmudic discourses in the historical-empirical aspect of it. A leading researcher of that approach, David Halivni3 not only recognized a literary design of a given Talmudic discourse but also paid due attention to some artificial, perhaps symptomatically artificial features of that design. He approached some of these features as symptoms or indications of hidden thinking processes in composing that the resulting design does not directly reflect. Halivni reinvents (“conjectures”) these thinking processes behind the constructed scenes of the discourses of the rabbis. The goal of his analysis of Talmudic discourses is to reconstruct these thinking processes. Halivni takes advantage of traces. However carefully M might have masked thinking behind the end-result composition, still against M’s will, these thinking
Through Talmud Criticism to the Talmud as Thought and Memory 49
processes left traces in composition in the form of symptomatically artificial or stretched elements of design. Similarly to Friedman, Halivni concludes from the agency to the existence of the agent, from the fact of there being thinking processes hidden behind the final form of design, to the existence of the thinker, the T—the redactor of the discourses of the rabbis. Let me first use Friedman’s terms to reformulate what Halivni claimed. For Halivni, the thinking processes behind the scenes of the discourses happened in response to some problems, specifically some failing arguments that M must have observed in the initial or “earlier” materials from several connected discourses of the rabbis. Halivni reconstructs these problems, as well as the thinking processes applied to resolving them. Similarly to Friedman’s logically nonapodictic contention, Halivni also holds that because there is a rationally creative thinking processes hidden behind the design—even if this thinking is expressed only indirectly, namely, as a symptom, for example, as mismatches of parts, displacements, or patches in the composition, which only a critical eye can discern—there therefore must also have been a thinker (T) behind that thinking, the agent behind the agency. However, if in Friedman’s approach, M could either empirically exist or alternatively be a purely literary instance par excellence, for Halivni, T must exist in the reality of history. M was a direct and in that sense transparent literary counterpart of design, just as an author is a literary counterpart of the novel, rather than a historically real writer of that novel. Just as an author is always an counterpart of reality constructed in a literary work, so also M was both a direct function and an implied part of the literary reality, which the discourses of the rabbis “represent” or construct. In contrast, T cannot exist in the literary reality alone. He cannot because in literature T’s presence is marked by no more than a symptom; T is not a merely a direct counterpart of a literary composition, because T thinks in ways that the composition does not represent, but only symptomatically indicates. Respectively, T must reside both in literature and somewhere else, thus in history. Not only must Halivni place T in history rather than only in literature, but he thereby also needs to grant T the reality of historical existence. However, because no proof of that existence comes from the external empirical data of history, it must come from a tacit philosophical argument: If there is thinking, a thinker must exist. In sum, because, T is neither a direct part (a character, named or unnamed) nor the counterpart (the author) of the literary reality that the design “represents,” Halivni must place T in the outside
50 Stakes
of the literary reality. This of course means for him to place T in the reality of empirical history. However, to be a part of historical reality means to exist. Lacking any empirical proof of T’s existence, Halivni establishes that existence along the lines of Cartesian argument: T thinks, therefore T empirically exists. He claims historical existence and supports it by the fact that T thinks, which means T posits intellectual problems (which are legitimate according to the rules of rabbinic discourse as Halivni understands them following Maimonides) and solves them by designing or redesigning a literary composition of discourse in one way rather than in the other. Halivni thus draws a chain of reasoning from the constructed nature of the discourses of the rabbis to symptomatic inconsistencies in these constructions to the hidden thinking processes behind the scenes of these construction to the historical existence of the thinker, or T, who, against T’s will leaves symptomatic displacements and mismatches of discourses in the resulting design. As already mentioned, reasoning from a symptom in design to a thinking process of which this symptom was an indirect indication and then from there to the existence of the thinker, Halivni follows the Cartesian path. For Descartes, “I eat” or “I walk” or, if you wish, “I design a literary composition” does not yet prove that “I exist.” However, “I think” does. Hence, in a Cartesian view, even if a designer might be a fiction, the thinker, unlike M, must exist. Yet Halivni follows the Cartesian path not simply because he lacks sufficient empirical historical proof of the existence of T. Had historical evidence of the existence of T been available, it would still not explain T’s unique intellectual standing between literature and history—both in and behind the discourses of the Talmud. The demand is that T must exist not only historically but also literarily, as a thinker standing both in and behind a literary scene. The Cartesian path serves to satisfy this demand. It explains not only why T must exist in history (I think, therefore I exist), but also how T is present, specifically absent in literature: through the process of thinking and the symptomatic traces thereof in design. The most important reason for which Halivni follows the Cartesian path and posits T rather than M is that T must remain both internal to the Talmudic discourse and at the same time external to it. T must be internal because for Halivni T is part of the chain of the rabbinic tradition. T is external because T is not a part of the literary reality that the rabbinic discourses represent, but neither is T merely a literary counterpart of this reality, and in particular, its author. What makes T an insider of the Talmud is that T thinks according to the rules of Talmudic
Through Talmud Criticism to the Talmud as Thought and Memory 51
discourse as Halivni understands them (namely, according to the rules of infallibility of any argument of any of the listed speakers in the Talmud). However, his problem is that T is not one of the characters listed or otherwise mentioned in the discourses of the rabbis. Therefore T must exist both inside and outside of that discourse, that is to say, both in literature and in history. Thereby the Cartesian connection between the thinker and thinking helps T to satisfy both demands to be an agency in literature and an agent in history at one and the same time. Needless to say, Halivni never explicitly says that stamm’aim or the anonymous redactors/designers of the Talmud thought and therefore must have existed. Rather, it is I who discerns a tacit Cartesian move in his argument. Explicitly he does no more, but no less that the following. He identifies the problems (dh.akim, “pressures”) in the composition of a given Talmudic discourse, ascribes raising these problems to the thinking of the redactors, who, as he assumes, solved these problems by composing the discourse in one way as opposed to another. Then, based on that literary-critical analysis of the text, he proposes a historical-chronological identification of the redactors as a historical group or groups of people. I see this as a Cartesian move from thinking to the existence of thinking person. Notably, Halivni insists on this historical-chronological identification of stamm’aim, but keeps it independent of empirical-historical data about rabbinic authorities of a given period of time, which comes from the sources outside of the Talmud. As his research progresses, he moves historical-chronological identification of the stamm’aim to later and later times in history. At some point the alleged life of the stamm’aim coincides chronologically with the period of savor’aim, of whom historians have a more certain empirical knowledge from the sources outside the Talmud. However, Halivni demonstrates reservations about identifying the stamm’aim with savor’aim, the latest generations of Talmudic authorities. Instead he insists that chronological coincidence does not yet prove social identification. Without going into detail of his position on the question “Are the savor’aim the stamm’aim of the Talmud?” it is enough to say here that Halivni (a) sees the stamm’aim as a historically real people, and (b) derives their chronological placement (and thus reality) from the literary- critical analysis of their thinking rather than from any empirical-historical evidence of their existence. It also goes without saying that Cartesianism in Halivni’s construction of the stamm’aim not only entails thinkers whose names happened to be
52 Stakes
unknown and whose existence is proven by virtue of their thinking alone; it also has to do with the stamm’aim’s impersonal character. They have no other personal or individual traits but rational thinking and being there, which is not enough to differentiate one individual from another. This means my association of Halivni with Descartes reveals not only the implicit workings of the “cogito” in Halivni’s model of stamm’aim but also involves disembodied, indeed dehistoricized rationalism, implying that rational thinking essentially has no body, much less a historical-empirical body, but is connected to body extrinsically only. Paradoxically this ill fits Halivni’s and to a certain extent Maimonidean understanding of how thinking in the Talmud works, let alone who thinks in the Talmud. As a result, contemporary Talmudic scholarship, à la Halivni, is haunted by a modernist embarrassment about the “stretches” (dh.akim) of Talmudic casuistry, which it addresses in part, and which also interferes with, the attempt to use those stretches as analytic tools to discover the historical thinking subject implicit in the text. (I thank Jonathan Boyarin for pointing out that connection.) Neither is it necessary to say that my claim that a researcher thinks in the ways of Descartes does not represent a charge against the results that the researcher achieves. Much less is it a call to reconsider. Instead, it is a statement of a specific intellectual discipline or of a model of thinking that the researcher follows either explicitly, or as is the case with Halivni, tacitly, but no less rigorously. In reading Halivni as a Cartesian thinker, my task is to specify and identify (to make explicit, and thus accessible for critical reflection) Halivni’s assumptions about the relations between language, evidence, and person, rather than trying to show that Halivni is “wrong.” True, Cartesianism is not the only model of thinking. Neither is it chronologically closest to the late ancient texts of the Talmud. However, discerning the fact that Halivni uses that model helps the reader appreciate the complexity of the thinking practices in the Talmud either by limiting the application of cogito sum or by discovering its broader intellectual foundations in medieval philosophy, both Jewish and Christian.4 Finally, discerning a Cartesian approach to thinking practices in the Talmud helps us establish that in the Talmud we deal with production of ideas and the solution of intellectual problems per se, rather than with literary composition alone. These factors collectively make the task of analyzing the thinking practices in the Talmud worth of pursuing. This task, however, is far from a destructive critique of Halivni’s work. Instead, it is a task of limiting the results of the application of cogito sum
Through Talmud Criticism to the Talmud as Thought and Memory 53
to understanding the thinking practices in the Talmud. For this task I do not reject cogito sum. Rather, I am starting from what in Halivni becomes clearer about thinking in the Talmud: that there is theoretical-rhetorical thinking per se, not only a process of literary composing. I then proceed in a hermeneutical movement from what is clearer to what is more obscure. This means progressing from modern, (post-) Cartesian and medieval (post-) Maimonidean views of Talmudic thinking to the thinking practices presented in the archive of texts known as the Talmud. Showing limits in what seems to be clearer opens up a space for asking about what is more obscure. It means only that analyzing thinking practices in the Talmud requires first to apply cogito sum to understanding the practices of Talmudic thinking, as Halivni does, and then to ask about the limits of this application by disconnecting cogito from sum, and more importantly ego from cogitare in favor of analyzing the thinking practices in the Talmud well before they are either ascribed to a person (present or absent, named or nameless, represented or suppressed in representation) or circumscribed in the pattern of refuting and defending, or both ascribed and circumscribed, as the case is in Maimonides’ Introduction to Mishneh Torah as well as in Halivni’s revision thereof. The scholarship of Jeffrey Rubenstein and his students continued Halivni’s literary-realist approach leading to a new, cultural-literary approach. This new approach did not question, but rather promoted, Halivni’s conception of the stamm’aim even further through reading homiletic and narrative texts of the Talmud as a way of reconstructing the “cultural” reality of a rabbinic academy run by the stamm’aim.5 Is this cultural reality literary or historical? I leave this question without an answer, thus inviting further discussion. The above differences between competing literary-formalist and literary- realist approaches translate into critical readings of specific Talmudic texts and therefore call for an illustration. Such an illustration is both necessary and too technical for readers without prior experience with Talmudic texts. I therefore defer providing such an illustration to the appendix to the book, in which I highlight and analyze how Friedman’s and Halivni’s conceptual differences translate into their diverging text-critical readings of the genesis of one and the same Talmudic text. Here, instead, I proceed directly to the first part of the book, in which I analyze the structure of the question of who speaks in the Talmud and discover the implied question of who thinks in the Talmud, which text critical scholars of the Talmud have de facto answered by positing redactors or designers of Talmudic discussions.
part two
Who Speaks?
preamble
The Virtual Author
“Who speaks in the Talmud?” is not a question I asked, nor is my goal to add to the answers given to it. Rather, the question arose in contemporary Talmud criticism. My purpose in this part is to understand the question and its hidden structure. To that end, I attend to the answers the scholars of Talmud criticism have given to the question and pay even closer attention to a question these answers implied. As I will argue, that implied question was “Who thinks in the Talmud?” The text-critical scholars answered that implicit question using a certain notion of authorship; I explicate that notion as that of virtual author. As I argue, in their scholarship, the notion of the author has a structure of a virtual agency. While other characters in the Talmud belong to either an imaginary reality constructed as what is represented in literature or to the alleged reality of history, the virtual agency of thinking in the Talmud defies that distinction, emerging instead as a virtual agent and more specifically a virtual author of the Talmud. Yet, having no proper lens to discern virtual agencies, the 57
58 Who Speaks?
text-critical study of the Talmud ran into impasses, which I highlight and use as heuristic tools. Bypassing those impasses demands a different theoretical lens of analysis. To build such a lens, I study the impasses of Talmud criticism more closely, first in negative terms as impasses and then in positive terms as heuristic suggestions. I consequently use theories of Talmud criticism in order to approach the virtual agency in its proper terms, that is, as a site of multiple and heterogeneous processes of thinking employed in the complex processes in which the characters in the Talmud participate in making sense of the earlier authoritative texts. Their participation, however, does not necessarily mean full control of all thinking involved by the persona of one author. What it does mean, however, is that the thinking processes that Talmud criticism finds in the Talmud can and should be analyzed without any rigid connection to a thinking person, who, as a character, might participate in thinking, but might not own it. Even if such disconnection of thinking from the thinker defies the logic of Talmud criticism, dissociating thinking in the Talmud from a thinking person provides an interpretive ground for going through text-critical studies of the Talmud’s authorship to address thinking and remembering in Talmudic texts in their own right, that is, as virtual acts of thinking, rather than as acts defined by their agents, either imaginary or real. Implied in Talmud criticism’s dealing with authorship in the Talmud is a certain notion of the virtual, and in particular of a virtual agency, an author- like figure of either redactors, composers, or even often virtualized figures of rabbis in the Talmud. I consequently explore that particular virtual agency in its relationship with other particular notions of the virtual in what always already existed, indeed from the very first moment of emergence, as mutually intertwined traditions and bodies of text and thought known as philosophy, rhetoric, and the Talmud. Chapter 4 introduces the model of the “The Author” as a virtual agent. It consequently allows for a more complex but also more exact framing of either “redactors” or “composer” in Talmud criticism in terms of “The Author” as a virtual agent. Chapter 5 addresses the ways of human existence in the literary reality of the Talmud shaped by “The Author” and other characters, speaking or silent. Chapter 6 exemplifies ways in which the characters participate, each in her role, in a collective performance of a well orchestrated and yet never finished dance of making sense of an authoritative text in the Talmud as a way of remembering that text better.
four
Thought and Memory in the Talmud: The Ambiguous Status of “The Author”—and Beyond “Tom!” No answer. “Tom!” No answer. “What’s wrong with that boy, I wonder? You Tom!” No answer. —m ark twain, the adventures of tom
sawyer
1:1–5
“Of the two portions of the Scripture in the mezuzah the [absence of] one invalidates the other; indeed even one [imperfect] letter can invalidate the whole.” “Obvious!” “Rav Yehudah said that Rav said ‘It [sic] was only needed because of the tittle of the letter yod!’ ” “This is also obvious! Rather [it should be answered with] another tradition of Rav Yehudah quoting Rav that says, ‘Rav Yehudah said that Rav said: Any letter that is not surrounded on all four sides by a margin of parchment is invalid.’ ” —babylonian talmud, menachot 29a
Who speaks? Who in the first text is calling Tom? And who says “No answer”? And where and when does this action take place?1 Who in the second text exclaims “Obvious!”? And who is responding? And where and when does that conversation occur? 59
60 Who Speaks?
Luckily, the first text has all we need to answer those first questions. It has a preface signed “The Author, Hartford, 1876.” It lets us know it was “The Author” who said “No answer.” As the text of The Adventures of Tom Sawyer continues, “The Author” introduces Aunt Polly and the other characters. Aunt Polly is first introduced in her little house as she calls for Tom, a character who already has been explained in the preface as “a combination of the characteristics of three boys whom I knew” and who “therefore belongs to the composite order of architecture.” A signed preface, carefully introduced characters, fairly well defined scenery, time and place, “The Author,” and even his name, “Mark Twain,” on the book cover are true luxuries we enjoy. What else do we need? Yet there is more. From external historical evidence, we even know the real name of the writer, Samuel Clemens, who invented Mark Twain, Aunt Polly, Tom Sawyer, and the other characters, or, as some people would choose to see it, Samuel Clemens and Mark Twain were the same person. Yet, none of this information is given for the second text. We must rely on the probabilities of inference only. Maybe the first person speaking is a recitant who proves his or her recollection of a clause of the law by reciting it by heart. Maybe the second person is another, more advanced student who says “Obvious!” and is thereby attacking the recitant for saying obvious things instead of reciting the clause authentically. (By implication, “obvious” means having no point, while a good law must always have a point.) Maybe the third speaker is yet another (or maybe the same?) advanced student who now defends the recitant against the attack. This advanced student uses a quote from earlier authorities in an attempt to explain that “it” (contextually, meaning the recited law)2 is in fact not obvious but rather has an important point: “It” is needed, if only because of the tittle of the letter yod. Maybe in saying “This is also obvious!” the teacher shows a weakness in such an attack and defense and instead demonstrates a more elegant way of doing the same exercise. Ironically, or perhaps with a didactic purpose, the teacher acts along the same lines: to attack by saying “This is also obvious!!” and to defend using a quote from the same earlier authority, but citing a thematically more remote place from the same authority, Rav Yehudah,3 whom the advanced student had chosen to cite.4 We are not told. Maybe the dialogue between the teacher and the apprentices is staged in the courtlike setting of a house of study—or is it a part of a law school curriculum? We are not told that, either.5
Thought and Memory in the Talmud 61
However, do we really know much about Mark Twain as an author, either? Who is he? Is he just another character, perhaps more real than Aunt Polly and Tom Sawyer, but still halfway in the world of fiction? Alternatively, is Mark Twain as real as Samuel Clemens? In the preface, “The Author” assures us that the characters are drawn from reality, and we are, of course, to believe him—well, as long as we believe that “The Author” is a real, rather than a literary, figure. Does “The Author” belong to history or rather to literature? For that, we are not given a clear answer. Of course, this is all minor compared with the abundance of information we get about other questions, but this lack of certainty about “Mark Twain” as “The Author” of The Adventures of Tom Sawyer disturbs more when we think about authorship in the second text, in which nothing at all is certain. Who is the “Mark Twain” of the second text, and even more intriguingly, who might be the “Samuel Clemens” of it? The second text is of course from the book called “the Talmud.” Contemporary scholars of the Talmud have posed the above questions about the second text and similar texts, and have used different approaches to answering them, but they all have encountered a series of paradoxes or antinomies. These antinomies have posed a limit to the theories in which they arise. In this chapter, I will analyze the array of these approaches and the antinomies to which they lead. In the chapter that follows, I will take a different route and try to learn from the fact that there are antinomies. To that end, I will consider the texts of the Talmud as an archive of exercises of memory, an approach that presents the Talmud in such a way that these particular antinomies either do not arise or reveal something about “The Author” of the Talmud as a thinking and remembering existence, performing and performed in the Talmud. To develop a theoretical framework of analysis with which to approach the text of the Talmud in this way, I will actively engage the resources—the approaches and theories—of Talmud criticism while broadening the context of inquiry to include conflicting rhetorical and philosophical understandings of what it means to remember and what role thinking may play in memory. The analogy with Tom Sawyer, Mark Twain, and Samuel Clemens will help illustrate divergent approaches to analyzing the authorship of the Talmudic texts and the antinomies at which they arrive. The same analogy will also help introduce the rhetorical and philosophical views of thinking and remembering by way of understanding the relation between the figure of “The Author,” on the one hand, and thinking and remembering, on the other.
62 Who Speaks?
The Name of “the Talmud” (First De-essentialization) As a first step, let me gloss and thereby de-essentialize the name “the Talmud.” “The Talmud” is a name retroactively applied to a set of formalized and thereby idealized protocols of legal, ethical, and religious conversations or exchanges staged in a rabbinic academy in Babylon at around the fifth century C.E. that took its present shape no later than the second half of the first millennium. The discourse recorded in the Talmud is centered on the Mishnah, a third-century code of instructions for rabbinical courts. The discussions in the Talmud follow a protocol that begins with one person, a tann’a, reciting a short section of the Mishnah from memory. Others then exercise heuristic attacks and apologetic defenses of the speaker’s citation in an effort to reaffirm the validity of the passage from the Mishnah, which would otherwise have to rely on merely mechanical memory of those who are reciting the text by heart.6 What is predominantly at stake in these conversations is the reliability and meaning of the quotation for application in court practice. The text is unusual: It is not a narrative addressed to an outsider, nor is it signed by any author. The only speakers in it who are identified by name are the quoted teachers, the amor’aim or earlier Talmudic masters, and the people connected with them. The Talmud recounts the sayings (dicta) and deeds of the amor’aim, either directly in their names or in the names of others who witnessed them or witnessed the witness who witnessed them. In most instances in which the word or deed of a master is cited, there is intense discussion of what precisely the master has done or said. The word “said” means not only “said” in the sense of being an accurate citation of the master’s words but also what the master “thought,” “taught,” or “claimed” in the sense of what an interpretation of those words might be. Another part of the Talmudic protocol includes no amor’aim. Instead, it presents a pure line of polemical soliloquies (in the form of a conversation with oneself) in which one and the same speaker heuristically refutes but ultimately defends the Mishnah. Sometimes a polemical issue in a soliloquy is resolved by citing a master; other times, a Talmudic soliloquy reaches a conclusion with no master to cite. The texts in the Talmud are both a record of and a protocol for the above types of conversations, which, due to their rhetorical nature, always remain intrinsically open for expansion through new refutations and defenses. After the protocol was either interrupted by extraneous forces of history or
Thought and Memory in the Talmud
63
proclaimed to be “sealed” by subsequent generations of rabbis, who thereby could use it as a stable source of authority, these conversations continued to be recorded, and at a certain point, the Talmud became a book. But it was never intrinsically finished. From the time of its completion, the text has been integral to Jewish study and scholarship. As a way of analysis, the Talmud has also “continued” itself through the medieval and modern periods in an extensive body of post-Talmudic scholarship. Let me now go back and reread the example of Talmudic text quoted above as a fragment of such a protocol of refuting, defending, and ultimately remembering. This text presents a short and concise example of a unit of Talmudic discussions, which are often much longer and more elaborate. The technical name for any such unit is a sugya, roughly interpreted as an “argument.” Such units or sugyot (in plural) are building blocks of larger forms of Talmudic discussions—tractates and, ultimately, the whole corpus of texts called “the Talmud.” This example of a sugya is concise, but representative, even if other sugyot may include components that are not represented in this example. For one thing, this example does not include homiletic narratives. However, other units will include all the components this example includes. The hitherto predominant idea of the rhetorical structure of a sugya was expressed in medieval Talmudist notions of refutation and defense (קשיא “—ופירוקrefutation and counterrefutation,” or “—קשיא ותרוץa refutation and an escape from it”). This idea is also shared by modern scholars of the Talmud. In the example above, the exclamations “Obvious!” and “This is also obvious!” rhetorically function as refutations. In turn, the responses “Rav Yehudah said” and “Rather [it should be answered with] another tradition of Rav Yehudah” rhetorically function as defenses. Understood in these terms, the rhetorical structure of the sugya comprises two (in other cases more) rounds of refuting and defending. However these rounds do not stand alone. Rather, the second round builds on refuting the first, so that the structure of refutations and defenses is cumulative, rather than simply sequential. This example of sugya also contains the named characters Rav Yehudah and Rav. Based on other occurrences of these characters in the text of the Talmud, medieval genealogists of rabbinic authorities explained that Rav Yehudah belongs to the second generation of authoritative rabbis of the post-Mishnaic period, called amor’aim, and Rav (the student of Rabbi Yehudah ha-Nasi, the “composer” of the Mishnah, and the teacher of Rav Yehudah) belongs to the first generation of amor’aim in Babylon. If genealogy is translated into
fup-dolgopolski-04.indd 63
10/10/12 3:53 PM
64 Who Speaks?
chronology, this would take us respectively to the earlier and later periods of the third century C.E. In the medieval genealogical literature, Rav is also recognized as the founder of a rabbinic school (academy) in Sura, and his student, Rav Yehudah, is recognized as the founder of a rabbinic school in Pumbedita. Both of these amor’aim are remarkably absent from the stage of refuting and defending. Instead, they, or more specifically, their dicta, are quoted (or recalled) for the purposes of defense. By whom? To answer this question is first to note that this sugya contains other characters who are not explicitly identified by either name or status, but only by the fact that they speak and by the way in which they speak. They say “Obvious!” and also say “Rav Yehudah said,” as well as “Rather [it should be answered with] another tradition of Rav Yehudah.” Based on the roles they perform in the discussion, I propose to call these speakers respectively students and teachers in a rabbinic academy. In terms of genealogy, these teachers and students are from later generations and lower in status. In terms of chronology, they belong to a later time than the absent characters, Rav Yehudah and his teacher, Rav, who both are absent in the “now,” but are assumed to belong to the past. Yet in literary terms, these titleless and nameless speakers are the only ones who act on the stage. The next question to ask about them is what these titleless characters do. According to the medieval interpreters of the Talmud, these characters heuristically refute and ultimately defend laws of the Mishnah (and in other examples, also dicta) that they recall and discuss. Without contesting that interpretation, I propose to include it in a broader framework of understanding. This broader framework has to do with the more nuanced, refined, and competing rhetorical and philosophical notions of memory, such as recollection, remembering, forgetting, verifying, inventing, and refuting. By heuristic refutation and ultimate defense of what had initially been only a recalled and recited formulation of law, these characters also take memory to a much higher level—from the mechanical memorization of the law at the start of the sugya (“Of the two portions of the Scripture in the mezuzah,” etc.), to the rhetorical verification and reinvention of it and thus to a better understanding of that law by the end of the sugya. This higher process of remembering is complex. The three components it involves—refutation, (re)invention, and verification—collectively contribute to an outcome that is a more precise remembering of the law and of its
Thought and Memory in the Talmud 65
applied meaning. In terms of the discipline of rhetoric, the ultimate work of these characters is that of remembering by refuting. This process includes inventing—that is, both discovering (finding out) what was the case about something and inventing (creating) something new about it—the practice embedded in the rhetorical notion of inventio. These characters collectively contribute to this higher activity of remembering, but none of these characters owns or orchestrates that activity. Although each of these titleless and nameless characters participates in the process of remembering, none of them controls that process. Consequently, the next question about this sugya is whose memory is it, or who remembers in it? In the later chapters of this book, this will in fact prove to be a different question: What way of being in the world is involved in, or performed by, remembering? For now, however, I will deal only with the initial form of the question: Who remembers in the sugya? To begin answering this question in this form is first of all to recognize the presence of yet another character on the stage. Equally nameless and titleless, this character is indefinite in terms of grammatical number. It therefore can be said to be numberless. This character performs very differently from the teachers and students. The function of that character is neither to speak nor to be quoted, but to let other characters do what they do; that is, its function is to orchestrate the dance of discussion, the rounds of increasingly refined remembering. If there is anybody at all who controls or owns this complex process of remembering, it would be this hidden character, the one who orchestrates the memory dance. Who is that character?7 Whose is the higher level of memory in this text? To begin answering these difficult questions, let me come back to the analogy of this sugya with the text of The Adventures of Tom Sawyer. Using this analogy to answer the question of who the character is who orchestrates the sugya as the performance of memory involves answering the seemingly simpler question of who the author or composer is in these two texts. For that purpose, the analogy with The Adventures of Tom Sawyer will prove both useful and limited. By this analogy, the question of who remembers in the sugya would be reduced to the question of who its author is, and I will explore that question first. The question thus is: Who is the Mark Twain of the Talmud? The scholars of the Talmud of the last century developed a number of elaborate text-critical approaches to deal with this question. These scholarly
66 Who Speaks?
approaches help reveal both the richness and the complexity of the figure of “The Author” and also demonstrate the methodological limitations that need to be removed in order to examine it. Two predominant and mutually competing text-critical approaches to the Talmud contribute to understanding the figure of “The Author.” Following the suggestion of earlier twentieth- century academic scholars of the “exact Talmudic science,”8 representatives of these approaches maintain a binary division of the text into its form (text) and its content (the reality and ideas represented in it). These scholars consequently need to establish and analyze the form of the text first, well in advance of drawing any conclusions either about its content or from that content.9 I describe these two paradigmatic (and competing) text-critical approaches as literary-formalist and literary-realist. The cardinal difference between them is between a literary versus a historical conception of the production of texts, which appears in their different conceptions of the placement and nature of the figure of “The Author.” The scholars of the literary-formalist approach, of which Shamma Friedman is representative, place the figure of “The Author” outside the content of the text and treat it as a figure of the literary production of the form of the text, a figure that may or may not have a historically real equivalent. In contrast, for scholars who employ the literary-realist approach, such as David Halivni, the figure of “The Author” is located both in the reality that the text represents and in the literary representation of that reality, that is, both in the history of text’s production and in the content of the text. Thus, the scholars of the literary-formalist approach emphasize only the literary rather than the historical figure of “The Author” as a producer of texts. The scholars of this approach talk about individual “designers” of compositions in the final text of the Talmud (Talmud shelefanenu, “the Talmud in front of us,” in Friedman’s terms) while only allowing for, but not requiring the existence of historically “real” equivalents for these figures. The role of the literary-formalist scholar is to describe transformations of the content of the Talmud from its original form, by virtue of the work of designers, to the text’s final form. The figure of “The Author” as the “designer of the final composition” serves these scholars as an instrumental presupposition for critical analysis of the text. However, the construction of the figure of the designer of the composition does not become the subject of critical analysis. By contrast, the scholars of the literary-realist approach insist on the historical existence of the authors, or as these scholars call them, the
Thought and Memory in the Talmud 67
“redactors,” of the final form of the Talmud and also consider them to be disguised coparticipants in the historical reality that the Talmud as a literary text represents to its readers. The role of the literary-realist scholar is therefore to help the readers of the Talmud see the disguised actions of these historically “real” redactors. I have introduced these approaches in very general terms, but I will now illustrate basic elements of each of these two approaches using our sample sugya as an example. Although to the best of my knowledge, this particular sugya has not been directly addressed by the scholars of these approaches—in part because they have attended to much more complex cases—I employ it here to help build an abstract, comparative-theoretical model of the basic elements of each of these text-critical methods. In the literary-realist approach, pioneered by David W. Halivni, text- critical scholars look for so-called “forced interpretations” of the dicta of the earlier masters or amor’aim in the texts, which the scholars consequently ascribe to the “anonymous redactors” of these texts. The more precise formulation of the task of the critical scholar is thus to unveil the hidden work of the anonymous redactors of the dicta of amor’aim (in this example, of Rav as reported by Rav Yehudah). As the figure of “The Author,” the redactors are construed as actively thinking intellectuals with an apologetic agenda of their own: to assure the efficiency of the dicta in the arguments of either refutation or defense. According to these scholars, the goal of the redactors was to make sure no dictum fails to meet the redactor’s standard. The amor’aim never fail as effective refuters or defenders, and if they seem to fail, the redactors must “force an interpretation” in which the dicta can be defended. They can use several methods of forcing an interpretation. The redactors can rescue a threatened dictum by creating a new sugya or a fragment of a sugya in which this dictum can be inserted to play a more satisfying role in the opinion of the redactors. The job of a literary-realist scholar is to find the forced interpretation and then to reveal the work of the redactors. For example, the “it” in the defense of the first round’s “Rav Yehudah said that Rav said ‘It was only needed because of the tittle of the letter yod!’ ” can be seen as a sign of such a “forced interpretation.” Indeed, in this dictum of Rav’s, the “it” surely refers to a formulation of some law ascribed to the sages of the Mishnah. Yet that dictum gives no direct indication of the reference to the particular law on “two portions of the Scripture in the mezuzah,” discussed in the sugya. This discrepancy alone might indicate the presence of a forced interpretation.
68 Who Speaks?
The redactors, of course, do not necessarily directly change a dictum they want to rescue. Instead, they may reinterpret that dictum by generating a discussion ascribed to anonymous teachers and students, of which the dictum would be a part. In such a discussion, the teachers and students speak on behalf of the amor’a or—as the redactors might have done in this and similar cases—create a new sugya in which the dictum would play a more satisfying role. In our sample sugya, the dictum is used in the first round of refutation, and this refutation is refuted in the second round, thus serving as a necessary, but intermediary step on the way to the “right” defense. This organization indirectly suggests that the first dictum of Rav might originally either belong to a different polemical context or refer to a different law—which would thematically relate more directly either to the graphical elements of the letters or perhaps not to mezuzahs but rather to either phylacteries or to Torah scrolls. In addition, unlike the final context in which it appears, in its original context, the first dictum of Rav might have referred to the law by way of clarification only, that is, without any polemical context whatsoever.10 Yet whatever this “original” context might be, the redactors transposed the dictum of Rav into the new context, our sample sugya, in order either to make that dictum polemical, rather than merely interpretive of the law, or to make it more effectively polemical than it would otherwise have been. In this approach, “the redactors” or authors were programmatically constructed invisible participants in Talmudic discussions,11 acting in the disguise of anonymous teachers and students and simultaneously as the historical-biographical creators of these discussions. Based on their apologetic program of making the amor’aim effective in polemics (rather than letting the amor’aim be mere elucidators or interpreters of teachings of the Mishnaic sages), the redactors made these anonymous teachers and students act as they do. The redactors thereby function as both Mark Twain (they act in the Talmud) and as Samuel Clemens (they at the same time produce the Talmud). In sum, in the historical-literary approach, the difference between Samuel Clemens and Mark Twain is generally either not recognized or is regarded as trivial. In turn, in the literary-formalist approach paradigmatically developed by Shamma Friedman, the scholars eliminate the figure of “The Author” as a participant in the text and instead conceive that author as only the designer of the text. These scholars look not for the activity of the redactors in the text, but rather for transformations of the text of the Talmud from its original or “old” form (Talmud kadum, lit., an “anterior” or “early” Talmud), which
Thought and Memory in the Talmud 69
included only apocrypha and dicta, to its presently known “new” condition (Talmud shelefanenu, lit., “the Talmud in front of us”), which also includes a highly developed new form, the discourses or fragments of discourses by nameless students and teachers. In addition, scholars of this approach pay close attention to migrations of texts and materials from one of these three forms—apocrypha, dicta, and discourses12—to another. The starting point of their theory is that the sugya is not a record of a live discussion in a rabbinic academy, as some uncritical readers might claim, but rather is a carefully designed literary composition. These scholars then reason from observing the design to the existence of a designer—either in the sense of a literary figure or perhaps of a historical person or a group of people working in either one or several generations.13 For example, in this approach, our sample sugya is a well-designed literary construction, in fact one applied more than once in talking about similar, but distinct topics—mezuzahs and phylacteries (Babylonian Talmud [BT] Menachot 29a and 34a). The design of the sugya is expressed, first of all, by strict sequence of the rounds of refutations and defenses. The second round is not possible without the first, and the first is not persuasive enough without the second. What is more, there is some slight evidence of the migration of the material, if not between forms, at least between rhetorical functions. In the “old” Talmud,” the dicta of Rav served as an interpretation of two apocryphal traditions of the sages of the Mishnah, which in the final “new” form of the Talmud appear still as two apocrypha at 29b, although they change their rhetorical function: They now refute Rav Yosi’s interpretation of the dicta of Rav. For scholars of this approach, both the carefully crafted literary structure of the sugya and the observed reassignment of forms, or, as in this example, of the rhetorical functions of texts, serves as evidence of there being an intelligent designer behind this design. This designer is at the very least a literary figure in the production of the text, a figure that does not appear (or hides) as a character therein. This designer is analogous to “The Author,” in particular, because the literary existence of the figure of the designer welcomes the existence of a historical equivalent, even if it does not require it—a historical group of people who are analogous to Samuel Clemens. In sum, in the literary-formalist approach, it is fully understood that the designer of the discussions or “the Author” is clearly different from other characters. Because, in the literary-formalist approach, “the Author” is separated from other characters, the difference between Mark Twain as a character and the “historical” Samuel Clemens is no longer
70 Who Speaks?
trivial: The former is a literary counterpart of the composition, and the latter is a historical-biographical subject. Yet because both are separate from other characters in the composition, the difference between Mark Twain and Samuel Clemens is recognized but still is not strictly required. As a result, the scholars of both text-critical approaches concentrate their critical work on analyzing the production of texts as representations of content—the life of a rabbinic academy. As a necessary instrument of analysis, these scholars theoretically construct the figure of “The Author” as either “redactors” or “designers” who produce the final form of the text. These scholars further use this construct as a critical tool in analyzing the texts of the Talmud as literary representations of reality, in comparison with how that reality might look like, if the authors’ impact on representation is critically analyzed. However, they do not make their constructed figures of “The Author” an object of critical analysis. Their findings about the construction of the texts are therefore as indispensable as they are insufficient. They help us understand a significant part of the text’s production, but do not critically clarify the figure of “The Author” that they construct and use, but do not analyze. Therefore, answering the question “Who is the Mark Twain of the Talmud?” requires us to analyze the construction of the figure of “The Author” by using proper critical tools, and these tools need to be found elsewhere. To begin this critical work, we need to observe that the figures of “The Author” that the text-critical scholars have constructed both reveal and occlude the thinking processes expressed in the text of the Talmud. These scholars help reveal that thinking process by virtue of the fact that the constructed figures who occupy the role of “The Author” in the two different approaches do, indeed, think, albeit in different ways: either as implementing an apologetic agenda or as constructing the literary design of the Talmud in the format of refuting and defending. Yet these figures of “The Author” also occlude that thinking expressed in the text, because each simplifies and essentializes the figure of the author and each reduces that figure to the sole agent, owner, and originator of what is thought. In sum, their use of the figure of “The Author” occludes the complexity of both this figure and its relationship to thinking (not to mention to remembering), a complexity that is expressed in the construction of the text. To use the results of text-critical scholarship to answer the question “Who is the Mark Twain of the Talmud?” I will need to pay much more
Thought and Memory in the Talmud 71
attention to the complexity of that figure, or, in more general terms, I will need to de-essentialize that figure. Despite the great contributions of the text-critical approaches to the Talmud, in particular in terms of constructing different versions of the figure of “The Author,” the question of who “The Author” in the Talmud is needs to be asked afresh and anew, with the text of the Talmud understood as a form of expression of the processes of thinking (and as I will further argue, remembering), and not only as a form of the represented content.
The Ambiguous Status of “The Author” Who, then, is the Mark Twain of the Talmud? In the light of the shortcomings of the text-critical approach, the first question about this question is where to look for an answer: in the historical reality of the text’s production or in the world created and represented in the Talmud as a literary text? To approach that question, we need to clarify the way in which the status of “The Author” is uniquely ambiguous, situated as it is between the Scylla of literary representation among the other characters of the Talmud and the Charybdis of the empirical-historical (“biographical”) production of the text. One ideal type of scholar (call them “realists”) would suggest that “The Author” is the historical producer of the text: Mark Twain = Samuel Clemens, who created Tom and Aunt Sally. Another type (call them “personificationists”) would consider “The Author” to be a literary character among others, so that Samuel Clemens created the characters Tom, Aunt Sally—and Mark Twain.14 Both “realism” and “formalism” in Talmud criticism have strong points to them. Of course, anyone (or anything) occupying the position of “The Author” appears to be more real than other literary personae, since (despite a popular claim that writers make) anyone (or anything) occupying that position and his or her or its thinking has more control over the characters than the characters have control over “The Author.” However, the relatively greater degree of reality of “The Author” does not necessarily make that figure more real in historical terms. “The Author” might still be only a function (and fiction) of the literary work: Not Samuel Clemens, but Mark Twain “authors” The Adventures of Tom Sawyer. The reading public can readily ignore the difference between Clemens and Twain, but cannot forget it.
72 Who Speaks?
There always remains a place for arguing whether or not Mark Twain is historically real. Remarkably, any empirical tool or experiential model will hardly ever be found satisfactory to give a final answer to the question of whether Clemens and Twain represent two different figures, or just one.15 This question is thus above and beyond empirical-historical judgments about reality. Yet even more remarkably, the transition between Clemens and Twain, both literary and historical in its nature, can never be fully explained in exclusively “realist” or exclusively “formal-literary” terms. The picture of the relation between Twain and Clemens becomes much clearer owing to the fact that however empirically real (in particular, however graspable in the imagination) they both might be, the “realistic” Clemens and the “literary” Twain remain mutually exclusive rather than complementary. It is either/or: Either Clemens is the full historical equivalent of Twain, or Twain is a literary fiction that Clemens created and as such is no more real than Tom Sawyer. Therefore, the divergence between “realists” and “literary formalists” cannot be resolved in empirical terms, either. There will always be room for at least two competing empirical approaches to placing Twain between Clemens and Sawyer: Twain will be seen as either closer to Sawyer or as closer to Clemens. Accordingly, there will always remain a place for “realists” and “formalists” to disagree: Unlike either Clemens or Sawyer, Twain will be put either on the outside of the text or inside it, and for that reason, neither of the two approaches can prevail. Therefore, any attempt to answer the question “Who speaks in The Adventures of Tom Sawyer?” falls into this aporia. The same problem of undecidability applies to the question “Who speaks in the Talmud?” Hence the appropriateness of a more modest question, “Who is the Mark Twain of the Talmud?” Who occupies the ambiguous, liminal position of “The Author” in that text, a position in the text, but not of it?
From Speaking Author to Thinking Author Since, on the level of the empirical adjudication of the figure of Mark Twain, neither the literary constructivists nor the realists can prevail, understanding the divergence between these two approaches should reach beyond the empirical plane alone. Indeed, between these two approaches there is an even more subtle, albeit more important and less empirically graspable
Thought and Memory in the Talmud 73
issue at stake. It concerns the core notion of both “formalist” and “realist” approaches, which is the notion of “The Author” as a thinking being. By virtue of thinking, “The Author” designs the characters in “the composite order of architecture”16 and controls their actions. Yet, thinking is not an empirically graspable virtue, because it is not a fully empirical phenomenon. This is particularly important for understanding the figure of “The Author” in the Talmud. Of course, the question of placing “The Author’s” thinking either outside the text or inside it is foundational for understanding the “realist” and “formalist” approaches to any literary text. Yet this question becomes particularity important for understanding the difference between “realists” and “formalists” in studying the Talmudic texts. As both “formalists” and “realists” would require, in the Talmud, there theoretically might (or some would even say “must”) be a character of “The Author,” a Mark Twain. However, unlike in The Adventures of Tom Sawyer, he happens to have direct representation neither in the text of the Talmud nor in a preface, much less on a signature page, which the Talmud does not have. As I have already shown, in Talmudic scholarship, we deal with different notions about positioning the Talmudic equivalent of Mark Twain between Samuel Clemens and Tom Sawyer, or, in other words, with placing the Talmud’s author(s) (whether the “redactors” of the teachings of the earlier masters or the “designers” of the compositions of Talmudic discussions) between literary and historical realities. This positioning stands in direct connection to the competing understandings of the role of “The Author’s” thinking. How can either a literary theorist in general or a literary researcher of the Talmud in particular discern that thinking? It goes without saying that neither “formalists” nor “realists” allow for a separation between whoever occupies the position of “The Author”17 and his or her thoughts.18 However, what “goes without saying” can often define the end result in advance. In this case, it does. For both “formalists” and “realists,” Mark Twain thinks as surely as Clemens does. Thus, the scholars who assume either of these approaches remain faithful to the Romantic notion that thinking must belong to its author, or simply put, must have an author, so that neither the “realist” exclusion from the literary text of the author and his or her thoughts nor the “literary constructivist” inclusion of the author and his or her thoughts in that text affords any possibility whatsoever to establish the reality of either the author or the narrating agent separately from his or her thoughts. The fact of existence of “The Author”—in either the “real” world
74 Who Speaks?
or in the world represented in literature—remains totally dependent on the fact of the thinking activity ascribed to that author. However, such a rigid connection between whoever occupies the position of “The Author” and his or her thinking activity leaves important possibilities out. These possibilities include thinking that emerges in the text with the author left outside the text, or, alternatively, thinking that is located outside the text with the author encompassed in the text. Yet instead of considering these possibilities, scholars who foster either “literary constructivism” and “realism” or both rigidly (and naturally, for modern and medieval texts) connect the notions of having a thought and being the author of that thought. The stakes involved in keeping a rigid linkage between thoughts and authors are so high that they never have to articulate that linkage, much less expose it to questioning. For both “formalists” and “realists” the linkage between authors and thinking has seemed too obvious and questioning it too dangerous for the integrity of the broader Romantic idea of “authorial intent,” which is a particular version of the association of thinking with a thinking subject. I use the term “Romantic” in the general sense that posits a direct and unquestioned link between thoughts and personality. In this sense, both “literary constructivists” and “realists” remain faithful Romantics, contesting between themselves about where to put Mark Twain and “his” thoughts on the scale between Clemens and Sawyer. They would not allow any weakening of the strength of the rigid linkage between authors and thought processes.
Thinking beyond “The Author”? (Second De-essentialization) However, beyond either the modern tradition of Romanticism or its continuations in text-critical scholarship of the Talmud, the connection between thoughts and their authors has been far from either obvious or even welcome. For example, the second-century church father Irenaeus, the bishop of Lyon, carefully kept his thinking separate from any claims of his personal authorship. In presenting himself as orthodox, he defined heterodox heretics as people who claim themselves to be the authors of their thoughts. Authorship of a thought meant for him heterodoxy and thus heresy. Irenaeus positioned himself as someone who claimed no thoughts as his own. Instead, he wrote on behalf of the authority of the divine.19 To contrast Irenaeus with a
Thought and Memory in the Talmud 75
seemingly similar modern example, Hegel defined himself not as an author who generates thoughts, but rather as a thinker (meaning for him a rational philosopher, as opposed to an imaginative—even if perhaps rationalizing— religious person) who follows the thought process. He construed his own thought process and hence his writing as the emergence and evolution of the Spirit (which for him was both an emerging thought and an evolving being) according to both a phenomenology and a logic of transformation that encompassed Hegel as a thinker. Unlike Irenaeus’s separation between thoughts and personal authorship, Hegel’s way of detaching thinking from the subject is possible only because thinking is reattached to a different subject, “Spirit,” which in itself is a major topos of Romanticism. That Romantic topos allowed the minor nineteenth-century American Transcendentalist poet Jones Very to claim that it was not he who wrote his poems, but the Holy Spirit through him, which prompted one of Ralph Waldo Emerson’s few known witticisms: “Cannot the Holy Spirit parse, and spell?”20 In contrast to these modern examples, for Irenaeus, the divine does not speak instead of him or through him, but rather Irenaeus writes with the authority of the divine, unlike those “authors” or “heretics” who write and think on the basis of their own authority.21 Like Irenaeus, the rabbis also separate authorship from authority. The Talmudic Tractate Menachot, in a discussion that follows our sample sugya, cites a famous homiletic narrative that shows just that. In this example, the authority of “the laws given to Moses on the Mount Sinai” comes through Moses, who, however, is so much not the author of those laws that he cannot even understand them. Moses leaves understanding of the laws to Rabbi Akiva, who lives generations after Moses. Yet Rabbi Akiva does not claim the authorship of what he understood, either. Instead, he positions what he has invented as already belonging to the law given to Moses. Moses delivers the law, but does not understand it; Akiva understands the law, but does not own it. By the same token, neither of the two is the author. And G-d is not the author of the laws delivered to Moses and explained by Rabbi Akiva, either. To avoid any possible misunderstanding of this point, let me recall another homiletic narrative about the rabbis, in which the daily preoccupation of G-d is to study and to understand the rabbinic inventions in law that rabbis generate in their academies day by day. Like the rabbis, G-d is busy learning the latest interpretations of the law that G-d gave to
76 Who Speaks?
Moses. The divine is thus an authority but is not an author. In Irenaeus’s terms, this would mean that G-d is neither a heterodox nor an author. As these examples suggest, in addition to ascribing the thinking process to subjects represented in the text (Mark Twain, Tom, and other characters in The Adventures of Tom Sawyer) or to the subject who produced the production (Samuel Clemens), there is yet another approach to thinking. In that approach, thinking is treated simply as thinking, rather than as a property ascribed to an author as the subject of thinking. It is to that approach that I turn in the next chapter.
a In the search for a subject who thinks—in particular, who redacts or designs texts—scholars in the text-critical tradition were hardly ready to grant thinking anything but a personal basis. They instead construed thinking in rigid association with a person (character, author) found either in historical reality or in the reality represented and/or constructed in the text. As a result, either programmatically or de facto, in addition to the aporia between the “realist” and “literary constructivist” positions on the position of “The Author” in or outside the text, which might be called the “in/of” aporia—that is, the whether the redactors of the Talmud should be placed inside the reality represented in it (Mark Twain) or outside that reality, as those who created it, but are not part of it (Samuel Clemens)—these scholars also confronted two further and related aporias in which the tension between “realism” and “formalism” has ramified. The first of these is the aporia of the emendation versus the interpretation of the texts of the tann’aim in the Talmud—that is, of the redactors’ “real” implementation of an apologetic agenda versus the rabbis’ construction of the literary design of the Talmud. The second is the aporia of Talmudic culture as either the actual environment from which the Talmud historically emerged, that is, as a rabbinic academy in Babylon, or the environment that the text of the Talmud, as text, implicitly represents and constructs. As we have seen, construing the anonymous redactors as either “Twains” or “Clemenses”—whether in literarily represented or “real” worlds—is predicated on proving these subjects’ existence by their thinking. Likewise, both the act of emending and the act of interpreting must feature “The Author’s” intention and thus assume the rigid connection between
Thought and Memory in the Talmud 77
an author (the subject) and the thoughts that the author might have. And the notion of a Talmudic “culture,” that is, the culture of the “anonymous redactors,” is predicated on the existence of the redactors as thinking subjects and thus does not provide sufficient ways to decide whether the reconstructed “culture” in which the redactors live is a part of the historical world or a literary construct. Each aporia is thus based on the theoretical assumption of the existence of a thinking subject. In the more detailed exposition of these three aporias to follow, I will trace that assumption at work, attempt to remove it, and propose a different model for Talmudic discourse, which I will explore in the chapters to follow. I base this model on a descriptive-theoretical rather than on either a “formalist” or a “realist” empirical approach to the text of the Talmud. That model disconnects thinking from the subject (its negative claim) and reconnects thinking to memory, now understood as the main activity performed in the Talmudic text (its positive claim). However, one consequence of my analysis thus far of the text-critical scholarship of the Talmud in the light of the aporias that characterize them should by now be clear: We must modify our question “Who is the Mark Twain of the Talmud?” and ask a related, but quite different, question. If it is undecidable whether the redactors or of the Talmud should be placed inside the reality represented in it or outside that reality, then to ask, “Who is the Mark Twain of the Talmud?”—that is, to ask, “Who occupies the position of ‘The Author?’ ”—is to ask a question whose answer cannot be definitively decided. Instead of the presence, real or literary, of a thinking subject in that text, we need to ask, “What kind of human existence do we find in the text of the Talmud?” This, as we will see, entails the question of what it means to be a thinking being without being a classical Cartesian subject.
five
Human Existence in the Talmud: Thinking as Multiplicity and Heterogeneity
In the last chapter, I identified three aporias in which, given the assumption that in the Talmud thinking must occur in a person (character, author) found either in historical reality or in the reality represented and/or constructed in the text, it is undecidable where the position occupied by what I have been calling “The Author” can be located. In addition to (1) the “in/of” aporia (the undecidability of whether the redactors of the Talmud should be placed inside the reality represented in it), there is (2) the aporia of the emendation versus the interpretation of the texts of the tann’aim in the Talmud, and (3) the aporia of Talmudic culture as either the actual environment from which the Talmud historically emerged, that is, as a rabbinic academy in Babylon, or the environment that the text of the Talmud, as text, represents. Each of these internal aporias remains irresolvable unless the discussion switches to a new theoretical framework in which thinking in the Talmud can be dissociated from the figure of the thinking subject—“The Author,” the 78
Human Existence in the Talmud 79
“anonymous redactors,” or the “designers.” In what follows in this chapter, I explore these aporias in greater detail in a negative way, in terms of the antinomies or internal impasses that characterize them, that is, as problems that cannot be solved in the particular theoretical frameworks in which they are originally formulated, and as a way to introduce the possibility of a new, broader framework of thinking about thinking in the Talmud. In the chapters that follow, as I attempt to build such a framework of analysis, I then treat these aporias and their respective theoretical frameworks positively by using them as resources for that new framework of thinking about thinking and remembering in the Talmud. To probe that approach, I begin by reassessing the internal antinomies of the competing text-critical approaches to the texts of the Talmud. I reframe these antinomies as points at which the tight connection between thinking and the figures represented in the text can be loosened. As a result, the rigid linkage between thinking and thinking subjects in the Talmud gets undone. And while in both text-critical approaches the text is taken to be a representation of the life of the rabbinic academy, identifying these weaker points in the linkage between thinking and thinking subjects allows me to disentangle the thinking process from its occlusion in the figures of this representation, such as thinking authors or characters in the Talmud. It thus leads to a way to begin to answer the question “What kind of human existence do we find in the text of the Talmud?” I consequently search for a broader theoretical framework in which to access the thinking in the Talmud as thinking, rather than as the activity of a certain person. I broaden the framework in which to analyze thinking in the Talmud by introducing the rhetorical and philosophical approaches to thinking and practices of thinking in late ancient texts, discerned through (and despite) the lens of our contemporary rhetorical and philosophical understanding of what it means to think. One obvious heuristic significance of including the philosophical and rhetorical perspectives of analysis is that in the philosophical and rhetorical practices of thinking in late antiquity, associating thinking with a thinking subject often was either not a viable possibility or was dismissed as either a misunderstanding or oversimplification of what thinking (or, more specifically, being a thinking being) involves. For example, in Proclus’s fifth-century account of Amelius’s late third-century commentary on Plotinus, we learn about the complex connections between thinking and intellect in Plotinus.
80 Who Speaks? Amelius represents the demiurge as a triad and as three intellects or three kings—existing, possessing, and seeing. The division between the three intellects is such that the first intellect is in its being (ontos) what it is. The second [intellect] is what is thought of (noeton) in it; since this one completely participates in the first, it is called second. The third intellect is what is there in the first and at the same time emerges as the second. Namely, the whole of the intellect is identical with the thinking intellect, which copulates with it. Behold, it possesses what is there in the second intellect and sees the first intellect. Therefore the greater the distance, the weaker is the possession. By these three intellects, Amelius means three demiurges, who are three kings in Plato (Epist. II 312 e 1–4) and three in Orpheus, i.e., Phanes, Uranus, and Chronos.1
Thinking is thus a compound of an intellect and of seeing what is in that intellect. What it takes to think has a complex structure: to be an intellect or thinking being, to grasp representations, and to be the two in the one. To think properly is thus a double activity: simultaneously to see the intellect grasping representations and to be that intellect and that grasping. In this view, a simple understanding of thinking that reduces it to the activity of an agent distorts the true understanding of the complexity of what it really takes to think.2 In Amelius’s terms, the question to ask about the text of the Talmud is no longer “Who speaks?” or “Who thinks” or “Who occupies the position of ‘The Author?’ ” in it (that is, “Who is the Mark Twain of the Talmud?”), but rather: “What is the compound structure of thinking expressed in Talmudic text?” It is not “Who thinks in the Talmud?” but rather “What does it take to think in the Talmud?” For the Talmud, the answer to this question may be different from that of Amelius for Plotinus.3 However, the complex structure of thinking that this question implies simply cannot be addressed by any reductive connection of thinking to the activity of an agent, much less to the representation or figure of such an agent. What the example of Amelius illustrates is that in order to understand the complexity of the thinking process in the Talmud, I will have to complement the binary logic of representation (the representation of characters versus what/who is represented) used in text-critical scholarship of the Talmud with a more complex logic, which I call—by way of both borrowing and renegotiating a notion from Gilles Deleuze’s reading of Spinoza—“the logic of expression.” In Deleuze’s version of it, the logic of expression, which he identifies in Spinoza’s ontology of substance, includes three elements of expression, rather
Human Existence in the Talmud 81
than two elements of representation:4 not the act of representation (which requires an agent in the sense of a thinking subject) and what is represented thereby, but what is being expressed (“substance,” for Spinoza), what is actually expressed (“attribute,” for Spinoza), and the expression itself (“modus,” for Spinoza). I will heuristically use that logic for reading the texts of the Talmud and will consequently renegotiate Deleuze’s understanding of the structure of expression in application to these texts. In these terms, thinking will reemerge either as expression or as the reading of a textual expression, so that the texts will be approached not only as representations of characters, but also as expressions of—as I will argue—apersonal and heterogeneous forces of thinking. As the example of Proclus’s commentary on Amelius also illustrates, broadening the theoretical framework in which to access the thinking in the Talmud does not mean reducing the thinking practices in the Talmud to those in rhetoric or philosophy. The difference between Proclus’s commentary and the text of the Talmud needs no explanation, especially as far as the weight given to citing or, broadly speaking, remembering the specific dicta is concerned. In the Talmud, recalling and quoting a dictum is intrinsic to the discussion. In our example from Chapter 4, a student recalls and cites a dictum of Rav Yehudah about a dictum of Rav and uses the dicta to carry out his own thinking in defense of the Mishnah. Proclus, in contrast, directly interprets the teachings of Amelius, not Amelius’s words per se. Yet, my goal is not to draw any direct comparisons or contrasts. Rather, a broadened rhetorical-philosophical perspective on thinking is needed only for asking the question of thinking in the Talmud in a more precise way, even if and precisely because this perspective does not yet allow us to answer it. Getting closer to the answer will require both using and renegotiating the tools of contemporary theories in philosophy, rhetoric, and in the text- critical Talmud studies in order to read the ancient texts of the Talmud to discern their practices of thinking as both comparable to and distinct from what we can learn from philosophical and rhetorical theories of thinking. Without getting ahead of myself, I can only say that the Talmud demonstrates a unique configuration of the elements of thinking, refuting, remembering, and mnemonic memorization, in a way that other texts of late ancient culture do not. Even if other texts of late ancient culture employ the same elements, they configure them in different ways. Approaching the texts of the Talmud as not only forms of representation, but also forms of expression and a display of thinking forces5 at work
82 Who Speaks?
complements the existing view of the text as a form of represented reality (the life of the rabbinic academy) by treating that representation as an expression of something else—something that informs the representation, is immanent in it, but is not quite represented in it. But to make room for that different but complementary view of what is to be found in the text of the Talmud, we must first get beyond the antinomies or internal impasses that characterize the existing views of the nature and status of “The Author” in and/or of that represented reality, that is, clear away the problems that cannot be solved in the particular theoretical frameworks in which these impasses were originally formulated.
In/Of The “in/of” aporia has an empirical and a theoretical component. A first sign that the redaction theory leads to an aporia was an empirical and historiographical problem that David Halivni’s students have pointed out:6 the lack of any external historical-biographical evidence of the existence of the Talmud’s alleged anonymous redactors.7 However, in Halivni’s theory, the problem is not merely empirical but also theoretical. According to the theory of the Talmud’s redaction, the redactors’ role in the Talmud was to represent the teachings and sayings of the earlier Talmudic masters, the amor’aim, in such a way that they do not suffer in the Talmudic process of refuting and defending. To that end, the redactors tacitly reframed the dicta to make sure the dicta function well in either refutations or defenses. According to this approach, the redactors are construed as agents in the literary reality that the Talmud portrays and in particular implies. However, the redactors are not among the named or nameless characters that the text either mentions or even implies in Talmudic discussions. Rather, it takes a critical scholar to detect their existence by reading the text against its grain. In this respect, the redactors are to be understood as the redactors of the Talmud and thus to belong to a reality outside of the literary reality of the Talmud. In sum, the redactors thus exist in the reality of the Talmud or in a different reality altogether. Both propositions in this antinomy assume a mixed—historical and philosophical—notion of existence. Taking the “different reality” to mean historical reality, the literary-historical scholars take the “realist” approach and immediately are confronted by the lack of external biographical-historical
Human Existence in the Talmud 83
evidence of the redactors’ existence. However, initially in the theory of redaction, the notion of existence is defined not by the historical-empirical or biographical notion that it exists because it is reliably documented, but rather by the philosophical notion that “they historically must have existed, because their thinking has been traced in the text in addition to the named amor’aim and the anonymous characters who dispute in the Talmud.” That is, analysis of the Talmud, discussion by discussion, demonstrates the thinking process of the redactors as they attempt to ensure that the amor’aim’s dicta serve well in scenes of refutation and defense. As these analytical readings make clear, that thinking process is different from the thinking of the amor’aim or from that of the anonymous disputants in the Talmud, because it happens on the level of composition, and not on the level of participation. However, because these scholars rigidly connect thinking with persons who think, they quickly proceed from thinking to thinking individuals, who therefore must have historical existence. The result is an uncritical combination of historical-empirical and philosophical (Cartesian) ideas of existence: They thought, therefore they must have existed. Would a less “realist,” that is, less empirical-historical or less documentary- biographical, notion of the reality of the Talmud’s redactors remove that antinomy? Could an idea of a reality other than the biographical-historical be helpful? Hardly so, for the antinomy would still remain in place. In other words, redaction theory cannot deny the greater degree of reality of the redactors. Neither “formalism” nor “realism” help remove the antinomy. They cannot, because both approaches revolve around the notion of personal existence, either individual or collective, as the conditio sine qua non of thinking.
Emended or Interpreted? The impasse involved in the second aporia emerges when, in the course of refuting and defending, a character in the Talmud says that the original phrasing of the Mishnah “was missing something.” This character then follows up with a new phrasing to fill in the gaps. A scholar of the Talmud consequently asks: “What was the intention of that character—to emend the wording or to interpret the meaning?” The aporia of emendation versus interpretation is formulated in the philological-historical approach in the work of Jacob N. Epstein. It has to do with the scholar’s programmatic
84 Who Speaks?
declaration of the inability to decide on the thinking demonstrated in the text of the Talmud: whether the intention of the anonymous thinking subject (the character) was to “emend,” that is, to correct, a (presumably corrupted) text or to “interpret” it without offering any textual change.8 What Epstein both takes for granted and does not take into account is that characters who rephrase the sayings of the Mishnah do so by way of a defense against a refutation, never by way of an attack. What is more, whoever composes the Talmudic discussion often makes the same or other characters recall and quote an apocryphal tradition from the sages of the Mishnah (baraith’a, lit. “an external [text],” is a term in Talmudic parlance to refer to such an apocryphal text) following the defense to provide additional verification. That additional support always comes after the defense, not before it, which suggests that in the composition, the defense that is accomplished by thinking triggers support done by mechanical recollection. To take this suggestion one step further: Just as the verification (by a baraith’a) comes by way of recollection, the initial defense comes by way of memory, albeit not by way of mechanical recollection, but rather by way of a rhetorical attack on and ultimately a defense of the initial formulation of the Mishnah. Here, memory, thinking, and attacking-defending are all one and the same process, the outcome of a composition that involves at least two characters, one who attacks and one who defends. However, each character is only a part of the larger figure of thought in the Talmud, which involves both attack and defense. The second character surely has the function of defending the Mishnah but (according to an account such as Epstein’s) only probably has any self-contained single intention—either to interpret or to emend. This character in fact does not have intentions, but only a role in the compound figure of thought and memory. In this case, the function of the character is exclusively to deflect an attack against the Mishnah. Everything else is subordinate and should not be used outside of the context of the defense. The line of the argument used by the defender is similar to “If you attack the Mishnah in such a way, I defend it by reading the Mishnah in such and such way, which is also supported by how this topic is presented in a baraith’a.” The claim of the defender is thus conditional, rather than absolute; contextual, rather than ontological; and responsive, rather than assertive. Therefore, only if a researcher of the Talmud wants to invest the character with self-contained and conscious “intentions,” it becomes possible to ask what is being asserted: an emendation or an interpretation of the Mishnah.
Human Existence in the Talmud 85
Additional support for this reading comes from a broader observation about other defenses in the Talmud. They operate neither by emending or interpreting, but always by a refined and verified understanding/remembering of the content of the Mishnah attained by refuting and defending the Mishnah, often helped by mechanical recollection of a baraith’a, as well. Once again the root of the aporia of emendation versus interpretation is in the notion of a character as an isolated thinking and therefore existing subject.
Is “Culture” Literature? The aporia of Talmudic culture arises in a recent continuation of Halivni’s “realist” construction of the Talmud’s anonymous redactors. This approach proposes to elucidate the character of the anonymous redactors by explicating information about the “culture” of their life in the rabbinic academies. This information is assumed to be implied in the text of the Talmud. Proposed by Jeffrey Rubenstein,9 this approach emphasizes the historically “real” culture of rabbinic academies, rather than their literary construction in the Talmud. The category of “culture” in this approach is meant to bridge the gap between “literary constructivism” and “realism.” To do so, it introduces “culture” as a middle term that is both implied in the text and historically real. That approach turns attention to hitherto relatively underappreciated narrative fragments of the Talmudic discussions. Because the goal of this approach is to reconstruct the cultural environment in which the redactors in/of the legalistic discourses could have acted and thought, Rubenstein attempts to reconnect what was rigidly split in two parts, starting from commentaries on the Talmud in the Middle Ages—homiletic narratives and legal discussions. Using the narrative components of the Talmud to learn about the creators of the legalistic narrative is as innovative as it is fruitful. However, reconstructing the culture implied in the narrative parts of the Talmud leads to the impasse involved in the third aporia. The aporia is that the cultural environment implied in the Talmud is undecidably either a literary fiction or the historically real environment of the anonymous redactors. Reading Harry Potter with techniques of cultural reconstruction would also allow us to restore the maps, ethos, and other elements of the cultural environments implied in the novel. In fact such reconstruction might have already been undertaken. However successful, it
86 Who Speaks?
tells nothing about the real as opposed to the fictitious characters of the environments implied in and explicated from the five books of Harry Potter. Of course, no one who is not delusional would claim that the “culture” that can be adduced from the Harry Potter books is “both implied in the text and historically real.” It may have been “inspired” or “based” on something “historically real,” say British schools of the twentieth century; but so is virtually all fiction, in some way or other. Similarly, inferring the “historically real” culture of the rabbinic academy from what is implied in the texts of the Talmud does no more and no less then reconstructing the implied rather than explicitly described “culture of the Babylonian Talmud.” Therefore, unlike Harry Potter, what is implied (rather than explicitly described) in the text of the Talmud as “culture” has a stronger claim for “historical reality” than the nondelusional reader of fiction could ever have. The third aporia thus can be seen as a version of the “in/of” paradox. However effective the resulting cultural reconstruction might be, it does not yet tell us anything about the real, historical existence of the anonymous redactors. Instead, this approach takes the existence of the alleged redactors of the Talmud for granted, seeking to provide more knowledge about the reality and environment in which they acted. It places the redactors outside the Talmud, but inside the culture of the Talmud as evidenced by the text, thus preserving both the undecidability of whether the redactors are characters in the Talmud or the historical producers of it and the conception of the redactors as thinking subjects producing their interpretations of earlier masters in an environment that that approach reconstructs using the narrative parts of the Talmud. As in the literary-realist approaches, only the assumption of the existence of the redactors as Cartesian subjects underpins the historically real, rather than literary-fictive status of the culture reconstructed from the narrative parts of the Talmud. This analysis of the culture of the Talmud also has continued the tradition of taking for granted the partition of the narrative and legalistic discussions, precisely in that it is assumed to be something that needs to be bridged. Although it innovatively has proposed to repair the split between narrative and legalistic elements of the text, the partition between them remains axiomatic, rather than problematic. Instead, the cultural approach has productively used the fact that the narrative and legalistic discourses are interspersed in the same book. This approach, however, has made no attempt to explain the significance of that interspersion. The question of why the Talmud has these two kinds of discussions at all has remained unasked. The same applies
Human Existence in the Talmud 87
to the question of what the contribution of these two interspersed elements is to the Talmudic discourse as whole, which has remained outside of the cultural approach.10 Once again, the notion of the reality of the thinking subject, this time a collective thinking subject, or “the anonymous redactors” both turns on the problem and makes it irresolvable. As an extension of the “realist” approach in Talmud criticism, cultural reconstruction faces the same aporia of placing the cultural reality as the context of thinking in either a historical or a purely fictional but stylistically “realist” mode of the existence of “the anonymous redactors” of the Talmud.
a Despite rather sharp differences between them, undoing the rigid connection that all three aporias share between thinking and the subject, in particular between the design of the Talmudic discussions and the redactor or intelligent designer behind them, leads to a different question than “Who is the Mark Twain of the Talmud?” that is, to the question of a thinking existence in the Talmud. It thereby invites an answer that takes us beyond the impasses, antinomies, and aporias I have just explored. Attaining this answer requires reformulating the question in its proper terms: “What kind of human existence do we find in the text of the Talmud?” As I have already begun to show, both asking and answering that question have to do to with clarifying the relationships between thinking and remembering beings in the Talmud.11
Beyond the Aporias: Remembering, Not Recalling What in the framework of the text-critical theories of the Talmud has emerged as a set of aporias may become a heuristic tool when viewed in a more general theoretical framework. To lay the groundwork for the elaboration of such a broad theoretical model in the next chapter, and for understanding these aporias in positive, rather than negative terms, requires an account of the increasing disconnection between memory and thinking in the Talmudic, rhetorical, and philosophical disciplines from late antiquity to the medieval and modern scholarly disciplines. The main features of that
88 Who Speaks?
disconnection are the separation of thinking from remembering, the privileging of thinking over remembering, and the demotion of memory to the lower echelons of the structure of subjectivity. Understanding the relation of thinking to memory in late ancient texts cannot be done without putting in question the rigid connection between thinking and the subject in the concept of thinking subject and interrogating the tie between subjectivity and memory. If attached to the notion of a thinking subject, the notion of memory almost inevitably coincides with that of personal recollection, which in turn reduces the role of memory to the merely auxiliary process of managing data. Defined that way, memory has, of course, very little to do with thinking, if the latter is understood to be the core feature of the intellect. Similarly, if rigidly connected to the notion of the subject as both the agent and foundation of thinking, the intellect is no longer an independent faculty, but only a property of the thinking subject, either human or divine. However, if these connections are regarded as loose, as they indeed were in the late ancient schools of rhetoric, philosophy, and Talmud, then thinking can regain its status as an independent faculty, the status it already had in Aristotle’s teaching of the active intellect. Alternatively, thinking can come to the service of memory as the main path of both knowing and being wise, as in the Platonic schools of both philosophy and rhetoric. In yet other ways, thinking can come to the service of memory in the discipline of Talmud and other disciplines of rhetoric. In any of these scenarios, the figure of “The Author” is no longer a privileged—much less, sole—site of either thinking or remembering. Instead, “The Author” becomes a figure of participation in thinking and remembering, rather than the figure of a privileged location. It becomes multiple and heterogeneous.12 This idea of the homogeneity of thinking does not change, regardless of whether the author is one person or a united plurality13 of individuals. In the first approximation, “multiplicity” would involve variety in what’s numbered, while “plurality” would just involve numbers. An example of plurality would be a “collective author.” The “anonymous redactors” constitute a figure of such “collective authorship.” At least in Halivni’s understanding of them, the redactors never disagree one with another. Instead, they delegate that job to other characters. Granted, Halivni’s redactors model polemical discourses between the amor’aim in the Talmud and often even create anonymous characters for the purposes of that modeling, but the redactors themselves do
Human Existence in the Talmud 89
not polemicize.14 As the collective author, they remain immune to polemics. Shamma Friedman’s vision of the author of the Talmud as the literary designer of the compositions and discussions does not change that attitude: The designer may be either plural (“collective”) or singular, yet neither heterogeneous nor intrinsically polemical. The homogeneity of “The Author” is also compatible with changes in “The Author’s” positions. This can be illustrated by an example of plurality without multiplicity, that is, without variety, which is the figure of “The Author” in Hegel’s philosophical system: “The Author” as Spirit.15 As I have argued above, Hegel’s way of detaching thinking from the subject is possible only because thinking is reattached to a different subject, “Spirit.” The plurality in Hegel’s figure consists in Spirit’s ascent through different stages of the system. Yet because each new level is a “sublation” (Aufhebung) of the previous level, “The Author’s” and the reader’s changing positions neither depend on the polemics of the text nor adopt multiplicity: Differences accumulate or sublate in the unity of Hegel’s end result: Spirit coming to consciousness of itself. Furthermore, modern figures of “The Author” can allow for multiplicity, if not for heteronomy, in the structure of “The Author” and of thinking. Seen as a figure of “The Author,” Heidegger’s notion of Dasein exemplifies that possibility. If Dasein features one’s being there in the world, Dasein’s relationship to the world is structured comparably to how “The Author” relates to the world described in The Adventures of Tom Sawyer. “The Author” is present in the world, albeit in a way that is different from all the other characters. The latter must have at least somewhat articulated bodies or at least bodily features—hands, heads, hats, postures, and so on; and the objects with which the bodies interact—paintbrushes, apples, “golden” teeth, other bodies, and so on. In contrast, “The Author,” as a character, can be present without any articulated body. What is of even greater difference, other characters are not necessarily aware of, or responsive to, their being; instead they only act and function in direct response to circumstances and prompts. However “The Author” is different. “The Author” knows and declares that other characters are constructs. Some of them, like Tom Sawyer, are of a “composite order of architecture,” and some of a less eclectic design, but they all are constructs “The Author” ascribes to himself. By knowing about the constructed nature of other inhabitants of his world, “The Author” also knows that even if the
90 Who Speaks?
other characters are there, they might not have been. “The Author,” in other words, knows that their being there in the world depended on “The Author’s” work. Aware that the characters might not have been there, “The Author” thereby also implicitly appreciates their being in the world, rather than their mere functioning in response to the prompts. “The Author” therefore surely knows the difference between functioning, and being. What it means, however, is that “The Author” can apply, and implicitly has always already applied, that knowledge to himself and his own being in the world as well. “The Author” implicitly knows his being, which also means he at least implicitly understands a deferred possibility of his not being there as well. Dasein is thus structurally similar to “The Author.” Like “The Author,” Dasein by necessity has options to notice or not to notice, to appreciate or not to appreciate the being of other characters (and, by extension, of herself) in the world; to become anxious about nothingness implied in that knowledge of being, or instead, to became responsive to being. Like Dasein, “The Author” can also become bored (i.e., as Heidegger has it, both excited about things and anxious about the nothingness lurking behind them). Another route for “The Author” is to become entertained and, again, excited by things, thereby occluding being behind functioning, and existence behind entertainment. Even if, it may be argued, in The Adventures of Tom Sawyer, “The Author” ends up exploring that route, still, like Dasein, “The Author” could have preferred to respond to the fact of being—of the world, of the characters, and of himself. The way to do that would be to philosophize about being that enables the being of these beings. Yet “The Author” is not a philosopher, and at this point Heidegger and Mark Twain part their ways, because Dasein philosophizes and “The Author” does not. Yet the difference between Heidegger’s Dasein, and Clemens’s “The Author” is the matter of choice, not of structure. The structure remains homologous in both Dasein and in “The Author”; they both represent a multiplicity of incompatible yet mutually complementary choices. Indeed, among all other characters, “The Author” is the only one who by default has the possibility to respond to being, and thereby an ability to occlude being as well. Similarly, among other beings, Dasein is the only one able either to occlude being (in boredom, anxiety, functioning, entertainment, etc), or to respond to it in philosophizing about being, or in poetry. Yet, where does that unique and singular ability of Dasein come from? The singular positioning of “The Author” among other characters sheds
Human Existence in the Talmud 91
light on that question. The singularity of “The Author’s” position has to do with his unique situation between two worlds, the world of Samuel Clemens, and that of Tom Sawyer. I conceptualize that singular position in between and betwixt the two worlds as virtual. “The Author” is virtually there in Tom Sawyer’s world, and virtually there in the world of Samuel Clemens. As virtual in each of them, as he is, “The Author” actually and actively participates in both. This dual participation takes place even if the two worlds only virtually, and never actually, coincide. As already briefly noted, that unique, virtual position of “The Author” is a multiplicity. The virtual position of “The Author” is a multiplicity because it involves mutually incompatible and at the same time complementary positions toward being versus functioning: “The Author’s” position has linearly incompatible components: He is always already aware of his and other characters’ being, and at the same time has options to occlude being by getting anxious, or bored, or by indulging in either functioning or entertainment. What is more, “The Author” reveals something about that multiplicity, which Dasein does not. Explicitly literary (rather than historical) in genesis, the virtual position in between the worlds explains the multiplicity of the incompatible orientations of “The Author.” As already established, one is orientation toward the being of himself, the other characters, the world, and ultimately being itself. The other, however, is even more nuanced. It is an orientation toward occluding being behind either believing or disbelieving in the existence of “The Author” himself, as well as of the characters, and the world. That virtuality and nonlinear multiplicity of “The Author’s” orientations come precisely from “The Author’s” reticence to either fully believe or fully disbelieve the being of the world, the characters, and himself—a position that “The Author” so successfully casts to the readers. “The Author” is virtual not only because he is situated in between the worlds of Samuel Clemens and Mark Twain, but also because “The Author” is the site where a decision in favor of being or nonbeing can be bracketed, rather than made. By extension that virtual positioning of “The Author” explains Dasein as not only multiple, but also virtual, that is, as never making an absolute choice between belief and disbelief in the being of the beings.16 In that sense, just like “The Author,” Dasein does not fully belong to the world in which it is. It therefore can appreciate, question, and un-decide the being of the world, the people therein, and himself.
92 Who Speaks?
Again unlike “The Author,” Dasein might not have a second world (say, the biographically “real” world of Samuel Clemens) to which to belong. Perhaps this is why Heidegger’s Dasein does not refrain from deciding on being versus nonbeing, or more precisely does not bracket the choice of occluding being through functioning versus responding to it through philosophy and/or poetry. Instead, Dasein philosophizes by deconstructing boredom and anxiety, or openly confronts nothingness through the will to death. It thereby appreciates its own being, and by extension the being that makes all (other) beings be. Yet again, the difference between making a choice or refraining from choosing is of implementation, not of structure. Both “The Author” and Dasein happen as multiplicities, in which going one way is predicated on having the other, incompatible, option, even if the latter is needed just and only to avoid taking it. This site of embedded structural multiplicity in the positions of both “The Author” and Dasein is where their thinking becomes distinguishable from the possible plurality of contributors “collectively” generating “The Author,” who, however many, end up thinking in the same way. Yet this multiplicity is not heterogeneity. Occluding being and responding to being are two orientations, mutually incompatible and mutually necessary, and not open to any final decision about them. That introduces multiplicity, albeit again not heterogeneity, because both orientations stem from one and the same unique and singular relationship of “The Author” or of Dasein to the worlds to which they belong.17 There is more on the way from multiplicity to heterogeneity. In Heidegger, thinking as a multiplicity can no longer be fully attributed to a thinking subject, that is, to a subject who allegedly “thinks” and therefore “is.” Yet, even if Heidegger detaches thinking from subject, he reattaches it to what is no longer a subject as an “ego” (again, either individual or collective, and either named or anonymous) but rather still a subject as a foundation from which light of being is shed on all particular beings, which light Dasein is to gather and recognize, so that through Dasein the true subject as a foundation (that is no longer as an ego) shines forth. Heidegger thus introduces multiplicity into the notion of Dasein. Yet again, Dasein operates in solitude and its multiplicity is homogenic, rather than heterogenic, and in that sense, the “ego” is not fully removed. Left out of ego, in thinking remains solitude, and thus homogeneity rather than heterogeneity of thinking, however much multiplicity that thinking displays.
Human Existence in the Talmud 93
These examples of the figure of “The Author” in Halivni, Hegel, and Heidegger have at least two things in common, one positive and the other negative. The figure of “The Author” is characterized by a unity that includes plurality or multiplicity, but that plurality or multiplicity does not allow for heterogeneity. For all three of these modern scholars, “The Author” is a single figure, and the thinking it does is homogeneous, whether it is done by an individual or by many individuals acting either as one or in ultimate solitude. By sharp contrast, as the example of Amelius’s interpretation of Plotinus shows, it is possible to conceive of heterogeneity in thinking. Amelius finds Plotinus practicing and understanding thinking as a combination of intellect (nous) with what is perceived in that intellect (noeton). The single site at which thinking occurs both is that intellect and sees what’s grasped in it. As such, it is inherently heterogeneous. Amelius’s interpretation also draws close to the practices of thinking in the Talmud, in which thinking is not possible without a multiplicity of characters holding heterogeneous views in polemics with themselves (one character can propose both an attack and the defense against it) and with the others. These characters do not replace each other, but rather need each other in order to carry out the work of thinking. Thinking by means of refuting and defending each other, these characters at the same time and by the same token do their work of remembering the Mishnah correctly. Both Amelius’s and the Talmud’s thinking are sites of multiplicity and heterogeneity, each in its own way. In the case of the Talmud, however, heterogeneity is not only a constitutive condition of thinking but also is a programmatic practical way to think by refuting and defending. Thinking thus involves the substantial, rather than the figurative, participation of the characters. In the Talmud as a site of multiplicity, thinking is a dance of characters who never coincide one with another and never forget the polemical positions of the others. Even if this dance may imply a choreographer who composes it, thinking does not happen on the site where this choreography was designed, but rather on the site of the characters themselves.18 For example, in our sugya, the choreographer or composer might have strategically decided to move the discussion from a more plausible defense (by the first dictum of Rav) to a more effective defense (by his second dictum), but the concluding defense does not represent a valid point without the polemical processes that lead to it but are not “sublated” in it. It does not stand by itself. Instead, thinking in the Talmud occurs in order to refute. However, refutation occurs
94 Who Speaks?
in order to understand and to remember. This makes sense if and only if thinking is disconnected from the notion of a unitary and homogeneous thinking subject. The key to answering not “Who thinks in the Talmud?” but rather “What kind of human existence do we find in the text of the Talmud?” in terms of how that thinking occurs thus is the status of memory in the Talmud—and that, too, makes sense only if it is seen as disconnected from the notion of a unitary and homogeneous subject. The protocols of memory in the Talmud are both related to and disentangled from the protocols for the exercises of memory in the rhetorical- philosophical schools of late antiquity. In the latter protocols, memory was either reduced to mnemonics, or, on the contrary, understood, in the Platonic sense, as the path to knowing the ideal forms. In contrast to the mutually exclusive Platonic and mnemonic understandings of memory, the Talmudic way of connecting memory, thinking, and refuting is uniquely situated between other late ancient attempts, combining rhetorical and philosophical approaches to memory.19 In the late ancient disciplines of Talmud, rhetoric, and philosophy, thinking was only one of the vehicles of memory. Memory was not centered on the notion of the remembering subject; rather, it was centered on what is remembered—images, things, words, or ideas. However, if, as was increasingly the case in medieval and modern times, either a human or divine subject is assumed to be the foundation of the processes of thinking, memory often is reduced to recollecting data held in the mind of the thinking subject. As a result, remembering begins not with what is remembered, but with who recalls an image or thought. By contrast, dissociating thinking from subjectivity opens up room for understanding memory as an ultimate goal of thinking and arguing in late ancient texts, rather than as merely a tool for collecting data. The modern assumption is that recollection either controls or at the very least expresses a sequence of words and/or facts that provide data for thinking. Of course, having some data is essential for any thinking, but the data of memory are not privileged over other sources of data. Much fewer data are required to prove the existence of the thinking subject, whether human or divine. Rather, that proof comes from the activity of thinking, not from the data it uses. However, the role of memory in thinking—or, perhaps more precisely, the role of thinking in memory—was very different in the late ancient disciplines of philosophy, rhetoric, and Talmud.
Human Existence in the Talmud 95
That difference is as radical as it is subtle, and grasping it as difficult as it is important. Heidegger, for example, contrasts modern ideas about thinking and remembering to Plato’s and Platonism’s concept of remembering. For Heidegger, Plato’s concept of remembering implies a “catching-sight-once- again, [hence] the revealing, of beings, sc. [that is] in that by which they shine-forth.”20 He opposes this understanding of memory to what he holds as the reductive view of memory as the mere recollection of data. Hegel already had expressed a rough parallel of this difference between memory as a “catching-sight-once-again” and the modern idea of recollecting in his thesis that to remember means to understand, rather than to memorize in any mechanical way. A seemingly similar difference is found in late antique texts such as, for example, Quintilian’s curriculum of rhetorical exercises, Institutiones Oratoriae, a textbook of rhetoric. It is by our powers of memory that we must secure the easy flow of what we have formulated in thought, instead of letting it keep us from looking ahead by anxious backward glances and the consciousness of being absolutely dependent on what we can recall to mind.21
As Quintilian suggests, in delivering a speech, our memory should not be used to recall what we thought before. Rather, based on what we understand (rather than what we mechanically memorize), memory secures the easy flow of what we have formulated in thought, now as an argument that we present. Memory helps us process in speech what we have formulated in thought, not torment us with what we can or cannot recall to our mind. Memory for Quintilian thus should not be mechanical recollection (much less recitation). It should be the power releasing an easy flow of what we have formulated in thought. More generally, we thus have to use memory for thoughts, not for words. However, using memory for thoughts requires thinking and renders memory a skill associated with thinking, in particular, with transforming thinking into a line of argument, not with recollection. What seems to be a similarity between Quintilian, on the one hand, and Heidegger’s and Hegel’s interpretations (and appropriations) of Plato, on the other, should not mask an important difference in the role of human existence in remembering. A modern theorist of the art of memory, Patrick H. Hutton,
96 Who Speaks?
highlights that subtle difference. Hutton presents what can be seen as a new, modern version of Quintilian’s claim. Just as Quintilian argued that memory is less a mechanical recollection and more a way both to hold and to delineate a thought (and ultimately an argument) in which an orator grasps the matter, in the modern context Hudson dismisses the reductive psychological view of the art of memory as mechanical mnemonics—a set of techniques in support of recollection. The art of memory as it was understood in its classical formulation provided not only a useful skill but also a way of understanding the world. For some mnemonists the design of the structure of their mnemonic system corresponded to their conception of the structure of knowledge and so implied a vision of the world. The power of the mnemonist lay in his ability to interpret the world through a paradigm that would provide its initiates with a clavis universalis, a master key to the workings of the universe.22 From this perspective the art of memory was not only a pedagogical device but also a method of interpretation. It is this link between the art of memory and the making of paradigms of cultural understanding that suggests the larger significance of this topic.23 Two approaches to memory—mnemonics versus thinking about the world—form, for Hudson, two competing visions of reality and of the place of human beings therein. For a practitioner of mnemonics, the world does not need any interpretation. Everything is clear and uncomplicated, readily available to be stored in the palace of memory. Delivery therefore requires no fresh thinking, but only remembering and keeping things in order. In contrast, for a practitioner of philosophy, not all of the world is clear and uncomplicated. Rather, it constantly requires thinking and interpreting by way of revealing what is hidden in things. Therefore, interpreting the hidden is an intrinsic part of the philosophical practice of memory. Instead of either a purely mnemonic use of memory for recollection or a purely philosophic approach to memory as a way to uncover the occluded things in the world, for the rhetorician, according to Quintilian, memory is what allows the orator “to look ahead” in thinking, while recollection is what allows the orator to “look backward” into his or her inventory of words and images. Recollection retrieves data, but memory carries thinking forward and thus anticipates what is ahead. In his view, recollection and memory form two competing powers that control the orator, but his or her ability to control each is different. Perhaps the orator can control his or her recollections using
Human Existence in the Talmud 97
the art of mnemonics, but not the element of memory. In that case, the orator lets memory work without telling it in advance what to do. Memory allows the orator to follow what comes from “ahead,” to employ it by way of discovery and invention. In Quintilian’s desired scenario, the orator “looks ahead” and thereby follows memory and thinking, rather than trying to retrieve or control what is in the past, in recollection. Thus, his orator is not the site in which thinking (or remembering) originates. Rather the site of memory lies in what is “ahead,” not in recollection. Thus, for Quintilian, memory, as opposed to recollection, always deals with what is new. Memory helps the orator reveal by creating (inventing), and then uncovering what already is available (discovering). In fact, the Latin term inventio comprises the senses of both “invention” and “discovery” in one meaning. Quintilian thus removes the orator from the center of speech. Remembering as he delivers the speech, the orator does nothing more and nothing less than carry out thinking toward what is “ahead.” Unlike philosophical conceptions of memory, of “catching- sight-once-again” of what has gone before and is prior, the rhetorical conception of memory is prospective, always dealing with what will come next. How can one in practice exercise this notion of remembering? How can one teach a student to do this? That is exactly what the text that begins “Of the two portions of the Scripture in the mezuzah the [absence of] one invalidates the other; indeed even one [imperfect] letter can invalidate the whole” that I analyzed in Chapter 3 does: It remembers the Mishnah. It builds a line of argument in which the Mishnah’s rhetorical innovation “shines forth,” but by the logic of invention comes from “ahead.” It is oriented toward what will come next. The practice of memory in the Talmud is thus closer to the rhetorical conception of memory than it is to the conception of memory in the philosophical tradition since Plato. Using the technique of (perhaps artificially created) refutations of the Mishnah (“Obvious!” “This is also obvious!”) followed by defenses, it catches sight of the Mishnah once again, revealing and by the same token creating the rhetorical invention in it, either about the status of the letter yod, or, as it does in the final sally, about the margins that surround the letters. In the Talmudic curricular exercise of refuting and defending the Mishnah, the tann’a, or recitant, represents the element of mnemonic recollection, and the pair of refuter and defender represent the way that memory intellectually elicits and thus grasps, as Hutton would put it, the “interpretation” of the law. In addition, in the second text, the interpretation requires not only a student to attack and defend and the teacher
98 Who Speaks?
to refine these moves, but also perhaps a choreographer or composer to grasp the dance of their thinking and express it in a formal composition that bespeaks the orientation of the dance toward memory. In addition, the Talmudic practice of memory contains elements that both the rhetorical conception and its philosophical counterpart do not. The characters in the Talmud accomplish memory by refuting and defending, but Quintilian treats memory separately from refutation. The Talmudic characters use the rhetorical exercise of refutation and defense as a way of making the text of the Mishnah “shine forth” in a plurality of ways and thus be remembered not only mechanically, by declamation, but also intellectually, by interrogation, which does not fully fit either conception of memory. Seen in such a broader comparative perspective with the disciplines of philosophy and rhetoric in different time periods, the practice of thinking and remembering performed by Talmudic characters thus suggests a completely different approach to the study of texts, or more precisely, to the ways in which human existence can be understood to appear in them.
Human Existence in the Talmud If literature is one of the ways in which humans define and express their existence in the world, the texts of the Talmud should not be an exception from that general rule. Therefore, we finally are in the position of answer the question “What kind of human existence do we find in the text of the Talmud?” That is, what kind of human existence in the world is both defined and expressed in, by, and through the Talmudic texts? When it is put that way, this question is three in one and thus requires three answers. In the text of the Talmud, readers find representations of a rabbinic academy of the late ancient period in which anonymous teachers and their students argue one with another, often quoting the dicta of earlier authorities, amor’aim. By way of implication, the texts of the Talmud represent the existence of the authors of the text. And both through texts that explicitly represent people in a rabbinic academy in Babylon and through implicit representations of the authors in these texts, the creation of representations of humans in a world is also a way of being in the world. In that sense, to be in the world means to express a power that both creates and exceeds representations, both explicit (teachers and students) and implicit (authors.)
Human Existence in the Talmud 99
In the text, a careful reading helps critical readers elicit additional information about the environment and culture of the academy represented there, even if it is not directly described in the text. The resulting picture represents the world of the rabbinic academy as well as the way human beings exist in that world by means of—by default—unresolved polemical arguments. By reading Talmudic texts with a critical eye, scholars thus discover yet another human existence represented and defined in the text, if only by way of inference. According to one group of critical scholars, human existence of this kind includes anonymous redactors’ of the dicta of the amor’aim (the earlier authorities). To frame these dicta more effectively, these later redactors, stamm’aim generated discussions between anonymous characters in the rabbinic academy. By reframing the old dicta of the amor’aim in new discussions between anonymous characters, the redactors also implicitly signaled something about their way of existence in the world of the academy, as apologetic, engaged, and covert as it was. According to a competing group of critical scholars, the only implicit human existence in the Talmud is that of the designers of the literary forms of discussions in the academy. This theory does not entertain the existence of anonymous redactors, but instead explicates the implied existence of the composers of the entire discussions. In that approach, the implied human existences (the composers) are detached from either the represented reality of the rabbinic academy or of the people therein. Their existence is implicit because it is represented only by their activity of composing. More specifically, unlike the redactors, the composers do not necessarily have a definitive apologetic agenda of protecting the efficacy of the dicta in the discussion. Neither do the composers covertly participate in the represented reality by setting up forced interpretations of the dicta. Instead, the composers exist exclusively by way of producing the representation of the reality of a rabbinic academy, that is to say, as the authors of the composed text of the Talmud. And yet and still, regardless of the differences in these respective theories, both redactors and composers emerge as implied authors of the discussions in the academy, which critical scholars help explicate. Both the explicit representations of the rabbinic academy or of the people therein that are found in the text of the Talmud and the implicit representations of the authors found by implication in the analysis of it have to have been produced. The implicit author can create explicit characters, but
100 Who Speaks?
someone has to create the textual conditions that make that author implicit. Even if one solves that puzzle by saying the author has created himself or herself, implicitly, that self-created author would still have to perform two completely heterogeneous functions: to be a creator and to be the one who is created in one and the same respect. This suggests two ways of human existence in a literary text, including the Talmud. Both named and anonymous personae who act in the text; some of them also constitute “The Author” of the text. In the other way, “The Author” is only a designer of the text, but not an explicitly active character therein. Both ways of human existence require the “The Author” to be a person, either individual or collective. There is yet a third, less personal way of being in the world. This is a force that creates the figure of the author, which is very different from the result of creation. We find this third kind of human existence in the text of the Talmud. Shamma Friedman’s theory of composers or designers of Talmudic discussions provides a theoretical resource to understand that impersonal force. In Friedman, the composer of a given Talmudic discussion is not only a figure of authorship, but also a figure of genesis. The composer is not only an authorial counterpart of the given composition, but also a force responsible for making textual materials migrate between older and newer literary forms in the resulting text of the Talmud. The texts migrate back and forth between older forms (sayings of Mishnaic sages and dicta of amor’aim) and the newest form (dialectical exchanges between rabbis and teachers in the academy). Thus, for Friedman, the composer is not only a literary counterpart of the resulting text of the Talmud, but also a force of transformation on the way from the old to the new form of the Talmud. As a literary counterpart of a resulting composition the composer is a static figure of an “author,” that is, a particular implicit character in the text, which is not represented in the text, but expressed through the compositional form of that text. Yet, as a force of transformation and genesis, the composer is a figure of becoming. As such, it comes before and goes beyond any particular state of both implicit and explicit personae in the text—that is, before and beyond both the rabbis and implied “authors” expressed in the designed rather than spontaneous nature of discussions in the Talmud. As a force, the composer models a third way of human existence in the Talmud. It is not a person or a character but rather a force of transformation, by virtue of which both the composition and the composers come into being.
Human Existence in the Talmud 101
Far from being either blind or irrational, this force is the force of thinking, which again comes before and reaches beyond any personification of that force in explicit characters, or in the characters present in the implied stage direction of the dialectical dance the reader sees on the page, or in designers of resulting discussions in the Talmud. This force is the fourth, and perhaps most mysterious, way of human existence in the Talmud. As I will later show in more detail, this way of human existence is the way of memory and remembering, which is even less person-centered than the ways of both explicit characters in the text (rabbis and students) and the characters implied by that text (authors and/or composers). It is both hard and unproductive to imagine the power of transformation as a person plotting to move texts between forms, for example, to have decided to take some words away from the amor’aim and to put them in the mouth of anonymous disputants, or vice versa. Instead, even if, as unlikely as it sounds, the transformation is done by consideration of an individual, that individual is more likely to have acted based on the dictates of thinking understood as memory, that is, on the ways of gathering the arguments together by “memory for things,” rather than by either individual recollection or, even less plausible, by plot and manipulation. As rational as these textual transformations might have been, they must have come to whoever enacted them by way of memory, which is not centered on a person, rather than by way of independent person-centered thinking. Deleuze’s analysis of the tripartite division of expression as opposed to the bipartite structure of representation helps approximate that third way of human being in the world. As mentioned above, a representation is static. It involves what is represented and a representation thereof, thus making for two parts involved. In contrast, expression is a more dynamic (albeit, as I will later argue, still linear and homogenous) model. It involves three elements: what is being expressed, what is actually expressed, and the expression per se. The middle element, what is actually expressed, is always confined to the resulting expression, while the first element always exceeds that expression. In these terms, the force of thinking is what is being expressed in the resulting design of Talmudic discussions, yet the figure of the designer of that discussion is what is actually expressed. Yet, however useful these terms of expression could be to explain the difference between figures of genesis and figures of—both explicit and implicit—representation in Friedman’s theory of the Talmud, they are not
102 Who Speaks?
sufficient to grasp a possibility, which Friedman articulates, but shows no interest in developing. The possibility is that the force of thinking is apersonal, that is to say, has no rigid association with any particular person who participates in the thinking. Granted, this thinking has a discipline. As I will explain in greater detail in Chapter 8, that discipline might be a version of the discipline of “memory for things” as opposed to “memory for words.” “Memory for things” works by associating images of things (pragmata) with familiar places, whereas memory for words associates words or their parts with certain images, which remind of words regardless of the things that the words tell. That discipline would, of course, have to be shared by anyone participating in thinking, but that discipline would have no rigid connection to any particular person, agent, or, in more modern terms, subject, or any other participant. If approached through their literary creations, humans would, therefore, generally have at least three different ways of being in the world: as characters, authors, or forces of thinking. A person can be a character, in particular an author, but one can only participate in the otherwise apersonal transformative forces of thinking. These three ways of human existence amount to (1) an expression of one’s power to represent and the resulting explicit representations of characters and contexts, along with (2) the implicit representations of the authors of these representations, and (3) the transformative heterogeneous and no-longer solitary, and in particular no longer personal, forces of thinking in the ways of memory. That transformative thinking comes before and goes beyond any particular expressions, not to mention representations, which human being in the world creates. However, in the resulting representation, these forces of thinking are not only expressed but also, and by the same token, occluded by what is expressed and by the world represented through that expression. Therefore, the occlusion of what has been expressing itself in what is expressed in an expression is as inevitable as it is to be undone in the reading. What makes this task of undoing difficult is that occlusion cannot be remedied by any direct representation of the forces involved. The latter can only be expressed through representation, but cannot be represented as such. In sum, expressed through the act of representation, the force of human being in the world as the force of thinking is never represented in it. In application, it means that the texts of the Talmud are to be read not only as representations, as has been done so far, and not only as expressions, which is yet to be done, but also as workings
Human Existence in the Talmud 103
of a dynamic thinking and transformative force of memory, which Friedman both detects and conflates with a personification of composition of the Talmud in the static “authorial” figure of the designer of the discussion. These three answers—characters, authors, forces of thinking—have logical consequences. The first consequence concerns the role of a reader of the Talmud. Indeed, a work of literature not only represents and expresses, and either reveals or occludes thinking forces, but it also constructs a reader by preparing a place for that reader to occupy and to perceive the scene of represented reality. Occupying this place, a reader can naively believe in or even identify herself with what is explicitly represented or else critically detect what is implied, and perhaps even assume the viewpoint of the implied authors. Yet the reader can and should go further in an attempt to interrogate what is both expressed and occluded in the representation and in expression. Only by having these three regimes of reading in place—belief in what is represented; critical detection of what is implied, and assumption of the viewpoint of the putative authors of those implications; and attention if not to the forces of thinking, then at least to their occlusion—can the reader have a picture of the ways of being in the world she encounters in a given literary text. The second consequence concerns the theory of expression. If the transformative force of thinking in the ways of memory is heterogeneous rather than implying a solitary thinking subject (individual or collective, personal [including anonymous] or impersonal), then Deleuze’s reading of expression might be necessary but not sufficient. Unlike Deleuze’s reading of Spinoza’s theory of expression, what is being expressed in the representations of the text of the Talmud does not have to be an equivalent of the philosophical notion of a single and therefore also solitary divine “substance” that is expressed differently in different attributes, but remains the same in, above, and beyond all attributes and expressions, however different they might be. In contrast to that reading of Spinoza, the force expressed in expression does not have to be equal to itself; it could be ecstatic as well as static. Because it has no direct representation, it is not a theoretically fixed entity. In traditional philosophical terminology, that means that the human existence expressed in the text of the Talmud does not have to be either a “substance” or a “subject.” As an excess of expression, the thinking force allows no rigid personifying or naming. The third consequence is that even if this force of thinking is generically expressed in any work of literature, in different works it may be structured
104 Who Speaks?
in different ways. In every work of literature, that elusive force or power still has an expressed structure, which is by definition different from the structures of represented reality. This structure is specific for human existence in each literary text, and the structure of the power expressed in the Talmud needs to be analyzed in its particularity. The fourth consequence concerns the task of the analysis to follow. The hitherto predominant theorizations of both the expressed agencies and forces of thinking in the Talmud and of the constructed readers were formulated in terms of a human or divine thinking subject (both as the agent and the substance of thinking) who, among other things, also recalls, remembers, or forgets. The Talmud and perhaps other late ancient texts are different in the way in which the expressed structure of power is organized in terms of the relationship between thinking being, remembering being, and subjects (authors, or characters) who think or remember. Describing that structural difference in the relationship between remembering, thinking, and (human or divine) subject informs the task of analysis in the following chapters. In the light of these four consequences, if thinking is disconnected from the notion of a unitary and homogeneous thinking subject in the Talmud, and if understanding and remembering are likewise disconnected from the notion of such a subject, how do the literary characters in the Talmud who collectively, yet multiply, diversely, and heterogeneously contribute to thinking and remembering, make sense of the texts of the Mishnah? Or to put it another way, how do the characters in the Talmud think about and remember the Mishnah if that process is both apersonal and beyond the control of any individual character, including even the author, designer, or composer of any given discussion in the Talmud?
six
Sense in the Making: Hermeneutical Practices of the Babylonian Talmud
If thinking is disconnected from the notion of a unitary and homogeneous thinking subject in the Talmud, and if understanding and remembering are likewise disconnected from the notion of such a subject, how do the literary characters in the Talmud, who collectively, yet multiply, diversely, and heterogeneously, contribute to thinking and remembering, make sense of the texts of the Mishnah?1 To answer these questions, I both juxtapose and mutually complement medieval Talmudists’ rhetorical theories of sense in the Talmud with the logical approaches to sense in Bertrand Russell’s and J. L. Austin’s truth theories.2 I use the resulting theoretical lens to isolate and analyze complex interactions between the three distinct, yet compositionally closely intertwined, series of the making of sense in Talmudic discourse: refuting and defending, as described by medieval Talmudists; verifying truth by reference, as depicted by modern logicians; and inventing, testing, and refining references in order to construct the truth of a text, as practiced in the 105
106 Who Speaks?
discourses of the Talmud. I identify complex interactions of the three series as an open protocol of constructing truth, a protocol of which the Talmud is a paradigmatic example. Finally, I situate this protocol vis-à-vis more general understandings of what sense is and what it takes to make it. As discussed above, the Talmud is a collection of legal and homiletic discourses set in a Babylonian rabbinic academy at around the fifth century C.E.,3 but produced at a later period. Since the twentieth century, scholars of the Talmud have predominantly focused on the genesis of the Talmud’s text from its inception until the Talmud’s final versions in medieval manuscripts and early printed editions.4 In understanding the Talmud as the end result of this historical genesis, these scholars have relied on their practical-intuitive knowledge of the ways in which the text of the Talmud as a literary work makes sense to them. That practical knowledge, however uncritical, incorporated a centuries-long, constantly evolving view of the Talmudic discourse coined in medieval Platonic and Aristotelian methodologies of Talmudic study.5 These methodologies understood the basic process of Talmudic discourse as a series of rhetorical attacks on the formulations of the Mishnah, an early third-century code of instructions for rabbinic courts, followed by defenses. Taking into account both the influence of these methodological traditions and theories of the Talmud on contemporary Talmudic scholars and the literary-intellectual, rather than documentary-historical nature of the Talmud’s discourses, I propose an alternative question. In what follows, I ask: What can we learn about sense-making practices in the Talmudic texts? This question is not entirely new. Medieval theorists also asked about Talmudic sense making, but their answers were limited by their own paradigm of the basic processes of Talmudic discourse. Moving beyond their assumptions, I address the question of Talmudic sense making by means of a case study of a selected discourse in the Babylonian Talmud. I proceed by first isolating and then theorizing the Talmudic practice of making sense. To that end, I heuristically apply contemporary logical-philosophical theories of sense making, in conversation with and in contrast to the medieval ideas of Talmudic methodology.
From Russell and Austin to the Talmud I depart from the truth theories of referential correspondence6 developed by Bertrand Russell and referential correlation developed by John L. Austin
Sense in the Making 107
by introducing the notion of a pertinence criterion for truth. These theories define truth as either the correspondence of a proposition to a matter of fact or a correlation between a linguistically understandable claim about a matter of fact (“type,” in Austin’s terms) and the reference of that claim (a “situation”). “Type,” thus, is the sense an utterance linguistically makes regardless of whether or not that utterance is true, that is, whether or not it “correlates” with a “situation in reality.” Thus “there is a square circle” is a linguistically understandable utterance (it has a “type”), yet there is no “situation” to which that utterance can correlate. Therefore, that utterance is false. Both theories understand truth as a correspondence. For Russell it is a correspondence between the proposition and the “matter of facts”; for Austin, the correspondence is between “type” and “situation” in a given utterance. The difference has practical implications. For example, in Russell, there is no way to decide whether or not a statement “I always tell lies” is true or false. In general, the problem is that if the “I always tell lies” is true to the facts, it is false, and if its false to the fact, it is true. Austin avoids that paradox by placing it within the “type.” For him, because the “type” already embeds paradox; it corresponds to no situation. Therefore, whereas Russell has no way of determining the truth or falsity of “I always tell lies,” for Austin, this utterance is clearly false, because it cannot have a corresponding “situation.” As I will soon show, by asking what may be a specific legal issue a Mishnah addresses, and by contemplating different, even mutually exclusive answers, Talmudic discussion may raise several versions of the “type” for a given Mishnah, and to search for a respective correlative situation for each to prove the Mishnah true under each type. Instead of two elements of correspondence, we have three. Instead of either “proposition” versus “matter of facts,” or “type” versus “situation,” we have the language of the Mishnah, a set of possible interpretations of its “type,” and a set of possible “situations” for each possible type. That move makes the truth of the Mishnah less dependent on any particular type under which this Mishnah is linguistically understood. In the analysis in this chapter, however, I address an example of only one among two different “types” of linguistic understanding of the Mishnah, as that type becomes probed for a correlative situation and in that way discussed in the text of the Talmud. In Russell’s and Austin’s approaches, the lack of either correspondence or correlation renders a proposition or a claim false. In contrast, the notion of truth in the Talmudic discourse that I analyze in this chapter is defined by
108 Who Speaks?
a truth criterion that is new in the contemporary theoretical context—the criterion of the referential pertinence of the Mishnah’s instruction concerning the “situation,” to borrow Austin’s term. This criterion requires the consequent application of the Mishnah’s instruction to hypothetical references or situations that the student in the rabbinic academy proposes, applications that the teachers or other students disprove. As speakers in the Talmudic discourse, students and teachers in the academy construct and probe situations until they come up with those to which the Mishnah’s instruction pertains. The process of constructing and probing situations advances through the series of attacks and defenses these speakers make. By going through these motions, the speakers determine whether or not the Mishnah’s instruction pertains to a proposed reference or situation, and if one speaker proves it does not, that speaker or another one raises the possibility that it applies to another situation, which often appears as an adjustment of the terms defining a previous one. If one speaker attacks by outlining a series of logically conceivable possible situations and fails all of them one by one on the pertinence test, the other speaker defends them by reconceiving some of the rejected possibilities so that they can pass the pertinence test in the next round of attempted refutations. The result is a much narrower set of much more viable situations to which the Mishnah’s instruction may pertain. The general criterion by which a speaker either approves or disproves a situation is whether or not the Mishnah’s instruction pertains to it. Using that criterion, at the end of the process, the speakers collectively attain a refined formulation of the situation or situations and of the ways in which the Mishnah might pertain to it or them, thereby confirming that the type or the sense in which the speakers have understood the language of the Mishnah is viable. This practice is different from Austin’s approach to truth as a correlation between the type and the situation in two major ways. First, the lack of correlation with a reference does not prove that the Mishnah is false but only leads to a search for another situation. Second, the Mishnah remains true even if a type or a certain linguistic sense in the understanding of the theme of the Mishnah has no correlation with any situation. If such is the case, the type needs to be changed, but the Mishnah’s potential for being true remains the same.7 There is more. Pertinence serves not only as a criterion of truth, but also and much more importantly as a generative principle to create references,
Sense in the Making 109
that is, situations. If the test for the pertinence of the Mishnah’s instruction for a proposed situation fails, the result helps construct a new situation or adjusts the previous one to have a better chance of passing that test on the next round of refutations. That makes the pertinence criterion of truth a tool for actively generating truth, rather than a measurement for passively judging it. If a given reference fails the pertinence test, the speakers in the Talmudic discourse either look for a new one or conceive the failed one in a new way. In sum, unlike contemporary truth theories, proving that Mishnah instruction neither corresponds to or nor correlates with a proposed reference or situation does not render the Mishnah false, but rather invites the speakers either to construct a new situation or to reconsider the type. What this means, however, is that the speakers in the Talmud do not judge the Mishnah’s truth based on reference, but rather question the truth of reference based on the Mishnah. In Russell’s terms, this means that matters of fact must correspond to propositions, not propositions to matters of fact. The students and teachers of the rabbinic academy seek the truth in this way in part because their rhetorical assumption is that the Mishnah makes sense leading to truth even if they neither know yet what that sense is nor trust any proposition of what that sense (“type,” for Austin or “reference” for Russell) might be without testing it first. The speakers first prove that the sense of the Mishnah’s instruction is of a certain type and then entertain the possibility of a different type, with a respective set of situations or references for it.8 As a result, instead of the binary of type versus situation, the speakers of the Talmud apply a more complicated formula that involves two or more types discerned in one and the same text of the Mishnah and several situations under each type. By contrast, in Austin’s approach to truth as correlation, the truth of a claim originates in a bidirectional correlation between the linguistic sense or type of the claim and the situation that is thereby claimed to exist, and therefore, for him, the sense or type precedes the discovery of the truth, that is, precedes the correlation between the type and the situation. In this aspect, Austin’s theory of truth as correlation differs from Russell’s correspondence theory of truth, in which truth is a unidirectional correspondence of a proposition to its reference, that is, to a matter of fact. As with Russell’s correspondence theory, the sense-making practice of the speakers in the Talmud takes a unidirectional approach, but they verify reference by the pertinence of the proposition to it, not the proposition by the
110 Who Speaks?
reference. By the same token, the sense-making practice of the speakers in the Talmud differs from Austin’s bidirectional approach to truth as correlation. Unlike Austin, the students and teachers in the Talmud measure the truth of the situation only by seeing if the instruction in the Mishnah pertains to it. They do not measure the type by the situation, but rather invent a situation and measure it by the pertinence of the instruction to it. Yet an even more important difference between the practice of the speakers of the Talmud and Austin’s theory is that Austin allows the possibility of there being sense without truth—for example, a type that does not correlate to the situation. However, as we have seen, the Talmudic speakers heuristically hold the Mishnah to be true even before they explore the plurality of its senses—its types and the situations to which the types might pertain. Thus, like Austin, and unlike Russell, the speakers in the Talmud differentiate between truth and sense. Unlike both, however, they place truth before sense, not sense before truth.
From Medieval Methodologies to the Talmud Isolating referential pertinence as a truth criterion helps differentiate the Talmudic practice of sense making from its theorization in the medieval Platonic and/or Aristotelian methodologies of the study of the Talmud. Although these medieval methodologies also considered attacks and defenses to be the main form of the Talmudic discourse, they disregarded the generative role of attacks and defenses in the sense-making process. Instead, they either depreciated the attack-defense form as extrinsic to the legalistic content of the Talmud (Maimonides)9 or appraised that form as the main paradigm of the thinking Talmudist. In the latter case,10 the efficacy of Talmudic thinking was defined by a different criterion of truth, as compared with that of pertinence—the criterion of invention, or the novel. According to this criterion, an understanding of any given Talmudic dictum is true if it isolates the novel, or an invention, made by that dictum. Isolating what is new in a dictum requires defining an alternative understanding of the same matter, an understanding that either the contemporary audience of the rabbinic academy or a later student of the Talmud would have accepted until the dictum replaced that alternative understanding. Thus, an alternative understanding is what the dictum refutes.
Sense in the Making 111
Shaped in the early modern conceptualizations of the Talmud, this view promoted what the dictum refutes as a new truth criterion. I know the truth of the dictum if I know what it came to refute. By making the refutation of alternative understandings a central purpose of the dictum, and thus into the criterion of its truth, an early modern (or modern) student of the Talmud can contemplate what the dictum accomplished—the point it refuted and thus what the new understanding of things was (or more technically, what was the novel) that the dictum invented. Knowing truth thus lies no longer in correspondence, but rather in correlating the dictum with what it refutes in it. Although that approach provides a very fine and sharp appraisal of the intellectual value of refutation in the Talmud as a means of invention—and as a truth criterion—it presumes that the protocol of invention as refutation is the main protocol of sense making in the Talmud. Needless to say, this is a powerful answer to the question how Talmudic discussions make sense for a student living in post-Talmudic times. However, by providing such a powerful answer to the question of making sense, they thereby have closed the question and precluded any other approaches to it.11 Without denying the validity of this answer, my goal here is to pose the question of making sense in the practices of the speakers in the Talmud in its own right first, independent of any specific answer to it. In what follows, I try to show, through a case study of a text from the sixth chapter of the Babylonian Talmud Tractate Bab’a Metzi‘a, which is the second of the three parts of the Talmudic tractate on damages, Nezikin,12 that once the question of how sense is made in Talmudic discussions is reopened, what becomes clear is the role played by the criterion of pertinence in the rabbinic academy and the generative power that the use of that criterion exemplifies. I conclude the chapter by moving once again from a close reading of this text to a more general summary of the role played by the generative pertinence truth criterion in Talmudic discourse.
“One hired skillful workers. Then they deceived one another.” Our discourse begins with a reciter, or tann’a, quoting by heart from the Mishnah, a third-century Palestinian Hebrew text, which will then become the subject of a discussion in Aramaic between rabbis and their students in a rabbinic academy in Babylon. Yet, before I begin, a brief stipulation is in order.
112
Who Speaks?
However carefully constructed, Talmudic discussions are not designed for us, people of modern times. Therefore, directly translating these discussions into English does not serve. Instead, I will present the discussion as it unfolds in three distinct, yet intertwined series: the series of speakers, the series of refutations and defenses, and the series of constructed references or “situations.”
the series of speakers
The Mishnah, an early third-century instruction for rabbinical courts, reads:13 השוכר את אומנין והטעו זה את זה אין להם זה על זה אלא תרעומת One hired skillful workers. Then they deceived one another. They can have no claims against each other except for complaints.
This short paragraph in the Mishnah leads to an extensive discussion in the Talmud. What I have translated as three separate sentences in the Hebrew of the Mishnah coheres in one single phrase—indeed a short, but full, story. Linguistically, the first two sentences are its theme, or topic—what is being talked about. The third is the rheme, or comment—what is being said about it—which in this case is an instruction to rabbinical courts to use when a case fitting the theme arises in court. To use Austin’s terms, one might ask, What is the type of this theme of the Mishnah? The Talmudic discussion, a fragment of which I will analyze, considers two versions of that type. One is “the workers deceived one another”; the other, “the owner deceived the workers.” The Talmudic discussion begins with the first, linguistically more obvious “type.” The question about the type of the Mishnah (that is, who deceived whom) is, in effect, the first question addressed in the text I analyze, which is, like many other texts in the Talmud, a discourse set in an idealized Aramaic-speaking rabbinical academy in Babylon. As an institution of rabbinic training, the academy not only analyzes the existing cases, it more often creates new cases by which the students can better understand the application of a given law of the Mishnah. The Talmudic discussion I read here is an example of creating cases, rather than analyzing a given case in a “real” court.
fup-dolgopolski-06.indd 112
10/9/12 12:02 PM
Sense in the Making 113
The language of the Mishnah above is laconic, but not clear in reference. In third-century Palestine, the final sentence of the Mishnah, “They can have no claims against each other except for complaints,” might have sounded like a way of saying, in “type,” that the “deceived” party has no strong point in court. Even if that party might complain about injustice, it might not have enough to present a reasonably winning case in the court. However, a few centuries later, in the Babylonian academy, complaining about injustice had become a separate legal category, and thus a new “type,” as opposed to not having the right to complain at all, as well as to having a nearly automatically winning claim of unpaid salary to present in court. It would thus seem as if that separate legal category is the type in this example. Understood under this new “type,” the Mishnah thus instructs the rabbis that even if no claim of monetary satisfaction can be approved by the court, the court should still recognize complaints about injustice or “deceit” as rightfully made. However, as we will see, what the discussion that ensues puts in question and answers more precisely is what type and situation in fact pertain here. In this idealized setting of the rabbinic academy, the Mishnah is recited by heart and then is discussed by the rabbis and their students. As their discussion progresses, the speakers discover more than one viable version of the specific case or “situation” to which the theme or type of the Mishnah might refer and eventually even more than one type.14 Yet, under the first type (a worker deceived another worker), we are only told the following. A skilled worker was hired, no doubt either by another worker sent by the owner of a business or farm or directly by the owner, but by the end of the day, one of the parties “deceived” the other. Coming to court, the deceived party complains about injustice and on that basis claims money. The other party contests the rightfulness of the monetary claim and by implication denies the complaint about injustice. But who exactly deceived whom? And who did the hiring? Also, what were the particular circumstances involved in that hiring? In asking these questions, the rabbis and their students look into the theme of this Mishnah, “One hired a skillful worker. Then they deceived one another.” The rabbis attempt to determine the “type” in that theme and consequently ask to which “situation” that theme refers. A successful answer would tell how the rheme “They have no claims against each other except for complaints” pertains to that “situation.” In so doing, the rabbis and students also refine their understanding of the type.
114
Who Speaks?
The tann’a, or reciter, does not say anything further about all these matters. Instead, questions concerning the type and situation of the theme and the pertinence of the rheme to that situation drive the discussion as it moves through a series of apologetic attacks and defenses, all centered on that text in the Mishnah. The speakers first raise doubts and then establish a viable possibility—in fact, multiple possibilities—concerning the type and situation to which the Mishnah’s theme could refer and to which the Mishnah’s instruction would pertain. This carefully composed, multistep discussion reveals the practice of making sense to which I attend. Importantly, the discussion happens between two speakers of Aramaic.
the first aramaic speaker: what is the type?
A speaker of the Aramaic language opens the discussion by what might sound like an artificial attack on the reciter’s word choice. In Austin’s terms, the purpose of the first speaker is to come up with a type, under which the Mishnah can be linguistically understood. The target of this attack is the word “deceived,” which the reciter of the Mishnah prefers to the phrase “withdrew from the contract,” which is used elsewhere. The Aramaic speaker first rhetorically attacks and then ultimately defends that word choice.15 חזרו זה את זה לא קתני אלא הטעו זה את זה דעטעו פועלים אהדדי [The attack:] “The reciter did not say [a party] ‘withdrew from the contract.’ Instead he said the party ‘was deceptive’ to another one!” [The defense:] “Rather, it means the workers deceived one another.”
This attack seems quite artificial. After all, any word choice can be doubted if there is another possibility, and there almost always is. What, then, is the first Aramaic speaker’s purpose? We will better understand that purpose by paying attention to the defense that follows in response. As that response will suggest, the choice of “was deceptive to another one” as opposed to “withdrew from the contract” is indicative of the type of cases to which the Mishnah’s theme refers and to which the rheme pertains. Let us not ignore the lesson in semantics: The choice of words that otherwise can be synonyms can help clarify the type (thus, the type is about two or more workers, not
fup-dolgopolski-06.indd 114
10/9/12 12:03 PM
Sense in the Making 115
about an owner and a worker: “workers deceived each other,” rather than “the owner and worker deceived each other and the deceived party withdrew from the contract”). How? Before considering any specific answer to that question, let us appreciate what answering it strategically does. Having an answer verifies that the choice of words in the Mishnah was not arbitrary. It thereby also helps establish that using one expression (as opposed to another was in no way an imprecision in the memory of the tann’a. By the same token, it was not a sign of any lack of authority in the presentation of the Mishnah. In the answer, the first Aramaic speaker raises the possibility that the choice of “[workers] deceived one another” over “[either the owner or a worker] withdrew” might have been informed by a difference in type. The speaker aims to isolate what that difference was. Aramaic idiom helps. Translatable as “deceived one another,” the Aramaic reflexive form ‘it‘u (lit., “they deceived themselves”) serves an interlinguistic clue. The Mishnaic Hebrew causative form hat‘u (“one party tricked the other”) may or may not have a similar meaning but surely sounds similar. For a speaker of Aramaic, the phonetic similarity between Aramaic and Hebrew suggests that the Mishnah’s reciter’s choice of “[workers] deceived one another” as opposed to “[either an owner or a worker] withdrew” might have implied a semantic difference in reference to the parties involved. That is, it introduces the possibility that “they” refers to others in addition to the “one” who did the hiring and the “skillful worker” who was hired; or this is how the first speaker understands the type in the language of the Mishnah. So, as the speaker proposes, choosing “withdrew” signifies that the reference is to withdrawing from a contract between an owner and the hired worker, while either “tricked” in Hebrew or “deceived” in Aramaic would refer to the hired worker, who, at the request of the one who hired him, hires other workers, in which process an act of deception takes place. “Deceived” thus refers to a worker hiring others, while “withdrew” would refer just to someone hiring a worker. If so, the Mishnah’s type is “a worker hiring another worker” and the Mishnah’s instruction on complaints about deception would now pertain to a situation happening between these two workers. The proposed reference to a situation of a conflict (“deceit”) arising when a worker hires others is probabilistic and heuristic in nature, rather than rigorously logical or linguistic. The situation does not apodictically follow from the language of the Mishnah and the type presented therein. No linguistic analysis of the word “deceived” as opposed to “withdrew” can yield a reference to a worker
116
Who Speaks?
hiring others as opposed to someone hiring a worker. Instead, the transition from words to their references is synthetic, rather than analytic. That is, it first opens up room for a question about reference (that is, about type and a corresponding situation). The speaker then synthetically fills that room from the outside, rather than analytically deriving the content from the inside. Moreover, at the same time and by the same move, since asking about the word choice represented a rhetorical attack on the validity of the reciter’s memory, the answer or the situation is supplied by way of defense and is satisfactory insofar as it fulfills that role well. In more general terms, in the discourse, the search for a situation for a currently considered type is circumscribed in the protocol of attacking and defending. A success in defense serves as a criterion of truth for the proposed reference to a situation. The reference is initially acceptable if it successfully serves the defense. In even more general terms, this means that the protocol of the acceptability of a situation follows the protocol of the dependability of the type. The rule is that the reference can be true if it helps the defense and is surely false if it does not. In sum, an externally supplied reference does no more and no less than make the Mishnah appear to be rigorous, rather than arbitrary. The first Aramaic speaker interprets the Mishnah to refer to a type of “skilled worker hiring other workers,” and they “deceive” one another along the way. This speaker thus proposes a possible type and is in search for a possible situation to which the theme of the Mishnah refers: two workers deceive one another in hiring. Yet simply raising the possibility of a reference to workers deceiving one another is not enough. To defend the reciter’s memory, and thus the Mishnah’s authority, the first speaker needs not only to supply a type but also to find a situation in order to demonstrate that the instruction or rheme of the Mishnah might viably pertain to it. In some more detail, to make that possible type of reference viable, the speaker needs to justify it by finding a model, a possible factual situation of two workers, to which the Mishnah’s instruction about the claims of unfair pay would pertain. A question, indeed, an attack, by the second Aramaic speaker quickly makes that demand. the second aramaic speaker: what is the situation?
היכי דמי What does it look like?
fup-dolgopolski-06.indd 116
10/9/12 12:03 PM
Sense in the Making 117
Translated into the language of Austin’s theory of correlation, the question above means: Can we find the situation? Can we thereby justify the type?16 The second Aramaic speaker is faithful to the role of an attacker and uses the question for an attack. That speaker promptly asserts the need to propose a viable reference or situation for the type of a worker hiring others, that is, to find a situation to which the instruction in the Mishnah would pertain. The speaker then systematically shows that satisfying that need will be hard. The speaker first asks a rhetorical question: “What does it look like?” which actually asks “Can the Mishnah be understood under such a type?” Meaning also: “Can there at all be a situation, to which the Mishnah thus understood would pertain?” This speaker then unpacks the question though an elaborate logical argument in which a tree of possible situations for the type is considered and all its branches are cut off one by one. The immediate end result is that there seemingly can be no such situation of a worker hiring others for which the Mishnah’s instruction would pertain. The more subtle significance of that move, however, is that in the larger framework of the attack-defense protocol of discussion, the argument of the second Aramaic speaker serves as an attack on the type. The logic of this attack is as follows: If there is no situation matching the type, that type seems to be wrong, and the Mishnah should be understood in a different way, that is, under a different type. Overall, the argument of the second speaker is presented as a series of tactical defenses of the first speaker, all of which are undermined by new attacks. Branches of new possible situations arise, are tested for pertinence, and fail that test, one by one. In the next section, I closely explore the sense-making procedures in this argument.
all branches are cut
One branch of the possible situations17 arises if the owner sends a worker to hire another worker for four coins per day, but to take advantage of the market, the worker hires his peers for only three coins. A symmetrical branch of possibilities is if the master asks the worker to hire a peer worker for three coins per day, but given the market, the worker can hire a peer for no less than four coins. The second speaker dismisses the first branch from the start, because it does not pass the pertinence test: “Would you say that the owner told him to
118 Who Speaks?
hire for four but he told the workers “for three”? How can the workers justly complain about that? After all, they received and accepted the offer.”18 That means that the Mishnah’s instruction about justified complaints of unfair pay would not pertain, because there is no place for such a complaint here. That second Aramaic speaker therefore turns to the complementary branch of situations, in which the labor market is higher than the owner had assumed. “Should you say [then] that the owner sent the worker to hire others for three, but the prices were running at four?”19 The worker hires others at a market price that is higher than the owner has approved. Yet at the end of the day, the workers are paid only three coins. Can workers viably claim the fourth coin in court, or at least complain about injustice? A “yes” answer seems to be obvious. But is not that answer too strong? If the workers have a nearly automatically winning claim to receive the fourth coin, the other party should not even bother to go to court, or the judges would have no hesitation about how to solve the case. Because the workers can legally claim the money, there is no place for complaints about injustice. This means the Mishnah’s instruction (rheme) does not pertain.
a At this point in the argument, the second Aramaic speaker returns to the case (that is, the “situation”) of hiring workers for four coins when the owner had approved the price of only three, this time with the worker who does the hiring specifying that their pay was “on him/her.” The speaker quickly fails this on the pertinence test by arguing that the case is once again too strong for the Mishnah’s instruction to pertain to it. The speaker quotes an apocryphal text (in post-Talmudic parlance, a baraith’a), a text that by a silent consensus, that is, one that is followed in all Talmudic discourses and approximates Mishnah in its authority. Based on the baraith’a, the speaker infers that the worker must pay his peers four coins in full and only then approach the owner for reimbursement. (I leave the question of the nature of that inference outside of the current discussion.) That baraith’a undeniably grants the workers their full compensation, which clearly renders the Mishnah’s instruction to deny the monetary claims of the workers, yet to approve their complaints about injustice, irrelevant to the case. Thus far, it also leaves the students and the rabbis in the academy with no case or situation that has passed the pertinence test. Where either there is no place for complaining or there is a place for an
Sense in the Making 119
obvious winning claim, there is no place for the Mishnah to instruct whether or not the complaining is rightful. Yet the second speaker does not stop before considering and ruling out all logical possibilities for situations. In search of a logically possible case to which the Mishnah would rightly pertain, the second speaker moves to an alternative version of the last reference. The worker might have told others that their pay of four coins was “on the owner.” In contemplating that possibility, as the attentive audience might have already noticed, trusting an absent party (the owner) might sound even harder and thus less acceptable for the workers. However, they might still find it worth the risk if four coins is a price exceeding the average market price on that day. As a result, these workers take a risk for a possible gain, having not much to lose if, at the end of the day, they get an average market price for the work, instead. That again renders their complaints about injustice unsupportable, and hence the Mishnah’s instruction not pertinent, for in this case, too, the hired workers can have no complaint about the risks that they took. If the preceding case was too strong to pass the pertinence test, the current case is too weak to pass it. The Mishnah could not possibly instruct judges on how to proceed in case of a complaint that falls apart by itself. In sum, as the systematic efforts of the second speaker have shown, all logically possible situations for the type, under which the Mishnah has been understood (that is of workers deceiving one another in hiring) seem to fail the pertinence test. If the worker decreased the offered price, and the others accepted it, they can neither claim they were underpaid nor complain about injustice. If the worker increased the offered price in his own name, he has to pay, and thus there is no place for complaining. If the worker increased the price in the name of the owner, the workers could not complain either, because they knew they were taking a risk, but would still be paid at the level of market price. Hence, the Mishnah’s instruction (rheme) pertains to none of these possible situations in reference, and thus the Mishnah’s theme (or the type thereof) cannot refer to them, either. At the end, the second speaker renders all apodictically conceivable ways to refer the Mishnah to workers deceiving one another as possible, but not pertinent, and in that sense proves that the Mishnah, if understood under this “type,” is false. To present the work of the second speaker in even more general terms, I might say that through attacking the first Aramaic speaker, the second Aramaic speaker considers the interpretation of the reference of the Mishnah
120 Who Speaks?
(“workers deceived each other,” rather than “the owner deceived the workers” or “the workers deceived the owner”) that the first speaker proposed. The second speaker then outlines all logically possible cases to which the Mishnah could have referred, followed by proofs that the Mishnah’s instruction pertains to none of these cases. Having dismissed all logical possibilities, the second speaker has thereby demonstrated that the first speaker interpreted the Mishnah in such a way that it can apply to no case whatsoever. It has no viable reference. Indeed, as it is interpreted by the first speaker, the theme of the Mishnah might have several possibilities of reference, but its instruction (rheme) can viably pertain to none of them. In a purely logical world of apodictic possibilities, as if the Talmud were a chess game, the above might mean that the second speaker put the first one in checkmate, because the second speaker has demonstrated that the Mishnah has no viable reference to which it could reasonably pertain. This argument helps the second speaker show that the first speaker’s interpretation of the Mishnah is false. Following Austin, one might say that the second speaker proved that the first one interpreted the Mishnah in such a way that the Mishnah describes the type of a situation (a descriptive claim: workers deceive one another), but there is no factual situation to which that type would demonstrably pertain. By the end of the speech, the second Aramaic speaker has established that the first speaker’s interpretation of the Mishnah as “the worker deceived other workers” is logically false, which means it is defeated.
the first speaker again
As it is known from Hölderlin,20 salvation comes from the site of disaster. Instead of accepting defeat, the first Aramaic speaker makes use of the final result of the second speaker. The first speaker adopts one of the failed situations and invents it anew. Far from giving up when confronted with a logical checkmate, the first Aramaic speaker defends the type of the Mishnah by raising new possibilities of pertinence, precisely where the last logically conceivable reference failed the pertinence test. To do so, the first speaker takes a theoretically different approach as compared with that of the second speaker. Instead of constructing a possible situation of reference followed by a pertinence test, the first speaker makes use of an already refuted reference to raise a new possibility for pertinence. The result is that an already
Sense in the Making 121
constructed (and failed) possibility of reference passes the pertinence test. Using that strategy, the first speaker defends his initial defense of the Mishnah’s type as workers deceiving one another. Stunningly, the two speakers not only fight a fight, but dance a dance. Their polemics actually constitute constructive cooperation in sense making, rather than a destructive contest in which only one wins. The second constructs the references, but fails them on the test. The first builds on their ruins, precisely where the second has cleared the ground. The first uses the last of the references, but conceives new possibilities for passing the pertinence test. The movements of two speakers form a strophe and an antistrophe, as if the moves made by them were choreographed by the rules governing two series of movements. The first series is that of polemics: thesis, attack, defense, attack, defense, and so on. The second is that of constructing a reference, failing it on the pertinence test and constructing another one on its ruins. Remarkably, the two series do not coincide, nor do they coincide with the distribution of the speakers’ roles. The role of the first speaker is to defend the Mishnah, but instead, that speaker starts with an attack on it. The role of the second speaker is to attack the first speaker, but instead, that speaker starts with a heuristic defense of the first speaker. Constructing references, failing them on the pertinence test, and attempting to adjust them occur both in the role played by the second speaker and in the second appearance of the first one. In this complex interplay of roles and series of moves, the Talmudic discourse is now at a point where the second speaker adjusts the last, failed reference of the first one so that it would be good enough to pass the pertinence test. How does that specifically work? How does the new pertinence test work better than the last one? It does so by challenging a hidden assumption about the state of the market. In the second speaker’s discourse, the last failed possibility of reference was to a worker hiring others for four coins with only three approved by the owner. The worker stipulated that the payment was the responsibility of the owner. The workers accepted the risks of not receiving the fourth coin because of the chance of a better reward. In failing that possibility on the pertinence test, the second speaker silently assumed that the going rate in the labor market was three coins per day. To prevent the case from failing, the first speaker now challenges that assumption. If work at both three
122 Who Speaks?
and four coins per day was available, the workers who accepted the offer of four, with responsibility for payment “on the owner,” but at the end of the day received only three, might be successful in court in either claiming the money or at least in complaining about injustice. They can argue that if they knew they would have received only three, they could have tried to get another job offer for four coins per day. In this case, both their claim and their complaint can reasonably either win or lose. Their argument is neither too strong nor too weak. Having a claim that is neither too weak to win nor too strong to lose means that the Mishnah’s instruction (to deny the claim, but to approve the complaint) is pertinent to the situation. Without the Mishnah, the judges would not have any obvious way to decide. Thus, the deceived workers’ argument, “There were offers of three and of four, and they said to him, had you not said ‘four’ to us, we would try harder and find an offer of four,”21 passes the pertinence test. Pertinence of the instruction to the situation ensures the type of its theme. The situation is that a worker agrees to work at four coins per day “on the owner,” and with the price of three coins per day preapproved by him, and with the labor market at both three and four. This situation now matches the theme of the Mishnah, that is, its type, interpreted as the “workers deceiving one another.” That validates the first speaker’s interpretation of the Mishnah’s type as referring to workers deceiving each other, even if in a remarkably narrower situation.
the speakers’ shadows
The success of the first speaker thus brings a kind of closure, but the narrowness of referential basis thus attained opens the question again. Other Aramaic speakers in the academy—shadow speakers—look for other ways to vindicate the instruction, ways that the first Aramaic speaker might have explored, but did not. I called these others “shadow speakers,” because they act on behalf of the first speaker and attempt to accomplish what the first speaker has already done, albeit in a different way, proposing other possibilities of situations to pass the pertinence test. Exploring other venues on behalf of the first speaker is stunning evidence of the open nature of the academy curriculum. The “finished” discourse allows for expansion and thus for continuation. Let me take a closer look at the structure of the shadow speakers’ discourse.
Sense in the Making 123
One shadow proposal suggests, with a surprisingly frank awareness of its probabilistic nature, that it was the owner of the field who hired others to work on his field for three coins and who then went to work in another field, taking the chance of getting four coins per day himself, rather than three: “If you will, the reference was to an owner [who hired others to work on his field and sold his own labor for a better price]. They said to him, had you not said ‘four’ for us, I [the owner of the field] would not have done such a disgraceful thing [to hire others for three coins to work on my field while doing work on your field for the same pay!].”22 That owner might be right to claim underpayment or at least to complain about the injustice of it. Thus, the Mishnah’s instruction would pertain to that case, as well. The other shadow proposal is symmetrical with this one, albeit more elaborate. Unlike its twin, it involves not only constructing a situation and disproving it through the pertinence test but also through an internal series of attacks and defenses. To that end, it invokes a shadow of the second speaker, as well. This shadow proposal is as follows, “If you will, the reference was to [ordinary] workers [those who used to work for the owner—Rashi, ad locum]. The workers said to him, ‘Because you have told us “four” we have done an excellent job!’ ”23 This situation passes the pertinence test because the workers have what seems to be a viable argument that at the same time is not too strong. However, an attacking voice on behalf of the second speaker argues that this in fact might not always pass the pertinence test, because “what if their work was digging a drain,”24 which implies work of unknowable quality. The shadow voice of the first speaker deflects the attack: “Even in such a case, the quality of the work might be known.”25 At the end, the shadow voice of the second speaker suggests, “The drain might have been all full of water, and thus the quality of work was unknown.” This means the argument of average skilled workers complaining that they have done a job that is better than what they were paid for is not strong enough to pass the pertinence test, because their argument is very hard to support. Unlike its twin, the outcome of this polemic remains unclear. The proposed situation either fails completely or needs to exclude one of the typical jobs—digging a drain or other jobs comparable to this, as far as the knowability of how well it could be done is concerned. Yet even the defeat of this proposal makes the previous one look stronger: The situation of a field owner who went to work for another owner represents a pertinent possible
124 Who Speaks?
reference for the first speaker’s interpretation of the Mishnah. It thereby adds to the strength of the first speaker’s defense of the Mishnah’s type. Yet even here, the shadow voices of the first two speakers do not become silent, which is once again evidence of the open nature of the protocol of the Talmudic discussion. The shadow voices now return to the point at which the original second speaker considered the reference to be to a worker who decreased the price at which he hired other workers from four coins to three. As the audience is assumed to remember, the second speaker failed this reference on the pertinence test for its being too weak by arguing that the workers cannot have any complaint about a price that they have both considered and accepted. However, the voice shadowing the first speaker acts in much the same way as the “real” first speaker and offers to amend the reference so that it would be able to pass the pertinence test. “If the worker will say to them, ‘You have considered and agreed,’ they can respond with a phrase from Proverbs, ‘Do not take away from the owners what already belongs to them!’ ”26 As obliquely as it does, the quote suggests to the worker “when the work is already done, you should not take from us what you promised, because the promised is ‘already’ ours.” This argument has strength, but is not strong enough to win without a decision of the court, because using Proverbs to solve a monetary dispute is strong enough for one litigant to make a viable claim for money. However, it is not strong enough to make that claim legitimate in the eyes of the other litigant. This is where an intervention of the court is pertinent. Following the instruction of the Mishnah, the court does what the litigants cannot do without it: approves the viability of the claim and denies monetary compensation for it, because and only because the Mishnah tells what to do in such case. It means the situation pertains to the Mishnah’s instruction. That makes the first speaker’s defense of the Mishnah’s type as referring to two workers deceiving each other even more viable than before. As far as the overall composition of this discussion is concerned, this latter victory compensates for some weakness in the argument that immediately precedes it. From that point on, the Talmudic discussion continues with yet another shadow voice. Like other voices, it picks up the discussion between the first and the second Aramaic speakers. This shadow voice expands the tree of possibilities that the second speaker constructed in order to refute, and the first speaker ventured to defend. However, the shadow voices sound weaker. They only expand existing branches of the tree and run them through the
Sense in the Making 125
pertinence test, but no longer affect the outcome of the main conversation— between the first and the second speakers. The Talmudic discussion continues and will soon reverse its direction to support a completely different type of the Mishnah in which the Mishnah refers to what the first speaker dismissed, that is, to an owner or a worker changing his or her mind. However, I will stop here to to isolate, in more general terms, a protocol of sense making that the discussion reveals by performing it.
the open protocol
The fact that the first and the second speakers have shadow voices is an excellent illustration of the open nature of the protocol of making sense in the Talmudic discussion. First, the speakers build on each other’s contentions, rather than attempting to vanquish each other by means of them. The series of thesis-attack-defense does not coincide with the series of two speakers, nor does it coincide with the series of constructing references and submitting them to the pertinence test. The first speaker starts with an attack and continues with a defense. The second speaker aims at undermining the defense of the first by attacking it. Yet, that speaker does not start from a direct attack, but rather from a thesis. The speaker heuristically and synthetically supplies the interpretation (type) that has been advanced by the first speaker with a possible reference (situation) in which a worker has hired another while decreasing the price. The speaker thus proceeds with the series of thesis, attack, and dismissal of the thesis. As the discourse of that speaker progresses, another series, that of constructing references and failing them on the pertinence test, is created. In turn, when the first speaker responds to that, the format of the response is a continuation of the latter series, replacing the refutation with an approval, which builds up the same series by offering a positive solution for the problem that the second speaker resolved in a negative way. By working against each other, the speakers work together. Thus, the series of thesis, refutation, and defense serves as a generative sense-making process. Had rhetorical apologetics been the only role of this series, it might simply function as an infinite feedback loop of attacks and defenses. However, it does not, because it serves yet another role, that of constructing and testing the references as a way to make sense. Thus, an important product of the overall argument is the synthesis of an increasingly
126 Who Speaks?
refined reference for the Mishnah through a series of pertinence tests. Second, the Aramaic speakers do not have personalities. Rather they function as placeholders defined by the difference in their choreographed roles, not by their identities or by any content or structure of their arguments. The first defends the Mishnah (by heuristically attacking and ultimately deflecting the attack). In turn, the second destroys that defense. After that, the first responds. Because the two speakers are placeholders, they can be populated with agents producing a variety of arguments as long as they fit the placeholders’ roles. This leads to generating various shadow agents or shadow voices that expand and enrich the argumentation of the placeholders. While these shadows may rhetorically threaten to change the overall dynamic of the initial exchange between the placeholders, they in fact end up reapproving the overall structure of the initial argument by immunizing it against yet another possible attack. Their ultimate role is that of reinforcement.
a Third, the true reference of a theme of the Mishnah is one that passes the pertinence test, which determines if the Mishnah’s instruction (rheme) pertains to that reference. References for a theme in the Mishnah are synthetically supplied from the outside, rather than analytically derived from the Mishnah itself or even from its linguistically conceivable types. This allows me to draw some fine lines between the Talmudic sense-making process and Austin’s views of how sense is made. To begin with, the types involved in making sense of a situation can be many, and no type can correlate to the reference unless that reference passes the pertinence test. In addition, there is a certain requirement that a reference must meet to pass the pertinence test. To have the Mishnah’s instruction, or rheme, pertain to the reference of its theme, the supplied reference (theme) needs to be neither too strong nor too weak. In such an approach, therefore the criteria of truth include not only referentiality, but also pertinence. Establishing the truth of an interpretation of the Mishnah requires that the situation must correspond to the theme of the Mishnah, rather than the Mishnah to facts. Unless textually corrupted, the Mishnah cannot be wrong. The facts (that is, the situations or references), however, can be wrong. One might argue that this difference from theories such as Austin’s and Russell’s exists only because the Mishnah is a normative text, rather
Sense in the Making 127
than a descriptive text. However, normative texts in general could also be approached using the standard correspondence or correlation theories of truth. Of course, normative texts not only describe, but also prescribe. However, both prescribing and describing spell out the facts, either as they are or as they should be. Thus, both deal with situations without necessarily entering into matters of pertinence. The correspondence theory is predicated on the notion of the world’s being or, in case of normative texts, having to be one way or another. But the Talmudic speakers’ conception of truth is predicated not on facts that are one way or another, but rather on the facts’ ability to pass the pertinence test. What is more, the pertinence test attempts to apply instruction to facts, which the instruction neither describes nor prescribes, but only regulates. This leaves either the actual or potential existence of the facts outside of the consideration of the Mishnah. As for Austin, the situation makes sense not because it does or does not exist, but because it correlates or does not correlate to the type. Unlike in Austin, however, in Talmudic discussions the criterion of that correlation extends beyond the Mishnah’s type to the Mishnah’s instruction (rheme). Of course, the potential to exist is necessary for any fact to qualify as such, but this condition is not sufficient. The demand of sufficiency is satisfied only by passing the pertinence test. This means that even though the Mishnah is considered a normative or prescriptive text rather than as a descriptive one, the protocols of sense making do not completely follow either the correspondence theory or the correlation theory of referential truth. Instead, the speakers repetitively attempt to make sense of the Mishnah through reading the theme of the Mishnah as a “type.” They consequently use rhetorical techniques of refuting and defending to propose, test, and finally narrow down a “situation” to which every type they try can pertain. This entails a rather particular theory of relationship between sense and truth neither Austin nor Russell have. The plurality of the ways to read the “types” and to identify a situation to which each type would pertain dissociates sense from “truth.” To be true no longer requires making one and only one perfect sense. There can always be more than one way in which the Mishnah makes sense, and this no longer undermines but rather reinforces a demonstration that the Mishnah is true. This relationship between sense and truth differs from modern thinking. Instead of limiting themselves to finding only one “true sense” of the Mishnah as an utterance about what is, or in this case, what should be done in the court, the speakers in the Talmud consider a variety of types under which
128 Who Speaks?
the Mishnah may make linguistic sense and be true. They work very hard to both discover and invent a situation corresponding to each of the types, which would render each type “true.” Both unlike and in terms of Austin’s theory of correlation in a true utterance, as an utterance, the Mishnah can have more than one type correlating to a situation. Thus, the Mishnah can be a true instruction in more than one way. Thus, the truth of the Mishnah is no longer confined to the correlation between type and situation, but rather transpires both despite and precisely due to the variety of types that can pass the pertinence test, or be correlative, in Austin’s terms. What is more, the speakers in the Talmud follow a protocol and are no longer seen as the origin or cause of the thinking they do; rather they only participate in thinking as replaceable and interchangeable agents. Neither can any particular speaker in the Talmud claim control over the protocol of thinking, which that speaker performs together with others. Instead, they participate in a very complex dance of thinking, which both establishes and verifies a diversity of possible types in which the Mishnah makes sense and correlates with a situation. The characters participate in the process of thinking on the way to the truth, but do not orchestrate that process. Their thinking happens as a dance of refutations and defenses, and this too cannot be done with only one dancer. In brief, the thinking the characters collectively do cannot be done with only one thinker. Therefore, if there still should be one thinker behind the thinking in which each of the characters onstage is only a part, that thinker should be found offstage. And if every character on the stage is a necessary but insufficient part of the thinking processes displayed through the dance, who can claim the entirety of thinking in that dance? If this is no one of the characters, then whose thinking is it after all? And if this is not one of the characters, then, who thinks?
part three
Who Thinks?
preamble
The Virtual Subject
From the question “Who speaks?” we arrive at an answer Talmud criticism gave to the implied question “Who thinks?” Answering both questions with their respective notions of “The Author” as either “redactors” or “composer” of Talmudic discussions, scholars in Talmud criticism assumed a strong and rigid connection between thinking and a person who thinks. By the logic of that connection, a thinking person has to be the center, the source, the root, and the moving force of thinking, in which that person participates. That rigid connection allowed the scholars in Talmud criticism to prove the existence, albeit ultimately virtual, of the “redactors” or “composers” of Talmudic discussions. That proof of existence was based on a thought process the scholars discovered in the text, which they ascribed neither to a named character in this text (Halivni’s view) nor even to any anonymous character in this text (Friedman’s view), but invariably still to a character—to a persona, either individual or collective, construed as “redactors” or “composer.” Thus, from observing unattributed thought-forms in the Talmud, the scholars of Talmud 131
132 Who Thinks?
criticism automatically concluded to the existence of a character who thinks, to whom they attributed these through-forms. The rigid connection between thinking and the person who thinks led Talmud criticism to impasses, rooted, as argued, in missing the virtual in the dichotomy of the real versus the imaginary nature of “The Author” as the site of the thinking in the Talmud. If not to overcome, then to bypass these impasses, an extended framework of thinking about thinking in the Talmud is needed, which would reach through and beyond Talmud criticism. As a first step in building such a framework, the hitherto implicit question “who thinks in the Talmud” is to be addressed explicitly. The complex structure of the question “who thinks” calls for detailed analysis. Such an analysis would either lead to a more precise answer to the question of who thinks in the Talmud, or would make it possible to reconsider the question, thereby also leading to a better understanding of the thinking processes in the Talmud. My focus in the third part of the book is the core structural element of the question “who thinks”: the notion of the thinking subject and a construction of the virtual agency it implies. Extending a body of work in the history and archaeology of philosophy to the history of Talmud’s formation and interpretation, I argue that the notion of the thinking subject is grounded in medieval theological and philosophical agendas and their continuations in modernity and that it informs Talmud criticism. What is even more important, the notion of the thinking subject represents a historically specific version of the virtual, as an element of reality that is weaker than actual, but stronger than potential. Ramifications of that notion of the virtual in such constructs as “redactors” and “composer” of the Talmud, I will argue, both animate and limit the possibilities of Talmud criticism as a way to think about thinking in the Talmud. To think through these ramifications in philosophical terms, I discuss shortcomings of the notion of the thinking subject and of the model of the virtual it implies. This discussion helps explain the necessity by which Talmud criticism implicitly interpreted “The Author” of the Talmud as a virtual site of thinking rather than as a factual, imaginary, or historically real person who thinks, even if such an interpretation went against the explicit tasks and/or agendas of Talmud criticism. The complex structure of the question of who thinks in the Talmud calls for detailed analysis, because, if better understood through such an analysis, the question “who thinks in the Talmud” would either lead to a more precise
Preamble: The Virtual Subject 133
answer or would make it possible to reconsider the question, thereby also leading to better understanding the thinking processes in the Talmud. Chapter 7 introduces complexity and undoes the obviousness of the notion of the thinking subject in the history of philosophy and by extension in the history of Talmud as a discipline of thinking. Chapter 8 reintroduces Augustine as a resource for thinking about thinking without the thinking subject by highlighting the role of memory in Augustine, and engaging resources of Talmud criticism to negotiate an extension of Augustine’s model of thinking in remembering to apply to Talmudic discussions.
seven
Who Thinks in the Talmud?
If person-centered models of the authors as well as other—named and unnamed—characters in the Talmud cannot suffice to answer the question “Who thinks in the Talmud?” what can?1 To approach this question, I again simultaneously invoke the traditions of scholarship in the Talmud and in philosophy in a mutual hermeneutics of the Talmudic and philosophical traditions. Juxtaposed, these traditions provide an opportunity to illuminate each other’s conceptual foundations and in that way create a new approach to the question of the thinking processes in the Talmudic and philosophical traditions. From the point of view of Talmudic studies, this project entails using studies in pagan philosophy in a way similar to how studies in Christianity have been used to understand rabbinic texts better. From the point of view of studies in philosophy, the advantage of this project lies in expanding the textual basis of studying the thinking processes in late antiquity. Introducing the understudied practices of rabbinic thought as displayed in the texts of the period complicates the map of the strategies of connecting or disconnecting the notions of the subjectivity of the subject and thinking. 134
Who Thinks in the Talmud? 135
The methodological status of the project of a mutual hermeneutics can best be represented by examining the concept of anachronism. Although studies in the Talmud have always followed the rule of anachronism that forbids using later texts to explain earlier ones, another form of anachronism— conceptual anachronism—has often been almost explicitly and axiomatically accepted in different disciplines in the field of Jewish studies, for example, in the literary study of the Bible. However, researches in other disciplines, such as intellectual history or the history of concepts, are much more likely to question using later concepts to explain earlier texts. My position is that the choice is not whether to employ conceptual anachronism, but rather whether to employ it uncritically or critically. My methodological choice is critical conceptual anachronism. The “critical” component comes from the above context of mutual hermeneutics of the traditions and disciplines of thinking. The conjunction of the notion of “thinking” with that of a “subject” from the Middle Ages through modernity established a new philosophical notion, the “thinking subject,” but that conjunction also created a productive (mis) understanding of the thinking processes in the Talmud, a (mis)understanding that has endured within the post-Talmudic tradition. In this chapter, I focus specifically on that (mis)understanding. I once again analyze the twentieth- century theory that unreflectively constructs the notion of anonymous redactors in and of the Talmud, using the notion of the thinking subject as a model, this time as a point of departure for looking further at intellectual practices in late antiquity in both philosophy and the study of the Talmud. This study is, of course, a separate task of inquiry, yet that task would not be possible without preparing the ground for it by applying recent philosophical critiques of the thinking subject to critical analysis of the Talmud as a text, as a thinking process, and as a theoretical object. That viewpoint complicates both the philosophical understanding of the birth of the thinking subject and the hitherto dominant view of the Talmud as a product of anonymous redactors living in late antiquity. To that end, what follows undoes the erasure of the thought processes in the Talmud from the intellectual map of the West. That erasure began in the Middle Ages, with the union of Trinitarian theology with a philosophy of thinking, including intellectual and political agendas both responding to and resisting the newly created union between theology and philosophy in the Jewish, Christian, and Muslim worlds. The new union left no place for
136 Who Thinks?
the set of dynamic intellectual practices of thinking inscribed in the Talmudic text. As a result of this erasure, the protocols and practices of the Talmud were interrupted. What had been seen and used as an open text of rabbinic discussion became a closed archive, available for reopening only for the purposes of commentary, analysis, or imitation or even treated as permanently closed to serve, instead, as a source of laws progressively reformulated in a comparatively more static, more systematic, and more accessible style. Among other places, that systematic style flourished in rabbinic legal institutions, where an open-ended style of thinking became combined with the doctrinal and systematic nature of either practical legal or doctrinal conclusions. Maimonides’ Guide and Mishne Torah are examples of this stylistic change, this closure. So is Aquinas’s Summa in the Christian tradition. A preliminary, but decisive question to ask here is how a union of Trinitarian theology and Aristotelian philosophy could entail the erasure of the protocols of thinking that characterized the Talmudic text. The answer to that question lies in the notion of the thinking subject and in the (im)possibility of that notion. However, we will not be able to begin entertaining that answer, much less to undo the erasure that the notion of thinking subject both entails and explains, without first looking at the enigmatic birth of the thinking subject (and of the concept of being for human beings, in that sense) in the union of theology and philosophy from the Middle Ages through the modern period. The recent work of Alain de Libera will be my guide here.2 Only with that enigma of the birth of the thinking subject at the center of our attention will we be able to start understanding and criticizing the erasure of the thought processes in the Talmud for the purposes of undoing it.3
Apology for Undoing It is time to apologize—indeed, to do so in the classical philosophical sense, exemplified, among other instances, by Plato’s Apology of Socrates, but also in the sense in which a Talmudic discussion is an apology for the Mishnah, the third-century code of laws for rabbinical courts. For Socrates, apology was a way to fight for the power of his own positions against the accusations leveled against them, not for his own life; for the validity of his teaching, not for its universal acceptance; for justice in the society of which he was a member, not for winning his own case in court. In a Talmudic discussion of the
Who Thinks in the Talmud? 137
Mishnah, the stakes are no less critical: reaffirming the Mishnah’s otherwise merely mechanical memorization as a legally reliable source of authority, not making one’s interpretation of that Mishnah win over another; reconfirming the Mishnah’s power as a source for both legal and ethical decision making, not agreeing on what exactly the “true and only” interpretation of it is. The work of undoing requires a similar apology. Undoing is a multistep task. Here, it aims at the realistic goal of tracing the erasure of the thought processes in the Talmud, not at a utopian return of the erased but at a mature formulation of the problem, not a premature solution to it. In addition, and perhaps unlike the above ways of apology, the procedure of undoing implies not only questioning the birth of the thinking subject, but also asking about what constraints the notion of birth may exhibit. It questions not only and not primarily the notion of the thinking subject as a concept of a quite enigmatic birth, indeed a hybrid, a monster, a contradiction in adjecto, but also the notion of birth itself, which paradoxically combines two ideas: the idea of a historical emergence in reality of what has already been there in potentiality and the idea of coming into existence for the first time. Let me then undertake that apology. Undoing is a relatively new strategy. It became known to philosophers and critical theorists only with the development of phenomenology at the beginning of the twentieth century. It first emerged under different names and in different aspects: in the name of the “phenomenological reduction” of the natural belief in the existence of objects and subjects of human experience, which, as the mathematical metaphor has it, become “reduced” to the “what” and the “how” of what is given in experience, regardless of the subjects’ and objects’ alleged being (or nonbeing). Another name for “undoing” was also appropriated from the language of mathematics: the “bracketing” of the natural belief in the existence of the objects in experience, yet another “de-struction” (more widely known as “deconstruction”) of the natural belief in things seen, or as Joan Stambaugh renders it in English, the “de-structuring”4 of what would otherwise seem to naturally or historically exist. Offered as a way out of the then-current crisis in accounts of the positive scientific knowledge of what is, as opposed to what is not, the phenomenological method of undoing allowed philosophers to differentiate a rigorous description of what is given (in this case, without defining to whom) from the uncritical belief in the existence of what either is or is not there. However, the undoing undertaken by that method stopped at the natural belief in the
138 Who Thinks?
existence of objects and subjects of experience. Thinking was not entertained without the assumption that it is an act committed by an agent, and in that sense by a subject, that is, without there being a thinking agent inherently present in any process or procedure of thinking. Even the most radical claim of “thinking without a thinker” still depends on what it denies—the idea of a thinking subject. In short, everything could be de-structured or bracketed except for the agency inherently present in any process—indeed, any act—of thinking, or for that matter, in any case or instance of experience. By contrast, de Libera’s work offers a way to break philosophy’s dependence on the notion of the thinking subject, a dependence that otherwise endures even in its negation. He does not propose an alternative to phenomenology or any of its continuations, but rather undoes the notion of the thinking subject by asking a radically new question: How was that subject born? De Libera’s approach to undoing the notion of the thinking subject becomes critical for understanding the thinking process in the Talmud at the time before the notion of the thinking subject erased Talmudic thinking from the intellectual map of the West. Thus, a closer look at de Libera’s answer to the question of how the thinking subject was born becomes necessary.
Philosophical Archaeology and the Birth of the Self De Libera argues that the thinking subject should not be taken for granted, that is, treated as if it were either always already born or at least conceived. Instead, he teaches us to ask questions that involve things well before and despite the point of the “always already”: to interrogate the birth itself, rather than its alleged result. He does so via the project of a “philosophical archaeology” that both builds on and differs from the “archaeological” investigations of Michel Foucault. Foucault’s method of archaeology is a combination of historical attention to developments in texts over time and philosophical attention to concepts. Using this method, Foucault asks how, in historical changes in the practices of confession, the concept of the self was conceived as an obscure area in the human being calling for hermeneutical clarification by confessing one’s sins in public. Foucault called these practices “a technique of the self.” De Libera’s project has a finely tuned, but very important difference, in that Foucault, as de Libera shows, still assumed the existence of an always already existing thinking subject approaching itself
Who Thinks in the Talmud? 139
through various techniques, from the pagan technique of confession as a transparent and self-regulatory and thus private way to regulate mistakes to the monastic technique of “eximologesis,” or publicizing the self in front of the other in an effort to cleanse the self of sinful thoughts, however obscure and even unknown they might have been to anyone involved in the process.5 De Libera’s analysis challenges Foucault’s view of eximologesis by attacking the underlying notion of that theory: the notion of the thinking subject. Foucault’s account of confession in the Middle Ages assumes the existence of a thinking self that cleanses itself from any sin in its thoughts by making those thoughts public. It is the publication alone that makes one clean of any sin in one’s thoughts, whether or not the person who confesses or the recipient of confession is able to figure where exactly the sin sneaked into thinking. For example, for Foucault, if someone confesses to himself or herself in private: “I took the Lord’s name in vain fourteen hundred times,” this would not do. The sinner must make that statement in public, in the presence of a designated recipient of that confession. Even a simple account of all the thoughts a monk had during the day, if offered in public, that is, in the presence of a recipient, cleanses any sins he may have committed, even if those sins remain unknown. In contrast, in de Libera’s account of confession, thoughts and the self are not assumed to be intrinsically connected one with another in the first place: Thoughts can come from the outside, contaminating the self, and therefore can be purged back to where they came from through confession. For de Libera, however, as we will see, such a purging would mean purging all thoughts altogether, whether or not a sin is among them, so that in the result, the cleansed self would not have any private thoughts of its own, sinful or not. How are we to understand this conception of a self purged of all thoughts—a self without thoughts? To do so entails understanding the way in which the self can be seen as separate from thinking. De Libera isolates two initially disconnected networks of philosophical concepts that came together in the Middle Ages to create a single, but still heterogeneous, network, a symbiosis between the heritage of pagan philosophy and the theoretical needs of Trinitarian theology. The first network of concepts is that of subjectivity. It had to do with the philosophical notions of subject, support, substance, and person. The subject of an action, however, does not have to be the same as the substance that, by definition, underlies that action. For example, when a person dies, the language falsely suggests that person is not
140 Who Thinks?
only subject to dying, but is also the subject of dying. That person, however, is not necessarily the “agent” who supports or carries out the act, yet the act is ascribed to that person, as if he had committed it. When the victim of a hit-and-run dies, for example, it is the driver who is held responsible; but the death is ascribed to the victim. This takes us to the second network of concepts, the network of, as de Libera renders it, subjectity, or being subjected to an action, rather than being a subject of that action. That network has to do with the philosophical concepts of action, attribution (attributing the action to a subiectum), inherence (the presence of the subiectum in the action), and denomination (using the subiectum to give the action a name). An action may involve the support of a person, yet this does not make it necessary that the action be attributed to that person, or that the person is inherently involved in it, or that the action must carry the person’s name. For example, when the state executes a murderer, it is not the hangman to whom the action belongs. In this example, the hangman is the subject of the action in the sense of being the only support or substance without which that action might not have taken place, yet the action in question does not inherently require that specific person, much less have to be named after that subject or attributed to him/her/them. In short, the distinction between these two networks shows that an action and the person who carries it out do not have to be intrinsically interconnected. Even more important is the relationship of these two heterogeneous networks to understanding a third category, that of the thought process. How does the notion of the thought process relate to the notions of action and/ or to a person or subject in and to that action? De Libera insists that at least in the philosophy of Aristotle, thinking is considered a faculty (dunamis), as opposed to an action (energeia). To express the distinction in English, I term thinking seen as a faculty “the faculty of thought” or, more briefly, “thought.” In this approach, thought is either not an action, or, if it is action, it is very different in kind from the action of a subject in its interaction with an object or matter. In the latter interaction, both change. However, when the faculty of thought is used, it either does not change at all or changes in a completely different sense than the subject changes when it acts or is acted upon. De Libera concludes that for Aristotle, the disconnected networks of subjectivity and subjectity are in turn conceptually disconnected from either the faculty of thought or its use. Therefore, as he emphasizes, for Aristotle, the modern notion of the thinking subject would be a contradiction in terms.
Who Thinks in the Talmud? 141
In linking the subject to the faculty of thinking, the modern notion of the thinking subject conflates subjectivity and subjectity. Consider the expression “I think.” This seemingly simple statement implies the following series of claims:
1. I am the one who does the thinking (my thinking is an act, and I am the agent who does it); 2. I am central for my thinking: My thinking process cannot go on without me (I am inherent in my thinking); 3. I own my thinking (my thinking is attributable to me); 4. I yield my name to my thinking (my name serves to denominate my thinking).
The above claims of (1) active agency, (2) inherence, (3) attribution, and (4) denomination imply that thinking is an act or action in the sense of an active process. In addition, however, the expression “I think” also involves a rather different series of claims:
5. I am the vehicle or carrier of my thinking; without me my thinking has no supporter to bear it (support); 6. I am the substance that underlies my thinking (substance); 7. I am a person to whom the thinking occurs (person); 8. I am subject to my thinking: I follow the logic or obey the flow of my thoughts (subject).
The latter claims of (5) support, (6) substance, (7) person, and (8) subjection imply a notion of the subject in the passive sense of a substance (substratum) of thinking, or a person subject to his or her thought process, rather than in the active sense of an agent in the first four. In contrast, although the sense of the subject in the first four claims is active, in the last four, it is passive, yet regardless of that contrast, all eight claims above have one thing in common: They assume a connection between the notion of the subject (either active or as the basis of action) and that of thinking. Yet, conceiving of thinking as a faculty, rather than as an act, as Aristotle did, undoes the linkage between thinking and the subject—between thinking and both subjectivity and subjectity.6 It disconnects thinking both from
142 Who Thinks?
the subject as an agent in whom thoughts are inherent, to whom they can be attributed, and whose name they can bear and from the subject as the support, substance, or person who is then subject to those thoughts. As a faculty, thinking entails my ability to think, and to be able to think does not mean to have all my potential thoughts in me, as in a reservoir of possibilities. Instead, it refers to my ability to generate new thinking. This ability to think is very different from any action either potentially contained in it or that actually results from using it. As a faculty, not an action, thinking does not depend on a subject, passive or active. Thinking informs the acts of the subjects and forms the ways in which they move, but as a faculty, thinking does not move or act.
The Abstraction of the Thinking Subject As de Libera shows, prior to the rise of the Trinitarian theology in the Middle Ages, with its allegiance to the notion of persona as an inherently present agent and thus as a subject in and of any thinking, Aristotle as well as his Christian (indeed also Jewish) followers in late antiquity associated the thinking process with the figure of the sage, that is, with someone who has and uses the faculty, the capacity, the dunamis, to think in every act or in every energeia. The sage is neither a subject of or even an agent of thinking.7 The sage, or for that matter his or her persona, is radically extrinsic to any thinking that he or she enacts. The sage thus is neither inherently present in the process of thinking nor posited as an agent of this process, no matter whether passive or active, for example, as either a demiurge or as the matter or material of its work. Instead, for Aristotle—and for Augustine—thinking was seen as the realization of the sage’s ability to think, an ability that does not alter, change, or become exhausted in the process. Similarly, in Rabbi Akiba’s famous image, thinking, or in his case learning the Torah, is just this: a flame lighting from another flame. The new flame increases, but the old does not fade. The new flame does not alter the old one; nor does it use up any of it.8 We therefore need to appreciate what has been erased in both the Jewish and the Christian understanding of the nature of argumentation and thinking, in particular in the thinking processes in the Talmud. Recent scholarship has begun to use both pagan and Christian philosophical thought to understand Jewish texts and Jewish thought in late antiquity, and in what will follow, I will seek to reconnect the traditions or practices of Talmudic
Who Thinks in the Talmud? 143
and philosophical thinking of that period.9 However, here I am investigating thinking processes, not the characteristics of either rabbis or philosophers as alleged thinking subjects. The question “Who thinks in the Talmud?” was hardly either asked or implied before Maimonides, but once it was, the fundamental answer was always the same: a thinking subject. Maimonides, in the introduction to his Mishneh Torah answers that question with a name: Rav Ashe, from one of the latest generations of the amor’aim. Six centuries later, Mosheh Chayim Lutsatto in his Ways of Reason answers the same question not with a name, but with a title: “the composer of the Talmud.”10 For him, the composition embodies the composer and is a paradigm of the ideal of enlightened reason. From Maimonides to Lutsatto, from the twelfth century to the eighteenth, the implied assumption of a thinking subject remains in place. That assumption does not go away in subsequent developments. In the twentieth century, in Chanoch Albeck’s work,11 the question of the thinking subject becomes historicized in a notion of the “anonymous redactors of amoraic traditions,” whom Albeck conceives as immediate students of amor’aim, or Talmudic masters, in their respective generations. David W. Halivni disagrees, but only with the chronological and generational placement of the redactors. He explains the thinking self of the redactors using terms not unlike those used to analyze human sexuality: The thinking process of the redactors is subjected to hidden (but, of course, at any rate legitimate) drives of both converting divergence among earlier authorities into disagreements between them, and making sure that none of these authorities loses. Managing these contravening and at the same time closely connected tasks leads to “forced interpretations,” which in turn calls for the hermeneutical analysis that Halivni and his followers conduct. These drives inform the thinking processes of the anonymous redactors, who, in Halivni’s version, are the implied but ultimate thinking subjects of/in the Talmud. Unlike Albeck, Halivni claims that the redactors lived several generations (and centuries) later than the masters they redacted, and, of course, they were reducible neither to a single person nor to a set of anonymous students associated with individual masters. But despite all these important developments, the redactors still have remained what Rav Ashe was for Maimonides: the implied thinking subjects in and of the Talmud. However, if we take a closer look at the transition from Maimonides to Halivni, we will see a shift in emphasis in conceptions of the thinking subject
144
Who Thinks?
from the subject to the thinking itself—an abstraction of the subject into its thought. Maimonides answered the question of who the ultimate thinker in the Talmud is with the name Rav Ashe, characterizing his style of thinking as that of refutation and counterrefutation ( קשיאand )פירקה. As he conjectures, “And Ravina and Rav Ashe are the last of the [authoritative] Torah scholars [ חכמיםlit., “sages”] in the Talmud; it was Rav Ashe who wrote [חיבר, lit., “composed”] the Babylonian Talmud in the Land of Babylon, about a hundred years after Rabbi Yoh.anan wrote the Jerusalem Talmud.”12 If Rav Ashe—the last master of the Talmud, the last of the amor’aim, whose chronological time, if calculated according to the internal genealogical relationship of the sages in the Talmud, would be the fifth century C.E.—is the one who “composed the Babylonian Talmud,” what was the intellectual style of that composition? Maimonides addresses that question, if only in passing. The matter of both of the Talmuds is interpreting the languages of the Mishnah; narrowing down the legal matters to which those languages apply, as well as discussing legal inventions made in the rabbinical courts in the period from our holy rabbi (the composer of the Mishnah) up to the moment of composing of the Talmud. . . . I saw fit to write what can be determined from all of these works in regard to what is forbidden and permitted, and unclean and clean, and the other rules of the Torah: Everything in clear language and terse style, so that the whole Oral Law would become thoroughly known to all; without bringing refutations and counterrefutations, but rather clear, convincing, and correct statements, in accordance with the legal rules drawn from all of these works and commentaries that have appeared from the time of Our Holy Teacher to the present.13
Maimonides replaces the Talmudic composition’s intellectual style of “bringing refutations and counterrefutations” with “clear language and [a] terse style.” Just as this new “clarity of language” absolutely requires having a single composer, in this case, Maimonides himself, the “old style,” that of the Talmudic composition, would, of course, have a similarly single composer— either individual or collective. In Maimonides’ version, that composer is Rav Ashe. Yet the picture becomes less straightforward if we account for Lutsatto’s no less philosophical view of the Talmud. In Maimonides’ view, Maimonides and Rav Ashe are different in their style—Maimonides’ philosophically rigorous rational exposition of what is permitted or forbidden versus Rav Ashe’s
fup-dolgopolski-07.indd 144
10/9/12 12:04 PM
Who Thinks in the Talmud? 145
rhetorical and nearly sophistic style of refutations and counterrefutations— while in the view of Lutsatto, the philosophy of enlightened reason is said to be similar in intellectual style to the Talmud. One way of explaining that striking contrast is to recognize that for Maimonides, the emphasis is on the subject: The thinking subject is divine and is only partially represented through the thought processes of individuals. (In fact, Maimonides’ own thought processes are the best representation possible.) In contrast, for Lutsatto, the emphasis is on the thinking subject’s thinking, so that for him, divinity and humanity become of the same importance, as far as the thinking process is concerned. Thus, for Maimonides, the style of bringing refutations and counterrefutations signifies philosophical disarray and an unnecessarily chaotic plurality of opinions that can be replaced with an exposition from just one philosophically sound point of view. In contrast, for Lutsatto, the style of refutations and counterrefutations is not at all chaotic. Rather, it is a vehicle for sharpening one’s rational position, so it does not really matter if a refutation is articulated by somebody else or by the speaker himself. Let all roads lead to the same place, that is the truth in the statements achieved by arranging them one against the other. The law of refutation that a person brings against a statement of his peer and the law of refutation that a supporter of the statement brings against himself are exactly the same. . . . For we should look not at the people, but rather at their statements themselves.14
Lutsatto’s language indicates both of the points on which he differs from Maimonides: his emphasis on the thought or “statement,” rather than on the person—the subject or “supporter” who thinks—and his appreciation of refutation and counterrefutation as the style proper for enlightened reason. Lutsatto’s emphasis on the thinking of the thinking subject sets the stage for the next step in the ongoing abstraction of the notion of the thinking subject. Two centuries later, in the work of Albeck and Halivni, the subject part of the thinking subject fades even more, almost to the point of disappearance. Yet it does not disappear. In Halivni’s and Albeck’s work, the personal traits of the redactors of the Talmud almost vanish as the thinking subject becomes anonymous—both in the sense of being unnamed and in the sense of acting because of a hidden drive to construct disagreements between earlier
146 Who Thinks?
authorities and protect the authorities from losing their positions. That drive also permits an unnamed subject, whom Halivni hermeneutically discerns, attends to, and describes. Nevertheless, however close it may be to fading, disappearing, or vanishing, the unnamed subject must remain in the work of these later scholars, just as it was and had to be there in the work of Maimonides. A careful and unhurried reading of Halivni’s hermeneutical analysis will help explain the seemingly natural necessity of the subject to remain in accounts of the thinking subject in the texts of the Talmud and help undo the connection between thinking processes in the Talmud and the medieval- modern idea of a thinking subject. To undo that connection is to ask the question of the thinking processes in the Talmud independently of either endorsing or rejecting the assumption of a thinking subject. To begin this reading, I must return to the example of a Talmudic discussion in terms common to Maimonides, Lutsatto, and Halivni, that is, in the style of “bringing refutations and counterrefutations.”
Refutation and Counterrefutation: The Rhetoric of the Talmud I see no point in retelling even a small portion of this Talmudic discussion. Such an account will always be poor, compared with the original text. Instead, my goal is to use a magnifying glass both to enlarge and to bring into focus what underlies the text, as well as to bring to light the idealized setting that the literary form of the text implies and creates, but does not describe. The task of describing that setting is what any reader, including myself, will have to conjecture by restoring a context or a setting to the phrases of the text: the rhetorical context of the Talmud. One way to restore the implied scene is to see it opening as a tann’a, a master of recitation, stands in front of a gathering of teachers (amor’aim) and students (talmidei h.akhamim) in a rabbinic academy in Babylonia. The tann’a (recitant) declaims a portion of text. The caveat in his role is that he is supposed to have a good mechanical memory and the will to exercise it, but no other elements of authority. Thus, in a setting remote from the Mishnaic sages in both time and place, the tann’a quotes a fragment from the Mishnah before an assembly of amor’aim in the house of study. A historian would say that the tann’a recites an early third-century instruction codex for rabbinical courts called the Mishnah. A linguist would
Who Thinks in the Talmud? 147
specify that each segment of that body of instruction is also called a Mishnah and that, as we have noted in the preceding chapter, unlike modern languages, the phrases are structured as a theme and a rheme, an affirmation about the theme, rather than in a subject-verb-object structure, as in a sentence. A theorist in rhetoric might note that the Mishnah prescribes how a court should proceed under the following type of litigation circumstances, rather than describing or representing a state of affairs.15 Two [litigants] hold to one and the same piece of a cloak (.t alith). The first claims, “I have found it!” and the second claims, “I have found it!” The first claims, “It is all mine!” and the second claims, “It is all mine!”
Under such circumstances, the Mishnah instructs the court: The first [litigant] should swear that s/he at least owns a half of it, and the second should swear the s/he owns at least a half of it. And they should split it.
In a slightly different situation, the instruction is different, too: The first claims, “It is all mine!” and the second claims, “A half of it is mine!” The one claiming “all is mine” should swear s/he owns at least three-fourths of it, and the one claiming “a half of it is mine” should swear that s/he owns at least one-fourth of it.
When the tann’a is finished reciting what he memorized, the people in the courtroom or in the house of study interrogate its textual content, acting in what Maimonides called the style of “refutations and counterrefutations.” Here, a heuristic attack (or refutation of the validity) of the Mishnah’s instruction comes in the rhetorical form of a soliloquy. Linguistically, in a soliloquy, an argument is presented from a first person conversing with himself or herself. Rhetorically, the power of that form is in inviting the audience to observe and to identify with the soliloquist who goes through a dialogue with himself or herself. In this particular case, the soliloquist (with whom, again, his audience rhetorically identifies) raises doubts about the accuracy of the instruction’s language as recited. In linguistic terms, all the questions
148 Who Thinks?
that the soliloquist raises are identical in their pragmatic meaning: They all are asked from the rhetorical standpoint of the audience’s identification with the speaker. As a result, the silent audience is persuaded that it is as if it were asking these questions itself. What is the role of a personal name, then? Although it is always possible, introducing (or supplying) a personal name for the soliloquist is extrinsic to that genre of persuasion. For maximal rhetorical effect, the name of a soliloquist is effective only if the person is already famous or at least known to his or her audience. Otherwise, the name is not important. A name can also be rhetorically effective when the soliloquist is trying to emphasize his or her authority or uniqueness. Otherwise, the soliloquy does not require the name of its speaker. Therefore, if imitated, it can work effectively without the presence of the original speaker. Ultimately, it neither requires nor implies any original speaker—neither now, at the time of its performance, nor in the past. Yet it surely represents an intensive process of thinking that has more than one virtual (and fundamentally nameless) person involved. This is not the soliloquist, but rather those who constructed both the soliloquy and the soliloquist as part of it while not necessarily being identical with the soliloquist they construct. Just as the name of the soliloquist is not intrinsically important for the genre, so, too, the name(s) of those who constructed the soliloquy are not as important as their thinking leading to that construction, so that they naturally remain unnamed. In this specific case, as in many other instances of Talmudic soliloquy, the speech starts and proceeds directly, with no name introduced. A soliloquist asks himself or herself a “Why would I . . .” question. [If I were the tann’a] why would I present the claims [in the Mishnah] so nonlinearly? “The first claims, “I have found it!” and the second claims, “I have found it!” The first claims, “It is all mine!” and the second claims, “It is all mine!” The claims should instead be presented one by one, like this: “One claims, “I have found it, and it is all mine!” and the other claims, “I have found it, and it is all mine!” And yet the instruction does not present claims one by one!
After some further pondering over that question, the soliloquist feels dissatisfied with his own search for a direct answer. As a result, in the eyes of the learned audience, the Mishnah is now under attack. The style of refutation and counterrefutation requires a defense.
Who Thinks in the Talmud? 149
The defense comes from the audience, which cites the names of the amor’aim. The audience recalls a tradition attributed to the names of either Rav Papp’a or Rav Simi, the members of the older and therefore more authoritative rank of post-Mishnaic authorities, amor’aim. Their dictum about the Mishnah is offered as a counterrefutation and answer to the soliloquy. Said Rav Pappa, or according to some Rav Simi the son of Ashe, the instruction in the Mishnah said so [i.e., nonlinearly] because its first part deals with the claims of having found [the cloak], and the second with the claims of having it bought from a merchant.
Regardless of the current context of its emergence, in this quote, the earlier masters, or amor’aim, found the instruction to refer to two different categories of litigation, one concerning the ownership of a thing found and the other the ownership of a thing bought and sold. In the current context, the teaching of the amor’aim successfully addresses the concern that the soliloquist had about the order of instruction. It suggests the instruction did not present claims one by one because it in fact refers to two different litigations, one between two people claiming they found the cloak, the other between the two people claiming they bought it. That allows the soliloquist to conclude that a seemingly inexplicable nonlinearity has a reason: The Mishnah refers to two different kinds of litigation. And so he and the audience conclude: “Hence it [i.e., the nonlinear order of instruction] was necessary.” At this point, the first round of a refutation and counterrefutation is over, but the soliloquist’s conversation with himself (or rhetorically with his audience) is not. A new round begins. The counterrefutation becomes a target of a new refutation, an attack that, as the style requires, will be followed by a counterattack. But by this point I have explained enough to prepare the stage for analyzing Halivni’s reading of the Talmud.16
who forced the redaction?
Halivni explains the soliloquist’s attack on the Mishnah as an artificial course of action forced by a certain line of thought. He simultaneously and perhaps too quickly attributes that line of thought to a person, whom he will find
150 Who Thinks?
facing a difficult choice: to create a rather artificial question for a ready- made, but otherwise useless, answer (as I will explain, that of Rav Pappa or Rav Simi the son of Ashe) or to give up that answer altogether, thereby undermining the authority that the answer must represent. He further posits persons, subjects, agents—in a word, the anonymous redactors—as the cause of that line of thought. In doing so, he fundamentally (and axiomatically) considers that line of thought as an act of thinking, a difference that is greater than it may seem, for, as I have noted, thought was seen in Greek philosophy as a faculty (dunamis) that only extrinsically connects to an agent who possesses it.17 In contrast, thinking was seen as an action (energeia) that absolutely requires an agent to act it. Thus, we deal here with three different paradigms. To take “Ego cogito ergo sum” as an emblem of a modern paradigm, it axiomatically considers “ego cogito” to be an act and the human “ego” to be an agent or subject both behind and in that act. Medieval rationalists, both in philosophy and in the study of the Talmud, would differ from that model on only one point—in attributing thinking to G-d as the ultimate agent or subject of any thinking, rather than to a human subject. They would consequently limit the intellectual role of the human to reading divine thoughts written in a human mind, now construed as the privileged site of human thinking. But the fundamentals remain the same: Thought is an act of thinking and therefore requires an agent or subject (either divine or human) in and behind it. However, there is a third paradigm, and in that paradigm, rather than being intrinsically an action of thinking, thought is a faculty or ability that different individuals may possess in different degrees. Halivni sees the soliloquy as resulting from both an action and thinking, rather than from thought. Thought as a faculty (dunamis) becomes exclusively reduced to the action of thinking in the sense of energeia, as what thus must be attributed to a subiectum, to a hupokeimenon, or to some underlying matter that consequently becomes personified in his theory as the anonymous redactors of the Talmud. His automatic choice of looking at thought in terms of thinking, rather than in terms of a faculty, or dunamis, that does not act, but rather makes agents act, conceals the key to the mystery of Halivni’s concept of the anonymous redactors who produce forced interpretations of the teachings of earlier masters. Halivni sees in the actions of the anonymous redactors that he posits (1) the activity of subjects in an active sense as agents in whom (2) thoughts are inherent, (3) to whom they can be attributed, and (4) whose name (the
Who Thinks in the Talmud? 151
name is “anonymous”) they can bear; he also sees the redactors as subjects in the passive sense of the (5) support, (6) substance, or (7) persons who are therefore (8) subject to those thoughts. These eight heterogeneous elements, which I have already introduced above in two distinct groups (elements of action: acting agency, inherence, attribution, and denomination; and elements of being acted upon: support, substance, person, and subjection) work simultaneously in what seems to be a single act that Halivni finds the redactors to be forced to make in their thinking. He explains their move as a response to a need to supply a new refutation to a counterrefutation that was unsuccessful in its original context. Supplying a new refutation is, of course, an action. But this is not enough. He therefore posits an instance (in de Libera’s technical terms, a matter involved in an action—a subiectum) that underlies that action. Furthermore he attributes that action to that instance, assumes that the underlying instance was always inherently there in the action, and uses that instance to give the action its name—a set of moves not foreign to either Greek or Latin philosophy of the subiectum, or matter, that in Greek or Latin philosophy underlies any action. Yet he also does something that a medieval theologian likely would do, but that a pagan Aristotelian philosopher would not. He connects the thinking behind the action to the subiectum of that action, making that subiectum into a person who thinks, thus into a thinking subject, whom he will then quite logically designate as a thinking subject without a name. In his final and most effective move for us, as moderns, he attributes both the thinking and the action to that personified subiectum—the anonymous redactors of the Talmud, who, remarkably, must have thought of or “conjectured” the forced interpretations in exactly the same way as Halivni did. The origin of that personification lies in his projecting the moves he discovered when confronting the text onto someone whom he assumes to be behind it. There’s more. Halivni emphasizes both the difference and the strong connection between the soliloquy as a refutation of the Mishnah and the tradition of the amor’aim used as a counterrefutation for it.18 His first step is to claim that the soliloquist’s attack on the Mishnah is a strange decision to make. The Mishnah’s nonlinear style could have been explained as a mere convenience of memorization, yet that answer would deny the very necessity of being linear. He thus opens his source-critical analysis of the soliloquy by presenting an alternative approach to the Mishnah. In that approach, the Mishnah relates
152 Who Thinks?
to only one case of litigation—that of two litigants claiming they found the cloak. That helps eliminate the seemingly nonlinear character of the instruction by conceiving it as linear through and through: What seemed nonlinear in fact proceeded in a syntactic order from general to particular. “I found it, and it is all mine!” is just one claim with two parts, one more general and one more particular. The Mishnah presents the general parts of the claim of each litigant first, followed by a series of their possible particular parts. Halivni finds support for that approach not only in a modern Eastern European commentator on the Talmud, Rabbi Manasseh of Ilya,19 but also in the syntactic connections that he supplies to the language of the Mishnah. With the addition of syntactic connectors (presented here in {curly brackets}), the Mishnah now accords with Rabbi Manasseh’s linear reading of it. Two [litigants] hold to one and the same piece of clothing. The first claims, “I have found it!” and the second claims, “I have found it!” {the decision is that if} [t]he first claims, “It is all mine!” and the second claims, “It is all mine!” {then} the first [litigant] should swear that s/he at least owns a half of it, and the second should swear that s/he owns at least half of it. And they should split it. {Yet, if} [t] he the first claims, “It is all mine!” and the second claims, “A half of it is mine!” {then} [t]he one claiming “all is mine” should swear s/he owns at least three- fourths of it, and the one claiming “a half of it is mine” should swear that s/he owns at least one-fourth of it.
Read in that way, the Mishnah still can allow implications for a different litigation case, one that involves buying and selling, rather than finding. Yet those implications would no longer be an intrinsic or direct part of the case to which the Mishnah is referring, Halivni emphasizes. What was a reference to deal with becomes an implication that might be, but does not have to be taken into account. In the new reading, the Mishnah no longer refers to litigation over buying and selling but can yield implications about it. In Halivni’s and Rabbi Manasseh’s linear reading, unlike what the soliloquist makes the amor’aim say, the Mishnah does not directly instruct on litigation about buying and selling, but refers only to a dispute about things found. Yet the alliance between Rabbi Manasseh and Halivni is not complete. Rabbi Manasseh simply dismisses the soliloquy as making no sense; Halivni takes the lack of sense as a symptom indicating that there were hidden forces at work.
Who Thinks in the Talmud? 153
Rabbi Manasseh did not distinguish between stamm’a de gemarr’a (the anonymous soliloquy, in my reading) and the amor’aim (whom the soliloquist has engaged in that soliloquy). Neither was he concerned with the fact that the amoraic tradition marked as “Said Rav Pappa, or according to some, Rav Simi the son of Ashe” means that the opening of the Mishnaic instruction refers to finding a thing, but the continuation of it to buying and selling it, because and only because of the power [i.e., in the context] of the question that the soliloquist proposed at the start of his speech, “How can I explain the reason for which the instruction presents claims nonlinearly: The first claims, ‘I have found it!’ and the second claims, ‘I have found it!’ The first claims, ‘It is all mine!’ and the second claims, ‘It is all mine!’ It should instead be presenting claims one by one.”20
The Lithuanian rabbi Manasseh was sure that the Mishnah refers to only one general case of claims—things found—followed by several possible variations on that claim, as far as the extent of claimed ownership is concerned. Thus, a litigant can claim generally, “I have found it,” and supplement it by either, “It is all mine,” or “Half of it is mine,” depending on his or her feelings of justice in dividing the found thing, which might be different from case to case. With this linear reading, the Lithuanian commentator leaves no room for the question of nonlinearity in the Mishnah. For him, the Mishnah is linear through and through, and both the amor’aim and the soliloquist have simply overlooked that. At this point, though, Halivni’s argument diverges from Rabbi Manasseh’s. He claims that the amor’aim did not overlook the possibility of a linear reading but simply had to reject it silently, due to persistent polemical pressures that the amor’aim felt, that the soliloquist knew, but did not mention, and that Halivni has helped discover. These pressures had to do with the issue of acquiring a thing for someone else without acquiring a part of it for yourself, an issue on which the amor’aim as a group irresolvably disagreed elsewhere, as witnessed in the Talmud. Acknowledging that unresolved disagreement in the camp of the amor’aim (no matter whether it was a retroactively constructed disagreement or not, although I will return to that question in a moment), the soliloquist, as a faithful adept of all the amor’aim, had to support those who might see the Mishnah as relevant to the disagreement at hand. The soliloquist therefore calls on
154 Who Thinks?
the stage those of the amor’aim who would read the Mishnah at hand in the light of that disagreement. Therefore, the soliloquist rejects a possible linear reading of the Mishnah even before he starts speaking, for how can one speak while undercutting the source of his or her own authority? The soliloquist therefore prepares his audience to read the Mishnah as strongly as possible (which in this case means nonlinearly), that is, as referring not only to finding things in the street but also to buying them in a store. In sum, according to Halivni, the anonymous soliloquy was constructed by the redactors in response to an ongoing disagreement among amor’aim elsewhere in the Talmud (I will have to say more about that broader disagreement later in the text) in which one of the parties, represented by the names of Rav Pappa or Rav Simi the son of Ashe, was about to lose. To prevent that, the redactors were forced to create an artificial soliloquy that engages this Mishnah in that disagreement. That soliloquist has read and presented the Mishnah in such a way as to keep that disagreement alive. To do so, the soliloquist strategically creates a perspective on the Mishnah to make it a foil for the argument of at least one camp of the amor’aim in the disagreement and does so with the sole purpose of introducing the Mishnah in the field of the ongoing disagreement, not for putting an end to that disagreement. To that end, the soliloquy is created in order to dismiss a more obvious, linear reading of the Mishnah as not relevant to the amor’aim’s disagreement and in order to create an artificial attack on the Mishnah, resolving it by introducing a reading of it that fits the perspective of one camp in the disagreement—the weakest camp. All these considerations and strategies of the anonymous redactors become clear to Halivni as a result of his conceptual a priori decision to draw a distinction between stamm’a de gemarr’a (the “anonymous tradition,” a term that he reads as an anthropomorphic personification: stamm’aim, the anonymous redactors) and the amor’aim. Therefore, for Halivni, Rabbi Manasseh’s dismissal of the original attack on the Mishnah as merely erroneous is a result of the rabbi’s (and the tradition’s) failure to draw that a priori distinction, assuming instead that in the diegesis (that is, in the represented content) of the Talmud, what the soliloquist says is a direct and naive continuation of the thought process of the amor’aim. Indeed Rabbi Manasseh sees neither the legitimacy of the problem that the soliloquist presents against the Mishnah nor any discontinuity between the soliloquist and the amor’aim, whom the soliloquist diegetically cites. Of course, Rabbi Manasseh might have agreed that on the level of diegesis, the soliloquist and the amor’aim
Who Thinks in the Talmud? 155
whom he cites are obviously not the same, but he sees no difference whatsoever in their positions or considerations. The soliloquist would for him be at best a later adept of the amor’aim who expresses their view of the Mishnah at hand, not someone who created or invented these views. Halivni’s position is radically different. He goes beyond diegetic relationships between the amor’aim and the soliloquist and discovers the behind-the- scene forces—the allegiance of the soliloquist (or rather of the redactors who must have invented him and his speech) to the amor’aim as a larger group of masters in disagreement, an allegiance that justifies reading the Mishnah in the light of and for the sake of not letting that disagreement drop.21 That decision is clearly not a part of the diegesis. Rather, it comes from Maimonides’ view of the intellectual style of the Talmud as refutations followed by counterrefutations. The soliloquist silently decides to account for the presumably retroactively constructed disagreement between the amor’aim, letting that disagreement inform the way in which the Talmudic argument is developed, or rather, he creates room for such a development. For Halivni, this silent decision is not only a fundamental factor working behind the scenes (and behind the diegesis) of the Talmudic discussion; it also radically informs the direction that the Talmudic discussion takes. However, although the decision is clearly not a part of the diegesis, Halivni in fact introduces the strategic decision the alleged redactors made behind the scenes into the diegesis of the Talmudic discussion. This is the only way for him to differentiate his position from that of Rabbi Manasseh and to defend his a priori conception of redactors. As a result, the old traditional term of stamm’a de gemarr’a, of an “anonymous tradition,” is used to represent a soliloquist who is remarkably transformed from a rhetorical position into an instance of independent thinking, indeed, into the redactor of the Talmud, “The Author,” as we have been calling whoever or whatever occupies this position, who acts both behind the scenes and on stage. (I will return to this issue in greater depth in Chapter 8.) In fact, that position even becomes an independent, yet hidden authority, as opposed to representing a direct teaching inherited from a group of the amor’aim, contrary to what both Rabbi Manasseh and the previous rabbinic tradition would hold. The difference between Halivni’s and Rabbi Manasseh’s positions deserves an even closer look in terms of the question “Who thinks in the Talmud?” Again, according to Rabbi Manasseh, had the stamm’aim seen the Mishnah the way he did, there would be no question of nonlinearity. Therefore, if asked to
156 Who Thinks?
explain why that discussion in the Talmud takes place, Rabbi Manasseh would be forced to say that it happened either by mistake or because his way of parsing the Mishnah syntactically was overlooked by both the amor’aim and the anonymous soliloquist alike. As powerful and insightful as he is, the modern Lithuanian rabbinic commentator would thus surely see no difference between the amor’aim and the soliloquist, for all of them would be at fault. Not so is Halivni’s much more nuanced sense of the case. He draws a distinction between the amor’aim and the soliloquist, who becomes both a speaker on the stage and the redactor behind the scenes. The soliloquist/ redactor knew about the polemics and disagreements in the camp of the amor’aim, and in order to preserve their strength, chose to read the Mishnah as a foil for that disagreement, that is, from the perspective of those of the amor’aim who would see the Mishnah as relevant for the polemics concerning acquiring a thing for another without acquiring it for yourself and thus from the perspective of “Rav Pappa, or according to some, of Rav Simi the son of Ashe,” that is, of the amor’aim for whom this Mishnah would be helpful in their polemics, which they are about to lose elsewhere.22
What Is Called “Thinking”? What is conceptually at stake here is no less than understanding the role of the subject in thinking. As we have just seen, acting behind the scenes and in both continuity and rupture with the amor’aim, the redactors are said to have decided to create a rhetorical character, the soliloquist, and to pair the answer that the Mishnah gives with a question that the soliloquist offers. Halivni addresses the element of rupture with his notion of forced interpretation. In fact, that notion is the core of Halivni’s method of analyzing the Talmud. The notion of stamm’aim as empirical-historical subjects to whom to attribute the interpretation is only a superstructure built on top of it. To make the leap from tracing forced interpretations to the concept of an entity behind them, however, requires subscribing to the problematic and theologically charged concept of the thinking subject, either in whole or in part. Halivni thus follows the tradition of “attributivism,” in de Libera’s terms,23 which requires that any act of thinking be attributed to an agent in and behind it and that the agent be identified as both the substance and the subject of that thinking.
Who Thinks in the Talmud? 157
However, the desire to preserve the disagreement between the amor’aim that Halivni (or Rabbi Manasseh of Ilya) discovers does not necessarily mean that a decision to preserve it, conscious or unconscious, was made. It could simply mean supplying an attack for a defense that is not employed anywhere else. Yet for that, no deciding agency is required, only a careful working up of a question to which this decision might be an answer. Just as it would be “absurd to speak of a builder as being altered when he is using his skill in building the house,” as Aristotle put it, it would be also “wrong to speak of a wise man as being ‘altered’ when he uses his wisdom [i.e., learns things].” Hence, for Aristotle, knowing or learning is either not to be considered a passive process of being acted upon or “altered,” or else the “alteration” that occurs when a wise man learns or understands a new thing doesn’t mean what it does in the case of teaching or in other cases of being acted upon. A sense of “alteration” proper for learning as opposed to teaching is “the development of an existent quality from potentiality in the direction of fixity or nature.”24 In more general terms, thinking and learning have nothing to do with the subject, according to Aristotle and in the Greek tradition of thinking at large.25 As I have just shown, Halivni’s a priori notion of anonymous redactors, like the figure of “The Author,” requires them to stand in both rupture and continuity with the amor’aim—in rupture behind the scenes of the Talmudic refutations and counterrefutations and in continuity on the stage. The long tradition that Halivni follows, that of supplementing the thinking process with a subject, first registered at the time of Thomas Aquinas and Maimonides, thus follows Maimonides and Lutsatto in two ways: in seeing the perpetuation of refutation and counterrefutation as the default and almost de rigueur style in the Talmud and in subscribing to the tradition that associates thought with a subject, thus erasing the classical concept of thought under the name of a thinking subject.26 How to understand that erasure of thought by the notion of the thinking subject? Answering that question requires looking into late ancient traditions of associating, and in particular dissociating, the notions of subject and thinking. As the next chapter will start showing, crucial for an answer will be an understanding of the complexity of the relationship between thinking and memory.
eight
The Hand of Augustine: Thought, Memory, and Performative Existence in the Talmud
The thinking process in the Talmud has nothing to do with the modern notion of the thinking subject. Instead, as this chapter will show, it is a process of the collective rational reinvention of the memory of tradition produced through the conversations that the characters conduct in the text. How are we to place the thinking and remembering practices embodied in these conversations? How do these practices connect to other practices of thinking and remembering in late antiquity, in the disciplines of rhetoric and philosophy? As we will see, contemporary text-critical views of the Talmud once again can help us explore these connections, and once again, I address these questions through a critical engagement of the intellectual work of David W. Halivni and Shamma Friedman because their work contributes to a comparative understanding of theories and practices of memory in late antiquity. I particularly look at how these questions pertain to current assumptions about what kind of text the Talmud is. This analysis foregrounds the broader 158
The Hand of Augustine 159
theoretical consideration of the issues of thinking, memory, and temporality framed in premodern and postmodern perspectives. The task of this book is to discern the Talmudic text as it comes to us from late antiquity as an intellectual project, a form and a body of thought, rather than as a text in the sense of an object of philological analysis in which a text is reduced to a medium for a certain content. As already briefly discussed, and as I will show in greater detail here, contemporary scholarship of the Babylonian Talmud inherits the Maimonidean perception of the Talmud as predominantly a series of refutations and defenses designed on the way to formulating the Oral Law. This philosophy-driven view of the Talmud not only lends the Oral Law to Maimonides’ translation of it into a universal and rational code, but also presupposes the Talmud as a body of text and thought in which universally accessible knowledge of the Law (halakhah) takes precedence over the wisdom intimated by poetical/individual narratives of the sages (haggadah). In one single move, this approach marginalizes the narrative texts of the Talmud and ascribes the Talmud as a whole to a rationally thinking and therefore intrinsically disembodied human agency working through the series of refutations and defenses. This agency then becomes either personalized in the names of Ravina and Rav Ashe or, in a more recent view, becomes depersonalized as the Talmud’s anonymous redactors or composers. In this approach, refutation becomes an ultimately dispensable instrument of knowledge at the service of justice, rather than an act of wisdom in the face of what can’t be apodictically known. With the philosophical (mis) translation and the consequent codifications of the Talmud in hand, if one needs to access the Law, either memorization of the Mishnah or understanding it by means of complex Talmudic discussion becomes unnecessary. In this chapter, I undertake to reveal the construction of the precritical object that contemporary scholars of the Talmud have critically analyzed and to conceive that precritical object in a new way, once again by combining the theoretical resources of these scholars with resources developed by studying the complex dynamics of relationships between the philosophical and rhetorical traditions of thinking in late antiquity and later periods. When understood as an intellectual project, the Talmud has much to tell us about the ways in which thought and memory can be conceived—and performed—without recourse to the concept of subjectivity in the Cartesian tradition. Discerning the text of the Talmud as it comes to us from late antiquity involves both critical and positive approaches. The critical approach engages
160 Who Thinks?
with a philological predicament in the academic or “scientific” study of the Talmud. It prepares the ground for exploring the nature of the Talmud as an intellectual project and revisits philological tools and concepts already available for undertaking that exploration. The positive approach uses these “scientific” tools and concepts to broaden the horizon of the exploration by returning the Talmud to the context of other intellectual projects in the complex relationships between the rhetorical and philosophical traditions of thinking. I depart from J. N. Epstein’s program, which proposes to establish the text of the Talmud first before even attempting to establish the logos of it, which, for him, would be the second and therefore deferred task. This was a powerful—perhaps too powerful—principle of text-critical Talmudic scholarship. Epstein formulated that program at the very onset of the academic or “scientific” study of the Talmud in the beginning of the last century. His principle still actively works in contemporary philological, historical, and literary studies of the Talmud, both in Israel and in the United States. As productive as it has proved itself to be, this principle implies a reductive understanding of “logos” as the content of the text and of the text as the form of this content. This rigid dichotomy prevents us from critically addressing the form of the Talmud as an intellectual project and instead takes this form for granted. In this approach, the form in which this text is first given for understanding by a critical scholar is assumed to be self-evident and self- explanatory, even before the critical analysis of it begins. This reductive implication is both necessary and correct, because otherwise, scholars could not consider the philological critique of the text to be the first step and the “logos” the second. One simple reason for that reductive understanding of logos is that a critic must have always already conceived the object of the critique—even before any critical analysis of this object—in this case, of the text—can begin. Before the text can be criticized, it must make some (even preliminary) sense and thus adhere to a certain standard of making sense. This must happen. However, the process of making sense can and should be open to critical analysis, such as I have attempted to undertake in the preceding chapter, just as the text itself is. There, I showed how the speakers in the Talmud (re)construct truth using the rhetorical strategy of apologetic refutation and defense. Here, I explore further the consequences of that analysis. Such a use of refutation and defense also implies that memory or remembering is much more than accessing a depository of data, but
The Hand of Augustine 161
instead serves as a form of thinking—a form of thinking that is performative and, as we will see, “virtual.” Given the text-critical scholars’ preference for beginning by critically establishing the text, rather than by critically analyzing their taken-for-granted understanding of that text, the scholars of the Talmud have had to ask the question of genesis at the very start of their critical work. They have framed that question as “Who composed the text of the Talmud?” To answer this question, the scholars of Talmud criticism have predominantly used historical and literary frameworks of analysis. I have discussed these in the chapters above. In this chapter, I both revisit and expand these frameworks by asking this question of genesis from a literary-philosophical perspective, which I have already introduced in analyzing sense-making processes in the text of the Talmud. As I showed in Chapter 4, the Talmudic process of refuting and defending involves a technique of memory. This technique combines rhetorical and philosophical elements. The rhetorical element of memory comes from late ancient rhetoricians, who used mnemonic devices, for example, the “memory palace,” to control the flow of their recollections, which helped them remember both words and things. The philosophical element of memory comes from the Platonic philosophers who—following Plato’s metaphor of the cave—treated remembering as the foundation of genuine knowing and forbade any use of mnemonics. Theoretically, these mnemonic and epistemological approaches to memory mutually exclude each other, but they can combine in the practice of thinking. To understand the unique character of this combination in the Talmud, I suggest disentangling the Talmudic process of refutations and defenses from the embodiment of that process on the stage of a rabbinic academy. Instead, I locate refuting and defending in the memory of the composer of a Talmudic discussion—call him “The Author”—who, according to all contemporary scholarly accounts, constructs or composes the discussion onstage. If described in terms of rhetoric, “The Author” choreographs the dance of refutations and defenses in which the characters analyze the words of a given tradition, for example, of the Mishnah. If described in terms of philosophy, “The Author” recalls/discovers the correct content of the Mishnah by logically sequencing the process of refuting and defending it. I conclude here by asking about the nature of “The Author”—or of any such entity—that is, the nature of a site of both remembering and knowing from which the action “onstage” in the text is seen and controlled.
162 Who Thinks?
I begin by reading a text of Augustine in which, as Frances Yates argues, rhetorical (mnemonic) ideas of memory and philosophical (Platonic) ideas of cognition meet each other and combine to create a unique approach to mnemonics as a technique suitable for the Platonic cognition of what is true. Augustine writes: And I come to the fields and spacious palaces of my memory, where are the treasures of innumerable images, brought into it from things of all sorts perceived by the senses . . . . When I enter there I require what I will to be brought forth, and something instantly comes; others must be longer sought after, which are fetched, as it were, out of some inner receptacle; others rush out in troops, and while one thing is desired and required, they start forth, as who should say, “Is it perchance I?” These I drive away with the hand of my heart, from the face of my remembrance, until what I wish for be unveiled, and appear in sight, out of its secret place. Other things come up readily, in unbroken order, as they are called for; those in front making way for the following; and as they make way, they are hidden from sight, ready to come when I will. All which takes place when I repeat a thing by heart.1
These words from the tenth book of Augustine’s Confessions describe how, in his search for Godly things, the writer enters the mnemonic palace of memory to do the philosophical work of remembering. He enters as a skilled master of the rhetorical art of memory, ready to manipulate the remembrances that come his way from the walls and corners of the palace, to sort them out using not his hand—manus—but the “hand of his heart.” The phrase “the hand of my heart” is truly obscure and in need of explication. Even if we accept that “the hand of his heart,” in context, means also “the hand of his mind,” or that by way of an even more general contextual consideration that “heart” and “mind” were synonymous in late ancient faculties psychology, these generic explanations would not be accurate enough to explicate the phrase as it is used here. It is particularly necessary to explicate the phrase more precisely both because of the strong dichotomy drawn since the Enlightenment between “the heart” (the passions) and “the mind” (reason) and because it will help draw an analogy between what this phrase means and what I am arguing here about the Talmud. In the context of the quote in the Confessions, any simple redefinition of the heart as the mind, if based either on a general synonymy
The Hand of Augustine 163
between these terms in late ancient faculties psychology or even on, again, a general contrast with a medieval concept of the “eye of the mind” are not precise enough. To explicate this phrase with greater precision, I propose to use the analytical terminology of performatives. A performative is a device or condition that grounds and orients processes—in this case, thinking and remembering, even before they are articulated in terms of what is being thought or remembered, or else in terms of who thinks and remembers. For example, as discussed in the Introduction, a blind but framed gaze always embodied in a photographic portrait and underlying the possibility of that portrait is a performative. Variably attributable to a photographer, or to a viewer, or even to a camera, that gaze in fact belongs to none, and therefore cannot be defined by its subject or bearer either. This gaze, however, is both always presumed and never thematically included in the photograph; instead it enables the process of photography and arises in it only if critically sought for. This gaze may be said to be a virtual and organless condition or ground of a photograph, rather than an actual element, object, or subject in photography. Never actually on the photograph, it has nevertheless predefined the dynamics of relationship between the objects in the photograph, the photographer, the camera, and the viewer. Such a blind but framed gaze, as bodiless and virtual (more than merely potential, but less than actual) as it is, is an example of a performative existential. If the “hand of the heart” can be explained as such a performative existential, then there is no need to ascribe Augustine’s processes of remembering to any subject or agent either behind them or even in them. I propose to explain the “hand of the heart” as such a performative—that is, to illuminate the obscure meaning of the “hand of the heart” not by what it is or to what/whom it can be attributed, but rather by what it does in the text and how it relates to the elements of reality described or implied in that text. As Augustine says, the hand of his heart sorts out remembrances until the right one comes to his memory. As such, this hand is, of course, not a really existing or even imaginable organ, but rather a virtual, even if still bodily agent. Notably, there is nothing more remote from the Cartesian model of thinking centered on and stemming from a subject who allegedly thinks than this model of the “hand of the heart” performing its job in the palace, if not to say in the theater, of memory. It also seems clear from the context that Augustine has in mind something other than “mind” or “eye of the mind” when
164 Who Thinks?
he comes to speak of “the hand of my heart” here: He first discriminates between mere “mass” and what “animates” it—“life,” which even horses and mules have—and then between mere “life” and the “power whereby I imbue with sense my flesh,” that is, not just “life,” but perception: “I, the one mind,” make sense out of the sensations of the five senses. But even animals have this power, he says. So “I will pass beyond this power of mine also, for this also have the horse and mule, for they also perceive through the body.” This is all part of a structural gradatio of “ascent” toward knowledge of God through attributes that Augustine, as a soul in a body, possesses, and the next step beyond mere “mind” is memory: So ascending beyond mere perception, “I will pass then beyond this power of my nature also, rising by degrees unto Him, who made me.” But in the end, that’s not going to get him all the way to God, either. He might never get out of Plato’s cave. As I understand it, the irruption of “heart” here in this gradatio of the characteristics that define being human is consistent with Augustine’s overall use of the term as an orientation or yearning of the soul toward God. The hand of his heart bats away what’s inessential, “until what I wish for be unveiled, and appear in sight, out of its secret place,” a claim that’s stated in the language of theological revelation, not rhetorical mnemonics or philosophical epistemology. That orientation toward knowledge (or the orientation toward that kind of knowledge) is neither merely rational nor merely affective, but both affective and rational—part of the ontotheological problematic of identity/difference involved in having been made in God’s “image.” The whole extended passage thus repeats, on the level of memory, the epistemological predicament of the soul, rather than that of the mind. After a long examination of what’s available in memory, Augustine concludes: “Where then did I find Thee, that I might learn Thee? For in my memory Thou wert not, before I learned Thee. Where then did I find Thee that I might learn Thee, but in Thee above me? Place there is none; we go backward and forward, and there is no place.” That pitches him back into his distance from God, into his status as “flesh” and then into all the temptations it’s heir to. From there, it’s on to humility, confession, Christ as mediator, and so on. What it means, however, is that if we still need a name for the visitor in the palace of memory, then it is clearly not a mind, much less a thinking subject, but rather a soul entering the world as the palace of memory to attempt to remember and in this and only this way to think of God.
The Hand of Augustine 165
With all his broader theological agendas in play, in this passage Augustine acts as a skilled master of the rhetorical technique of mnemonics, of which the palace of memory is a classic component. What is even more important, he also proceeds in the manner of a Platonic philosopher who uses dialectical tools to ascend from the world of appearances and thereby recall in memory the world of ideal forms—ideas. He thus combines two mutually exclusive approaches to memory: as an object of manipulation and as a precondition of the human encounter with the divine. His metaphor of the “hand of my heart” is very precise. It reflects his double allegiance to the rhetorical manipulation of memory (in mnemonics) and to philosophical reverence for it (in cognition). Talmudic practices also combine the elements of mnemonics with those of the philosophical uses of memory. First, in the text of the Talmud, a reciter, tann’a, uses one stand-alone part of the rhetorical art of mnemonics, “memory for words,” in order to deliver the teachings of the sages of the Mishnah. As Yates explains it, even rhetoricians (not to mention philosophers) generally perceive the use of “memory for words” as less trustworthy than the other part of mnemonics: memory for things, including ideas. Yet, just as memory for words has it place in mnemonics, memory for the words of the Mishnah and other traditions has its place in the Talmud. In the Talmud, there are characters who specialize in remembering, word by word, the words of the Mishnah and other materials, such as Midrashic (exegetic) compilations. These characters are somewhat ambiguously labeled tann’aim, a term that refers to both the memorizers and to the great authorities of the past whose words the memorizers remember and transmit from one generation to another—by memory. Just as philosophers and some rhetoricians distrust “memory for words,” the other characters in the Talmud do not trust these mechanical memorizers and reciters of the words of tradition either, not until what is remembered can be verified in some way, often by arguments for and against, or more precisely, first “against” then “for” what is recalled by being recited. These other characters instead practice the mnemonic art of the memory of “things” (ideas, matters, legal cases), in order to accept these words as the words of the tann’a. In the process of verification, these characters seek to recall and sort out particular things that correlate with the words of the reciter. This Talmudic process of verification is unique. In other rhetorical schools, these two mnemonic techniques are strictly separated, and in Platonic philosophical schools, which used reasoning as the vehicle of memory/
166 Who Thinks?
knowing, mnemonics were not permitted at all. But the Talmud combines the memory for words with the memory for things, using the Platonic understanding of knowledge as memory attained through reasoning. The Talmud is thus an almost impossible combination of the rhetorical device called the “palace of memory,” employed for recalling both words and things,2 and the philosophical devices of knowing as remembering through argumentation. Thus, while Augustinian thought, like the Talmud, combines the otherwise mutually exclusive practices of memory—rhetorical and philosophical, the Talmud goes beyond Augustine by bridging the two parts of mnemonics, using the philosophical practice of memory to mediate between words and things. By recalling and sorting out ideas to which the remembered text might refer, the characters in the Talmud ultimately verify the results of the memory for words with the data they mine; using their memory for things, they connect to these words. That rough description is only an approximation, however, and a more nuanced reading is needed to understand the Talmudic, philosophical, and rhetorical practices of memory in the light of each of the others. In order to construct a lens through which to explore the Talmudic practice of memory more precisely, I engage contemporary historical, literary, and philological theories of the Talmud. That will involve an interdisciplinary approach. Hitherto, the dominant approach in the critical study of the Talmud has been at best only multidisciplinary. Critical scholarship has explored different ways of conceiving the Babylonian Talmud anew. It has embraced increasingly broader cultural and intellectual contexts—from Iranian culture, to the Greek, Latin, and Syriac philosophical, rhetorical, and literary schools in late antiquity and the Middle Ages, to Christian and Islamic cultural, religious, and intellectual milieus of the respective periods. In analyzing those contexts and connections, historical, literary, and cultural approaches have been dominated by the notion of “influences,” including the euphemistic replacement of the notion of influence with what is called “mapping out” or comparative typologies. Even if comparative typology no longer claims historical causation (“influences”), it still has retained the assumption of impartial historical observation sub specie aeternitatis, as if the researcher were rooted in no intellectual tradition among those he or she critically studied. What follows explores a different route to an understanding of the literary- intellectual construction of the Babylonian Talmud. I follow the strategy of
The Hand of Augustine
167
minor inquiries conducted on the basis of multiple perspectives rooted in the multiple traditions involved, of which none can claim any kind of domination: the mutual hermeneutics of the Talmudic, rhetorical, and philosophical texts and traditions in the context of the competing and conflating rhetorical and philosophical understandings of memory in late antiquity, interrogated in the light (and thereby sometimes against the grain) of their medieval and contemporary understandings.
“Without refutations and defenses”: A Traditional View of the Talmud Moving hermeneutically from what is more clear to what is more obscure, what is clear is that both contemporary and medieval interpretations of the Babylonian Talmud treat it as predominantly a series of heuristic refutations of given authoritative/normative texts, followed by apologetic defenses against these refutations (קשיא ופירוק, Maimonides’ terms).3 In the preface that serves as raison d’être for his Mishneh Torah, and specifically for its task of codifying the Oral Torah, Maimonides describes the Talmud in a way that coins what has become the traditional medieval and modern view of the Talmud: Ravina and Rav Ashe stand at the end of the generations of the masters of the Talmud; and Rav Ashe is the one who composed the Babylonian Talmud in the land of Shinar, approximately a hundred years after Rabbi Yoh.anan composed the Jerusalem Talmud. The matter of the two Talmuds is interpreting4 the words of the Mishnah and clarifying its depths, as well as inventions of every rabbinical court from the days of Our Holy Rabbi (who composed the Mishnah)5 until the end of the composition of the Talmud. From the two Talmuds, as well as from [the parallel versions of the Mishnah in] Tosefta, from the [midrashim] of Sifra and Sifri [respectively on Leviticus and Deuteronomy] and from additions [to them]—from all of them, it becomes clarified what is forbidden and what is permitted, pure and impure, obligatory and exempt, kosher or unusable, as the authorities were copying each other from the mouth of Moses on Mount Sinai. [In the Talmud,] it also becomes clarified which of the above rules were enacted by the sages, or prophets in each generation, for the sake of building “protective walls around the Torah,” as they have explicitly heard from Moses “Therefore shall ye keep My charge”6 (Leviticus 18:30), which suggests “you should protect my commandments.” Similarly [in the Talmud] it becomes clarified which of the
fup-dolgopolski-08.indd 167
10/10/12 3:53 PM
168 Who Thinks? above are customs and legal adjustments enacted or customarily performed in a certain generation, as a rabbinical court of that generation saw fit; because it is forbidden to deviate from them [too], as it says, “thou shalt not turn aside from the sentence which they shall declare unto thee, to the right hand, nor to the left” (Deuteronomy 17:11). And [it] also [becomes clarified regarding] the criminal laws and rulings, which were not received from Moses, but the Great Assembly of that generation decided on them using [analogical] rules by which Scripture can be expounded, and which were consequently approved by the elders, who passed down a tradition that the ruling is such and such. Rav Ashe composed all of it in the Talmud, [covering the period] from the days of Moses until his own days.
This description talks about the “depths” of the Mishnah that the Talmud “clarifies.” What exactly are those “depths”? And what does “clarification” entail? “Depth” implies a “surface” of the text beneath which the “clarification” must reach. In the description, the task of clarification is dual. It intrinsically enfolds two heterogeneous components: a timeless idea and the time-defined placement of it in a certain rank of authority. This clarification is thus analytical and genealogical at the same time and in the same respect: It both clarifies timeless ideas and places them in the timelike genealogy of authorities, ascribing specific timeless ideas (for example, laws or rulings implied, but not stated explicitly in the texts) to Moses, or to the prophets, or to the sages, and so on. To clarify thus means to isolate an idea and ascribe it to an authority in a way in which the idea is not possible without the authority. Indeed, without this genealogical rooting, the “depth” of the text has no authority to support it, and without that depth, the authorities devolve to empty placeholders on the “surface” of the immediate or explicit meaning of the texts, without any eidetic content in them beyond that surface or beyond the text’s immediate textual meaning. This renders clarification an exercise in apologetics: The clarified ideas come from the interpreter’s “here and now,” yet rely on genealogical authority from the past, and belong to that authority. What this means, however, is that clarification is not a linear process of retrieving an idea from the past to the “here and now” of the interpreter. Instead, clarification is an intellectual work performed by an interpreter who moves from the present (understanding an idea) to the past (placing it in the rank of authorities). Plumbing the “depths” of the Mishnah thus involves neither a timeless logical analysis of the text nor a historical reconstruction of ideas in it, because just as “depth” implies “surface,” “clarification” implies an
The Hand of Augustine 169
initial “obscurity” or “opaqueness” of that textual “surface” both in terms of the ideas expressed in it and in terms of the authority to which this ideas must be ascribed. Eliminating that implied opaqueness or obscurity is the task of the interpreter, who both describes ideas and ascribes them to an authority of a certain rank. Or rather, I should have said that the interpreter “describes ideas by ascribing them” to an authority other than the interpreter. The interpreter thus reconnects timeless ideas that he finds in the text to a ranked authority. Needless to say, such an interpretive procedure leaves the question of the historical genesis of these ideas outside of the realm of “clarification.”7 This approach to the “clarification” of the “depths of the Mishnah” informs Maimonides’ description of the Talmud and consequently what Talmud criticism addresses as the precritical view of the Talmud. In this precritical view, the Talmud is as an apologetic project that is neither an empirical-historical reconstruction of the history of Jewish law nor a logical- theoretical contemplation of legal ideas, but rather precisely what the intellectual idiom of “clarification” reflects—reading texts to produce ideas and to rank them by ascribing them to different authorities in the past. Maimonides’ description of the Talmud also defines another, indeed central, element of the precritical view, which is an already discussed philosophical hostility to the rhetorical element of refuting and defending in the Talmud. Maimonides finds this protocol of refuting and defending in the Talmud “unbearable” and proposes to dismiss that process altogether. To this effect he writes: For this reason, I, Moshe son of Rav Maimon the Sephardi, found that the current situation is unbearable; and so, relying on the help of the Rock blessed be He, I intently studied all these books, for I saw fit to write what can be sorted out from all of these works in regard to what is forbidden and permitted, and unclean and clean, and the other rules of the Torah: Everything in clear language and terse style, so that the whole Oral Law would become thoroughly known to all; without refutations and defenses or diverging arguments, but rather clear, convincing, and correct statements, in accordance with the legal rules drawn from all of these works and commentaries that have appeared from the time of Our Holy Teacher to the present.8
As I have already began showing in Chapter 7, Maimonides’ project of Mishneh Torah both negates the literary form of the Talmud (“refutations, defenses,
170 Who Thinks?
diverging arguments”) and translates its legalistic content into a new, philosophically rational form of “clear language and terse style.”9 What the Talmud “clarified” rhetorically, which for him means irrationally, Maimonides comes to “sort out” philosophically or “rationally.” His task is thus to “sort out” what was “clarified” in the Talmud about what is forbidden and permitted, and so on. It goes without saying that in his new project, Maimonides proposes to replace the complex and, at least for some, “obscure” rhetorical analysis in the Talmud by “sorting out” the laws in a more “clear” outline of the Oral Law in the style of philosophical rationalism, that is, as he has it, “without refutations and defenses,” but rather by formulating the law in the rational and universally accessible form of a logically or rationally graspable codex. By dismissing the Talmudic form of refutation-defense in favor of the philosophical form of systematic reasoning, Maimonides left his faithful heirs with only two options: to reject the form of refutation-defense or to prove its rational-philosophical validity. That choice meant an elision. It elided looking at the “surface” of the Talmudic text, its literary form of refutation and defense, in connection with the larger programmatic element of Talmudic discussion. That element is memory—memory for words and memory for things, including the ideas behind these words, as ancient manuals of rhetoric have it.
Criticizing the Precritical: Current Constructions of the Precritical Talmud The elision of the “surface” of the Talmudic text, of refutation and defense as a literary form, has led the heirs of Maimonides, in their efforts to plumb the “depths” of the text, to treat as illusory the traditional view of the Talmud as a recording of heuristic refutations of the Mishnah, followed by apologetic defenses of it in the form of synchronic conversations between the amor’aim and their students, interspersed with homiletic teachings of either the amor’aim or the sages of the Mishnah. As Wilhelm Bacher has it, the scholars have understood that the questions that “clever students” offer in these conversations were formed during approximately three centuries after the time of the amor’aim, to whom the questions are literarily addressed. Thus the amor’aim only seem to respond to questions, which in fact were formulated centuries later. As David Halivni has it, the dicta of the amor’aim were modified and reframed (“redacted”) by
The Hand of Augustine 171
a hidden agency—anonymous redactors who made sure the amor’aim have sustainable answers to the “clever” questions or that these answers withstand the series of refutations and defenses without losing the authoritative role they are expected to play. For Halivni, this process of redaction must have happened after the three centuries to which Bacher refers. As Shamma Friedman critically notices, Bacher’s (and by implication also Halivni’s) critical picture is only a projection of the illusion of synchrony, that is, of conversations occurring together in one era, into the dimension of diachrony, that is, into a different time in the past, three centuries later. Halivni’s enrichment and complication of that picture by introducing anonymous redactors of the amor’aim’s dicta still remains in the main chronological dimension. The latter, however, was only a diachronic projection of the literary illusion of the synchronic discourse of the amor’aim with their “clever students,” now mediated by the no less clever anonymous redactors of the amor’aim’s sayings. Friedman radically disengages from the illusions of either synchrony or diachrony by introducing a new, completely heterogeneous dimension of time: the time of textual transformations. In that time dimension, the illusion of coherent synchronic conversations is created by the process of the migration of texts between the three main literary forms in the Talmud: the dicta of the amor’aim, the apocrypha of the sages of the Mishnah, and the refutations and defenses in the form of the speeches of their students, which Friedman terms the “anonymous talmud” (stam talmud). Friedman sees that migration as a result of the intelligent design of a “master of sugya” (baal hasugya), a designer who orchestrates (or at the very least symbolically instantiates) the migrations of texts between the forms in order to create the illusion not only of synchronic conversations but also of coherent and plot- driven exchanges of refutations and defenses between the “clever students” and the amor’aim. The result of these textual transformations is the illusion that the Talmud is a record of conversations happening in the “here and now” of a rabbinic academy in Babylon. However, as Bacher, Halivni, and Friedman all invariably insist, that illusion of synchronic conversations in the rabbinic academy persists. It persists in precisely the same way that the illusion that the sun “rises” and “sets” persists. In the rotation of the earth on its axis in its revolution around the sun, the sun of course does not move, so the “rising,” and “setting” of the sun, in this taxonomy, is “illusory,” but that illusion is also persistent. There is thus a double determination of how a day passes. And just as there is a double
172 Who Thinks?
determination in the persistence of the illusion that the sun “rises” and “sets” and in the fact that the earth rotates on its axis, there is a double determination involved in the putative activity of the designer, the baal hasugya—and not only in that of the designer but also in that of critical researchers of the Talmud. The designing activity of the baal hasugya is a paradox: Making the audience believe the illusion that she or he creates, the baal hasugya has to both implement the process of design and forget about that process, just as a painter must both pay close attention to technicalities involved in creating the illusion she or he creates but at the same time, as a viewer, believe in it. Critical researchers engage in the same double determination in their account of the textual transformations that the baal hasugya undertakes and thus face the same paradox. They, too, must pay close attention to the technicalities involved in creating the illusion of synchronous refuting and defending, but at the same time believe in the illusion. These scholars thus analyze, but do not cancel that illusion, whether by projecting illusory synchrony into historical diachrony (Bacher), or by further complicating that diachrony by the invisible anonymous agency of redactors residing in it (Halivni), or by introducing a completely heterogeneous, indeed parallel line of time, the time of the migration of texts between forms, which, orchestrated by the designer of refutations and defenses, the baal hasugya, creates the illusion of synchronic conversations (Friedman). Yet again, in all these critical approaches, the illusion of synchronous refuting and defending remains. Just as, for an astronomer, the illusion of sunset persists despite all her knowledge, so too the illusion of synchronic conversation in the Talmud remains in force despite all the knowledge about the Talmud’s composition a text-critical researcher of the Talmud develops. There is yet another reason for the persistence of that illusion. It serves as a stepping-stone that the critical research must constantly overcome: The illusion remains there as the “old” leading to the “new.” Without the “old” (the illusion of synchronic reading of the conversations) the “new” (the critical analysis of the genesis of these conversations) is pointless: The negation (the critical view of the Talmud) cannot exist without the negated (the precritical view). Thus, the illusion of synchronic conversation in the Talmud persists (a) despite and (b) due to constant denying it as an illusion. Notably, none of the above-mentioned scholars directly ask, much less answer, the question why that illusion persists, even when explained as illusion. Bacher’s criticism only transfers the illusion into the diachronic
The Hand of Augustine 173
dimension. Halivni’s only further complicates it with the anonymous agency of the “redactors.” However, Friedman’s theory of the designer of a sugya is associated with the “real” time of textual migrations between forms, as opposed to the illusory time of the resulting composition. Compared with the other theories, it therefore gets much closer to the question of why the illusion of synchronic conversations persists. Ironically, as already shown, Friedman both comes close to an answer and misses it due to the methodological or conceptual heritage of the medieval Talmudists embedded in his approach, which has to do with the notions of thinking subject and memory. That said, in thinking about the question why the illusion persists, Friedman’s theory is the most promising starting point and thus requires further consideration. For Friedman, texts migrated between the literary forms of the text in the time of the Talmud’s design, which he associates with the work of the baal hasugya, the master/designer of the sugya, rather than in the time of the resulting illusion of the conversations in the Babylonian rabbinic school, whether synchronic or diachronic. The time of these designed textual migrations is for him the real time of the genesis of the literary composition, as opposed to the illusory literary time—either synchronic or diachronic, and either created (Bacher) or discerned (Halivni)—in that composition. The real time of the genesis of the literary form of the Talmud, which is the time accessible to both the researcher and the designer, is as completely heterogeneous with the time of the persistent illusion of the conversations in which amor’aim synchronically respond to the questions of their students as the astronomical time of the earth’s rotation is to the perceived time of the rising or setting sun. Friedman, however, does not look at—and given his subscription to the medieval understanding of the Talmud as primarily a series of refutations and defenses, cannot look at—how the migration of texts between literary forms perpetuates the illusory synchronic exchanges of refutations and defenses between the “clever students” and the amor’aim. He considers the illusion to be persistent, but does not ask how the time of the textual migrations (between the forms of dicta, apocrypha, and anonymous dialectics) and the agency of the baal hasugya help to perpetuate it. One reason for this has to do with Friedman’s focus on the technicalities of creation in his uncompromising removal of the baal hasugya from the resulting discussions. For Friedman, the baal hasugya, or the designer of Talmudic discussions, does not necessarily appear as a character among other characters in these discussions. But could the designer be more intimately
174 Who Thinks?
connected with the resulting design accomplished by his or her technical work? Could the baal hasygya’s role in creating the resulting illusion (and particularly in perpetuating the persistence of that illusion) be understood in terms of connection with, rather than in terms of separation from, the literary form of the composition of the Talmud? Answering these questions requires reevaluating the nature of the persistent illusion itself: Refuting and defending in the Talmud need not entail synchronic (or diachronic) conversations in the form of dicta, apocrypha, and the “anonymous Talmud.” Instead, these three forms can be seen as parts of another and much larger literary form, of which the designer/composer of a sugya would also be a part in the same way in which “The Author” is part of a novel. Such a larger literary form might be envisioned as a memoir, an exercise in memory, a work of memory designed, orchestrated, and enjoyed not in the persistent, but illusory “here and now” of the rabbinic academy of the amor’aim, but rather in the “real” (but no less literary) “here and now” occupied by the baal hasugya, that is, by “The Author,” as I developed the nature of that liminal figure in Chapter 4, a figure whose trace appears in the text, but who is not of it. If we approach the overall literary form of the Talmud as a memoir in which the role and the status of “The Author” is central, then once again, the question “Who thinks in the Talmud?” in fact becomes the question “Who remembers in the Talmud?” And thus, once again, thought and memory appear as something other than the activities of the post-Cartesian subject. Understood not only as a designer of the text, but also as a practitioner of the art of memory,10 the baal hasugya is no more the site of such a subjectivity than “The Author” is such a site in The Adventures of Tom Sawyer. Yet a site that figure still is. In the Talmud, that figure is the site of the unavoidable, but also nonobjectifiable, ground of any experience: the ground, to which all experience is attributed.11 That ground is a bodily ground, but not in the sense in which “body” is commonly understood. In our example of the double determination of the day in terms of the rotation of the earth and the rising and setting of the sun, what makes that doubleness possible is the bodily positioning of the observer on the earth, even when reflecting on earth’s rotation and even when actually viewing it from space. This doubling occurs because, as Husserl argues,12 even if the body of the observer can travel in space, the earth as the ground of observation (as opposed to the earth as a geometrical-astronomical object)
The Hand of Augustine 175
also travels with the observer. What is observed from space as the earth rotates is still “sunrise” and “sunset” on earth. Needless to say, in this phenomenological account, “the body of the observer” does not mean a geometrical, physical, or medical object, any more than “the earth” is a geometrical-astronomical object here. Instead, “the body” represents the perceptual ground of experience, whether it be the sunset or, for that matter, any other persistent illusion, which remains persistent precisely because of the persistence of the ground of its experience.13 Seen in this light, the Talmud as memoir is simultaneously constructed by, centered on, and grounded in the “bodily” positioning of the baal hasugya—or, for that matter, of the researcher who discovers that position without occupying it— as the necessary bodily ground of any experience. Seen in this way, the literary form of the Talmud no longer requires that there have been face-to-face interactions of refuters and defenders on the stage of the rabbinic academy at a certain chronological time. Instead, that form only follows the sequence of refutations and defenses, no matter what their chronological genesis might have happened to be. To put it another way, specific refutations and defenses have no intrinsic time of genesis. A defense might be constructed before the refutation against which it defends, even if in the rhetorical sequence in the Talmud defenses come after the refutations. This is similar to a situation in which we have an answer before we find the right question for it.14 That means that synchronic, or diachronic, or any other order of time is extraneous to the refutations and defenses. And just as refuting and defending do not intrinsically have to represent a sequence of actions in time, the bare being-there or bare act of “memory for things” accomplished through refuting and defending is not intrinsically located in any chronological time, either, because remembering is done by refuting and defending,15 a rhetorical sequence that does not have an intrinsic time sequence attached to it. What I have described as the persistence of the illusion of synchronic conversations in the Talmud thus has nothing to do with synchrony or diachrony. It results instead from the bare being-there of the “memory for things” (which includes “thoughts”), “embodied” by the organless body of the figure of the baal hasugya in Friedman’s account and by the figure of “The Author” in mine. The bare being-there of “the memory for things,” liminally in the text, but not of it, is exercised through the active process of refuting and defending constructed on the basis of the normatively given results
176 Who Thinks?
of “memory for words.” In the Talmud, the bare being-there of memory acts in and through the process of remembering at the site occupied by “The Author,” that is, by the baal hasugya. The bare being-there of memory thus does not require the existence of any historical human being of flesh and blood. It does not have to exist in chronological time, either. Instead, it is a literary function of the text of the Talmud.16 The bare being-there of memory (also construable as the bare act/affect of memory)17 gets lost in the various incarnations that the scholars of the Talmud have afforded to it, from the figure of Rav Ashe in medieval post- Talmudic literature, to stamm’aim in Halivni, to the account of the baal hasugya in Friedman. However, since the bare being-there of memory is a literary and theoretical object, not an empirically accessible “entity,”18 contemporary philosophical thought actually is better equipped to deal with that entity with greater precision. Contemporary philosophers know modern versions of that bare being-there, such as Deleuze’s “body without organs,” as did late ancient philosophers such as Plotinus, with what Pierre Hadot calls his “simplicity of vision (regard ).”19 If such conceptualizations point, as they do, to a virtual, rather than actual, body, the question is how exactly to conceive the virtual body involved in Talmudic practice of memory? The organless body that occupies the liminal and ambiguous site of “the Author” in the Talmud thus functions much like “the hand of my heart” in Augustine’s account of memory. In that account, importunate memories “start forth,” and the hand of his heart waves unwanted ones away. Here, “The Author” both responds to the appearance of interpretations and dismisses those that are unwanted. These interpretations come from the fields of memory, rather than from any ego-centered process of speaking, thinking, or even composing/designing. They are, therefore, not generated by the thought process of any particular person. To put it another way, the organless body that occupies the liminal and ambiguous site of “the Author” in the Talmud is neither “real,” in the sense of a historically existing entity, nor merely “literary,” a textual fiction, but virtual. In the light of the preceding analysis “virtual” means both a site of transformation from the old Talmud to the new Talmud, and a relatively stable place occupied first by the composer or designer of the sugya, and then by its readers. A persistent illusion of either synchronic or diachronic conversations on the stage of a rabbinic academy persists precisely because that site of transformation (migration of texts between textual forms) coincides with
The Hand of Augustine 177
the site of stasis (the master of the sugya). As a site of both transformation and stasis, it should, however, be reduced to neither. Rather, and precisely because of that irreducibility to its particular populations, this site is virtual. As such a virtual site, the position of the master of the sugya is better approximated by Augustine’s “hand of his heart” in the palace of memory. A comparable place of both thinking and remembering, this site is not dominated by or centered on any particular “ego,” which might come to populate it, be it the ego of baal ha sugya, or of the audience, or, for an example of a less-egocentric model in Augustine’s work, the model of thinking as no less and no more than walking into his palace of memory. Both exceeding and enabling its particular, often mutually exclusive, populations, this site is not merely a site of thinking, in particular, of designing, as conceived in Friedman, and not merely a site of remembering and knowing, as conceived in Augustine. Instead, if I can still use the visual metaphor of a “gaze” to think about what happens completely outside of the discipline of seeing, but instead in the discipline of memory and thinking, that site can be said to be the “gaze” of memory. Similar to the photographic gaze, it precedes, grounds, and does not coincide with any of its particular populations: the photographer, the camera, the viewer, passing and/or posturing objects (and in particular subjects). No population can claim exclusivity in a photographic gaze. The photographer looks at the object from the point of view of the viewer, and vice versa, and the posing subjects do so in view of the photographer, and they all follow the laws of the posture as well as the more general discipline of vision in order to decide what does and what does not constitute a good frame. By that analogy, no population can claim exclusive possession of the “gaze” of memory, which as I will further argue, orients thinking in both transformation and designing Talmudic texts. Reducible to no particular population, and grounding all possible populations, as heteronomous as they might be, the “gaze” of memory is therefore neither merely potential nor fully “real” or “actual,” but, instead, virtual. On the way from anonymous redactors as Cartesian disembodied thinking subjects who exist by virtue of their thinking alone, to designers of Talmudic conversations who transform “old” textual materials into the “new,” current, form of the Talmud and who are in turn only literary functions of the result of the transformation, we arrive at seeing the Talmudic compositions as a display of the work of memory. That trajectory leads to the next
178 Who Thinks?
hypostasy of the who-question: from the question Who speaks? through the question Who thinks? to the question Who remembers? Unlike thinking, memory is always embodied. Yet, the body that engages memory and/or is engaged in memory is not always an actual, empirically graspable body that a historian could describe. Rather this body may be virtual. That means it has actual effects in reality and is therefore no less real, even if again not actual. This body has power to engage (in) memory even if, and in the case of Talmudic memory, precisely because it cannot exist as a Cartesian disembodied thinking subject. Thinking displayed in the Talmud is done in the ways of memory and by virtue of memory, for which thinking and even knowing are only means. Asked in the scope of memory, rather than in the narrower contexts of either speaking or thinking subjects, the new form of the who-question now both allows and calls for analyzing the notion of the virtual remembering body in a broader context of philosophical, rhetorical, and other relevant intellectual traditions, both in late antiquity and in their consequent history of interpretation. Such broader scope both makes possible and demands reevaluating medieval- and modern philosophy–driven notions of the virtual, predicated as they are on the notion of a thinking subject. In response to that demand, in the next part, I set out to explore the who-question about the virtual bodies of memory in the Talmud in the broader context of which the Talmud and the tradition of its study have always been a part. Who remembers, or whose memory is that, after all?
nine
What Is the Sophist? Who Is the Rabbi? The Virtual of Thinking Who is the rabbi? —t he babylonian talmud ( pes . 115a,
yom
. . . and make plain by argument what he is. —p lato,
39a,
qid .
the sophist
15a)
218c
On computer screens, where current perceptions tend to locate virtuality, the virtual limits itself—practically and normatively—to only one, frontal view at a very limited distance and with exposure for only a very short time and/ or with a very high speed of comprehension. However, this is only one kind of virtuality. There are other, much broader and older kinds that can perhaps teach us important lessons about new virtual times and spaces, while the knowledge of the new can help us better discern the old.2 The older phenomena of virtuality are not necessarily to be wed to any visual mediation. They work on different speeds and hence both assume and require different regimes of attention, but they still follow the laws of the virtual; perhaps precisely because of the slower speeds they display in their older sites, they are more accessible for analysis and understanding. Therefore, the older forms of the virtual give us a chance to understand virtuality before virtuality “understands” us—that is, before a perhaps more obvious but also less reliable view of the nature of virtuality takes control of how we understand the virtual. What follows in this chapter is an attempt to discern the older forms of the virtual using the lens of its newer forms, approaching the distortions and aberrations that the new can create more as heuristic tools and less as prohibitive 185
186 Who Remembers?
obstacles on our way to a mutual hermeneutics of the virtual, a mutually hermeneutical illumination of the relationships between philosophy and the Talmud that an attention to virtuality in both can create. Through the lens of this specific problem of virtuality in the contemporary study of the Talmud, I look at the role that virtual arguments play in philosophy as an intellectual genre and see what that role can help us learn about virtual arguments in the Talmud. Using the issue of virtual arguments, I also make a more general claim. My approach here is to situate Talmudic discourse vis-à-vis the oppositions between dialogue and oration, philosophy and sophistry, orthodoxy and heterodoxy, and being and nonbeing. Plato’s project of philosophy as opposed to sophistry privileges the form of the dialogue, a form in which a philosopher helps his partner give birth to the right opinion about the truth of what really is. Plato prefers it over the form of the sophistic oration, in which a sage, in his view, only manipulates right and wrong opinions without any extra value given to what truly is and therefore, Plato claims, sinks into the depth of nonbeing. My principal argument is that any choice between being and nonbeing, dialogue and oration, philosophy and sophistry does not allow disagreement to be what it is in the Talmud—that is, a goal, and not only a means. This disengages Talmudic discourse from both philosophy and its privileged others—sophistry, to name one. Such a disengagement transforms the question of the relationship between the Talmud and philosophy and its others from the hitherto dominant chronology-based comparative approaches into a different kind of question altogether. As I will explain, in approaching the relationship between philosophy and the Talmud from the perspective of a mutual hermeneutics of virtuality, we will encounter a number of road signs that point to the paths by which we can arrive at that question. My task is to follow those signs and to go where they lead: to the larger realm of inquiry to which the Talmud belongs, along with both philosophy and its others, which include, to start expanding the list, not only sophistry, but also philodoxy3 and orthodoxy—the embrace of opinions or arguments as such, regardless of their truth, and the combating of heresies.
a What follows in this chapter is a study concerned with virtual characters in or of both Talmudic and philosophical arguments, an étude on virtuality
What Is the Sophist? Who Is the Rabbi? 187
in the Talmud, philosophy, and the privileged others thereof, among which sophistry is the first. It will illuminate the way in which a virtual agent both acts more specifically and more importantly speaks. What this means, however, is that we are moving toward a broader literary-philosophical, rhetorical, and ultimately Talmudic theory of virtuality, a more general theory of virtuality that deals with speaking and also arguing without being.
Virtuality in Philosophy and the Talmud: A Mutual Hermeneutics Studies in philosophy and studies in the Talmud have long been mutually detached. If this detachment had not happened, both fields might have looked different. That detachment calls for both analysis and undoing. I will be examining different philosophers and philosophies from various perspectives of the Talmud or its learners, and at the same time, I will examine the Talmud from various perspectives in philosophy. This will not only explore Talmudic perspectives on philosophy and philosophical perspectives on the Talmud or its commentary, but also, and more important, will interrogate the boundaries both traditionally and academically established between these two fields. Such an inquiry concerns the virtual (or, as it has hitherto been predominantly interpreted, the “anonymous”) character of the faces or voices behind arguments featured in so many pages of the Talmud and its commentary, but also, as I will argue, in a philosophical dialogue such as The Sophist. The first question in the title of this chapter, “What is the sophist?” is a translation from Plato’s The Sophist;4 the second, “Who is the rabbi?”5 cites the Talmud. Hitherto taken separately, those two questions have invited quite familiar answers: One asks for the personal name of a rabbi behind a given statement in the Mishnah; the other for a definition of the sophist. However, if put together, these questions create further questions about themselves: Does not the title “sophist” also attach to a personal name, or is it not at least a title given to a man? And does not the Talmud also seek to connect a definition (or rather, a given position) and a name, or rather, a title? Both call for the elucidation of a particular kind of subject position. The two questions can illuminate each other and thus perhaps also invite a very different and less familiar answer.
188 Who Remembers?
a Whether the relationship between the Talmud and philosophy has been predominantly approached in terms of a direct comparison or as a question of appropriation/rejection of one of the traditions within the framework of the other, an underlying assumption has almost always been that there were two separate traditions to compare or to appropriate. This assumption has taken for granted that the two entities are different from each other, not only in terms of their histories but also in terms of the intellectual characteristics empirically derived from their historical analysis. Remarkably, the assumption of two traditions has never allowed looking at the two traditions symmetrically, either in terms of their histories or in terms of their conceptual formations. Philosophy has been approached as a living tradition ramified in a complex, but still linear history of unique and often self-sufficient masterpieces, which can therefore be detached from a chain of the history of philosophy and stand firmly on their own, while the Talmud has been predominantly treated as one collective masterpiece, a record of arguments (with neither a pronounced nor a historically attested author, either individual or collective), surrounded, but never surpassed by its commentaries, which always remain attached to it. (Perhaps exactly because of the presumably secondary character of the commentaries, historians of the Talmud have carefully differentiated the commentaries from the masterpiece, whereas the traditional adepts of the Talmud have insisted on an unbreakable continuity between them.) The path of mutual hermeneutics is different. Instead of a direct comparison between philosophy and Talmud, and instead of looking at the appropriation or denial of one tradition by the other, it approaches the traditions of Talmud and philosophy from their own perspectives, asking how and to what degree each tradition can discern the other. A way to realize such an approach is to ask two mutually related questions, which are indeed only marks of a larger set of questions about the relationship between Talmud and philosophy. The questions are, What can the Talmud tell us about the tradition of philosophy? and What can philosophy tell about the tradition of Talmud in the Talmud? A more hermeneutically specific way to address that set of questions is to ask what light, or possibly what darkness, can be found in a foundational text of philosophy, such as, for example, Plato’s dialogue The Sophist, if that text
What Is the Sophist? Who Is the Rabbi? 189
is looked upon from the perspective of both the academic and the traditional study of the Talmud. Another question if not necessarily a symmetrical question, would be what does a Talmudic discussion entail if looked upon from the perspective of philosophers, sophists, and, most important, from the perspective of the question of being that The Sophist introduces once and for all, both in philosophy and in its history? Asked in that way, these questions are by definition not chronological- historical. I am asking not what Plato can tell us about the Talmud but what his treatment of the question of being and of sophistry can teach us about it. Likewise, I am not asking what the Talmud or the tradition of its learning tells us about Plato’s philosophy, but what Talmud as an intellectual practice, indeed as a version of hermeneutics, can teach us about Plato.6 The answer to these questions is best approached by examining what I will call the virtuality of arguments and characters in Plato and the Talmud. The issue of virtuality in the Talmud has to do with the character of both the names standing behind Talmudic arguments and the arguments themselves. Virtual characters in the Talmud have been addressed as arguments made in an anonymity that was not so much declared as carefully masked by the names of earlier Talmudic masters. As we have seen, deciphering this palimpsest has been among the central issues in contemporary academic discussions of the Talmud. In the Talmud, arguments of many Talmudic masters are made not directly by them, but rather on their behalf, so that the reader, or rather the learner of the Talmud, often gets different, indeed, conflicting versions of what an argument of a certain Talmudic master was in his polemic with another master. This part of Talmudic discussion represents a very substantial part of the Talmud and appears even more in the commentaries on arguments either between Talmudic masters or between virtual representations of their arguments in the Talmud. Similarly, the sophist, who is the main target of The Sophist, is present there only virtually. He never becomes present on the stage of the dialogue, nor is he absent from it, in the sense of being somewhere else. In that sense, he is like “The Author” in The Adventures of Tom Sawyer. That, however, in no way prohibits the sophist, which indeed is a title and not a living man, from speaking in the dialogue and from contesting many of the arguments about him. What is more, the sophist is far from the only virtual character in the dialogue. The virtual or hypothetical student of Parmenides is another. The list can continue with other titles or representations of intellectual
190 Who Remembers?
trends or positions that the Stranger, the most active character, both criticizes and dismisses.
a What can Plato’s dialogue, inhabited by such important virtual characters, teach us about Talmudic arguments made in virtuality? To answer this question, we need to pay attention to the role and importance of those virtual characters for Plato’s dialogues, that is, the role of virtual argument made on behalf of someone who is programmatically absent from the scene of a philosophical dialogue. We need to investigate what significance such a programmatic absence has for the progression, and indeed, for the success of the argument. The virtuality of some of the principal protagonists in Plato’s dialogue can help us understand at least two problems involved in the study of the Talmud. These two problems are actually two sides of the same problem. The first is discussed in contemporary literary-historical-comparative scholarship. The second, unfortunately, precisely due to the literary-historical-comparative premises of that discussion, has remained unattended to. The first problem concerns historical existence of the hypothetical unnamed people behind the virtual arguments of the Talmudic masters. Again, to name the unnamed, contemporary scholars have had to invent a name, calling these people “those without name,” or in modern Hebrew, stamm’aim, a concept Halivni used to theorize “The Author” of the Talmud. The second problem is the problem of being, or the problem of the ontological status of Talmudic arguments. Without turning to Plato, only the first of these two problems has been and could have been considered, because it has been posed in the literary-historical-comparative perspective, which simply did not leave any room for the second problem, much less allow us to look at the two problems as one. In The Sophist, the question of the existence of the sophist is related to the question of being, because it is related to the speech of the virtual sophist. The connection between Plato’s discussion of being and nonbeing and his attempt to identify the sophist proves to be much stronger than has hitherto been recognized by historians of philosophy. The choice of being versus nonbeing for which Plato advocates is connected to the choice between dialogue and oration. Plato chooses submission
What Is the Sophist? Who Is the Rabbi? 191
and cooperation in a dialogue instead of contest and competitiveness in orations. His choice belongs to the same set of choices that he makes between philosophy versus sophistry and truth versus falsity. The virtual arguments of the sophists present in the text, as well as of the no less virtual students of Parmenides and other virtual arguments and/or characters in the dialogue, tightly connect with the Stranger’s elaboration on nonbeing. In fact, working only with virtual arguments, which, of course, is very different from participating in a dialogue, is what enables the Stranger to elaborate on the question of being versus nonbeing. This is not merely a matter of composition but also, and even more so, a matter of strategy—for the sophist, as the target of the dialogue, cannot be its direct participant, yet he still has to be able to argue, even before his identity or even his existence is established. Without his argument about nonbeing, no targeting of him can succeed. The question of nonbeing thus becomes the theme of the virtual defense of the sophist, just as it becomes the main stake in the Stranger’s strategy of catching the sophist in a definition of who he is. The question of the being and nonbeing of the subject who is present virtually in the text should also be asked in regard to the Talmud. To discern that choice better, we will need to take an even closer look at what is at stake for Plato in allowing for the virtual speech of the sophist. Doing so not only directly affects the outcome of the effort to identify the sophist but also concerns both the identity and definition of the philosopher—both what he is and who he is. Curing the fear of nonbeing becomes crucial for Plato if he is to identify the sophist or to answer the seemingly naive question of who he is. The question implies that the sophist exists, while the virtual sophist in the dialogue does not necessarily accept even his own existence, lurking instead in the realm of nonbeing. That allows Plato to deal with the issue of being and nonbeing as a way of identifying the sophist in the first place. To do so, he has first to trap the sophist in the question of who he is. Then, when this most difficult part of the game is over, all he needs to do is explicate the answer to the question, as he does at the end of the dialogue. Yet, in doing so, what Plato identifies is not a man of flesh and blood, but a personified concept—indeed, a title—which must be allowed to speak even before it is proven to exist. The sophist must speak without being—speak without any person present in speech; speak without being in speech. Plato thus identifies the sophist as a “what” that must also become a
192 Who Remembers?
virtual “who” in order to participate in the dialogue, using virtual speeches about nonbeing and truth that the Plato’s Stranger must make him generate. Not only must the sophist produce his virtual speech as an apology for nonbeing in general and for his own nonbeing in particular, but he must also argue that no speech can be false, but only true. At least as long as a speech makes some sense, there must be some truth to it. What this means, however, is that what is (being) is no different from what is not (nonbeing). By implication, the sophist can easily argue, but he cannot be—thus successfully avoiding the Stranger. The Stranger’s response to that analysis of nonbeing becomes crucial, not only for catching the sophist in the net of the question of who he is but also for establishing the radical difference between the sophist and the philosopher. The question of nonbeing is the site at which the difference between the sophist and the philosopher is established. Plato cannot possibly establish this difference without his own deliberation on nonbeing, which remains dialogical in form, but which in its content is a refutation and indeed a contestation of the position of the sophist. Within the dialogical form, therefore, it is the Stranger who must perform this task. Far from being only a literary device to catch the sophist in the question of his own identity without either persuading him or entering into a dialogue, Plato’s discussion of being versus nonbeing marks a pivotal argumentative step for Plato himself as he delineates who the sophist is and who the philosopher is. Catching the sophist in the question of his own identity lets the philosopher produce an account of his own identity. The crucial difference between the sophist and the philosopher is that the sophist hides in the “darkness of nonbeing,” where falsity is impossible and thus also indistinguishable from truth. The philosopher, of course, dwells in the “light of being,” and therefore both the sophist and the philosopher are very hard to see, especially for the masses, who easily satisfy themselves with no more than convincing opinions about things, without worrying about their truth or falsity. In a different text, Plato pejoratively designates the ways of the masses as “philodoxy”—that is, as already noted, embrace of opinions or arguments, not of the truth or falsity that the arguments might entail. While putting a real philosopher and a virtual sophist at the two extremes of the spectrum, Plato needs a representative of philodoxy in the middle. The virtual character of the sophist thus becomes indispensable for creating that spectrum. In one and the same move, the virtuality of the sophist makes
What Is the Sophist? Who Is the Rabbi? 193
the philosopher real, puts him in the light of being, and leaves the philodox Stranger mediating between the two. What does that connection between the virtual character of the sophist (or of his speech) and his dwelling in the darkness of nonbeing teach about the virtual arguments in the Talmud? Can the Talmud’s virtual arguments be ontologically qualified as a case of sophistry, rather than of philosophy? Or rather as philodoxy? Or, what is not very different from the latter, are they a mere play of imagination, as, for example, Maimonides would want us to think? To understand the complexity of these questions, let me proceed step by step on the hitherto unpaved road from the question “What is the sophist?” to the question “Who is the rabbi?” That road leads from Plato, via some important milestones, to the Talmud, to Maimonides, and beyond. Philo of Alexandria will help us make our first, short, but very important step on that trip. Despite his admiration of philosophy, Philo is very different from Plato’s philosopher. Plato’s philosopher primarily discusses issues, not texts. He is interested in what is, in being, not in speeches that can entertain the possibility of what is not, or nonbeing. In contrast, in his allegorical reading, Philo discusses texts as a gateway to discussing issues that a philosopher might also have discussed. This seemingly insignificant difference has a huge implication: Plato’s philosopher, by his contrast with the sophists, both initially and ultimately targets what is, or being, as distinct from what is not, or nonbeing. So even if Philo associates his reading of Scripture with the philosophy of being, Plato would still have to place Philo in the ranks of either the sophists or, at best of the philodox, for Philo has to assume the truth of the text, something that a sophist might allow, but a philosopher would hardly accept. Whereas Plato thus would identify Philo as a sophist, Philo could hardly agree with that identification. The blind spot between those two is telling. In particular, it helps us approximate a larger realm or place in which we also can locate the Talmudic tradition. Not entirely unlike Philo’s writings, Midrash as an exegetical expounding of Scripture and the Talmud as an exegetical interrogation of the Mishnah also discuss issues, albeit necessarily through the gateway of a given text from which they begin, proceed, and via which come to an end. Again, the most important difference lies in the implication: As forms of exegesis, neither Midrash nor Talmud necessarily take the discussion in the direction of what is (being) versus what is not (or nonbeing), but rather remain in what Plato classifies as the darkness of nonbeing and sophistry. To complicate things
194 Who Remembers?
even more, Midrash and Talmud never lose sight of the given text: Unlike Philo, they use it not only as a gateway but also as a target, and often as a tool. Like the discourse of the sophist, Talmudic discussion works hard to disallow the falsity of a given speech of the Mishnah. But in Talmudic discussion, the given nature of the speech is assumed to make it exist or to be and removes it from the realm of nonbeing. The practice of Talmudic exegesis, therefore, can no longer belong to nonbeing alone. Plato, therefore, would again almost inevitably classify exegesis as the practice of sophistry, for it starts from the shaky ground of a text that is either presumed to be not false or reconfirmed as true, instead of discussing the issue or matter directly. Philo, by contrast, classifies his exegesis as philosophy, that is, as a prioritized discussion of being to which nonbeing must be subordinate. In short, Midrashic expounding of Scripture and Talmudic interrogation of the Mishnah always start from a text, but unlike both Philo and Plato, they also dwell within it, precisely in the dark realm of nonbeing, as Plato (and perhaps even Philo) would see it. Given the gap between Philo and Plato, the identity of the rabbi in question is, therefore, hardly that of a philosopher, but even less that of a sophist. He both emerges from and disappears into the gap between Philo’s subscription to Plato’s philosophy of being and Plato’s implied dismissal of Philo as a sophist. Much less is that rabbi a philodox, a mere middle between the two extremes. Yet the option of philodoxy requires a closer look, which we would be much better equipped to take not with Philo, but rather with Irenaeus, a second-century church father who, with others, prepared the ground for Christian orthodoxy to emerge. He can be of great help for discerning any orthodoxy and/or philodoxy in the rabbi, for after all, both orthodoxy and philodoxy concentrate on doxa and thus can help illuminate each other. Is Talmudic exegesis a case of philodoxy? What makes the help of Irenaeus indispensable in answering this question is his performance of orthodoxy in its basic sense, as a true opinion that arises not from stating that opinion directly, but rather primarily from refuting the opinion or, in Irenaeus’s terms, “knowledge, falsely called so,” also known as heresy.7 Our way from Plato to the Talmud therefore might become easier or at least more predictable if we continue it with Irenaeus through his notion of heresy versus orthodoxy. Irenaeus is an example of a Gentile, that is, of somebody who wants to belong to the covenant with the G-d of the Jews, but who pointedly does not
What Is the Sophist? Who Is the Rabbi? 195
want to become a Jew. What allows him to engage with the covenant without becoming its direct subject? His philosophy-oriented refutation of sophistry and philodoxy as heresy is one part of the answer. In performing such a refutation, his own orthodoxy becomes possible. But it is only one part of the answer. A combination of his theology of G-d as logos with an ontology of G-d as being rather than nonbeing in his critique of “false knowledge” is the other. But perhaps even more important is the way in which he resolves the inevitable tension between two imperatives. One is the philosophical imperative to treat the issues, and thus what is, versus what is not. The other imperative is exegetical, which in the eyes of Plato would most likely involve either sophistry or, at best, philodoxy. Philo resolved a similar tension between exegesis and philosophy by discerning and treating the issues as existing above and behind texts, thus in the way of allegory, yet he neither had to position himself as a non-Jew nor was his allegorical reading outside the conventional lineage of Jewish tradition. He did not need any new covenant for himself. This, however, is not sufficient for a Gentile engaged with a covenant between G-d and the Jews. Why? Because such a Gentile cannot limit himself or herself to just reading already established texts philosophically. Instead, he or she needs to reestablish the texts anew, as a source of a new authority for his or her unique position as a Gentile newly engaged with a covenant already designed for those who either are or are willing to become Jews, not for those who want both to be in the covenant and to preserve their non-Jewish identity at the same time. Differentiating between philosophy and philodoxy becomes very useful for handling this task. It is precisely knowledge of what is that makes philosophy different from philodoxy. Therefore, the Gentile has to strive for true knowledge, knowledge of what is. At the same time, he or she cannot be Plato’s philosopher, for philosophers do not do exegetical work. Instead, he or she has to find a way between philosophy and philodoxy. The result is orthodoxy, a position that can arise only from refuting heterodoxy as heresy, not from any direct philosophical imperative to seek knowledge, for such an imperative would inevitably remain outside of exegesis and thus outside of the Gentile’s task. This might explain why the concept of knowledge comes to the forefront of Irenaeus’s intellectual world and why, indeed, he construes and constructs Gnostics as the main and inevitably virtual target of his refutation. By claiming that the Gnostics have knowledge, which, however, happens to be false
196 Who Remembers?
philosophically, he can prove their knowledge to be false exegetically. Thus, in a negative way, the only one he can use, he establishes his own position as by implication orthodox. As we learn from Michael A. Williams, he quite artificially constructs straw men, exegetes of the texts for whom philosophical knowledge was the ultimate value.8 He even calls them by the name of this most important and most problematic of philosophical concepts— “Gnostics,” or “those who know”—those who know what is behind what is already revealed in the text. In this way, Irenaeus creates orthodoxy by refuting the philosophically false knowledge of Gnostics, as well as by ridiculing their teachings as mere philodoxy. The resulting orthodoxy, however, is neither philosophy nor philodoxy. In fact, his project undermines the difference between the two. This is accomplished in two moves. The first move is a critique of philosophical knowledge of what is, or being, versus what is not, in terms of its dangerous claim to know being beyond what is already revealed in Holy Scripture. The second is his critique of personal authorship. Arguing from their personal philosophical authority, heretical authors create illegitimate additions to what is already revealed in revelation. That second move sounds like a philosophical critique of sophistry, whereas the first sounds like a sophistical critique of philosophy—if there can be such thing at all, for did sophists ever criticize philosophy, or could they ever have been interested in that? All this, however, happens for the sake of approaching holy texts, a task that neither philosophy nor sophistry either actually recognized or was potentially able to recognize. The resulting orthodoxy (for example, the affirmation that the creator of the universe and Christ as Savior are one and the same person in two versions) inevitably calls to mind a similar kind of critique, for it also represents certain knowledge of being and also extends revelation beyond what is “already revealed” (i.e., beyond Scripture, of course, unless the Gospels are also considered a revelation, a concession that Irenaeus is ready to make, thus allowing the concept of a New Testament to emerge). Eradicating the difference between the philosophically poor philodox and the antiphilosophically rich sophist, Irenaeus helps us visit a radically new place that Philo’s encounter with Plato did not reach. Instead of the blind spot between Philo’s admiration of philosophy and Plato’s implicit disqualification of Philo as either a philodox of the holy texts or even a sophist, Irenaeus reveals another blind spot, that of orthodoxy, which is surely not a
What Is the Sophist? Who Is the Rabbi? 197
philodoxy, for he at least theoretically differentiates false opinions from true ones and, refuting the false, reserves room for the true. That truth, however, is maintained not by philosophical knowledge of what is, but rather by a sophistically performed refutation of the opinions of the others for their reflection of what is not. Orthodoxy thus opens a new, empty space created by the sophistical refutation of philosophically false knowledge. This space exists only as far as and only because there still is something to refute. Does the rabbi reside in that space, as well, a space constructed with the blueprints of orthodoxy? If he does, he either maintains it without content—as it should be, for, when encountering a heretic, the rabbi limits himself to refuting him—or he fills the room with the opinions of his peers, who disagree one with another. Whether his orthodoxy is empty or full of disagreeing arguments, the rabbi advances the question of being and nonbeing, and thus of virtual arguments, to the forefront of the discussion again.
Language and Virtuality, Being and Exegesis So far, our inquiry has shown how the notion of being, or what is, helped Plato establish himself as a philosopher of light and of being versus either the sophist, located in darkness and nonbeing, or the philodox, the lover of mere opinions and plausible arguments, regardless of their ultimate truth or falsity. In the taxonomy of the philosopher, the sophists, the philodox, and the orthodox, being, or what is, thus proved to be a concept with a power greater than even Plato’s own (I almost said “orthodox”) notion of forms or ideas. The relationships between philosophy, philodoxy, sophistry, and orthodoxy have thus not only been established through the notion of what is, or being, but also have continued in the same way, even when a new (or, from the sophist’s point of view, only partially new) element, that of the exegesis of holy texts, entered the game. The relationship between philosophy and its default others—sophistry and philodoxy—was thus both challenged and complicated in a variety of ways when a new component, that of a holy text, with a being of its own and thus with the question of its ontological status (of either being, or nonbeing, or else of opinion) entered the game, as also did the new practice of relating and interpreting texts with such a problematic ontological status, the practice of exegetical reading.
198 Who Remembers?
As we have seen, for Philo, this new component or the holy text called for an allegorical interpretation, which Philo himself deemed philosophical, whereas Plato would by necessity deem it sophistical, or at best, philodox. What is more, the very notion of the holy text as speech that cannot be false is very close to the sophistical notion of the impossibility of the falsity of any speech whatsoever, provided that speech can make sense. A remarkable, symptomatic, and unbridgeable gap exists between these two thinkers: Philo appeals to Plato, while Plato dismisses Philo. This gap marks the existence of a blind spot between the exegetical and philosophical traditions. It has played itself out in a variety of ways. We have seen it at work with both Irenaeus and the orthodox rabbi who is in disagreement with his peers. One other way in which it has played itself out involves the notion of language itself. Looking at this makes us return to both Philo and Irenaeus, if only for a short time. For Plato, language mingles with both being and nonbeing, thus allowing the sophists or the philodox to tell lies. For Philo, language (at least a certain language) is holy and thus, just as for the sophists, cannot tell a lie (albeit, importantly for what will follow, that language can still be understood incorrectly). However, like Plato and unlike the sophists, Philo believes that the holy text must mingle with being only, not with nonbeing. If Plato sees telling no lie as a requirement to tell what is as opposed to what is not, and the sophists see language’s inability to tell lies as a testimony for nonbeing’s parity with what is, Philo allies with Plato. Hence, for Philo, nonbeing is a place in which the holy language is read wrong. This place is a trap. Once in that trap, the reader can no longer discern what the Scripture tells about what truly is. Instead, the reader reads in the Scripture about what is not. Irenaeus cannot be fully content with that, for if the holy text belongs to being, philosophical knowledge prevails over its holiness, a luxury that he, as orthodox, simply cannot afford. In the Talmud, the relationship between philosophy, philodoxy, being, and language is complicated by the Talmud’s exegetical task. That task applies not only to holy writing but also to the presumably authoritative and yet oral and thus not only authoritatively divine but also intrinsically human tradition of the Mishnah. Since the Mishnah allows and even requires competing interpretations, the Talmud might seem either to take the side of philodoxic indifference to the truth or the falsity of knowledge or execute a sophistical freedom of nonbeing. Either of these options allows Talmudists to authorize
What Is the Sophist? Who Is the Rabbi? 199
and reauthorize the oral tradition of the Mishnah through the process of its ongoing sophistical interrogation without the demand for philosophical validity and thus for any connection with being. However, the Talmud also has an element moving it in the direction of truth and thus away from philodoxy. Irenaeus can help us discern that element better. In Irenaeus, we saw a radically different configuration of philodoxy, philosophy, and the exegetical task, which could not even be compared to the configuration of the Talmud. He distances himself from knowledge of what is, and thus from philosophy, for philosophy represents a danger to his exegetical strategy of treating holy writing as revelation, which, in his view, is already given and thus cannot be extended further—not even through philosophically authenticated knowledge. To disallow any application of philosophy to the task of revealing what is, Irenaeus uses the sophistical strategy of refutation. Yet, due to his obligation to revelation in the holy text, and thus to distinguishing truth from falsity, he cannot position himself as a sophist or a philodox. Therefore, he cannot satisfy himself with the refutation of philosophically false knowledge alone, much less with any disagreement with his own peers or friends to whom he writes. Just as the Stranger talks to Theaetetus, not to the sophist, Irenaeus writes to his friends, not to the Gnostics. The latter are as virtual for him as the sophist is for the Stranger. Irenaeus’s solution is to establish a new position, that of orthodoxy—a correct opinion guaranteed not by philosophical knowledge, as in Plato, but by the critique and dismissal of any incorrect knowledge about a holy text or what it reveals. Eventually, his design of orthodoxy as an empty space created by means of refuting his virtual Gnostics almost inevitably leads to the establishment of a positive opinion about the revealed truth, which immediately puts that opinion or its interpretation under the fire of orthodox critique again, for it is constantly haunted by the virtual heretic who stands behind the whole process. This tension, or rather obsessive repetition, resolves itself in Irenaeus’s search for authority, as opposed to any kind of authorship. However different he is from the rabbis, Irenaeus indicates something very important about the Talmudic tradition. Like him, the rabbis in the Talmud distance themselves from philosophy or its use for furthering revelation as a way of fulfilling the exegetical task, yet, unlike Irenaeus, the Talmud does not univocally deny a continuation of revelation by means of, if not philosophy, then sophistry and philodoxy. Both in approaching
200 Who Remembers?
Scripture in Midrash and in approaching Mishnah in the Talmud, the rabbis do what Irenaeus does not do: approve the truth of a text through a refutation followed by a defense. By contrast, Irenaeus satisfies himself with refuting heresy alone, using the positive effect of the refutation to anticipate and welcome orthodoxy, even before establishing its positive content. Eventually, the empty space Irenaeus created fills in with a positive “knowledge,” or at least with a text, followed by a sophistical fight against the heresies that arise in reading it. In the case of the Talmud, the same place of orthodoxy fills neither with doctrinal or positive statements of faith about G-d nor with the truth of his revelation, but rather with disagreements between and about the rabbis themselves, including the disagreements about who they might—virtually—be. Not fully unlike the case of the sophist, the virtuality of the rabbis’ argument becomes indispensably part of who they are, or rather a function of who they aren’t, for any given argument in either the Mishnah or the Talmud might eventually hearken back to any of several possible rabbis, none of whom has to be real, but all of whom are virtual. A difference between the Palestinian Talmud and the Babylonian Talmud will serve as a magnifying glass showing the virtuality of the rabbi’s being. The earlier version, the Palestinian Talmud, is easier to compare and to juxtapose to Irenaeus’s creation of the empty space of orthodoxy surrounded by the ruins of what is first a virtually constructed and then a realistically refuted heresy. The example will show that like Irenaeus in Against Heresies, a rabbi in the Talmudic record, Rabbi Simlai, produces the empty space of orthodoxy with regard both to heretics and to his students. A later version of the Talmud, the Babylonian Talmud, will show how virtual argument in the Talmud fills in that empty space with refutations now directed not to heretics, but rather to peers. Under such circumstances, the arguments (and the addressees) of the refutations and defenses become no less virtual than they were with the heretics and peers in both Irenaeus and the Palestinian Talmud. Translated into English, the text in the Palestinian Talmud, PT Berachot 9:1 reads as follows. (The truly fascinating content of it should excuse the length of the citation.)
What Is the Sophist? Who Is the Rabbi? 201
PT Berachot 9:1. —How many divinities created the world?—heretics approached Rabbi Simlai with this question. —Are you asking me?—He said—Go and ask First Adam! As it says, “if you will ask about first days” etc. but it does not say “[from the day on which] Elokim created [bar’u—in plural form of the verb] Adam on the Earth.” Rather it is written there nothing else but “[from the day on which] Elokim created [bar’a—in singular form of the verb] Adam on the Earth. —Yet it is written, “In the beginning Elokim [literally, in plural: “G-ds”] created [bar’a—in singular verb form]”—they responded. —But is it written “[they] created” [bar’u—in plural verb form]? Nothing is written but “created” [bar’a in singular verb form], he answered. Rabbi Simlai said [to his students], “Every time the heretics attacked, the answer was found on their side.” —If so, why it is written, “Let us make Adam in our image and in our likeness?”—asked the heretics again. —It is in no way written here, “And they created Adam in their image,” but rather “And [he], Elokim created [waybr’a—in singular] Adam in his image.” [When the heretics left] his students said to him, “You kicked those out with a stick, but what are you going to answer to us?” He said to them, “Earlier, Adam was created from the Earth and Eve was created from Adam, but from Adam and on, ‘in our image and in our likeness’—a husband cannot be without his wife, a wife cannot be without her husband, and they both cannot be without divine dwelling between them.”9
Using the difference between alleged contextual readings of the verb “created”—singular versus plural—which in all cases spells “bra,” br’ the same way, with no grammatical indication of number shown in the way it is written, Rabbi Simlai disallows the heretics’ alleged use of vocalization to
202 Who Remembers?
determine the number and returns them back to the limits of how the verb is actually spelled, not to how it might be vocalized contextually to befit their assumption of multiple gods in creation. Thereby, he refutes their attempt to find a second divinity that the creation allegedly had involved. Rabbi Simlai is in refutation mode with the heretics and in defense mode with the students, but the space of orthodoxy remains empty with both. With the heretics, he creates readings supporting their assumption of multiple divinities in creation and finds the answers that refute these readings. With the students, he is responding (basically providing a support) to their already existing desire to be reassured that only one god created the world. In both cases, orthodoxy is empty. He is obviously not acting as a dialogical philosopher in his openly rhetorical (in Plato’s terms, sophistical) contest with the heretics, but he is not in a philosophical dialogue with his students, either, because he is following their desire for reassurance, not leading them to any new knowledge. He thus speaks without being. His two capacities— the refutation of heretics and the defense of the orthodox position in front of and for the sake of the faithful students produces opinions that do not intersect. What R. Simlai tells the heretics is not at all the same as what he tells his faithful students. Precisely because of that lack of intersection, Rabbi Simlai identifies with no specific opinion/argument that he makes. What that means, however, is that his speaking produces two heterogeneous effects—refutation and defense—in two heterogeneous environments, which intersect in no other place than his name. This is possible only if he belongs to none of these environments while still acting in both. He therefore acts as a virtual agent. Speaking from a position that is perfectly empty in both types of interaction, he, not unlike the sophist, speaks (and acts, i.e., refutes and/or defends) without being. He therefore remains virtual, as Irenaeus’s Gnostics do, and indeed as does Irenaeus himself, not to mention Plato’s sophist, or indeed his philosopher, who acts as no more than a symmetrical function of the sophist. In creating their respective virtual opponents, the Stranger, Irenaeus, and Rabbi Simlai also re-create themselves as virtual subjects. Virtual argumentation in the Babylonian Talmud, however, creates a different subject position for its participants. Translated into English, the Babylonian Talmud, BT Megillah 19b–20a reads as follows:
What Is the Sophist? Who Is the Rabbi? 203
Mishnah: All are qualified to chant The Book of Esther [both initially and after the fact of chanting, that is retroactively, even if initially, i.e., before the chanting, the person would not qualify?—S.D.], except for the deaf, imbeciles, and minors. Rabbi Yehudah makes a minor qualify. Talmud: —Who is the Rabbi [in the Mishnah] to say a deaf person does not qualify even after the fact? —“That would be R. Josi, for it is taught in the Mishnah, ‘The one, who is [already] chanting the Shema prayer and did not make it heard to his own ear, still fulfilled his obligation, said R. Yehudah. R Josi said he did not’ ”—Said Rav Mathn’a. —Why does it have to be R. Josi’s? Why does it have to disqualify [the chanting of the deaf] even after the fact?! It may well be Rabbi Yehudah’s; and disqualify the chanting of the deaf only initially, thereby intimating it works after the fact! —Such intimation should not even occur in your mind, because the deaf are mentioned just like imbeciles and minors are; and just as the latter two do not qualify even after the fact, so also the deaf do not. —Yet it still may be that each [in the group] comes out differently! —It hardly may. Because, in its last phrase, the Mishnah says, “Rabbi Yehudah makes a minor qualification,” which must have come from a premise that the first phrase does not make him qualify [at all, even after the fact]. —Yet only the last phrase is [explicitly] R. Yehudah’s, which intimates the first phrase would not be his! —Well, it still might very well be all by R. Yehudah. —It might not, for the first phrase in the Mishnah is not like the last one. The first forbids and the last permits. For then it would mean mentioning two different kinds of minors! —Oh, that only means the Mishnah would be corrupted; and here is what it would say as a matter of fact: “All are qualified to chant The Book of Esther, except for the deaf, imbeciles, and minors. To which minor does that refer? To one who is not yet ready for education; however, if a minor is already ready for education, he is qualified even initially, because Rabbi Yehudah makes the minor qualify.”
204 Who Remembers?
—Then, where did you end up placing the Mishnah?! In Rabbi Yehudah and after the fact! But we have a contrary teaching: “Yehudah the son of Rabbi Shimon Ben Pasi taught that a deaf person who can speak but not hear may set aside the teruma-offering”! —So, whose is the Mishnah then? Paradoxically, if it is of R. Yehudah, it makes [the deaf] qualify after the fact only, not initially. In turn, if it is of R. Josi, it disqualifies the deaf even after the fact! Rather, it is of R. Yehudah, and makes the deaf fully qualify even initially. —That would be quite contrary to the following Mishnah, “One should not say his grace after meals silently, and if one did, he did not fulfill his obligation”! —[Why is that contrary?] —For, whose can that be? Neither R. Yehudah, nor R. Josi can have it. For in Rabbi Yehudah, the silent grace should work even initially, and for R. Josi, it should not, even after the fact! —By all means, it would still be of R. Yehudah; and the silent grace after meals works even initially; and there is no difficulty here! The teaching [about teruma] indeed expressed the opinion of R. Yehudah, while the other [about the grace after meals] expressed an opinion of his teacher only, as it says in [an apocrypha of tann’aim], “R. Yehudah says in the name of R. Eleazar the son of Azariah: One who is chanting a Shema prayer, wants to make it heard to his ear, as it says, ‘Listen, Israel, the Lord is Our G-d, The Lord is the Only’ [which intimates it] makes what gets out of your mouth heard to your ears. R. Meir says, ‘. . . on which I command you today on your heart’ [which intimates that] according to the concentration of our heart is [the value] of our words.” —If you already go that far, you could equally say R. Yehudah [in the teaching about the teruma offering] argues just like his teacher, and what was taught [about the grace after meals] is coming from the name of [a different] R. Yehudah, the son of R. Shimon ben Pazi, and is in fact Rabbi Meir.
What Is the Sophist? Who Is the Rabbi? 205
This text is both a direct example and an immediate context for the question of what or who is, or was, or could be the rabbi. It goes without saying that the “is” or “was” or “could be” gets no distinct linguistic expression, but is only intimated. I will be interested not in these intimations, however, but in the more explicit “what” and “who” sides of the question. What kind of answer does this question invite in the Talmudic discussion above? In response to it, we find a series of possible names, or rather name tags, attached to the tradition recalled. None of the names becomes final, not even the one introduced last. The discussion leaves off with many virtual names possible precisely because none of them is definitive, and it leaves off with none of the names fully dismissed. Thus, for an answer, we get a host of virtual names, none of which can be fully dismissed and none of which can become the final or acceptable answer at the expense of others. Let us appreciate that. The answer to the question comes in a set of virtual names, the collective plurality of which does the expected work. It reaffirms the Mishnah’s authority, albeit not by finding a definitive and thus “real” rabbi behind it, but rather through a host of virtual, that is, strongly plausible, mutually exclusive name tags that thus satisfy the question “Who is the rabbi?” without claiming any philosophical knowledge of the “is.” With such a host of virtual speakers behind the Mishnah, and without any final and thus as it were “real” speaker to name, the Mishnah not only emerges as speech without being but also, together with the Talmudic discussion of it, loudly calls for a more careful look at the “who” and the “what” in all the other traditions of virtual speaking we have already observed. The “who” and the “what” in these traditions can now be approached afresh, with no imperative connection with either the “is” of the questions or with the being that it alleges. No longer dealing with the question of being or nonbeing invites a discussion of the personality, personal name, intellectual profile, or, if you prefer, the title of the sophist and/or the rabbi. That shift surely involves a mirror effect, because even more importantly, it invites a discussion of the questioner, the one who asks each of the questions. We should always keep in mind, however, that whoever looks in the mirror is never exactly the same as the virtual image he or she sees in it. The simple answer is that it is a philosopher who asks the question “What is the sophist?” and it is a Talmudic master or a Talmudist who asks the question “Who is the rabbi?” The “who” and the “what” parts of these questions can therefore be best approximated if the direction of our travel changes and we no longer go from
206 Who Remembers?
Plato and philosophy to the Talmud, as we did via Philo and Irenaeus, but rather proceed backward, from the Talmud to philosophy. Maimonides can hardly be surpassed as a guide on that path, because his intellectual trajectory goes in exactly the direction we need to follow: from the Talmud to philosophy. With him and his later discontents, both Talmudic and philosophical, we can attain a better grasp of the “who” and the “what” found both in and behind our questions. Maimonides helps us appreciate the personalities and/or intellectual profiles of both the Talmudist asking the question, “Who is the rabbi?” and of the philosopher asking the question, “What is the sophist?” He disqualifies the Talmudist as a person driven either by doxa, that is, by opinion and/or argument, or as, he has it, by “the power of imagination” alone. That philosophical definition, however unpleasant it sounds, will help us start approximating the profile of the Talmudist better. A distant intellectual relative of Philo and Plato, Maimonides exemplifies a new return from the Talmud to philosophy, now as a way of dismissing the Talmud (of course, without losing either its authority or its tradition). He dismisses the Talmud as sophistry and imagination, and thus as philodoxy, almost precisely in the ways Plato employs to dismiss the sophist. Again like Plato, he allies with being, or with what is, seeing nonbeing as only a part of being, something that language allows, but that philosophy must eradicate. Yet, also faithful to his exegetical imperative, Maimonides distances himself from the philosophical tradition of searching for truth, holding instead that the truth is always already revealed as Scripture, and people need only to understand it—using, of course, the tools of philosophy, not of the Talmud. Strauss suggested that Maimonides considered revelation a given truth to be understood with the help of philosophy and treated philosophers without revelation as people who are desperately searching for truth while it has already been fully revealed. This suggestion helps determine an important difference in how Plato would classify Maimonides in connection with how Maimonides classifies Plato. The difference involves the role that misunderstanding or perplexity plays for Maimonides, as opposed to its role for Philo as allied with Plato. Whereas for Philo, Scripture is misunderstood if it is not read philosophically, that is, allegorically, so that allegorical reading replaces misunderstanding by what is philosophically correct, for Maimonides, misunderstanding or perplexity is not an extraneous condition to be replaced, but the beginning of a true philosophical understanding of the names in the Scripture, or in a
What Is the Sophist? Who Is the Rabbi? 207
broader context, a starting point for understanding the revealed truth without substituting the truth of philosophy, but using philosophy as a helper. As we recall, Philo allied himself with philosophy, even if such a philosopher as Plato would rather call Philo a sophist, because Philo’s notion of holy writing (or writing without falsity in it) is what Plato’s sophist held about language in general: What can be said cannot be false. In contrast, Plato would hardly call Maimonides a sophist. Maimonides does allow that falsity (or seduction, delusion, and hence perplexity) exist in the language of the Torah, at least in the way in which people can intelligently misunderstand it. Such perplexity arises if people who are educated enough still do not grasp the full scope of what he calls “co-participant names” in the Torah, or if they do not discern a name as a “borrowed name,” or if they fail to isolate names that are dubious in nature and thus find it difficult to determine if they are of the first or the second kind. Most important, and what is most strikingly different from Philo, Maimonides allows and even requires educated misunderstanding or perplexity as a condition of the possibility of understanding in philosophical terms, for after all, those who are not educated in philosophy cannot be perplexed. The whole classification of names listed above designates both the conditions of misunderstanding and a way to overcome it. In addition, unlike Maimonides, Philo did not exactly position himself as a philosopher. Although admiring philosophy, he sought only to read or translate holy writing (admittedly a sophistical notion) into the language of philosophy. From all these differences, there appears an intellectual profile of the philosopher, dismissing, or rather replacing what Maimonides sees as the “less successful” image of the Talmudist. The profile of the dismissed Talmudist is thus closer to the sophist or the philodox, not to the “desirable” profile of the philosopher. This is, of course, not to suggest that Maimonides dismisses the Talmud as sophistry. On the contrary, he reestablishes the Talmud as a source of authority. Yet, he seriously considers the Talmudist’s proximity to sophistry—or, in his terms, to the faculty of imagination alone as opposed to the faculty of reason—to be a serious deficiency, inducing him to rewrite the Talmud (or for that matter, the Oral Torah) with the pen of philosophy. To that end, the Talmudist is to be reinvented by the philosopher, even if Maimonides wants to adjust the profile of the philosopher through the intervention of revelation (of which the Talmud in its pregiven form was not an obvious part). Maimonides’ negative portrayal of the Talmudist comes together with his allowance for and even requirement of initial perplexities in grasping
208 Who Remembers?
revelation. That helps isolate a different, in fact much more positive, perspective on the character of the Talmudist. That perspective is based on Maimonides’ allowance for perplexity that is then followed by a more precise understanding, which our next guide, Rabbi Izh.ak Canpanton, transformed from merely an allowance or requirement to an operative tool, indeed, the main device of a rationality that he conceived as both connected to philosophy and differentiated from it. Responding to Maimonides and, at the same time, allying with the Tosafists, the medieval commentators of the Talmud, who, unlike Maimonides, considered the Talmud to be not a collection of virtual arguments on different topics around the Mishnah, but rather one big argument, consistent and coherent in all its parts (despite all seeming incoherencies that the Tosafists both bring forward and overcome), Canpanton distanced himself from both Maimonides and philosophy by claiming Talmud to be a rational intellectual art of its own, which Maimonides, as a philosopher, missed in his philosophical thinking of the Talmud as merely a text to be understood or misunderstood. Against his own will, through his dismissal of the Talmudist in favor of the philosopher of revelation, Maimonides helped Canpanton discover the Talmudist anew, now as a type different from any other type (philosophers, prophets, and people of imagination) that Maimonides was able to identify and also different from any of the types (philosophers, sophists, the philodox, and the orthodox) that we have seen before. Help came from Maimonides’ (and Plato’s) recognition that language can mingle with nonbeing and thus with a genuine experience of failure (or perplexity), followed by the recuperation of truth. Talmud, for Canpanton, is therefore a case of neither philosophy nor sophistry. Of course, it is not a case of rhetoric/eloquence (terms for the philosophical interpretation of sophistic speeches), either. First of all, it is not a case of rhetoric, because philosophy already had defined and derogated rhetoric as eloquence. Unlike a piece of rhetoric (or sophistical eloquence), Talmud, for Canpanton, is a rational art of disagreement, not a nice speech. Second, the Talmudist is not a philosopher or sophist, for both of them subscribe to the goal of attaining agreement, either in the present or in the future, either on the basis of being and truth (as does the philosopher) or on the basis of nonbeing and the impossibility of falsity (as does the sophist). Unlike them, the Talmudist studies the Talmud by targeting the disagreements found in it not as obstacles to overcome, but rather as goals to attain. For the student, the disagreements in the Talmud come from the past and
What Is the Sophist? Who Is the Rabbi? 209
have the value of a fait accompli. A student of the Talmud is to preserve them at all costs, not to take them to a place of agreement. He must thus invest the Talmudic disagreements with the assumption of truth, an investment that he will also have to fund using the rationality of his argument on behalf of the disagreement. This investment is occasioned not by the fact of there being a disagreement, but rather by the student’s appreciation that they come from the past. They have the value of a fact and therefore cannot be and should not be changed, but only continued, which means “understood better.” It is their pasticity (if I may), that is, their simultaneous pastness and factness, that makes them into a goal of study. All the Talmudist needs to do to the Talmud is to invest intellectually in the permanent truth of the disagreements in it. That means, for Canpanton, that the Talmudist is neither a philosopher of being nor a sophist of nonbeing, for neither of them can grant a disagreement any truth whatsoever. That helps us give better answers to both the question “Who is the sophist?” and the question “Who is the rabbi?” As we have seen, both are functions of the virtual arguments and/or characters in the texts that pose them. Different from either presence or absence, virtuality (or respectively, the figure of a sophist and a rabbi) becomes indispensable for making the philosopher appear from the shadow of the sophist and the Talmudist appear from the shadows of the rabbis, philosophers, and Talmudists who are thus no less virtual than the virtual speeches they entertain. Most important for the question of virtuality, we now can better understand the role of the virtual personalities about whom both the philosopher and the Talmudist ask. Without asking about any biographically “real” sophist or rabbi, they ask about the truth of their intellectual positions, so that the name behind the question becomes a title, just as does the name to which the question refers. Personal identities and definitions of intellectual positions intermingle. What is more, the who—and the what—questions come to imply each other. In the case of the philosopher, the real question is then not “What is the sophist?” but rather Who is he?; and in the case of the Talmudist, the real question is also not “Who is the rabbi” but rather What is he? To which intellectual profile—or rather, as Menah.em Kahana has it, to which cluster of statements is the name of a rabbi to be ascribed? If so, and if what is virtual is not the same as either what is present or what is absent, what does that say about the real? Is the Talmudist a real sophist, as opposed to the virtual sophist that Plato produced? If virtuality is more complex than
210 Who Remembers?
the simple opposition between what is present and what is absent can allow, it perhaps can also teach us a lesson about reality. I thus conclude at a point at which the work undertaken in this chapter should continue, as it does in the next. The strategy of mutual hermeneutics that I have applied to the issue of virtuality in the Talmud and in philosophy helps us discern virtual arguments and/or characters, whereas the simple opposition between presence and absence (i.e., presence somewhere else) can no longer satisfy the demand for speech to make sense in which a virtual character and/or argument is created. The resulting irreducible virtuality of both the argument and the personality behind it generates a series of configurations between philosophy and its default others, on the one hand, and exegesis, heresy, orthodoxy, and the Talmud, on the other. Those configurations reintroduce philosophy into the hermeneutical horizon of the Talmud and the Talmud into the hermeneutical horizon of philosophy, and they do so in a radically new way in which the hitherto dominant contemporary intellectual divide between neo-Kantian anthropology (ramified in empirical philological, chronological, and sociocultural readings of the Talmud in terms of its “anonymous redactors”) and Heideggerian ontology (expressed in particular in Heidegger’s reading of The Sophist in the light of the question of being) shifts from prohibitive to productive. An investigation of such virtuality can therefore be done only in the way of a mutual hermeneutics of philosophy and Talmud, that is, in view of a different level of reality to which this virtuality invites us to refer. Is this reality transcendental? Is it more complex than what either philosophy (including transcendental philosophy) or its default others allow us to entertain? Neither a presence nor an absence, speech without participating in being, an investment in speech well before being, not an elusive ghost, but rather active speaking and arguing happening well before any more common visual expression of the virtual in visual terms, the virtual in philosophy and in the Talmud calls for further analysis, both Talmudic and philosophical. This analysis could hardly be done if these disciplines of thinking remain deaf to the virtual speeches that each of them so powerfully creates. Virtual speeches and the virtual agencies that these speeches embed in the text of the Talmud do not, however, testify to any reality beyond the literary—more specifically virtual—reality that they create. Much less do these speeches and agents testify to a person-centered Cartesian existence of the speakers and agents. By necessity, which the next chapter will help
What Is the Sophist? Who Is the Rabbi? 211
explain, analyzing a virtual body of who-what, embedded and engaged as it is, in Talmudic and philosophical models of memory extends to analyzing other virtual elements of realities represented in Talmudic and philosophical texts. If so, the question to ask is how modern ideas about virtual agencies might help understand both the construction of virtual reality and the forgetting about the virtual nature of that reality by its readers and researchers. I address this question in the next chapter by juxtaposing Halivni’s theory of the Talmud’s redaction to Sergey Eisenstein’s theory of film montage.
ten
The Talmud as Film I live with the ironic awareness that the very mode in which I delve into the Jewish past represents a decisive break with that past. —yosef hayim yerushalmi, zakhor , jewish history and jewish memory
It is the sorting out that makes the times, not the times that make the sorting. —b runo latour, cited in kathleen biddick, the typological imaginary: circumcision, technology, history
The virtual in the Talmud is not only an agency embodied in remembering but also the reality remembered. How is that remembered reality created and what exactly does it entail, if compared to the reality dealt with in philosophical thinking? This question is best asked by way of contrast to Cartesian thinking, which pointedly did not require embodiment. By extension, the contrast applies to post-Cartesian thought as well. Even though post-Cartesian philosophers1 underscored the importance of embodiment for thinking, they invariably did so by criticizing, and thereby building on, the Cartesian thinking subject. What is more, in approaching reality, post-Cartesian thought continued the traditional philosophical opposition of being and nonbeing; even phenomenology, as radical as it was, only bracketed that opposition while still leaving it untouched. Yet, if Talmudic compositions are, for the lack of a better word, memoirs, or more precisely workings of memory, the reality they encounter is the past. Unless misunderstood as praesens historicum, that past does not have to be defined in terms of presence, and in particular of absence; the past in Talmudic memoirs should not be misconstrued as “absent now” 212
The Talmud as Film 213
but “present before.” Excluded from the realm of presence, and in particular of absence, the past of the memoir can therefore be real only in the sense of virtual. Otherwise, it is no past. In which sense, then, do the Talmudic memoirs engage with the past as reality? If not presence, being, or existence, what exactly makes the remembered past real? To answer, and even to ask these questions with due precision, one needs a theoretical lens. In this chapter, I construct that lens with what I will immediately introduce as an anthropological inquiry. To that end, I borrow the modern notion of film montage and ask about the anthropological foundations of the construction of the virtual reality in the Talmud. The question concerns the anthropological foundations of the virtual in the strictest sense of the term, that is, in terms of how human being per se is understood as both a part and a foundation of the virtuality of the real and of the reality of the virtual in the Talmud. I approach these foundations through rereading Halivni’s theory of “redactors.” The chapter uses contemporary philosophical and anthropological theory to discern and eliminate tacit, but restraining, anthropological assumptions in his theory, and thereby to arrive at the foundations. I consequently reread his theory of redaction as a theory of montage. The resulting theory describes the topology of the virtual world of the Talmud and places the Talmudic processes and agencies of thinking and remembering within the context of that topology.
a Since their advent at the beginning of the twentieth century, the philological, literary, and, later, cultural study of the Talmud all have subscribed to the then-unquestionable ideal of scientific strictness and rigor that both the natural sciences and the humanities were expected to share. Thus, the new “scientific” philological study of the Talmud reserved the task of Talmudic interpretation to the traditional study of the Talmud, while it limited itself to a rigorous “scientific” analysis of how the text of the Talmud changed over time. The result was the birth of what I have discussed in the chapters above under the name of Talmud criticism. Shamma Friedman defined the separation between philological and literary analysis in the formula “let researchers research, and let interpreters interpret.”2 In contrast, as we have seen, the scientifically rigorous literary interpretation of the Talmud, for which Chanoch Albeck3 and later David
214 Who Remembers?
Halivni4 provide paradigmatic examples, was more interested in interpretation of the Talmud as a historical product of the literary activity of historical groups of people called stamm’aim, or the anonymous (and otherwise historically unknown) redactors of the Talmudic traditions. Literary analysts refused to look at the Talmud as simply and merely the historical evolution of the textual object of philology, but still insisted on historic chronology as a protocol of interpretation. As different as the literary and philological approaches are, they both have subscribed not only to the scientific ideal of rigor and strictness but also to the same ramification of that ideal in the criterion of chronological accuracy known as the avoidance of anachronism. In this chapter, I explore what this conflation of the ideal of scientific strictness with its ramification in chronological thinking both positively affords and also negatively affects. A heuristic contrast between Christian chronology and Jewish genealogy plays an important role in my analysis.5 As I noted, contrasts do not always mean mutual exclusions. Genealogical accounts convert themselves into chronologies in a smooth and persuasive way. Yet, some important things are easily lost in this conversion, and I am interested here in what is lost. In genealogies, their hierarchies are inseparable from their content.6 In contrast, chronological time can work as a pure form of a graphically represented progression and in this sense it is an empty representation that can therefore be filled in with any content, which can remain completely foreign to the line of time. The relationship between content and its chronology is like a birthday present and the wrapping paper in which it comes. The larger task behind my current investigation is an inquiry into what the difference between chronological form and the content it supplies means in the Talmud as far as chronology, genealogy, and different orders of time are concerned.7
The Addressee of the Talmud As early as the tenth century, in a genealogical account of the orders of authorities in the Talmud, The Igeret of Rav Sherira Gaon,8 the Talmud no longer appears as an ongoing process or living tradition but rather arises as something that, at least in a certain sense, had been finished, finalized, brought to a full stop, or formally closed to additions. Before that symbolic act of finalizing or “closing,” even the word “Talmud”9 must have meant
The Talmud as Film 215
something different, referring perhaps to “learning” in a sense not totally different from what was otherwise called midrash, a heterogeneous set of methods used for expounding the Written Torah without any direct subscription to the project of Mishnah as the Oral Torah.10 At any rate, sociologically, that symbolic closure of the Talmud was at the same time its first emergence in the completely new sense of a medium or even a source of authority that its adepts both created and used in their polemics with the Karaites, a sect recognizing the exclusive authority of Hebrew Scripture or Written Torah, and denying the authority of the Oral Torah of the rabbis. Although the Karaites claimed that the art of grammar provides a sufficient way to access, read, interpret, and implement the Written Torah directly, without the authority of the sages of the Mishnah and their Talmudic followers, the rabbis reestablished and reaffirmed the genealogical reliability, value, and consistency of their own authority by relating it to a past that has been marked, locked, secured, closed, and protected under the name of the Talmud. At that symbolic moment of closure, beginning from the Igeret, and through the subsequent generations of the rabbis, the Talmud has been genealogically associated with its own internal past in which, from the moment of the closure, the Talmud will always be understood to be born. Thus, for Rav Sherira, authority is not merely an external reason, in support of which, in his polemics with Karaites, he systematically retrieves the genealogy of the Talmud. His retrieval also is a way of securing the Talmud as a fait accompli of a past that can no longer be changed. More specifically, he undertakes a genealogical reconstruction of the orders of the sages of the Mishnah and their disciples, תלמידי חכמים, literally “the disciples of the wise.” In his genealogy, he effectively and almost transparently converts a rhetorically heterogeneous set of titles and references in which the Talmudic masters addressed the Mishnaic sages into homogeneous genealogical ranks of authority. The intra-Talmudic set of references that he transformed included the “rabbanan” / “h.akhamim” or “our teachers” / “the sages in the Mishnah,” “tann’aim,” that is, the sages themselves, or those who repeat their teachings, and the “amor’aim,” or the adepts of the sages, who in their orations creatively speak on what the sages taught. Before the symbolic closure of the Talmud, the word tann’aim, or, in the singular form, tann’a designated not only and not specifically the sages of the Mishnah, who perhaps more consistently were referred to as “rabbanan,” or “our teachers,” but also referred to trainees who excelled in mechanically memorizing the words of
216 Who Remembers?
the sages, but who themselves were not necessarily wise, only mnemonically gifted. This duality of the term and of its usage expressed a duality in the attitude that the amor’aim had toward the authority of the sages: The sages themselves were for them teachers who at the same time did not necessarily know (or at least explain) what they said. Rav Sherira put an end to this duality of attitude and ended up with a more homogeneous, less ambiguous, and, for that price, more organized and less questionable system of authority, which he called the order of tann’aim and amor’aim. His new genealogical system of reference represented simultaneously a lineage and a hierarchy of the generations of the teachers of the Mishnah, followed by both the generations and the ranks of the masters of the Talmud. One outstanding result of his revisitation of the Talmud was that the Talmudic term tann’a lost its complexity, no longer referring to both mechanical repeaters of the teachings of the sages and to the sages themselves. Instead, the ambiguity of the term was finally resolved in favor of the straightforward designation of the tann’aim as the sages of the Mishnah themselves, as opposed to the amor’aim, their successors and the recipients of their authority. Besides distinguishing the tann’aim or sages from those who repeat their teachings, Rav Sherira isolated two other genealogical categories: the masters of the Talmud, or amor’aim, and their less authoritative, but perhaps no less rhetorically gifted followers, or savor’aim, whom he also respectfully titles “our teachers” (rabbanan savora’i). In this tripartite partition of the symbolic genealogy of the Talmud, the Talmud comes from its own past in the three main orders of tann’aim, amor’aim, and savor’aim, a genealogical continuity of traditions that carefully masks a rupture in its symbolic time. For if thought of in terms of chronological time, rather than generations, the Talmud could never have existed before the symbolic act of “its” closure. However, after the closure and as a result of it, the Talmud acquired not only its future (with Rav Sherira as a part of it), but also (re)possessed its own internal past, a past that the Gaon genealogically (re)constructs. Rav Sherira thus began a complicated symbolic game that subsequent scholars of the Talmud in almost all but the most recent generations of the Talmud’s reception have had to play. If, following Rav Sherira’s thinking, we also look at the Talmud as if it were something symbolically finalized, finished, completed, and closed to additions and therefore open for both paraphrase and commentary, we can
The Talmud as Film 217
see the Talmud as a long, but complete phrase, something that has already been said. There is a general question that could be asked about any phrase, as Jean-François Lyotard proposes.11 With respect to the Talmud, the question is: To whom has the Talmud been addressed? Whom does the phrase of the Talmud, either the whole Talmud or any of its articulate parts, or sugiyot, anticipate as its addressee? Here, the time of the symbolic closure of the Talmud, its constitution as a single phrase, sets up for us an even more complex game. The addressee in question can no longer come from the Talmud’s past, that is, from the time before the Talmud was “closed,” for in order for a phrase to have an addressee, it must first come to its end or be, at least symbolically, finished. Therefore, in that sense, before “its” official closure, there was no Talmud at all, much less could there be the addressee for whom we search. In other words, there was no such thing as the Talmud before the Geonim declared it “closed.” Similarly, before the closure, no individual discussion (sugya) appeared as a permanently finished composition; instead, it functioned as potentially infinitely open, inviting new additions, extensions, or even more sophisticated transformations of the text d as the text circulated among also potentially infinitely changing audiences of rabbis and their students as both addressees and participants of discussions. If the Talmud’s addressee cannot come from the historical past before the Talmud was “closed,” then, as genealogy would suggest, the addressee would have to come from the Talmud’s internal past, a past to which only the Talmud itself is the witness. This seems to be the option that the twentieth- century scientific analysis of the Talmud takes for granted in researching the Talmud’s internal genealogical past as if that past were translatable into the language of history. In what follows, I analyze and evaluate this assumption and ask: If we understand the economy of the symbolic closure correctly, should we still look for the addressee in the Talmud’s past, either traditional or historical, or should we look for him or them not in the historical past, but in the future shaped by the symbolic (and thus not historical) closure from which the Talmud emerged? Put simply and briefly, my question is therefore this: Who is the addressee of the phrase that the Talmud became? I slowly and patiently begin approaching that question by analyzing the contemporary scientific theory of the Talmud, not to criticize that theory, either destructively or approvingly, but rather to release its heuristic power for approaching the question of the addressee of the Talmud.
218 Who Remembers?
History as a Protocol of Interpretation Today, historians discuss the question “Could the Talmud serve as a historical source?”12 Many incline to a negative answer to that question.13 I would like to ask a slightly different question: Does history, broadly conceived, provide a fully sufficient hermeneutical perspective for interpreting the Talmud scientifically? I focus my inquiry on an example of the scientific interpretation of the Talmud as expressed in the twentieth-century theory of Talmudic redaction, already introduced contextually in the chapters above. That theory posits the Talmudic redactors for whom the scientific theorists of the Talmud invented a new term, stamm’aim—literally, the anonymous redactors of the arguments of the earlier Talmudic masters, or amor’aim. Below, I argue that the stamm’aim theory inherited a very strong set of metaphysical premises, found perhaps for the first time, as far as the Talmud is concerned, in Lutsatto’s The Ways of Reason.14 As discussed, Lutsatto’s work interpreted the Talmudic “give-and-take” as a case of the dialectics of universal reason, recorded first in the mind of the Talmudic masters, then in the memory of their students, and finally on parchment and paper. Lutsatto’s agenda for Talmud study aimed to undo the disturbingly multifaceted forms of that multilevel history of recording and to help a student of the Talmud return to an allegedly original dialectical reason that, as Lutsatto shows, appears from behind the pages of the Talmud. A century later, in the cultural-historical schools of the Wissenschaft des Judentums, Lutsatto’s view of the Talmud as an instance of universal dialectical reason was supplemented by the idea of both the universality and the linearity of historical/chronological time,15 a time in which the Talmud, like everything else, was thought to be immersed. Finally, in the twentieth century, these two metaphysical premises—universal “dialectical reason” as expressed in the Talmud and the Talmud’s full compliance with chronological historicity—were connected to the idea of scientific strictness, resulting in the twentieth-century scientific theory of the Talmud’s redaction.16 Drawing on the example of that theory, with all the metaphysical premises it involved, I argue that the historical perspective of interpreting the Talmud should be supplemented by a strategy of interpretation that could be preliminarily designated as both rhetorical, for it deals with the rhetoric of the Talmud as a phrase, and ontological, for it analyzes the addressee that phrase brings into being.
The Talmud as Film 219
As I have argued, the rhetorical strategy of the Talmud’s interpretation, or, as an early modern Talmudic thinker, Rabbi Izh.ak Canpanton expressed it, the intellectual practice of Talmud as an art or techne similar to the art of rhetoric and also distinct from it (as opposed to the Talmud as an object of either history or tradition) was announced, perhaps for the first time, no later than 1463, in his work on the Talmud’s hermeneutics and rhetoric, The Ways of Talmud.17 However, nowadays, the intellectual domain that Canpanton discovered is also accessible through an approach that is chronologically closer to the twentieth-century theory of the Talmud’s redaction. This approach can be found in the early modern theory of filmmaking. Developed in the first half of the twentieth century, that theory bears on concepts of film montage18 that are in a way similar to the central concept of the scientific theory of the Talmud, the concept of stammaitic redaction in which the Talmud, as we have it today, was produced. A radical difference between editing a film using montage and editing the Talmud using redaction or, as David Halivni calls it, “conjecture,” lies not in the concepts of montage and redaction themselves, but rather in the metaphysical premises they imply. Film theory suggests that in filmmaking, instead of only one as it were “historical” time in which the film (or, by analogy, the Talmud) has been produced, there always are two other lines of time—the time of the representational content of the film, otherwise called its diegesis, and another line of time pertaining to connecting the film’s cuts together, otherwise called the film’s montage. A basic sense of the difference between the two times can be shown by one quick example: What in a film’s diegesis could last for days, years, or centuries, in montage can take only seconds. This difference between the time of representation and the time of a represented action is not totally unlike what has been already achieved in both the theater and the novel, beginning perhaps with Shakespeare’s plays19 and theorized in Bakhtin’s notion of the chronotope, as well as in the Russian Formalists’ distinction between sužet (plot) and fabula (story). Yet unlike works of literature created before films hit the screens, the dynamic of filmic montage and diegesis does not limit itself to the mere play of two or even more times. In films, the difference in times also results in the creation of completely new diegetic characters that may emerge in the eyes of a viewer, regardless of the director’s initial will or intention, but that can later be either controlled and manipulated or even intentionally made. The mere connection of cuts, even if they are unedited, can produce completely
220 Who Remembers?
new personae on the screen, new events in the mind of the film viewer, and even new ways of thinking, feeling, and looking at things that the original cuts of film never had, an effect called after the name of its inventor—the Kuleshov effect. Yet the difference between the time of the film and the time of montage affords Kuleshov even more. The Kuleshov effect includes more than using montage to create events which were never filmed. It also celebrates the emergence of a new gaze—the gaze of “the camera,” which, in his words, “explores the planet” (“Kamerah shagayet po planete”). Montage makes the viewers assume that new gaze of the camera as a new way of looking at things in the world. As a result, “characters” produced by the film montage include both (1) conventional characters, with their minds, bodies, and behaviors, portrayed in texts and on screens; and (2) human existence embedded in the text or in the film, but not described in it, just as a camera gaze is not described in a photo, but is nevertheless embedded in it. This human existence or gaze is a site occupied and “manned” at least twice, once by the photographer, and another time by the viewer, the photographer substitutes on her place. Similarly, the gaze of a film viewer as embedded by film montage is never directly described in the film, and it is also occupied twice: once by the “monteur” (to borrow the French term for the film editor) and another time by the viewer, whom Kuleshov uses to man the gaze of the camera he develops. Montage thus designs new characters not only in diegesis but also in front of the screen, in the sense that the viewer assumes the gaze of the camera constructed by the monteur, for whom the viewer consequently substitutes. This gaze now controls not only diegesis, but also thoughts and emotions the viewer should have. In the Kuleshov effect, the gaze of the camera becomes yet another character produced by montage: an “unnamed observer” in the world of the film who also occupies the seat in the movie theater. What is more, due to the difference in times, as was first suggested by Sergei Eisenstein’s method of cross-cutting that I will discuss in greater detail, cross-cutting in films can do much more than only show new diegetic personae and events. It also can construct a new, well-organized space for the film viewer to occupy. That space, call it, in terms of movie theater design, a sit, has preemptive features. Once in it, the feelings and thoughts of the viewer are in constant check. The space preempts these feelings and thoughts from going anywhere astray, except for taking a path constructed by the film. That process of construction is so intense that one might even
The Talmud as Film 221
say the implied viewer is also its effect. This is because the feelings, thoughts, and emotions of that viewer are fully controlled. This space is in a dark room and in the comfortable seat of a movie theater, in which the viewer, as if on a roller coaster, goes through different, but predetermined turns of thought and feeling that the film has provoked, not only and not primarily through diegesis, but also through cross-connecting the cuts with no diegetic connections at all. These feelings and thoughts of the viewer go far beyond the events that the viewer directly sees on the screen.20 As Eisenstein demonstrated, and as an example below in this chapter will illustrate, cross-cutting controls not only what a viewer considers to be the events in any given frame on the screen but also what the viewer thinks and feels after following a sequence of otherwise disconnected frames.21 Yet, however different the times of montage and of diegesis are, both of these lines of time belong intrinsically to the film (or, as we will see, by analogy, to the Talmud), not to the external history of its production. They are radically different from external, historical-chronological time, toward which the scientific interpretation of the Talmud has always been oriented. This means that instead of only one line of time, the chronological line of history in which the Talmud has hitherto been scientifically interpreted, we should deal with two other lines of time, neither extrinsically historical, both intrinsically Talmudic: a line of diegetic time (either chronological or generational, but not historical) pertaining to the representational context of the Talmud, and a different line of time pertaining to the Talmud’s montage or “conjecture,” the task that Halivni first ascribes to the stamm’aim and then, in another, more scientific attempt, undertakes himself. Let me explain the two intrinsically Talmudic lines of time, its time of diegesis and its time of montage, as they relate to the stamm’aim theory. One or another version of the scientific theory of the stamm’aim, the redactors of the Talmud, is by now accepted by virtually everyone in academic Talmudic studies, although the existence of the stamm’aim is neither attested by historical sources outside of the Talmud nor represented in the Talmudic genealogies, but only revealed itself as late as in the modern literary and philological analysis of the Talmud’s text. Among many other versions of that theory, a version proposed in Halivni’s Sources and Traditions is particularly inviting for a case study, for it only partially complies with the film-theory distinction between the representational context of a film, or diegesis, and the devices filmmakers apply in the process of editing, or montage. This linkage to the theory of
222 Who Remembers?
filmmaking provides the possibility of freeing the stamm’aim theory from the historical question of the stamm’aim’s existence or of putting that question in what a mathematician or a phenomenologist would call “brackets.” Such bracketing allows us to look at the theory of the stamm’aim as not rooting the Talmud in the chronological time-space of its external historical past, but rather as inscribing the historical element in the Talmud, in its representational content or diegesis, as opposed to the devices of its editing, or montage. Despite his envisioning the stamm’aim as a historically distinct group of individuals, Halivni described the acts of the stamm’aim in the purely functional terms of the recording, adding, changing, and so on that the stamm’aim did to the sayings of the earlier Talmudic masters, or amor’aim, so if one still wants to use the term stamm’aim for the potentially open list of heterogeneous functions of the Talmud’s production that Halivni has so impressively categorized under the name of “conjecture,” the stamm’aim would therefore belong no longer to the representational content of the Talmud, but rather to the making of the Talmud’s montage, which is always as intrinsically intra- Talmudic (and only extrinsically historical) as any representational content that the Talmud conveys. To analyze the Talmud’s montage in connection with and as differentiated from the Talmud’s diegesis is similar to and different from what Halivni has both envisioned and fulfilled. Of course, he has carried out this task in the historical perspective, not in the filmic perspective in which it intrinsically belongs. My analysis will thus work toward a new perspective to be opened not only on our hitherto scientific understanding of the Talmud but also on our hitherto philosophical understanding of rhetoric. Let me now briefly delineate the broader perspective in which Halivni’s philosophical premises of the linearity of chronological time, the orality of signification, and the priority of dialectics over rhetoric will be mapped.
Logic, Dialectics, and the Stamm’aim Theory The dual (and unstable) relationships between the Talmudic, philosophical, and rhetorical traditions were completely elided in the scientific, almost positivistic view of the Talmud shaped at the beginning of the twentieth century in the “science of the Talmud,” which was a continuation of the nineteenth- century project of the Wissenschaft des Judentums, adjusted to new notions of
The Talmud as Film 223
scientific strictness developed in the beginning of the twentieth century (for example, in natural sciences as well as in logic and phenomenology). The “science of the Talmud” revisited the traditionalist hermeneutic approaches to the Talmud, but redefined the Talmud as an object of science. Conceived at the beginning of the twentieth century, the project of the “strict science of the Talmud,” of which the theory of the Talmud’s redaction was an important part, belonged to a larger current of research in the humanities seeking methods and standards of scientific rigor comparable to those understood to govern mathematics and the physical sciences.22 Against that background, the shift in the study of the Talmud from the nineteenth- century Wissenschaft des Judentums to the twentieth-century “strict science of the Talmud” or from a “scientific” approach returning traditional texts to their history to another “scientific” restoration of these texts via philology meant that historicization was no longer a condition of the scientific analysis of a traditional text, but rather became a function of the scientifically strict analysis of the texts themselves. That shift, from the prevalence of historical contextualization to the strictness of philological analysis of the text, as revolutionary as it was, was nevertheless no surprise, nor was it astounding that the new “strict Talmudic science” shared metaphysical and more specifically ontological premises that all the other “human sciences” had in common, such as the axiom of signification (“oral signs come first, written signs come second”)23 and the axiom of time (“time always goes from the past via the present to a future” and is itself “timeless”),24 as well as an undeclared, but strong philosophical preference for the “apodictic” and strict disciplines of logic and dialectics (scientific philology included), as opposed to the only “probabilistic” or (even more dismissively) “eloquent” art of rhetoric. The true surprise, however, is that while all the other strict human sciences underwent a period of critique of their own metaphysical premises, the strict science of the Talmud did not, or, at best, did so in a much lesser degree. Consequently, it seems to be a good time to rethink the metaphysical premises of the strict science of the Talmud. For the purposes of my argument, I will relate to the metaphysical premises of time, signification, and dialectics as if they were in a sharp contrast to the tradition of rhetoric as has been pejoratively envisioned in the tradition of philosophy. I will examine the metaphysical premises of the stamm’aim theory in the light of the rhetorical form of what the scientific study of the Talmud has qualified as the stammaitic arguments.
224 Who Remembers?
Speaking intra-Talmudically, in terms of the Talmudic diegesis as Halivni reconstructs it, the arguments that the stamm’aim “recorded” and “redacted” originally belonged to the sages of the Mishnah and to the early Talmudic masters, or amor’aim. An advantage of Halivni’s version of the stamm’aim theory is that, perhaps unlike other theories of Talmudic redaction, his project seems not only to subscribe to the above metaphysical premises in full, but also to remain open, if only potentially, to a productive revisiting of them. But let me first establish that Halivni’s theory complies with the metaphysical premises in question, that is, that it obeys the axioms of time and signification, as well as following the preference automatically given to dialectics over rhetoric. Halivni assumes that the stamm’aim recorded what they initially “had” as the oral arguments of the amor’aim. He then conventionally qualifies the resulting exchanges of arguments in the Talmudic “give-and-take” as dialectics. However, what that qualification dismisses, as I have argued above, is that, first, the “record” might have been an act of innovative rhetorical invention (h.iddush), rather than a somewhat repetitious act of recording, and second, the result, “the give-and-take,” might have been a rhetorical exchange of refutations and defenses for the sake of getting the records at hand authorized by something more reliable than the mechanical memory of a tann’a, who is supposed to recall teachings, not to authorize them. As I have argued, the purpose was not merely dialectical competition for the sake of what dialectics always aims at—the victory of one tradition over another—but rather to provide the traditions at hand with the authority of something that was tested for rhetorical refutability25 and that had passed that test. While complying, in its present form, with the metaphysical premises of time, signification, and dialectics, in the light of the arguments that I have been making in this book, Halivni’s project seems to be, at least potentially, open to a new, perhaps more productive, but surely no less strict reformulation that arises from a revision of the above set of premises. Step by step, I will undertake that revision here, beginning from the broader context of his version of the stamm’aim theory and developing it further in comparison with the film theory, to which I have already referred.
From Talmud to Film and Back Let me begin where I left off, with the ontological dimension of the strict science of the Talmud that still informs Halivni’s approach—the axioms of
The Talmud as Film 225
time and signification and the automatic and strict preference given to logic/ dialectics over rhetoric. The axiom of signification says that the oral signs come first and the written signs come second, and the axiom of time establishes that time is always a line that always goes in one direction, from the past via the present to a future. At first glance, these axioms are not connected. However, even a slightly closer look reveals that without the linear progression of time, oral signs could not have come first, nor could written signs have come second. Nonetheless, the linear progression of time cannot happen without the process of signification. Indeed, the movement from past via the present to the future would be completely out of reach unless a historian revealed that movement for us through a reverse movement in which he or she reads the signs of history, thereby reading the historical past from what is available to us in the present.26 He or she reads whatever is available in the present to reveal the past that precedes it—reads the history just as anybody else would read written signs for the sake of their oral predecessors. Thus, tracking the historical movement of time is nothing but a semiotic process of reading and as such follows the axiom of signification. In short, in that nontrivial sense, the progression of time is exactly the reverse of the process of reading history and is thus complicit in the process of signification. Both oral signs and history belong to the past; both written signs and historical data belong to the present. Thus, the axiom of time and the axiom of signification are not two independent axioms. Rather they are two formulations of one and the same axiom, one in terms of signification and the other in terms of time. Therefore, in this sense, time and signification relate to each other neither as two completely separate kinds nor as a species subordinated to its own kind, but as two coextensive, even if quite heterogeneous definitions of one and the same kind with no other, more general name than these two names, “time” and “signification.”27 Having gone from time to signification and back, we are now closer to the third ontological premise of the strict science of the Talmud, the preference given to logic/dialectics, as opposed to rhetoric. Indeed, if linear time and oral signification underwrite each other and in that sense belong to one and the same kind with no more general name for it, that is possible only because both follow one and the same logic seen from two perspectives, as the logic of the sign (the oral comes first, the written comes second) and as the logic of time (progression from the past via the present to a future). In sum, if both chronology and semiology belong to one and the same kind with no specific
226 Who Remembers?
general name for it, this is because, as the terms “chronology” and “semiology” suggest, they represent two different, but not too different versions of logic, a logic of time and a logic of signification, and are therefore opposed to what is not logic. Since traditionally there is only one nonlogic—rhetoric—both chronology and semiology oppose rhetoric and ally themselves with all to which rhetoric was traditionally in opposition: logic, dialectics, and science. This general explanation of the connection between chronological time, oral signification, and the logical derogation of rhetoric prepares us for the next, more specific step. Since there is a dimension in which the historical progression of time, the process of signification, and logic/dialectics (as opposed to rhetoric) is the same, it makes full methodological sense to ask how the logic of time (chronology) and the logic of sign (semiology) in their alliance with dialectics connect to each other in analyzing one specific object of the strict sciences, the Talmud. My way of asking that question here is to look at the theory of Talmudic redaction in the stamm’aim theory as it was shaped both in compliance with and in revolt against Sherira Gaon’s traditional tripartition of the Talmud in the orders of tann’aim, amor’aim, and savor’aim that, beginning from the early medieval literature and on, dominated all prescientific and nonscientific receptions of the Talmud. While introducing a new, purely, and strictly scientific concept of stamm’aim, the scientific research of the Talmud tried to find the chronological place of the stamm’aim in the midst of that traditional genealogical partition, even if, beginning from the aforementioned Igeret of Rav Scherira Gaon, the partition was built not on the ascending chronological order of time, but rather on the descending genealogy of authority and only extrinsically coincided with the logic of time or chronology. What did this new intrusion of scientific chronology into traditional genealogical partition of the Talmud both afford and affect? A further comparison with film theory will help us address that question. In Halivni’s account, the “conjecture” the stamm’aim did, occurs via the following list of functions: adding, changing, extending, and, finally, filling in the gaps.28 That list of functions is not foreign to what a film theorist would include under the broader category of montage, which produces new representations, symbolic visions, or even whole new characters, not only in diegesis but also embedded in the film as the gaze of the viewer. By means of cross-cutting, montage puts together film cuts that represent not only different spaces but also different times and even epochs. It thereby gives
The Talmud as Film 227
the viewer completely new feelings and thoughts that the diegesis of the film cannot convey. Yet diegetically, they produce nothing, not even a new character on the screen. For example, in Sergei Eisenstein’s Alexander Nevsky (1938), a frame of Teutonic knights’ soldiers looking eastward to conquer Russia immediately associates with other frames in the then current documentaries of Nazi soldiers who are in the movie, if not yet in history, looking in the same direction and pose (in both the same geographical direction and the same direction across the surface of the screen). This does not create a new diegetic event, but rather generates a new viewer, that is, a viewer with a new sense of the continuity of historical time. For an example of even closer, and thus less associative montage, in his October, a sequence of a still life of Prime Minister Kerensky’s boots, the moving head of a mechanical peacock, Kerensky’s gloves, the peacock’s turning en face again, and a back view of the prime minister as he enters the door connects the frames one after another in a visual composition. Yet this creates no diegetic event whatsoever, but only a tragic-comic comparison between the prime minister and the mechanical peacock. That comparison, or, I would say, an almost Hegelian speculative statement, “Kerensky is a peacock” happens not on the screen, but precisely in front of it, in the space that the viewers both inadvertently occupy and intensely experience as they become the revolutionary proletarians that the film constructs in front of the screen. In contrast, Eisenstein’s earlier pieces of montage create their effects on the screen only, not in front of it. Recall the first frames of Battleship Potemkin, a both peaceful and disturbing romantic moving seascape of waters striking the shore. The waters induce the feeling of both uproar and worry that then immediately transfers to the next frame, the battleship in the sea, and then directly to the feelings of its sleeping but awakening crewmen, so that the viewer, with no words said, unmistakably sees what they feel. Both the seascape and the feeling of uproar completely belong to the diegesis of the film. Unlike the seascape in Potemkin, the peacock in October has no direct diegetic meaning whatsoever; it only helps create and direct the viewer’s feelings toward the diegetic image of Kerensky greeting his ministers. It shows not at all what Kerensky feels, but rather what viewers are to think about him. The place of the subject thus is no longer on the screen, but rather in front of it. The viewer is that subject, now addressed directly, as he, she,
228 Who Remembers?
or rather they are made to feel themselves in the position of a revolutionary proletarian, if not then in the diegetic time of 1917, then now, at least eleven years later, at the time of the film viewing. Montage thus creates a new viewer’s gaze full of a proletarian sentiment about the heroic past of the revolution without creating any new diegetic events on the screen. Similarly, a man (seen only via a shot of his head) goes from the far right of a frame to the camera’s closest left, while in the next moment, the viewer sees a woman in a different landscape and in a different diegetic time going from the far left of the frame to the camera’s closest right. For the film viewer, a simple connection of these two different frames in one single cut produces the effect of the man and the woman (knowingly or not) hurrying “to meet” each other, despite the geographical and chronological distances. Although this does not occur necessarily diegetically, it surely does so “existentially,” that is, in the feeling or thought of the viewer, who is, of course, the main character produced by the film. The most interesting result of such a device is the unique positioning of the viewer, who, due to the cross-cutting, is neither present nor absent in the film’s diegesis, but is certainly welcomed and even expected to occupy, or (as I have just said to myself) “to man” a preemptive position that the cross-cutting has prepared for the viewer: to see something that none of the film cuts could diegetically represent, or imply. In so doing, the viewer becomes the product of the Kuleshov effect, the locus of a performative gaze present not on the screen, but rather embedded in it, as if coming from the movie theater’s dark room. Montage thus creates a preemptive position that we as viewers are expected to take when we come into a movie theater, but not only then. Perhaps even more frequently, we fill this preemptive position of the viewer who not only sits, sees, and hears, but, as we will soon see, as someone who also thinks and feels when we are outside of the dark room, in the lights of a city sidewalk or behind the wheel. Sergei Eisenstein called the filmic montage in general and the cross- cutting in particular “ustroistvo vsech veshchei” (literally, “the creation of all things”). “All things” means, of course, not only objects, but also anyone who thinks of them, feels them, hates them, loves them, or in a word, experiences them. The human being or viewer’s self that cross-cutting generates is welcomed in the film without being either present or absent in it. And this unique situation is not totally foreign to yet another montage, which was produced, as Halivni describes it, by the functions of stamm’aim.
The Talmud as Film 229
Montage not only welcomes or makes a viewer occupy (or man) a preemptive position that the film has prepared, it also makes the viewer redefine that position and perhaps reveal, as film critics do, a new perspective that neither diegesis nor cross-cutting could “originally” convey. One result of this redefinition of a viewer’s presence/absence in the filmic diegesis consists in what a Talmudist would call making an invention in the Talmud, h.iddush, or an invention in the Scripture, midrash, developments that show things that “were there but have not yet been seen.”29 However, to explore further the significance for understanding the Talmudic economy of invention as a process in which, as in montage, the “viewer” is captured and/or created, I need first to investigate the ontological extension of the notion of montage that the notion received beyond the immediate scope of the film theory. Because of their rich ontological connotations—Eisenstein’s “ustroistvo vsech veshchei” (“the creation of all things”) is just one example—it is only natural that in their intellectual trajectories, the notions of montage and diegesis have gone far beyond films. Although born in film theory, the theoretical distinction between montage and diegesis has had a strong influence in other fields of the humanities. The most obvious (and productive) extensions of it proved themselves in literary theory, in which they were adopted for analyzing relationships between devices of writing and the representational effects they produce. However, perhaps the most methodologically powerful influence of this “filmic” approach is the light it sheds on our understanding of the mechanisms of representation in general. An analysis of the mechanisms of montage and especially of cross-cutting has opened up a new field in the ontology of representation, also called political ontology. Developed in the works of such theorists as Gilles Deleuze and Jean-François Lyotard the theory of political ontology has taught us that not only movies or literature but also other political or even intellectual constructions of reality and even human selves have their being-there formed on the basis of either a controlled or more frequently an uncontrolled montage or cross-cutting. As I have already noted, on its ontological level, cross-cutting makes very persuasive and, for the viewer, unbreakably “realistic” representations of things that otherwise are not or cannot be diegetically represented. That includes a “realistic” representation of the viewer to himself or herself in the aspect of being-there. Montage defines not only things represented but also human beings encountering those things and themselves in that aspect.
230 Who Remembers?
Rereading Halivni, we can now look at the Talmud as both a result and a process similar to such filmic, literary, or even political ontological cross- cutting. In more general terms, what would the Talmud be in terms of its intellectual and emotional production, as opposed to what it is as a representation? More specifically, what would the Talmud be in terms of its diegesis, as opposed to its montage, or cross-cutting? The methodological novelty (and advantage) of this approach is that it at least temporarily sets us free from the retrospective and therefore always potentially anachronistic investigations of the Talmud in any external positive, historical-chronological, or “cultural” contexts it might have. Unlike the externalism driven approaches that positivistic historicism exercised, the filmic approach places history, if any, in either the diegetic time of the Talmud or in its cutting, rather than placing cutting or diegesis in the space of history. Just as its two intrinsic parts, the cross-cutting and diegesis of a film, always come together in the film itself, not as external circumstances of its production, the cutting and diegesis of the Talmud always come together with it and in it. The two make the Talmud what it is, regardless of any external historical-chronological circumstances that may or may not have been involved. In other words, since not only diegesis but also the cross-cutting behind it always come together with the Talmud itself, only a strict filmic analysis can access the cutting of the Talmud without dependence on any external historical-chronological circumstances or historicist explanations. In fact, with his emphasis on the heterogeneous functions that the stamm’aim performed, Halivni exemplifies that filmic possibility really well. Therefore, my program of reading Halivni is to undertake a filmic and by extension literary and political-ontological analysis of the Talmud in terms of its diegesis and cutting. However, since the Talmud itself is clearly not a film, I should naturally proceed in such an analysis not by creating a direct analogy to film, but rather in the literary mode of film theory and then in the politico-ontological dimension of it.30
Halivni’s Literary Analysis of the Stammaitic Arguments As Halivni’s statement that the stamm’aim “had to rely on conjecture,”31 followed by his own scientific reconjecturing it anew, teaches us, the literary form that pertains to the Talmud par excellence is the form of the
The Talmud as Film
231
stammaitic arguments, the arguments that the stamm’aim “recorded” and “redacted.” Other literary forms involved in the Talmud, such as the Mishnah and midrash have their own independent life spreading far beyond the framework of the Talmudic arguments. It goes without saying, however, that unlike its other components, such as the Mishnah and even Amoraic traditions that may also have existed independently of the Talmud, the Talmud as an intellectual event (not to say as a historical object) does not exist without the stammaitic arguments that along with those of savor’aim shape its main discursive frame. Therefore, it is the issue of stammaitic and not any other arguments in the Talmud that should be at the center of any filmic analysis of the Talmud.32 In his Sources and Traditions, Halivni has quite clearly although not programmatically followed the above methodological distinctions of film theory. In an effort to ensure the compatibility of his and others’ new scientific notion of the stamm’aim with the medieval (Geonic) notion of the genealogical orders or ranks of tann’aim, amor’aim, and savor’aim, he divides the stamm’aim in two distinct groups according to their different functions. He calls one of the groups the “recorders” and the other group the “redactors.”33 However, the most important innovation is the emphasis on the differences between the redactors (stamm’aim) themselves, the anonymous recorders (stamm’aim ha-tann’aim) who, over time, transmitted the dicta of the masters of the Talmud (amor’aim), and the anonymous redactors (hastamm’aim ha-orkhim), who tested and investigated the explanations and arguments that caused the sayings and who attached explanations and arguments to them.34 The less significant, but better-known group, the recorders, fits into the traditional orders of the Talmudic masters smoothly enough without any disturbance, even translating itself nicely in intra-Talmudic terms, such as ( תנאreciter), ( תנןwe recited), and so on. However, the much more significant group, the redactors, does not have any assured anchor in the old ranks and can therefore be potentially projected onto different segments thereof, resulting in different scientific versions of the stamm’aim’s place . . . not in these specific genealogical ranks, but in a more abstract chronology, sometimes (although not in Halivni’s case) even without any insistence on associating the stamm’aim with any narrow chronological period of historical time. Although Halivni was consistently trying to isolate the chronological place of the recorders, other researchers made use of the concept of the redactors
fup-dolgopolski-10.indd 231
10/9/12 12:07 PM
232 Who Remembers?
even without having any clear chronological place for them, allotting for them just a chronological place, as potentially fixed as actually unknown. Without taking any stand between the different theories of placement of the stamm’aim in the ranks of the generations and chronologies (found in Albeck,35 Halivni, and others), I take the very fact of the sliding chronological and generational placement of the stamm’aim as opposed to the virtually unshakably firm positions of the generations of the tann’aim, amor’aim, and savor’aim in their ranks as a suggestion that we in fact may be dealing here with two heterogeneous orders of time, the progressive time of chronology and the descending hierarchical ranks of the generations and their authority. As I have noted, the order of generations easily, perhaps too easily, translates itself into chronology but not without losing its essential part, the hierarchy of the ranks by their authority, which also includes the possibility for a later adept of making his claim using the authority of an earlier master without being anachronistic. In other words, despite this tension between chronology and genealogy, to ensure a full convergence of the new, purely chronological concept of the stamm’aim with the traditional ranks of authorities, the science of the Talmud translates the descending time of authority into the ascending line of chronology, resulting in a compromise in the placement of both in the external time of history, even at the price of losing the dominant role of the hierarchical element of the traditional ranks. The stamm’aim do not claim any authority, unlike what the traditional ranks of tann’aim, amor’aim, and savor’aim do. To make the new concept of stamm’aim compatible with the old orders of tann’aim, amor’aim, and savor’aim, a scientist must translate or, better, convert the traditional format of the order of generations into the scientific format of the chronological order of time. However easily chronological precedence can pass for greater authority, unless authority is converted into chronology, the concept of stamm’aim can have no room whatsoever, because by definition, the stamm’aim have no recognized authority of their own. In addition to a difference in their chronological placement, the distinctive line between the first and the second groups of the stamm’aim has to do with the “conscious” or “unconscious” nature of the changes in the teachings of the amor’aim that the stamm’aim made. The first group makes only unconscious changes, the second—Halivni argues—is conscious of the changes. This is why very little evidence of their work was left behind for a modernist researcher of the Talmud such as Halivni. Here is how he introduces the
The Talmud as Film
233
opposition of conscious/unconscious and of visible/invisible changes in the sayings of amor’aim: In addition to the time of their activity, we should mention two other differences (two as one) between stamm’aim the redactors, and stamm’aim the recorders. . . . The first difference concerns consciousness of the stamm’aim about their activity. The second concerns the style of their work, as well as their own identity. The first difference is that stamm’aim the recorders, were not aware of their own activity. They did not feel that they extend, add, fill in, or change the sayings of amor’aim. . . . On the contrary, stamm’aim the redactors, were aware of their actions and knew that they had committed a change in redaction, for before them, [dialectical] give-and-take had not been transmitted to generations; in any case, not in an official or authoritative way. Their activity expressed itself in anonymous speeches and passages of which the Talmud is full. . . . The second difference is that activities of stamm’aim the redactors are easier to reveal. . . . They made a [n explicit] change and we can recognize their identity. Every giveand-take that comes after sayings of amor’aim is to be suspected of coming not from the amor’aim, but rather from the redactors.36
Because of their unconscious character, the works of the stamm’aim, the recorders, “have no explicit expression in the present text [of the Talmud], and we could only reveal their works by means of a deep and insistent analysis of the content of a Talmudic period [sugiya], and through knowledge of the sources that stamm’aim, the recorders used.”37 Conceptually associating ha-stamm’aim ha-tann’aim (the recorders) with an intra-Talmudic term “tann’aim” in general, Halivni links this group of the stamm’aim with what is mentioned in the Talmud directly as תנא, indirectly as תנן,תניא, and so on. According to him, these stamm’aim not only explicitly appear in the Talmudic diegesis as intra-Talmudic images but also—and even more significantly—go about their job of recording or echoing as a chorus, not as individual heroes. In short, thematically or diegetically present, or at the very least mentioned in their absence, the stamm’aim of this kind are explicitly there in the representational horizon of the Talmud. Unlike the recorders, the second group of the stamm’aim, the redactors, according to Halivni, are never mentioned in the Talmudic diegesis, not even by the work that they do, not even as people of unrecognized names. The Talmudic diegesis shows no trace of those redactors’ presence, nor does
fup-dolgopolski-10.indd 233
10/9/12 12:07 PM
234
Who Remembers?
it make even the slightest allusion to their absence. Yet, even though they are not mentioned in the Talmud at all, they are responsible, as he suggests, for the most important functions that any of the stamm’aim ever had. Halivni qualifies these functions as “conscious acts”38 of restoring, recording, adding the not yet recorded arguments behind the sayings of the amor’aim, and also changing the sayings themselves. However important and even central for the Talmud those redactors are, they have no diegetic image whatsoever, nor do they appear in the Talmud by any indirect mention. Only a scientist oriented toward the external logic of time more than toward the inner order of generations can reasonably detect their presence in the Talmud. There even is no intra-Talmudic term to which the stamm’aim orkhim, the redactors, can be associated, but nevertheless, this group of the stamm’aim is for Halivni much more important than the first one, and the enigma of their being there in Talmud without diegetic presence or even absence is most intriguing. It is not diegetic time but rather chronological time that gives him the key for a solution to this problem. To resolve the striking discrepancy between their absolute importance and absolute diegetic absence, absence even in the sense of absence alluded to, Halivni chronologically places these stamm’aim in the later, perhaps even Geonic generations, which are, for him, diegetically post-Talmudic, that is, postamoraic, but chronologically still “late-Talmudic.” Halivni not only gives a chronological solution to his hermeneutical problem, but also compensates for the tension between chronology and diegesis by connecting his concept of stamm’aim to the post-Talmudic concept of an anonymous tradition of the Talmud ()סתמא דגמרא. Explaining why there is a lack of intra-Talmudic diegetic terms for the anonymous redactors, Halivni associates them with what traditional post-Talmudic scholars, Geonim, and later medieval Talmudic commentators called stamm’a de gemarr’a, a term that connects by analogy to an intra-Talmudic term, סתם משנה, that in the Talmud refers primarily not to a reciter of the Mishnah, but to what or who is recorded in his memory. To ensure this association of stamm’aim orkhim with stamm’a de gemarr’a, Halivni, consistent with his main conceptual move, tacitly chooses to read the latter (the Geonic term) as referring not to either the amor’aim or savor’aim, whose names, as it were, merely did not happen to be preserved, but rather to the stamm’aim who recorded/restored the sayings of the amor’aim at the behest of the amor’aim themselves. Later, I will return to that choice. But for now, even the other choice, that of interpreting stamm’a
fup-dolgopolski-10.indd 234
10/9/12 12:08 PM
The Talmud as Film 235
de gemarr’a as referring to the stamm’aim, rather than to the amor’aim, whom the stamm’aim only “record” or restore, does not cancel the basic fact that the stamm’aim do not appear in the diegesis of the Talmud at all. In Halivni’s chronology-oriented approach to interpretation, despite all the tension between chronology and interpretation, he still uses the former as a means for the latter. However, if we translate (and transform) his approach into the language of film theory, the tension disappears. Halivni draws the distinction between the main event of intra-Talmudic diegesis, the movement of discourse from the sages of the Mishnah to the amor’aim and then to the stammaitic arguments, and the montage that the stamm’aim made in order to have this diegetic movement happen. In his description of the Talmudic diegesis, the recorders officially memorize and repeat (and, in that sense, record) the teaching of the sages of the Mishnah, followed also by the recording (and rearranging) of the sayings of the amor’aim. In turn, the redactors (stamm’aim orkhim) supplement these records with brand- new records of hitherto officially unrecorded arguments that, in diegetic time, the amor’aim always already “had.” Halivni thus persuasively shows that the main format of the Talmudic representation would not be possible without various kinds of montage or the cutting and rearranging that the stamm’aim made. The question that bothers him is this: If there was indeed a “third round” in the text—with Mishnah and Gemarah being the first two—why didn’t the redactors appear in it? To answer this question, Halivni once again has to assume a difference between the stammaitic conjecture/montage and diegesis. However, because he conceives this difference historically-chronologically and not purely functionally, he goes looking for the stamm’aim as a historical group of people acting outside the Talmud,39 in the space-time of its history.40 Halivni’s theory thus both translates into the internalist language of film theory and refuses such a translation. Yet, if freed from its obligation to scientific chronology and associated instead with other philosophical, rhetorical, and hermeneutical approaches to strictness, his theory translates quite straightforwardly into the terms of diegesis and cross-cutting, with the redactors, or stamm’aim orkhim, who, unlike the recorders, or stamm’aim tann’aim, never appear in the Talmudic diegesis,41 constantly embedded as the gaze, and sometimes even heard as the voice coming from an intellectual place in front of the screen to be “manned” or filled.
236
Who Remembers?
What is more important, however, is that without the implied, but never explicit economy of these filmic terms, Halivni’s theory of stamm’aimredactors and stamm’aim-recorders would not be possible. Indeed, the stamm’aim as redactors could have such a tremendously productive effect on/ in the Talmud only because they acted in the register of the Talmud’s cutting, not in the register of the Talmud’s diegesis, in which, as Halivni recognizes, they never appear. It is the conceptual positioning of them in cross-cutting, not in diegesis, that grants the stamm’aim-redactors their unique status of being there while diegetically being neither openly present nor even hidden or absent in any place in diegetic space and time. The unique status Halivni has theoretically granted to the redactors can work only because the redactors are nothing but functions of the cutting of the Talmud, the making of the new frames for amoraic materials, followed by their connection one to another, through the addition of new language, and finally through sounding as “a voice” with no name, no visible flesh, and no words to say on the screen. This set of functions (or if you still want to personify them, this set of redactors, if you still prefer to use that single name for the set of heterogeneous functions of montage) enables an event that, after the fact, in later reception, acquires, by the law of synecdoche, the name of one of its own diegetic parts, תלמודאor גמרא, the tradition, the Gemarr’a, the Talmud. In addition to many other much more important things that it does, such as designating the source of authority for all future generations of Talmudists and marking the act of symbolical closure or completion of “the Talmud,” this synecdoche of the Talmud tacitly reassigns the intra-Talmudic terms that had their specific references (best rendered as “a tradition” or “a saying the tradition preserved”) to designate a totally new thing or product in its whole, which today, in its totality, is called “the Talmud.” To apply the filmic analogy once again, the synecdoche marks the appearance of a new film, The Talmud, called after one of the objects shown in this film. The film has of course its own diegetic time and hence its own time of montage.42
Beyond Depersonification However, to complete the move from the chronological and personal stamm’aim to the impersonal functions of montage, it is not enough to depersonify the stamm’aim and to reduce them to functions, for in film, the
fup-dolgopolski-10.indd 236
10/9/12 12:08 PM
The Talmud as Film 237
functions of montage construct not only the film as seen but also the viewer of the film. Similarly, the functions of montage construct not only the Talmud as film but also the viewer or rather the learner as a human being in a space that preempts his, her, or their thoughts and emotions—including even self-perception or the sense of the self that the Talmudic montage prepares—from crossing the lines of what the implied and/or constructed film viewer or Talmud learner are supposed to have. That preemptive space both awaits and constructs the viewer or the reader/learner of the Talmud. Understanding this place is, of course, much more oriented toward the concept of human being-there than it is to the functions from which that concept results. The element of the human being-there in montage could be looked for not via the concept of the performers in the montage (e.g., the filmmakers), but rather in the concept of the recipients of the montage (the viewers), or those whom the Talmud addresses. That aspect of the constructed viewer in the film theory of montage not only applies to Halivni’s analysis of the Talmud, it also opens new ways to deal with one of Halivni’s major concerns: Why, unlike the recorders stamm’aim tann’aim, are the redactors not mentioned in the Talmud at all? Discovering how the film theory of montage can help to address this hitherto unanswered question comes with a complication: It will require us to note that there is yet another important discrepancy between film theory and Halivni’s historically oriented implementation of it. To introduce this discrepancy and to use its heuristic power more effectively, let me summarize the outcomes of my critical reading of Halivni’s position, for it provides me a point or rather a tension from which I depart. Halivni himself translates his own purely functional analysis of Talmudic montage or “conjecture,” into entities-oriented, or, as a Heideggerian philosopher would say, ontic terms of a probabilistic “history” of the montage performers. Thus, as we have seen, he somewhat artificially imports an “external history” of the Talmud into its internal diegesis. Nevertheless, the main grounds for his historical reformulation are not historical but rather literary or, if you will, “Talmudic” inferences/conjectures of what might and what might not have happened, for conceptually, those inferences could be made only from either the diegesis of the Talmud or from its montage, not from its external history. Not totally unlike his own stamm’aim, Halivni uses literary analysis to reconstruct the “history” of the Talmud, thereby trumping his scholarly predecessor’s and most significant other’s version of that history—Albeck’s.43
238 Who Remembers?
In fact, Halivni explicitly justifies that move by referring to conjecture. That is to say, he puts himself in the same position as the stamm’aim and produces a new conjecture/montage. In sum, while clearly working in terms of the cutting functions of extending and so on, Halivni converts these functions into the virtual historical images of their performers—human beings who diegetically become partly present and partly hidden or absent.44 Instead of analyzing the functions as such, Halivni converts them into the hidden, but still somehow and somewhere historically would-be-present anonymous individuals (conscious or unconscious about what they do) who perform the literary functions, but exist only in “history,” not in the Talmud.45 Therefore, however much I have highlighted the functionalist and filmic aspects of his approach, Halivni’s analysis of the stammaitic arguments is de facto not purely function-oriented. On the contrary, the paradoxical justice of his approach is that despite the severe theoretical difficulties he faced, Halivni was strongly oriented toward implying the virtual historical existence of the functions’ performers and thereby placed the stamm’aim as people, not only as functions, in the historical, chronological space-time outside of the Talmud, even at the price of leaving the territory of the literary analysis of conjectures and entering the space of positive history, with all the difficulties that a turn to positive history involves, difficulties that otherwise, in a purely literary-filmic or functionalist approach, he would not face. Despite the tension between his literary approach to the Talmud and the problematics of historical-chronological time, Halivni insists on embracing history and chronology, and he does so for a very good reason. To unveil that reason, we need to shift gears and analyze the ontological foundations of Halivni’s project of the stammaitic arguments. We will see how an ontological extension of film theory allows human beings to be involved with the pure functionality of montage, not in the space-time of chronology, but in a space that is occupied, in the case of film, by the viewer, and in the case of Talmud, by the reader or learner. To do so, we will need to take a closer look at the ontological premises of the stammaitic arguments.
Political-Ontological Analysis of the Stammaitic Arguments As I have shown, Halivni explicitly builds his theory of the redaction of the Talmud on the premises of both redactors and recorders as solitary
The Talmud as Film 239
self-conscious individuals who act sometimes consciously and sometimes unconsciously, with “conscious” versus “unconscious” actions one of the leading oppositions he uses to distinguish the recorders from the redactors. It is not completely by chance that Halivni’s terminology of “conscious” redactors and “unconscious” recorders in an “analysis that has to reveal the unconscious work of the recorders” sounds not unlike Freud’s analysis of Moses,46 in which Freud also makes a link between his analysis of the biblical story and the time of history and produces not only a new literary reading of the old narrative, but also a historical-chronological claim about the origins of Moses the person. Now is the time for a closer look at the link between Halivni’s (and Freud’s) obligation to history and chronology and Halivni’s ontological opposition between conscious and unconscious individuals involved in that history. First of all, it is only the opposition between conscious and unconscious actions that enables Halivni to introduce his new concept, the redactors, in place of the old Geonic term stamm’a de gemarr’a. Unlike Halivni’s reading of it, the latter term might very well have been a Geonic parallel to the (relatively rare) Talmudic term stam mishnah. Just as the latter term referred not to the tann’aim, in the sense of the recorders or reciters of the Mishnah, but rather directly to the sages of the Mishnah themselves, so also stama de gemarr’a in the Geonim might have referred neither to the recorders nor to the redactors of the early Talmudic masters, or amor’aim, but rather, speaking diegetically, however naive it may sound to a historian, to the Talmudic traditions coming from unidentified amor’aim themselves. Second, the opposition between the conscious and the unconscious enables Halivni’s main principle of literary analysis, search for “a moving cause,” which he applies both to his own reading of the stamm’aim and as an underlying principle of the work of the redactors. A “moving cause” makes both recorders and redactors do what they do. Hidden moving causes explain why the recorders unconsciously changed the teachings of the early Talmudic masters, thus contrasting with the conscious redaction that the redactors would make. Halivni also reveals a different set of moving causes that explains how the redactors consciously “restored” the (as it were) not- yet recorded parts of the teachings of the amor’aim, “their” “dialectical” give- and-take. His principle of searching for “a moving cause”47 is also Halivni’s own main methodological tool for analyzing the Talmud. In a sense, Halivni assigns himself the position of the last of the stamm’aim orkhim, the last of
240 Who Remembers?
the anonymous teachers/redactors of the Talmud, the only stamm’ai whom we, quite fortunately, know by name.48 At any rate, his opposition between conscious and unconscious stamm’aim necessarily leads him to project his literary (psycho)analysis of the Talmud’s montage into the time of history, into the chronology of the Talmud’s creation. Yet there is an even more serious factor contributing to this projection. It has to do with the ontological assumptions regarding the orality of the Talmudic arguments that Halivni’s theory of transmission inherently shares. Of course, it is much harder to depersonify what is assumed to be an orally transmitted tradition than to do the same to what is considered to be based on initially the much less person-related processes of writing (e.g., composing or recording). However, the real question is not what to foreground— the oral or the written—but rather what to use as a model for the process of thinking: the model of an oral conversation that has been committed to writing (including even the model of writing as composing, which perhaps even constructs and emulates an as it were original oral conversation) or alternatively, the model of the foundational elements of writing, which do not translate into the oral. I refer to sets of initial distinctions (différance, in Derrida’s terms) without which not only thinking but also speaking, not to mention writing in the basic sense of composing or recording, can be possible. If thinking can happen only within a set of distinctions, from which both speaking and recording or composing also stem, the real question about the nature of thinking in the Talmud and about instances involved therein has to do with that fundamental dimension of writing as the process of distinction making, rather than with the projections of writing in that fundamental sense into its applications in thinking, composing, or recording. The secret of thinking (as well as of composing and speaking) in the Talmud thus might have to do with writing in its fundamental sense.
An Oral Argument For Halivni, the redactors’ work was to record the not-yet recorded oral components of the arguments of the early Talmudic masters, or amor’aim. In that sense, the redactors did nothing new—they only continued the process of recording that the recorders began. All they recorded was considered to be as preexisting as any already existing records and as authoritative as their preexisting originals. If the unrecorded elements of the amor’aim’s arguments
The Talmud as Film 241
existed before being recorded, as they were supposed to, they of course must have existed in an oral form. However, as Halivni’s analysis teaches us, the issue of the oral character of the stammaitic arguments cannot be reduced to the simple fact that despite their recording in either memory or even ink, diegetically or intra-Talmudically, these arguments were oral. In fact, the initially oral character of the arguments of the early Talmudic masters (not to mention the sages of the Mishnah) is the ultimate condition of the recorders’ and redactors’ activity and, needless to say, their authority. However, this raises a question: In what precise sense are the arguments of the amor’aim oral? Of course, the traditional, “obvious” notion that their arguments had to have been oral has never been shaken, even by the merely technical fact that the “oral” Talmud has been recorded, printed, and, after its printing, primarily studied or learned in the form of a book or manuscript, not as an oral tradition. The fact that Talmudic learning intrinsically depends on a record of it (either in memory or on paper) never disturbed the self-evident image of the Talmud as oral, at least until the development of French poststructuralism in the late 1960s. Before that, for centuries upon centuries, the old view of recording as a merely technical enterprise had been an unchallenged standard in virtually all fields of the humanities except for the historical, cultural, or aesthetic analysis of calligraphy, handwriting, or printing. Talmudic studies were no exception to the mainstream, standard view of the recording of the oral as a highly important, but still merely technical enterprise. This view was in the best tradition of Platonism, in which recording ideally (but, as a Platonist would agree, not always in reality) has no significant effect on what it records. The recording is only to represent (re- present) the recorded, with as little disturbance or aberration as possible. In that traditional Platonist approach, everything that is recorded is supposed to have existed before the acts of “its” recording, or, to draw the moviemaking analogy, as if the drama or documentary on the screen preexisted the complicated process of recording and cutting that produced the movie. Just as a belief that the camera records, not creates, is an intrinsic part of so many films (as is often said, “Seeing is believing”), the belief that the recorded Talmud is only a secondary re-presentation of the preexisting “oral” Talmud has been a part of traditional Talmudic learning, at the very least since the time when the printed Talmud achieved predominance. Limited by the choice that this version of Platonism offers—either that the Talmud existed prior to its being written down or that it came into
242 Who Remembers?
being simultaneously with that writing—mainstream scientific studies (and Halivni’s work is no exception) chose to assume that the oral Talmud somehow existed prior to any effort to record it or at least that whatever was recorded existed before the recording itself. However revolutionary scientific research on the Talmud was in terms of questioning the validity and authority of sources, it still retained the traditional Platonist idea of the pure and true Oral Talmud that precedes its recording. No one tried to explicate the Platonic assumptions of either the traditional or the scientific study of the Talmud, much less to open them to a critique. Yet, from the late 1960s on, many thinkers in different fields have distanced themselves from the seemingly obvious Platonic approach to oral origins in general and to oral argumentation in particular. This is not the place for a detailed survey, much less for the analysis of these developments that go back to Nietzsche’s and Heidegger’s critiques of Platonism and forward to the post-Heideggerian tradition of French thought as it was allied with linguistic structuralism. I only mention here that in the post-Heideggerian intellectual context the question of how to isolate the oral character of the stammaitic arguments must be taken more seriously than before. Thinkers as different as Derrida and Deleuze who worked in that context have persuasively argued that in the traditional Platonist ontological equation of origin, original, and oral, there was also another, not always recognized element—linear temporality. Time, the fundamental condition of the oral’s precedence over the written and of its original character, was shown to comply with the laws of writing, not of orality. The ground of what seemed to be primary was found to follow the logic of the secondary. That paradox brought into doubt the hitherto obvious understanding of the record as a secondary copy of an oral original. As a consequence, the question of orality and temporality in stammaitic arguments should now receive much more attention. Naturally, the alliance with Platonism in both the traditional and the scientific study of the Talmud, however silent and unnoticed, had an enduring effect on the study of the inner form of the Talmudic arguments. The oral form of Talmudic discussion has been mainly approached in conjunction with another essential concept of Platonism—dialectics.49 Despite any possible differences between Talmudic parity-oriented “dialectics” and Platonic victory-oriented dialectics, the scientific study of the Talmud rarely, if ever, entertained any doubts about the dialectical form of Talmudic argumentation. For instance, Halivni writes: “At the time of the Mishnah and Baraitha,
The Talmud as Film 243
the explanations and arguments that made [the sages] conclude so and so, were not recorded for generations to come (and because these explanations and arguments were in the form of dialectics, mostly as refutations and defenses, with multiple subdivisions, we call them ‘give-and-take.’ ”50 Another Platonistic assumption has been that dialectical give-and-take is part of the art of logic, and hence researchers have sought some version of Talmudic “logic” to be retrieved from the alleged “dialectics” of the stamm’aim. Distancing myself from these Platonic assumptions, I argue that there is yet another, no less important aspect of the form of stammaitic arguments, namely, that those arguments are rhetorical to the same extent that they are dialectical or logical. If this is the case, then just as logic and dialectics arise from certain ontological foundations expressible as linear temporality and signification, the rhetorical form of the stammaitic arguments also should have an ontological equivalent that should be expressible in an alternative temporality that would not always repeat the logic of a linear progression of time. Specifically, I propose that reevaluating the art of rhetoric as the main form of the stammaitic argumentation also invites reevaluating the oral and preexistent character of these arguments. To reevaluate the putatively original oral form of the stammaitic arguments, we need not only follow these arguments as they were diegetically made by the amor’aim orally at the time before the stamm’aim put them on record, but we must also undertake an ontological analysis of their oral nature and of the temporality thereof. In other words, we need to look at the diegetic initial orality of the stammaitic arguments in the aspect of their montage. How is it that the oral character of the stammaitic arguments has something more to it than the seemingly simple fact that they had an oral existence before they were recorded and redacted? How should we access that difficult “more”?51 If their existence before either recording or redaction (both intra-Talmudic or diegetic and post-Talmudic or “historical”) must occur in time, and if the arguments must have preceded their own record, we need to take a more careful look at the ontological premises about the kind of time at work here. If the scientific theory of Talmudic redaction accepts that there were stamm’aim who edited and added their own comments, it would necessarily demand, as it does, that they were in fact part of the product and not mere transcribers of a preexisting form. This entails a linear temporality: First there is a work and then, later, people make changes that become part of a new work. However, the problem for these later “people” is that they cannot
244
Who Remembers?
admit the fact of their own existence as an authority. If they admit they are a part of the work, they thereby violate the authority of the amor’aim they represent. As a result, their own authority rhetorically and therefore existentially comes from their authorization of the past and from a virtual denial of their own presence, either in that past or even in their here and now. To follow this complex dynamic of time and presence requires an advanced understanding of the oral character of the stammaitic arguments. Reading the Talmud in terms of its diegesis and montage, we no longer need to follow the line of historical time, with conscious/unconscious individuals acting in it. Rather, we have two parallel lines of time, the time of montage and the time of diegesis, and only the second has any identifiable individuals in it. Respectively, we have two planes of oral element: One is an imitated oral element of diegesis in which the arguments of the early Talmudic masters preceded their stammaitic recording/redaction in time. The other is what can be called the nonthematic or nondiegetic, but still oral element, the element of montage. It is that latter plane, its time, and its oral element that require our attention. Unlike what Derrida found in his revision of the ontological premises of both the strict sciences (such as linguistics) and philosophy (including even Heidegger’s fundamental ontology), the oral element of the Talmud’s montage is not necessarily subsumed under the wider category of writing. But this is, of course, a theme for a separate argument. Here I offer only a brief remark on the relationship between the planes of time in montage and diegesis in an always ongoing event ironically called, in one of its diegetic or intra-Talmudic terms תלמוד, “Talmud,” a term that, as already noted, in the Talmud means anything but the Talmud in the modern referential sense of the word. The time line of the Talmudic montage is in no way a time of linear progression from one point of presence to the next, not even in the sense of going from a present in the past to what “now” stands for as the present in the present. If Talmudic argument, taken in general, and indeed the Talmud “itself,” is a post-Talmudic intellectual event in which we become characters in its montage, it can in no way have any present, historical or not, and is therefore oral in a sense stronger that the notion of the diegetically oral can afford. Rather, as also suggested in the notion of the Talmudic h.iddush, or invention, the intellectual event of the Talmud is a rediscovery of something that has always already been there, but that does not belong to any present whatsoever (not even to a present in the past), but rather belongs to our always open future, in which it waits for us to arrive.
fup-dolgopolski-10.indd 244
10/9/12 12:08 PM
The Talmud as Film 245
When seen as an event, the Talmud belongs simultaneously to the past and to the future, for the past of the sages is the future of their students. That, of course, pertains to the dynamic time of the Talmudic montage. In contrast, the static time of Talmudic diegesis is unidirectional and in that sense stable. Therefore, it always easily converts to a present in a past, or a future, or even a “now,” whatever that means. In the diegetic “now” of the Talmud, which by definition is not historical, but rather fictive, a tann’a is surrounded by the early Talmudic masters, or amor’aim, and their later adepts, or savoraim, who have come to refute, expound, correct, and eventually authorize what the recorder learned to repeat, without necessarily accepting only one victorious interpretation of it. A tann’a, a recorder of the teachings of the sages and/or a speaker in their name, as well as in the name of the sayings of the amor’aim; other amor’aim, who, unlike the tann’a can satisfy themselves with neither merely mechanical recording memories nor with orations, but rather employ dialogue-like exchanges—these represent the main groups that the diegetic content of the Talmud presents. Unlike the first group, or tann’aim, the amor’aim and savor’aim approach the tann’aitic records with new rhetorical refutations and defenses that, on the plane of montage, have to come from the amor’aim, resulting, on the plane of diegesis, in a contemporaneous final authorization of records that the tann’aim could not help but remember or narrate as something they draw from the diegetic past. All this happens in an ever ongoing present time of Talmudic diegesis, a diegesis that is never possible without a parallel line of time, that of montage. To conclude the analysis: Although the Talmud is not a film, the filmic distinction between montage and diegesis allows us to address the question of the oral character and temporal organization of the event of the Talmud in a more precise way by interrogating the oral character and temporal organization of the stammaitic arguments, both in their literary dimension and in their ontological and anthropological aspects. What in an intra-Talmudic diegesis appears as the classical Platonist ontology of originally oral dialectical arguments that later were recorded, in the intra-Talmudic montage looks quite the opposite. The face-value record of the sages’ teachings comes temporally first, not second, while its argumentative oral or rhetorical defense comes not from the past, but from the future, which always remains at stake. In montage, the oral arises in a rhetorical defense of an already given record, but it is not something that precedes that record. In addition, because of their intrinsically oral character, which cannot be finalized in any authoritative presence and therefore cannot be recorded, no recording of the oral
246 Who Remembers?
(rhetorical) arguments of the montage can be a diegetic part of the Talmud and can therefore in no way appear in it, except for working in the Talmud’s montage as something that lets this diegesis, its human reception, and its human recipient happen. Halivni actually reveals the traits of the Talmudic montage, although, because of his ontological premises (linear time and linear signification, conscious and unconscious individuals), he addresses that montage historically, in the time of history, not in the time of the montage itself, much less in its relationship to either diegesis or the addressee. But if the “stamm’aim” are a heterogeneous set of oral rhetorical functions of montage correlating with the recording process of diegesis, then the stamm’aim are not necessarily a hidden historical group of people, and the question of their activity should be asked differently, in terms of what Talmud those functions produce, not only and not primarily in terms of the historical chronology in which this production took place. Relying on history, Halivni offers a new conjecture of (the) Talmud, acting as a new stamm’a. The opportunity that he thereby creates, but does not fully exploit, is, instead of undoing the work of the stamm’aim, followed by redoing it in a new “scientific” way, to ask what the stamm’aim’s own project of the Talmud was “in its own terms.” To take this possibility even further is to recognize, as I have tried to show in preceding chapters, that the stamm’aim as a concept is heterogeneous to the stamm’aim’s own alleged project of the Talmud. If we thus look at the stamm’aim’s project as not historical, but rather ontological-rhetorical, it would mean that what Halivni began to describe as the stamm’aim introduces us to the much richer world of the functions of the rhetorical montage of the Talmud, not to the chronological world of people behind the Talmud, for whom we still may continue to look in the space-time of its external history. The Talmud’s addressee, the being-there that thinks in and in response to a record of the Talmudic arguments is indeed produced by the Kuleshov effect, which takes place not in the Talmud’s diegesis, much less in the Talmud’s historical chronology, as Halivni supposed, but rather in the Talmud’s montage. What we are left with is the main Kuleshov effect of montage: the reader or learner of the Talmud, who, as this book has argued, is not a Cartesian thinking subject sitting in front of the page, but rather belongs to a complicated topology of virtual agencies in thinking that, on the pages of the Talmud, collectively work at the service of the Talmud’s main endeavor—that of memory.
Conclusion
Let me conclude with a series of questions I posit that “you” are asking me, and my response will be a soliloquy similar to the anonymous soliloquies we have seen in the Talmud. Thinking aloud invites the audience to think along, attending to the next steps in exploring the relationship between thinking, memory, and the virtual that this book has prepared.
First Question Where, in broader terms, do the techniques of discussions in the rabbinical schools as presented in the late ancient text of the Babylonian Talmud stand vis-à-vis five rhetorical techniques conceived and taught in other ancient schools of rhetoric: memory, refutation, invention, example, and delivery? A: In contrast to the more generally practiced compartmentalization of these techniques in other schools, in rabbinic academies portrayed in the text 247
248 Conclusion
of the Talmud, these techniques are much more closely intertwined. In rabbinical schools refutation (including both attacking and defending) serves the purpose of memory. Through refutation, memory advances beyond mechanical memorization to making the Mishnah defendable and thus remembered in a more reliable way. Refutations are performed by anonymous speakers, resulting in rediscovery/invention of what was initially remembered in a merely mechanical way. Another rhetorical technique, examples, serves the same purpose of better remembering the Mishnah and other earlier teachings. Examples support refutations, defenses, or inventions. The fifth technique, delivery, serves the purpose of memory as well, albeit in an even more particular way. It creates the site or, to borrow a visual metaphor from the analysis of photography, a “gaze” of memory. Similar to the gaze, memory is also a performative. Unlike the “gaze,” memory follows not the discipline of vision but rather the discipline of speaking taught in the schools of rhetoric. Memory as a performative is embedded in the composition of the Talmud and defines the intellectual orientation in which that composition is delivered to the audience. Delivery orients all thinking done through refuting, defending, inventing, and exemplifying to one and only one direction, from mechanical toward more advanced forms of memory. Moreover, if the technique of delivery must also imply a character who would be the delivering agent, that character would be a silent narrator. In the text, the silent narrator speaks not, but instead performs by putting the elements of composition together in such a way that everything in the composition populates an otherwise empty frame, “gaze” (or call it a palace, or a theater) of memory.
Second Question What implications does the Talmudic performative of memory have for, or how does it complicate, philosophical and rhetorical conceptions of the relationship between thought and memory? A: Hitherto dominant approaches in both philosophy and rhetoric have either separated thinking from memory, or, at best, made memory serve the purpose of thinking in search of what truly is as opposed to what only seems to be. In contrast, as a performative, memory orients thinking to remembering, not the other way around; it thus, from the start, sets thinking free from the political and intellectual urge to decide between what truly is (or was) and
Conclusion 249
what only seems to be (or to have been). In previous conceptions, memory predominantly served thinking, which in turn most often was understood along the lines of knowing what is as opposed to what is not, or to what seems to be. In contrast, the model of the Talmudic performative of memory removes the urge to decide on being versus nonbeing in remembering, which even the phenomenological method of epoche would not have removed but rather only deferred by bracketing. In this model, memory acquires independence. It may be, but no longer has to be, a memory of what is or was as opposed to what seems to be or to have been. Therefore, exploring multiple and even mutually exclusive paths of remembering and understanding of one and the same text no longer undermines the validity of memory.
Third Question The book analyzed particular instances of the virtual as pertains to understanding the Talmud in the contexts of philosophy, rhetoric, and literary theory. Yet there are some more general questions about the virtual, which the particular analyses did not directly address. Thus, in the beginning of the book it says: “To understand digital virtuality critically is to understand and critically analyze the notion of the virtual in the first place, in its earlier historical manifestations, in general as a concept, and as a way of social and individual life, now and in the past.” Where do the particular analyses above take us in respect to this more general question? A: The question of the virtual is, of course, very general. It initially framed the inquiry in a very broad way, and now, with the particular analyses done, invites a continuation in an even more narrow and specific direction—that of the relationship between memory and the virtual beyond the ontological and political question of separating what is from what only seems to be. The local analyses in the book indicated a more precise future direction of exploring the virtual and did no more than scratch the surface of what such exploration involves. Yet here are some more general considerations and, by implication, tasks for future analysis. As a concept, the virtual arises in two mutually exclusive planes. One is that of ontology and epistemology. On that plane, the need to decide what truly is and what is not dominates. Memory serves as a tool for satisfying that need by knowing, that is, recalling, what truly is. The soul
250 Conclusion
remembers true kinds of things and distills these kinds from mutual mingling, thereby eliminating the main effect of mingled kinds—the nonbeing, which seems to be, but in fact is not. If, as the epistemologic-ontological model suggests, memory is submissive to knowledge, the virtual becomes a form of what is, which, however, escapes a simple opposition between what is and what is not. When it comes to the virtual in that sense, it is about nonbeing, which cannot be eliminated through distilling the kinds in knowledge, for example through enlightenment. Attained through knowing, enlightenment does not eliminate a certain kind of nonbeing, which is exactly the virtual. Instances of the virtual of this kind include persistent illusions (sunrise and sunset, capital as the source of wealth, psychotic deliria) which do not go away by sheer explanation of their nature. Marx’s notions of “false consciousness” and “inverted form,” Baudrillard’s “hyper-reality,” and Freud’s psychosis are conceptualizations of the persistent nature of these illusions and thus are particular theorizations of the virtual. A literary illusion that the Talmud is a direct recording of conversations between teachers and students in a rabbinic academy in Babylon belongs to that set as well. Explaining it does not make it go away, and making it go by force makes changes in the “real” world—an attempt to ignore the sunset disorients, to deny capital provokes a revolution, and to explain a neurotic symptom makes it come back in another form. Thus understood, the virtual does not belong to the actual but produces actual effects in things. Along the same lines, explaining the Talmud away as a set of illusory conversations may disorient both supporters and discontents of halakhic observance by creating a false, but nevertheless currently politically actual, dichotomy between either denying the Talmud’s authority or shielding the observance from critical study. The other plane of the virtual is that of memory. On that plane, memory is the main orientation in thinking. Rather than having thinking orient memory toward knowing what is, memory both orients thinking and makes it possible, just as the gaze both orients the photograph and conditions its possibility. Unlike other orientations of thinking, memory does not create any kind of specific anticipation, much less any kind of final grasping similar to what is thought of in thinking. In memory, thought is not wedded to what it is the thought of. A noema is no longer anticipated by the noesis, and the gaze no longer requires either what is gazed at, or the one who is gazing, although, of course, the gaze make all of the above possible. Looking comes before seeing, hearing before discerning, narrating before speaking,
Conclusion 251
and in the same sense, memory before thinking. In that sense, gazing, hearing, and narrating belong to the virtual. The Talmudic delivery exemplifies this model. In the Talmud, the task of memory is not to recall what is (or as a variation, what was) but rather to orient thinking to testing what is already remembered through heuristic refutations resulting in the rediscovery or invention of what actually was remembered. The memory here is never specific, much less final. It is not a memory of something anticipated, but rather a memory of something that is always yet to be invented/discovered. As such, this memory is not subservient to knowing what is or was, even if it may lead to inventive answers to that question as well. On that plane, in short, memory is an instance of the virtual, that is a performative. Other versions of it in the book have included the following: The sophist who is neither present onstage in the dialogue The Sophist nor absent from it (that is, not present somewhere else in the “real” world either), and yet speaks, thinks, and in fact almost dominates the dialogue on the stage. The empty orthodox, the rabbi who has no internal, that is to say, “real” position to take, but instead argues from within the arguments of those who argue with him. “The Author”: He belongs to both worlds, the biographical world of Samuel Clemens and the literary world of The Adventures of Tom Sawyer, but has no embodiment in any of the two “real” worlds. Nevertheless, nothing happens in these “real” worlds without “The Author.” In The Adventures of Tom Sawyer, he is a character without body, who both constructs and controls all other, embodied characters; and in the biographical world of Samuel Clemens, he is the identity that elevates Clemens’s biography beyond the ordinary. The Monteur: This is the one who putatively does the montage, following a strategy in putting the film-cuts together. She cultivates otherwise wild powers of film-montage working on the site of the viewer to be shaped through that cultivation. Montage is both a way to create diegetic events (either on the screen or in the mind of viewer) and a way to control, and, indeed, again, create the viewer. Montage also serves to create a figure of a monteur and to show that it is not the agency of position of monteur but rather that of the viewer that is the real site of montage.
252 Conclusion
The silent narrator in the Talmud: the one who delivers the Talmud as an intellectual project of using rhetorical refutation, invention, and exemplification as an instrument of memory, which in turns delivers in the ways of invention, and performs in, for, and by the audience, in which no one is a Cartesian thinking subject. However diverse the above versions of the virtual might be, none of them requires us to make a decision and to sort out what is as opposed to what is not, much less to what seems to be. They therefore represent historically new forms of the virtual, which the notion of the “new digital media” does not grasp. These forms of the virtual can hardly fall into the rubric of “new media,” because, however historically “new” they might have been at the time of their conception, “media” they were not. The rubric of “media” presumes, as Plato has it in The Sophist, mingling of kinds, so that one comes instead of the other, which in turn presupposes an ontological urge to decide on what is, and what is not, and what seems to be. And the instances of the virtual above do not belong to the plane in which such an ontological decision is urged to be made. This means that classifying the new contemporary forms of the virtual as “digital,” that is, by their “media,” precludes the possibility of asking how the new forms of the virtual relate to those historically new forms above. I have already addressed that question in terms of the relationship between thinking and memory. In that perspective, to ask the question whether or not “digital media” enhance or undermine our ability to remember in a proper way requires changing the terms of the question from “digital media” to the virtual in its complex situation between ontological and a-ontological planes. This takes me to a concluding remark about the relationship between thinking and memory explored in the chapters above. This remark also anticipates the direction of further inquiry. If thinking is modeled after knowing; and memory after thinking, then memory might be either mechanical memorization, or internalized knowing. In this approach, memory may thus entail memorized data, or Gedächtnis in the language of German philosophy, or “memory for words,” in the language of late ancient schools of rhetoric. In the same approach, memory may alternatively entail an interior space of individual recollection, Erinnerung, in German, or “memory for things,” in the language of the rhetorical schools. Not fully unlike the German version, “memory for things”
Conclusion 253
presumes interiority; more specifically an interior, for example, that of “a palace of memory.” The palace of memory, as the artisans of rhetoric have it, is an intimately familiar space, for instance, her house, which the orator populates with images of things she needs to remember for the speech in the order in which she placed them in the palace. Both mechanical memory and interiorized recollection presume a solitary individual, an interiority, and ultimately a solitude in remembering. However, as the texts of the Babylonian Talmud exemplify, if they are read in the context of philosophical and rhetorical schools of, and approaches to, thought and memory in late antiquity, thinking might relate to memory in a different way. In that approach, the workings of memory require an encounter with another person, and, thus, are no longer the affair of either a solitary individual, or of a unified group. Yet, in the Talmudic model, that encounter with another person cannot be initiated by you. Even before encountering another person, memory in general, and thinking as memory in particular, requires another person to encounter you. According to Levinas’s interpretation (and critique) of Heidegger in an earlier essay “Is Ontology Fundamental?” encountering another person, rather than dealing with either an object, or, in particular, with a subject can only happen through expressing that encounter to that other person through speaking. In the simplest case, this expression is a sheer saying hello. In his example, to encounter a person you have to say hello, including even to refuse to say hello. Marginal as it seems, refusing to say hello is, in fact, decisive in how Levinas draws a borderline between a person and an object or subject, which, in his view, both Buber and Heidegger missed. Levinas insists that the refusal to say hello can only be expressed to a person, but not to an object. You can, of course, say hello not only to a person but also, say, to a table. But would you refuse to say hello to a table? The refusal to say hello to someone is what ultimately makes you encounter that other one as a person. The refusal is what allows Levinas to claim that “human being” is no longer an essence (a subject, an object, or both) to know, but rather a being to encounter. Thus, Levinas, in both engagement and disagreement with Martin Heidegger and with Martin Buber, establishes a borderline between a person whom you encounter and a table, which you know. The person is thus the one to whom you can refuse to say hello to. Now, if, following Levinas’s model, memory too is an encounter with another person, rather than a vehicle of knowing things, then there is much more to memory than an encounter with the other: the other needs
254 Conclusion
to encounter you in the first place. For example, Phaedrus would never be able to remember his forgetting Lysias, whose words he remembered as his own, unless Socrates encountered Phaedrus and gave Phaedrus, and surely the audience of the dialogue, a chance to remember, if not what is forgotten, at least the forgetting. In remembering the forgetting, it thus all depends on the other person, Socrates who encounters Phaedrus, rather than on Phaedrus alone encountering Socrates as the other. That primacy of the other encountering you rather than you encountering the other concerns not only forgetting as a specific case of memory, but rather applies to memory as a performative in general. Just as forgetting proves to be irreversible, so too does memory. In this approach, memory is an irreversible condition, that of me being encountered by the other, rather than a merely epistemological phenomenon, of me knowing and controlling things. Memory has no absolute relationship with what is remembered, nor is it an anticipation of what is to be remembered. If you cannot remember it, try to forget it. Your memory is your condition of being encountered by the other, whom you by definition both cannot know but also cannot forget. Therefore, you remember. Memory is a condition of thinking that requires thinking to remember but does not anticipate, much less, grasp the memory of what it should be.
Fourth Question How does the theory of memory developed in this book in application to late ancient texts of the Talmud relate to the theory of disagreement developed in late-medieval or early modern disciplines of Talmud and Talmud study, which you have explored before? A: Let me use an example of such a theory of Talmud, which I have explored before in my reading of Rabbi Izh.ak Canpanton, for whom speaking was making sense only as refuting, and thus to speak and to make sense was always already to engage in a disagreement. The student of the Talmud, therefore, had to start from the outside, from grasping what the disagreeing parties of the past maintain according to the text, and to end up in the space of the disagreement, where there is no longer a place for the individual mind of the student. In that way disagreement was the goal rather than a means and the sense rather than a detail of the discourse. In the view of the late
Conclusion 255
ancient text of the Talmud developed in this book, such a conceptualization of the role of refutation in general and disagreement in particular is not to be denied, but rather to be included in a more general conceptualization of the intellectual project of the Talmud as that of memory. In that more general theory, mind yields to what Augustine has approximated as “the hand of his heart.” That theory of memory is centered not on an individual mind, not even from the start, but rather is conceived as a performative memory, in which refutation and disagreement are fundamental elements of performance.
Fifth Question Last but not least, where does the trajectory of the analysis in the book lead? A: Through “Who speaks?” to “Who Thinks?” to “Who Remembers?” we arrived at the past. It is no longer, as both post-Heideggerian and neo- Kantian traditions would have it, the past as a secondary element of time. It is not the past in the sense of the necessary fiction of a starting point, or a period of before, which it must be, if, as those philosophical traditions had it, the true source of the time is not in the past, but rather in having a future. What we arrive at as a result of the journey in this book is the past remembered, rather than the past recalled; and it is the past that is no longer connected to time in that modern philosophical sense of the time coming from the future. Perhaps it is not connected to time at all. It may instead be the past that arises in no time. This “no time” has several layers of meaning. In its weakest formulation, it means “in no single time” in the sense that the Talmudic conversations happen in no historically “real” time. More strongly put, it means, borrowing the language from film theory, that the Talmud requires a never consolidated difference between diegetical time and the time of montage, which creates a no one’s time and place, which in turn creates time and place for everyone else. Even more strongly, it means that the performance of the conversations in the Talmud is constantly driven by, and towards, an always open past. The power of that past is radically different at once from the power of either futuristic or chronological time and from what both post- Kantian and post-Heideggerian traditions of thinking can at best grasp as an “ontic past,” or the content of the past, which therefore is still a part of
256 Conclusion
time. If understood as ontic, and therefore an ultimately closed component of time, the past, as we learn from Heidegger, belongs to being, in both ontic and fundamental dimensions of it in Heidegger, who not only denies, but also continues neo-Kantianism. Instead, we arrive at the power of the open, to the open past before, above, and irrelevant to either (me)ontology, or to time. Discovering the power of the open past in the Talmud, we thus arrive at the open, which is no longer the Dasein, and is no longer related to being and/or time. Instead, the remembering practiced by the characters in the Talmud lets them stay in the open toward the past, which therefore becomes open toward them. In that light or rather in that darkness of the open past, the political implications of that power of the past, or rather its implications for understanding the political as such, as opposed to the economical, social, or institutional spheres, are yet to be claimed.
a p p e n d i x : t a l m u d c r i t i c i s m , a n a n a ly t i c a l e x a m p l e
“Composer” versus “Redactors”: David Halivni’s and Shamma Friedman’s Competing Readings of Baba Metzi‘a 76ab
The following is an an analytical example of what in Talmud criticism both enables and requires moving through the theory of redactors as the alleged “anonymous” thinking subjects to the analysis of the virtual site of the subjectless thinking, which is oriented, as the book argued, toward memory and remembering. This example addresses both readers with a background in rabbinics and related fields of academic research on the Talmud and those interested in taking a closer look at the highly elaborate techniques (often misconstrued as a “methodology”) with which Talmud criticism approaches a text from the Talmud. To that end, I provide a more detailed comparative analysis of two major schools in Talmud criticism than what I was able to do in the chapters above. This detailed analysis both supports and illustrates theoretical claims about text-critical theories of redaction and design of the Talmud in Chapters 3, 6, and 8. 257
258 Appendix
The Surface of the Text before Time Literary-formalist and literary-realist scholars in Talmud criticism have both sensed and too quickly dismissed the bare being there, or if you prefer, bare performance of memory, too promptly affording it flesh, blood, and even a name of either “the composer” or “the redactors” of the sugya, or a compositional unit of the text. It is important, therefore, to show how the scholars of Talmud criticism constructed the text of the Talmud as an object, which they consequently read through a critical lens. What these scholars criticized was a synchronic, and ultimately also a diachronic reading, on which the Talmud was either a direct or stylistically processed record of real conversations as they happened in a rabbinical academy in Babylon. Synchrony of conversations, on their view, was an illusion. In contrast, their critical reading insisted on genesis of the text, which they held true. On their critical view, the text of the Talmud results from editing or transforming in the course of time. However, the synchronic and/or diachronic reading they criticized (I call it a precritical reading) and their critical readings shared an important thing in common: a dimension of linear time, synchronic, diachronic, or, on advanced stages of analysis, even virtual. In contrast to using linear time as a framework for both critical and “naive” or a precritical reading of the text of the Talmud, the text can be read as a series of ordered refutations and defenses. In this alternative reading, I draw on the intrinsically timeless character of refuting and defending, which may, but does not have to be connected to a model of time—synchronic, diachronic, or even to the time of textual transformations: defense can be created before refutation, or vice versa, not to mention that, as Friedman makes clear, the texts may transfer between forms and positions in the resulting composition of the Talmud. What follows illustrates, and elaborates on, that general thesis using an example of a Talmudic text both Friedman and Halivni directly analyzed.
Competing Readings Shamma Friedman and David Halivni rarely happen to analyze the same text, but one such rare occasion is in the Babylonian Talmud bBM76ab. That they both analyzed this text offers us an opportunity to juxtapose their
Talmud Criticism, an Analytical Example 259
approaches empirically, rather than purely theoretically.1 Prior to attending to their interpretations, that is, to what they construe as the text’s genesis in time, I first explore the surface of the text, reading it in terms of refutations and defenses, independent of any “illusory” representation of time or place in synchrony, or else of the “real” time of genesis. Having already exposed the beginning of that particular Talmudic discussion in Chapter 6 in terms of how the characters co-participate in making sense of the Mishnah, I now narrowly focus on two fragments further down of that surface, on which Halivni’s and Friedman’s readings intersect. For convenience, I partition these two fragments into sections from A to D, and from F to J.
First Fragment: Upon Whom to Rely? In the context of the discussion in bBM76a, an unidentified teacher or student in an undefined rabbinic school, presumably in Babylon, asks a question. The answer comes by way of inference from a report on a relevant precedent from earlier court practices. The question was: A.2 [I]f the master told him [to hire other workers]3 for four, but he went and said to them “for three,” and they [accepted the offer] saying, “Whatever your master said,” how [to judge this]? If they [the workers] argue [at the end of the day—S.D] “You [deceitfully] assured us that such was the word of your master!” are we [the judges—S.D.] to rely upon his words, or perhaps on the words of the master? (bBM76a)
The case is that the master tells his worker to hire others for four coins, but the latter offered them only three, and they accepted, by saying, “Whatever your master said.” The question is how the rabbis are to judge that in court? Perhaps in considering the meaning of the words “as your master said” the rabbis should rely o the words of the hiring worker only. If this is the case, when the workers claim, “You have [deceitfully] assured us the master said ‘for three,’ give us four!” the court should apply the Mishnah’s instruction about the cases of deception to approve their complaint about deception, but deny their monetary claim (Rashi’s commentary ad locum). Alternatively, should the rabbis judge based on the words of the master and thus not apply the Mishnah, but rather grant the workers all four coins?
260 Appendix
A question behind this question is, “What counts?” The ultimate (but perhaps unknown) reference of one’s words or a contextually clear meaning of them? In this case, the ultimate reference was “four coins,” while the contextually clear meaning of “Whatever your master said” was “we accept the offer.” In more concrete terms, if a recipient accepts/approves a message (in this case an offer) through saying “whatever the sender said,” then what counts: the original words of the sender or their rendering by the messenger? The rabbis in the academy infer the answer from the precedent in Rav’s court reported in the name of Rav Nachman, which deals with a divorce situation that similarly involves a messenger changing the words of the sender. B. [She said] “Please deliver my bill of divorce to me [from my husband],” yet [the messenger said] “Your wife told me to accept [sic!—S.D.] her bill of divorce [on her behalf].” He [the husband] said, “Let it be as she said.” Rav Nachman said that Rava bar Abuha said that Rav ruled that even if the bill reaches her hands, she does not become divorced. (bBM76a)
In the precedent described, the wife asked someone to be a messenger to deliver the bill of divorce to her from her husband, so that she would be able to possess and thereby to accept the bill and thus to complete the process of divorce. Yet, the middleman rendered things as if she appointed him not merely to deliver, but rather to accept the bill on her behalf. Not aware of that discrepancy, the husband said let it be as she said. Rav, an early authority, judged that the divorce was not in effect, even when she possesses the bill of divorce. Upon having attended to this precedent as reported in the name of Rav Nachman, the anonymous rabbis in the academy inferred that Rav judged based on her words, i.e. on the words of the sender, not on the words of the messenger. Their reason was: Otherwise, when the bill reaches her hands, Rav would have acknowledged that the divorce had taken place. We read to that effect: C. By inference, [in the above precedent] Rav must have been relying upon her [i.e., on the sender’s—S.D.] words, for if you say Rav relied upon his [the messenger’s—S.D.] words, then, upon arrival of the bill to her hands, Rav would have acknowledged her divorce as complete, [but he did not]. (bBM76a)
Talmud Criticism, an Analytical Example 261
This inference means the court should rely upon the words of the sender, and thus, in the case above, the workers were to receive all four coins. What immediately follows in D in the Talmud is an invocation of Rav Ashe’s rebuttal of either the report about the precedent or of the inference made from it for the case of the workers in question.4 In an elaborate argument Rav Ashe claims that the report would have been just fine for such inference if the content of the original and its rendition had been switched, that is, if the wife originally asked the person to be a messenger to accept the bill, but the messenger rendered his role as that of a deliverer only. Rav Ashe thereby shows the unsuitability of this precedent for inferences of the sort shown above. D. Rav Ashe objected that [making an inference of this sort] would be fine had the precedent been reported in the opposite way: [The wife said,] “Please accept my bill of divorce for me,” yet [the messenger said,] “Your wife said, please deliver [sic!—S.D.] the bill to me.” The husband said, “Let it be as she said.” And Rav Nachman said that Rava bar Abuha said that Rav judged, “When the bill reaches his [sic!—S.D.] hands, she is [already] divorced.” That would indeed imply Rav relied on her words [as the sender—S.D.]. Alternatively if [Rav had judged,] “When the bill reaches her hands, she becomes divorced,” that would mean Rav relied on his [i.e., the messenger’s] words. However, the way the precedent was actually reported, Rav must have judged as he did [that the divorce did not happen at all, even after the bill reaches her hands] because he has completely eradicated his status of a messenger by saying, “She asked me to be a messenger for delivering the bill to her [on the husband’s part—S.D.]” thus not for accepting it [on her part—S.D.]. (bBM76ab)
Rav Ashe’s showing that the reported precedent cannot help answer the question of what is decisive—the original words of the sender or their rendition by a messenger—leaves the rabbis in the academy with their original question without any definite answer, which leads to further discussion.5
Second Fragment: “I would not have paid them!” Just as the Talmudic composition moves forward to take the readers away from the impasse of having no definitive answer, I also am moving fast
262 Appendix
forward to the second fragment of the same sugya, in which Halivni’s and Friedman’s interpretations will again intersect. In this fragment, to counterbalance the negative effect of an unanswered question at the end of a line of attacks and defenses of the Mishnah, discussed in Chapter 6 above, the rabbis now come back to their earlier elaboration on the meaning of the word hat’u in the Mishnah. Their earlier interpretation of that word had to do with reading it as (a) “workers deceived each other” as opposed to (b) “workers and/or master changed their minds.” The (a) led to at least one problem unsolved, and the rabbis now turn their discussion to option (b). E. Yet, alternatively you may prefer to say that the reciter of the Mishnah used hat’u in the sense of “either the workers or the master changed their minds.” (bBM76b)
The word hat’u in the Mishnah now means “the [hired] workers changed their minds and/or the [hiring] master changed his mind.” In the continuation of their argument, the rabbis in the academy further support this alternative meaning of the words of the Mishnah by quoting a text of a baraitha, or an apocryphal text. It is enough to say here that the context of that particular baraitha clearly suggests that hat’u means “they [either the hired workers or their employer—S.D.] changed their minds,” which renders that meaning of the term possible not only in the baraitha but also in the Mishnah as well—at least as long as the baraitha’s authority is unchallenged, or better challenged and shown to withstand the attack. Accordingly, in the next step of their argument, similarly to how the rabbis in the academy probe the validity of the recitations of the Mishnah, they test the validity of the recited baraitha as well. They do so by reporting a case of having this baraitha recited in the house of study of Rav. Their report suggests that Rav had objections to at least one clause of that baraitha. The clause in question rules that if the workers arrived at the workplace (e.g., a field) but found no work to do, they have to be paid their full wages. To show that Rav objected to, at least, that part of the baraitha, the following report says in particular: F. It [the baraitha—S.D.] has been recited in presence of Rav. “[The master] gives them their full wages.” Rav said to him [to the reciter], “Habibi [i.e., Rav’s
Talmud Criticism, an Analytical Example 263
uncle, Rav Chiya—Rashi, ad locum] used to say, ‘If I were [there] I would not give them but a wage of an idling worker!’ and you recite, ‘[the master] gives them their full wages!’ ”
In this report, Rav objects to the ruling of the baraitha (“I would not give them etc.”); his objection renders the entire baraitha unreliable, and, in the larger picture, puts the dependent reading of hat’u as “changed [their] minds” in danger. That reading is thus now under attack. To deflect the attack, the rabbis in the academy first doubt the accuracy of the report in which Rav attacks the baraitha. They ask rhetorically: G. But did not the baraitha [also] stipulate, “but a worker coming with full load is not the same as coming empty, or doing the work is not the same as sitting idle”?
This raises doubts about the “It” in the “It has been recited in presence of Rav” in the F above. Did Rav, and thus the “It,” really refer to the same baraitha?! (Perhaps he was responding to a different one!) Yet, the rabbis in the academy reaffirm that he did respond to that specific baraitha, but allege the report might reflect a situation in which the reciter interrupted the recitation in the middle. H. “They did not finish [reciting] it in front of him!”
However, this reconstructed scenario of interrupted recitation is not very convincing. In fact, it almost reduces the report about Rav objecting to the baraitha to an absurdity. As a result, the report no longer represents any serious attack on the baraitha. A stronger attack, and leading to a thereby more persuasive defense is needed. In response to that strategic need, an anonymous character—perhaps the rabbis in the academy—recall or invoke an alternative report of the same incident with Rav objecting to the baraitha. That new report naturally leads to a more effective attack and defense. In the newly recalled version, there is no need for any artificial interruption. I. In another report [of the same incident—S.D.] they did finish reciting it, but Rav said, “Habibi [i.e., Rav’s uncle, Rav Chiya per Rashi’s commentary ad locum]
264 Appendix used to say, “If I were [there] I would not pay such workers anything!” and you recite “[the owner is] to give to them as to an idle worker!”
This one still renders Rav as objecting to the authority of the baraitha, and the latter is again under attack. Yet, this time, the rabbis in the academy fully recognize the seriousness of he attack. They start by exclaiming: “This indeed refutes it!” However, they will deflect this refutation through a differential reading of the references in the baraitha as opposed to those of Rav. J.—This time [Rav’s position indeed] refutes [the baraitha]! —No, it does not! This [i.e., the exclamation of Habibi which Rav quotes—S.D.] might have referred to [workers] who did see that the field [was wet and thus unsuitable to work—Rashi] the night before; yet that [i.e., the baraitha refers] to [workers] who did not know [in advance] the field was wet.
For the participants of that conversation, Rav did indeed sound contrary to the baraitha, yet their points of reference might be different: either with or without knowing the conditions of the field (or by extension of the workplace) in advance. This provides a stronger attack on the baraitha and thus leads to a more satisfying defense, even if the new attack and defense builds on the previous one. As a result, a possible difference in reference found between Habibi and the baraitha now supports the authority of the latter, suggesting that Rav quoted Habibi to indicate a case that has been omitted in the baraitha, rather than in order to object to that baraitha. The version “changed minds” is now no less supportable then the earlier version “the workers deceived each other.” This concludes the walk on the surface of the discourse. Having traveled that surface, the implied reader observes how the characters have first considered and finally settled the counterevidence from the house of Rav and how that made the rabbis secure in their using the baraitha to support the second possible meaning of the word hat’u in the Mishnah. Thereby the rabbis in the academy achieved what they needed to at the current stage of their discussion: They validated the memory of tann’a, or the reciter of the Mishnah, by probing and failing to disprove both possibilities of references in these words: “workers deceived one another” or “[the master and/or hired workers] changed their mind.”
Talmud Criticism, an Analytical Example 265
Constructed as a sequence of refutations and defenses, the surface of the discussion of the rabbis in the academy continues; and, as we will see, Friedman’s analysis will involve some other fragments of that surface as well, yet I stop precisely where Halivni stops as he analyzes the text. I now turn to his critical reading of the surface of the text to show how from reading the surface with no time but only sequence involved, the critical scholars move to reading the text in terms of time. This is once again either an illusory synchronic time-place of the academy, or a “critically reconstructed” time of genesis, construed as the time of either “redaction” (Halivni) or “composition” (Friedman).
Above and Below the Surface: Illusory Synchrony versus Critical Genesis the stam: literary-realist analysis of the text’s genesis in halivni
Sources and Traditions is the title of Halivni’s multivolume project of critical readings in the Talmud to trace the work of its “redactors.” In this project he meticulously discovers, describes, and generalizes the works of anonymous redactors of the dicta of the named masters in the Talmud. The project is defined by three major concepts: makor—“original” and “source”; masoreth, or “tradition,” that renders the “original/source”; and stam, “the anonymous [redactors],” that is, a human agency causing a transition from the former to the latter. This should not be confused with a previously dominant linear model of transmission, in which preexisting sources were faithfully transmitted through the chain of tradition, with as little aberration as either possible or believable. Unlike that traditional approach (pun intended), for Halivni, a “tradition” of the Talmud as it stands now in the form of manuscripts and earlier printed editions results from a subtle and highly intelligent process of redesigning the initial “sources” into new discourses, driven by the apologetic program protecting the authority of the source. The process of redesign is technically called “redaction,” a term borrowed from book production. In turn, the term “sources” specifically refers to dicta attributed to amor’aim, post-Mishnaic authorities mentioned in the Talmud by names. Halivni ascribes this process of redesigning or “redacting” to an agency that he calls stam (“anonymous one,” or “hermetic one”).6 He further uses the modern Hebrew coinage stamm’aim (“anonymous individuals”) to embody otherwise
266 Appendix
disembodied workings of the stam, now understood as identifiable, albeit unknown, individuals or groups (“redactors,”) rather than an abstract bodiless but by definition highly intelligent agency of the process of redesigning. Despite a seemingly simple reference of the term “stamm’aim” in Halivni to a historically real and chronologically defined group of people, the nature of the agency of the stam or stamm’aim is complex. As I argued in Chapters 3 and 6, in Halivni’s analysis, the meaning of each of these terms is twofold.7 First, Halivni’s terms stam and stamma’im refer to literary personae in the Talmud—nameless and countless rabbis and students in the rabbinic academy who, on the pages of the Talmud, engage in discussion. By way of these discussions, these rabbis invoke the dicta of amor’aim. Often the anonymous rabbis covertly promote their own arguments while attributing them to amor’aim. Second, stam and stamm’aim refer to an agency that composes or recomposes the discussions of these anonymous rabbis, including the teachings or dicta of the amor’aim engaged therein. Stam in that sense is not on the list of the characters in the Talmud, either named or nameless. For brevity, I refer to the stam in the first sense as stam-in the Talmud and in the second sense as stamof the Talmud. However stam-in and stam-of are twin concepts. Even if in his analysis of the Talmud Halivni prefers to bridge the difference between the “in” and the “of,” rather than to emphasize it, the examples below will illuminate a complex dynamic of relationship going on between these two twins. In most instances, Halivni bridges these two notions of the stam by ascribing the process of composing to the listed personae in the resulting composition—either to nameless rabbis8 in the academy or, sometimes to later amor’aim. If the “in” refers to literary personae in the Talmudic discourses, the “of” refers to an agency responsible for composing these discourses. In the examples below, Halivni tries to bridge the “of” and the “in.” However, bridging does not always work and does not eliminate the heterogeneity of the two notions of the stam, but rather makes it more discernible and thus accessible for further analysis.
in/of: an example
I now turn to an example of Halivni’s critical reading of the fragments of the surface of the Talmudic discourse above. I will explore the “depth” of
Talmud Criticism, an Analytical Example 267
the genesis behind the surface together with Halivni, and, at the same time, reflect on Halivni’s analysis of that depth. Discerning the acts of the stam in bBM76ab,9 Halivni first posits a who- question. Who in the part C made the inference from the precedent in the court of Rav about the wife sending a person to deliver the bill of divorce, as reported by Rav Nachman via R. Abbahu? Who said, “By inference, [in this precedent,] Rav must have been relying on her words etc.”? Seemingly this was stamm’a de gemarr’a [i.e., the anonymous rabbis in the academy—S.D.]; and Rav Ashe responds to it. However, according to our approach, in which amor’a [a conventional title of any named post- Mishnaic authority—S.D.] cannot be responding to stam [i.e., to the rabbis in the academy—S.D.], we have to say that the speaker is Rav Nachman quoting Rav Abbahu [also spelled as Abuha—S.D.].10
The illusion on the surface is that the anonymous characters in C are making an inference from Rav’s words, and Rav Ashe responds to it in D. The truth below the surface is that in the genesis, Rav Nachman and/or Rava bar Abuha made that inference incognito. Because Rav Ashe is the one who responds to the inference—Halivni argues—in the literary setup of the text, that inference cannot have been made by the anonymous rabbis in the academy (or “stam,” as he contextually terms them). Who made it, then? It cannot be Rav, for neither Rav, nor anybody else for that matter would make an inference from one’s own ruling. If neither directly to Rav, nor to the rabbis in the academy, the inference thus should be ascribed to intermediary amor’aim—-to Rava bar Abbahu or to Rav Nachman who quoted him, Halivni concluded.11 Of course, ascribing an inference made from the words of earlier amor’aim to later amor’aim sounds straightforward, and on a par with the literary setup of the text. However, things get more complicated, and more telling, when Halivni turns to a similar example in Kethuboth. In that example too, on the surface, anonymous rabbis make an inference from Rav’s words, but in truth of the genesis, the inference was made by later amora’im incognito. However, in this case that explanation of the genesis leads to irregularity, indeed to absurdity. In the text of Kethuboth, responding to an inference made from his own ruling, Rav speaks directly in the first person singular! Again, could Rav have
268 Appendix
responded to an inference he made from his own words? Or perhaps more coherently the inference was presented to him by someone else, and then he responded? In any case that renders Rav aware of the inference. Halivni exclaims: “Behold, even Rav knows that the inference . . . had been made [from his ruling], and [thus] Rav could not have made it himself.”12 What it means, however, is that the literary integrity of the narration is disturbed. Readers have no way to tell if, as it was in the genesis in C and D, later amor’aim made an inference from Rav’s ruling and discussed it among themselves, or if, as it now follows, Rav personally responded to that inference made by someone in Rav’s court. The former option is no longer acceptable, and the latter is paradoxical. The paradox is that Rav must defend his own authority against an inference that is entirely dependent on that very authority! The inability to decide who and to whom is speaking distorts the compositional setup of the discourse. Halivni takes this discrepancy as a symptom, or—so to speak—as a scrap on the surface of discourse. That helps him trace the acts of the stam by reconstructing the sources and showing what the stam did to them. His reconstruction proposes that in the original source in Kethuboth someone in Rav’s court objected to Rav based directly on the authority of the Mishnah, rather than on Rav’s own words. Rav, of course, responded to that objection in the first person singular.13 At a later time, however, “the stam,” that hidden power in the genesis, took that original exchange and inserted it in the inference made from Rav words. The solution is found. Yet Halivni not only reconstructs the original source, or discovers the creative act of the stam-of the Talmud in Kethuboth and in Baba Metzi‘a alike. He also and no less importantly seeks to associate that act with one of the listed characters in the Talmud. His options are either (1) intermediary amor’aim, either in Kethuboth, or in B and C, or else (2) the anonymous rabbis in the academy, whom he also calls stam—in the sense of the stam-in the Talmud. Conceptually, he prefers connecting the stam-of the Talmud with the stam-in the Talmud, rather than with the amor’aim. This preference leads him to his final conclusion that the insertion of the, as it were, “original” conversation in the new inference (derived from Rav’s words in Kethuboth) was made by the anonymous rabbis in the academy, by the stam-in the Talmud. Halivni’s final verdict therefore reads: “It is possible that the insertion comes from the stamm’a de gemarr’a,” meaning from the anonymous rabbis in the academy, from the stam-in the Talmud.14
Talmud Criticism, an Analytical Example 269
Having moved from discrepancy in the literary “tradition” of the Talmud, that is, from an irresolvable problem in an otherwise smooth literary discourse that the sugya presents on the surface to the original “source,” from which the literary discourse must have been constructed, he concluded by establishing an association between the “of” and the “in,” as extrinsic as that connection might be. He went all the way from (1) discerning “a scrap” on the surface of the literary discourse, to (2) explaining that scrap as a trace of the action of the stam-of the Talmud in the genesis, to (3) returning that act of the stam into the literary setup of the Talmud through associating the stam- of with the stam-in, the genesis with the surface, the “of” with the “in.” Notably, his movement from (1) to (2) emerged from a stringent analysis of the text; yet moving from (2) to (3) was a matter of the general preference he gave to the anonymous rabbis in the academy over the later amor’aim, which comes from elsewhere. Yet, whatever option Halivni would have preferred, there remains an invariable. He shows a constant inclination to associate the stam-of the Talmud with the listed personae in the Talmud, either with an amor’a, or, as he prefers, with the anonymous rabbis in the academy—the stam-of the Talmud with the stam-in the Talmud, the pure and nameless “of” with the pure and also nameless “in.” The nameless agency of designing or “redacting”—the “of”—conflates with a listed even if nameless character, the pure and anonymous “in.” In more general terms, as Halivni reads the discourse, the term stam functions in yet two other separate senses. These senses have to do with a personal versus an impersonal quality of the stam. First is a compositional sense, in which stam-in conveys a sense of “anonymous personality.” It refers to nameless but active rabbis, recipients, and promoters of the “traditions” in the academy, who are also part of these traditions. The sugya literarily portrays their discussions, even if without giving out the names of the disputants. In that sense stam-in refers to abstract person(ae): teachers and students in the academy, who explicitly argue one with another, as they also invoke “traditions” they ascribe to earlier generations of authorities. Second is the sense of the “origin.” Stam-of refers to the “origin” of the actions that lead to recomposing the initial sources into presently given literary traditions. In the example above, stam-of as an origin of change entails an agency responsible for inserting an inference produced later back in the original refutation presented to Rav. Stam-of is thus not intrinsically personified, because actions might but do not have to be personified, neither need they be linked to a
270 Appendix
separate group of people. Therefore, in the second sense, stam-of is intrinsically disembodied and impersonal: an intelligent agency capable of making changes. In contrast, stam-in refers to personalities, to literarily characters, even if anonymous. What then is a relationship of personal and impersonal aspects in the concept of stam in Halivni’s view of the Talmud? For an answer I once again turn to Halivni’s final phrase in his analysis, “It is possible that the insertion has been made by the stamm’a de gemarr’a.” This verdict not only bridges a gap between the two separate senses of the stam but also indicates the mutual heterogeneity of these senses. Stam is both a nameless but personal and embodied character in the sugya and a disembodied agency responsible for modification of the original text. As important as it is, that modification is not a literary element of the composition. Thus, in the example above, Halivni ascribed the inference to the intermediary amor’aim, yet to him, the anonymous listed characters (rabbis and students) were responsible for presenting that inference as a part of the original source. However, he did not find the acts of inferring or inserting intrinsically connected to either personae. Of course, the rabbis in the academy are only one of the possible personifications and embodiments of the intrinsically impersonal and disembodied acts of making an inference or inserting it in the earlier source. The nameless stam-in gave body and personality to the stam-of, even if the stam-of was only an agency of inferring and inserting, which grounds the literary composition, but does in no way appear in it, except as a scar on the discourse. By way of yet another heterogeneity, stamof is disembodied, but stam-in refers to anonymous bodies of people. What is even more, the stam-in is a traditional character, nameless but capable of having a personal name and even personhood. The stam-of is an intrinsically impersonal disembodied function, for example, that of inserting a later inference in an earlier text. In an even more general sense, one might say that the difference between the two senses of stam is that of nameless but bodily personality (the “in”) versus depersonalized and disembodied function (the “of”), or else that of an anonymous presence in the composition of Talmud versus a complete elision from the composition, except as a scrap on the discourse. By any reading, Halivni’s argument uses the former to give both personification and embodiment to the latter. That charts a normal or balanced heterogeneous setting of Halivni’s conception of stam, in which every “of” is linked to, or
Talmud Criticism, an Analytical Example 271
even converges with, the “in.” Yet, what happens if his theory comes across an irregular case? As I will now show, in such as case, the “of” emerges in pure disembodied and faceless hypostasis, as both a pure action, and a moving force thereof—an action, and force, of montage, as Eisenstein would have perhaps called it. “An Unusual Case”: The “Of” without the “In”
The disembodied nature of the stam-of becomes even more evident as Halivni moves, as systematically as he does in this project, to the second fragment of the literary composition of the sugya, where the rabbis in the academy negotiate the traditions of amor’aim. That again has to do with the name of Rav. The resulting “tradition” is the text F through J. Halivni discovers a compound nature of that, which I graphically represent below with different typefaces. (I have used boldface for the source, underlining for initial insertions, and italics for insertions made later on.) Halivni’s analysis of this “tradition” isolates the ruins of the original “source,” as scattered all over the place in the resulting “tradition” as they are. Halivni also identifies insertions made on the way from the “source” to “tradition.” While this transformation is as multistep as complex, a graphic representation of the scattered elements of the original “source” in the resulting “tradition,” as well as of the insertions made in the process gives a bird’s-eye view, in which the surface becomes a tapestry. —It [the baraitha] has been recited in the presence of Rav, “[The master] gives them their full wages.” Rav said to him [to the reciter—S.D.]. Habibi [i.e., my uncle, Rav Chiya—Rashi] used to say, “If I were there, I would not pay but the wage of an idling worker,” and you recite, “is to give to them their full wage!” —But did not the baraitha [also] stipulate, “But a worker coming with a full load is not the same as coming empty, or doing the work is not the same as sitting idle”? —”They did not finish [reciting] it in front of him.” —In another version, they did finish reciting it, but Rav said Habibi [i.e., my uncle, Rav Chiya—Rashi] used to say, “If I were there, I would not pay anything,” and you recite, “is to give to them as to an idle worker!”
272 Appendix —This time [Rav’s position indeed] refutes [the baraitha]! —No, it does not! This [i.e., the exclamation of Habibi which Rav quotes—S.D.] might have referred to [workers] who did see the field [was wet and thus unsuitable to work—Rashi] the night before, but that [i.e., the baraitha] to [workers] who did not see the field wet.
His reconstruction suggests that in the original “source” Habibi said, “If I were there, I would not pay the workers who did see the field the night before.” If I can jump ahead for just one quick moment to Friedman’s view of this text, Friedman will be in disagreement with that, arguing that the distinction between having seen the field before and not having had a chance to is extrinsic for the baraitha in question and is borrowed by the composer of the sugya from the text that immediately follows, which is the dicta of Rabba (bBM77a), for the sole purpose of creating the continuity in the composition. Returning to Halivni’s reconstruction, Habibi argues that the workers must have seen the condition of the field the night before; they must have known the field was unlikely to be ready for work in the morning. By arriving at the field, they took their chances and lost. “Why should I pay them?” Habibi argues. For Halivni it means Habibi did not disagree with the baraitha, but rather he contemplated a situation in which the baraitha simply does not apply. However, later on, in reconceiving Habibi’s involvement with the baraitha, two concepts were borrowed from that baraitha: paying the wage of an idle worker and paying nothing. Respectively, “I would not pay” became interpreted in two different ways: “would not pay anything” as opposed to “would not pay but the wage of an idling worker.” These two interpretations ensure two different versions of how Habibi diverged from or even disagreed with the baraitha, as opposed to his “original” position of avoiding an application of the baraitha to his case. Additionally, to make him not simply diverge from the baraitha’s point of view but rather refute it, two complementary juxtapositions between his words and those of the baraitha have been made, “and you recite, ‘[the owner] is to give to them their full wage!’ ” and “and you recite, ‘[the owner] is to give to them as to an idle worker!’ ” respective to two interpretations. This enabled presenting his position as an attack on the baraitha, rather than as a divergence from its teaching. Now, Halivni observes that the conversion of what was a divergence into
Talmud Criticism, an Analytical Example 273
an attack is artificial and results from an impulse that resulted in a stretched or forced reinterpretation (dokhak). He wants to identify that impulse. For that matter, he approaches that forced interpretation the stretch as a symptom. He then both diagnoses and reconstructs the process of transformation leading to that symptom, resulted in the “tradition” above. The twofold nature of stam as stam-in and stam-of haunts Halivni on almost every step of this analysis. The heterogeneity of these two aspects of stam becomes even more apparent toward the end of it. To show this dynamic, I will now pay attention not to the result, but rather to the heuristics of Halivni’s analysis.
The Heuristics of Discovering the Pure “Of” Halivni started his analysis by observing a symptom already noticed in both medieval and modern commentaries on the phrase from the J, “This time [Rav’s position, indeed] refutes [the baraitha]!” First, Halivni allies with Ritva (Rabbi Yom Tov ibn Assevilli of thirteenth- and fourteenth-century Spain) who asked why the rabbis in the academy talk about refutation, instead of reading the story on Habibi as evidence of a simple statement of divergence between the positions of Habibi (or for that matter of Rav) and the baraitha? Why refutation instead of difference? Instead of answering this question directly, Halivni takes it as an indication of a change the stam-of made to the original source. He reconstructs what the source initially was and how it was subsequently transformed: “It seems that originally Rav said in the name of Habibi ‘if I were there, I would not give them.’ ”15 However, afterwards—he continues—versions differed in interpreting what exactly Habibi “would not give”—”anything,” as in the second version, or “but a wage of an idle worker” as in the first. Halivni finds a factual recognition of that change in the original source in the commentary on the Talmud in the seventeenth-century German pilpulist, Abraham Schapiz, Male Ratzon. However, Halivni’s evaluation of the extent of interpretive additions goes even further than Schapiz’s: “[Schapiz] did not feel that if so, both “and you recite, ‘to give to them their full wage!’ ” in the first version, and also “and you recite, ‘to give to them as to an idle worker!’ ” in the second, belong to these interpretations as well.”16 In more general terms, he thus uses the sensitivities of medieval and modern
274 Appendix
commentators about the text, but takes them in a different direction: from apologetic interrogation to critical analysis. To support that move, Halivni argues that without the interpretive additions the commentators felt and he diagnosed in a more precise way, there would be no evident link to the baraitha at all, and Habibi would sound like someone who simply suggests that the Mishnah simply does apply to the “there” or to his case. Habibi thus neither differs in opinion with the Mishnah, nor disagrees with it. Without these interpretive additions, Habibi would sound, as he did originally, like he “does not want to pay, and tries to justify it” by saying that the workers “did see the field was wet and thus unsuitable to work.” In sum, according to Halivni’s reconstruction, the original argument of Habibi had nothing to do with the baraitha at all. Instead, Habibi was looking for a way to escape applicability of that baraitha to his case, by saying, if I had been there, I would not pay the workers at all, because they must have inspected the field the night before and must have known it was wet and thus not ready for work. (Again, Friedman’s interpretation is at odds with this point, and thus with the whole reconstruction of the “source,” except for a general acknowledgment that a work of connecting disconnected text is going on.) Yet in Halivni’s view, what originally was a personal statement by Habibi to avoid applicability of the baraitha got chopped and cut and regrouped and added interpretations to in order to produce an attack on the baraitha and a defense against that attack. Habibi thus transforms from a practitioner of the law who argues the law would not apply to his case to a legislator who either has a legal opinion of his own (a sign of a higher authority), or attempts refuting the authority of a given regulation of the law (a sign of a lower authority). Halivni’s question is, “Who is responsible for that metamorphosis?” Who made Habibi transform from a practitioner of the law, as he is in the “source” to the legislator as he is portrayed in the “tradition”? Unlike what he did with the first fragment, in answering that question Halivni identifies no personal agency responsible for the transformation, at least not among listed literary characters in the sugya, named or anonymous. He limits his analysis to tracing and reconstructing the transformation, and he does not look for an agent of that transformation. This time, it all has to go in passive voice, for what is reconstructed is pure action, not the agent of that action. I would say, similarly to the rhetorical genre of centon, the
Talmud Criticism, an Analytical Example 275
clauses of old text or “source” had been used to produce a new composition or “tradition,” which, as the genre requires, is to be opposite to the original meaning. In this case it means from a practitioner to a legislator, and from a legislator of a higher rank to that of a lower rank. The new “tradition” made Habibi either change his position from that of the baraitha or even help Rava attack that baraitha. The passive voice was used again with chopped phrases of the original used to compose an attack and the defense; there was nothing, not even divergence, but only simple avoidance. However, unlike in the first, more usual case, Halivni points out nobody, no person—anonymous or not—to respond for that transformation. What he has is a pure action, a pure stam-of. This outcome of analysis is at odds with his previous case. Halivni discovered the transformative action or stam-of but unlike before did not link that action to any person in the Talmud at all, neither to stam-in nor to a later amor’aim. Having reconstructed the source and process of its change, Halivni came to the truly striking finale: “In front of us is an unusual case, a text in the Gemarr’a that looks like stam [stam-in, or the rabbis in the academy, on my reading—S.D.] but in fact is a part of a dictum [of an amor’a, Habibi—S.D.].”17 Usually the “source” transforms into a “tradition” by virtue of stam-of, and the transformation gets ascribed to a supposedly later group of people, stam-in, who do all they could to preserve the dicta. However, in this exceptional case the stam-of destroys the dicta and thus shows its wild nature. Therefore, in this unusual case Halivni must stop in the middle, because the transformation went against the standard literary roles of the stam-in, which was either discussing the dicta or the amor’aim, or developing arguments on their behalf. But going back to Halivni again, none of these roles tolerates transforming original dicta into new anonymous traditions, or more specifically allows transforming the dictum of Habibi into the direct speech of the rabbis in the academy. These same rabbis cannot foster the role, in which they not simply give a new look to a old source but also directly expropriate some words of an amor’a. In sum, because the dicta transforms in stam-in, the stam-in can not be held responsible for that transformation. As a result, as unusual as it is, Halivni makes no attempt to link the transformation of the source into a tradition to any listed personae in the Talmud at all. Instead, he must limit himself to describing the pure act of a disembodied stam-of, without linking it back to a flesh-and-blood stam-in.
276 Appendix
Halvini’s Notion of the Source (Makor) as Ursprung: The “Of” To see the stam-of destroying a dictum, and therefore not becoming a stam-in, is an “unusual” sight in more than one sense. First, Halivni shows that as a rule what rabbis in the academy ascribe to amor’aim as their “tradition” in fact results from a recomposition of the “source” by the stam, that is, as Halivni alleges, by these very rabbis. However, the case defies the rule. What compositionally looks like the tradition of stam, in the original source was a fragment of a dictum of an amora! Somebody must have taken the words of Habibi and given them to the rabbis. If “somebody” is stam-of, this is a rare moment at which the stam-of and stam-in disentangle and differ to a degree at which the stam-of creates the stam-in, so that they obviously can no longer be one and the same. Second, Halivni’s verdict is unusual for yet another, more subtle, but nevertheless striking reason, which comes as an implication of the first. The exceptional case helps explain the rule better. This unusual case reveals that the concept of “source” includes not only the original data, that is, the dicta of the amor’aim, but rather also, for a lack of a better word, a power, virtual as it is, of recomposing the dicta in a new form. “Source” thus refers not only to the original dicta, but also to the “source” of the transformation. Even if Halivni does not seem to be using this idiom of German philosophy explicitly, makor or “original source” emerges here as Ursprung, or the transformative force, a form of the virtual that reshapes the dicta into a new amoraic “tradition.” It means that the concept of the “source” of the “tradition” refers not only to the initial material later transformed but also to the moving—virtual—force of that transformation. Therefore, a truly “unusual” part is that the virtue or force of recomposing has no connection with any recognizable literary personality whatsoever, because no literary stam-in can treat an amoraic dictum with such a force as to abuse the authority of that dictum by detaching and expropriating the material thereof. No literary personae (or stam-in) can cut the dicta in parts and use the fragments to produce a new composition on behalf of themselves. The stam in the first and in the second sense, the “of” and the “in” never meet here, and this is what makes the case truly “unusual,” revealing the stam-in as part of the source, now conceived as Ursprung. The “unusual” only helps explain the general or “usual” better. Recomposing the fragments of an original dictum in order to invert the meaning of
Talmud Criticism, an Analytical Example 277
it (in this case to present Habibi as a legislator as opposed to a law practitioner) is truly the work of the stam in the second sense, the work of the “of,” now put in the open, because of having no “in” associated with it. This pure stam-of is a disembodied composing agency, force, virtue, Ursprung, linked to no particular character in the composition. This is why Halivni does not personify the virtue or the agency of this transformation. That “unusual case,” in which the “in” and the “of” never reconnect, helps reveal a more general rule, which consists of two parts: (a) The “source” of a “tradition” implies not only a dictum but also a disembodied virtue or quality of transforming that dictum into the “tradition,” which in turn gets discussed by the anonymous bodies in the academy; and (b) the “in” and the “of” are as heterogeneous to each other as the disembodied virtue of transformation and the flesh-and-bloods can only be. Last, even if Halivni’s conceptual apparatus goes away from a linear model of direct transmission from a source to tradition at hand, his refusal to link the “of” with the “in” in this “unusual case” shows that he still fosters the belief in the original dictum of an authentic amor’a, which in a way is a remnant of what Friedman criticizes as a fundamentalist view that the Talmud is a direct transcript of discussions in the rabbinic academies in late antiquity.
From Halivni’s “Stam-Of” to Friedman’s “Master of the Sugya” Juxtaposing Halivni’s reading of the same fragment of text with that of Friedman’s helps us understand both of the approaches better. In particular, it highlights that things that are “unusual” for Halivni are “usual” for Friedman—not because of any difference in the bare text they read, but because of the conceptual apparatuses they use. What is more important and promising is that Friedman’s and Halivni’s perspectives in reading the same text not only diverge but also intersect, thereby further helping approximate the “who” that “remembers” in the Talmud. Residing, as it does, in a place in which no personality and no ego are tolerated, while the thinking process is essential, Halivni’s pure “Stam-of” already approximates the “who” that remember(s) in the Talmud. In turn, in Friedman, paradoxically personifying cause and effect of the textual transformation in one and the same figure of “the master of the sugya,” removed as it therefore is from either on-or offstage personae in conversations, “the master” makes for another approximation.
278 Appendix
These scholars still share a subscription to Maimonides’ portrayal of the Talmud as primarily a series of refutations and defenses—either as the standard of apologetic thinking, in Halivni, or as an order, and in particular time- order, of the resulting composition. This fundamental similarity makes their diverging approaches not only competing but also complementary, even if, as I will show, their specific textual interpretations may be in conflict. With these two approximations at hand, one can describe the remembering “who” of the Talmud as an impersonal place of thinking—a place that is, as Augustine would have it, full of “hands of heart” rather than either “eyes of mind” or, more abstractly, of “thinking egos.” By necessity, which I will explain, this “place” happens in no time or in a time of the “who” that has no “place,” which helps discerning further details about the “who” that “remembers” in the Talmud. Let me begin with giving more necessary details about Friedman’s conceptual apparatus, including his “default” way of seeing the symptoms of transformations in the resulting text, as well as the way in which he treats deviations.
The Old and New Talmuds For Friedman, the result of designing is a composition of “the new Talmud.” The new Talmud comes into being as a result of literarily recomposing a thereby assumed antecedent composition or “anterior Talmud” (talmud kadum). The anterior composition must have included apocrypha attributed to the sages of the Mishnah as well as dicta of early post-Mishnaic authorities, amor’aim. Collectively these earlier elements form the anterior composition, a set of texts that were only loosely thematically associated with the Mishnah. That assumption of “the old Talmud” in Friedman is primarily a literary- formal concept and only secondarily a historical claim. Intrinsically, it plays only a functional role in Friedman’s and his school’s critical analysis of the text, which means it may but does not have to have extrinsic social-historical bearing. Yet, notwithstanding any projections of the concept of the Old Talmud into history, any access Friedman can have to it is only through a theoretical reconstruction of the literary process of its transformation into the New Talmud.
Talmud Criticism, an Analytical Example 279
Transformations: Migrating Texts between Forms For Friedman, Talmudic texts come in three forms: apocryphal traditions attributed to the sages of the Mishnah (baraitha); sayings attributed to the names of post-Mishnaic authorities, amor’aim (memra or dicta); and anonymous arguments of refutations and defenses (“stam-talmud” or “give-and- take” or “anonymous talmud” or, for brevity, “talmud” as opposed to “the Talmud” with a capital “T,” which in turn refers to either the New Talmud composed of apocrypha, dicta, and the anonymous talmud, or to the Old Talmud, which includes only the former two textual forms. Friedman is interested in isolating these three forms in the resulting composition, as well as in tracing migrations of the textual material from one form to another in the process of transformation. He isolates and undoes these migrations, resulted in reconstructing the Old Talmud and the way from the Old to the New. Texts migrate between forms. Methodologically, it means that the in-text or diegetical characters and their arguments never mingle with the standpoint of either the modern researcher of the forms or of the ancient “master”—the designer/composer of them. Both the designer and the researcher look at the forms of the text; however, the characters reside in diegesis of the text. Friedman is interested both in migrations of texts between forms and in the overall design embracing these forms in one composition. Focused on forms and migrations, he much unlike Halivni is therefore not interested in agencies and even much less in agents committing these migrations. Practically. it also means that, unlike Halivni, Friedman always looks at the forms in separation from their textual content, which keeps migrations of the texts between forms out of reach of any of the in-text characters populating the forms, and open exclusively to the researcher and “the master.” It also means that since Friedman rigorously differentiates between the texts and the forms they assume, he also differentiates between in-text positions and the position of the researcher of these forms. In contrast, Halivni is interested not only and not primarily in the forms or transformations of the material but also and more importantly in the agency and agent(s) causing those transformations, which he also often identifies with in-text characters, for, as I have already shown, strongly unlike Friedman’s distinction between the master or designer and the anonymous talmud, Halivni often associates the agency of transformation with in-text characters, stam-of with stam-in.
280 Appendix
And yet, and still, for Friedman, the designer of the Talmud must also be a literary character of its own kind. Perhaps there even is an affinity between that character and the modern researcher who discovers the fact of design and the transformations leading to it. On a theoretical level it means that Friedman implies continuity of his theory of rigorous differentiation between forms of the text with the practice of the composer of the Talmud. He undoes what he thinks the master must have done. Similarly, Halivni redoes what he thinks the redactors have done. Friedman argues that a formal distinction between the dicta of amor’aim and the apocrypha of the sages of the Mishnah is strictly maintained in the Talmud, even if there is no difference in content.18 Rabbis in the Talmud also show awareness of this distinction, he highlights. The reason for that, Friedman contends, is that “the student will feel transitions from a tann’a to an amora and from amora to the words of ‘talmud.’ ”19 Friedman uses the term “talmud” here not in the sense of a book, but in the sense of “words of expounding” done by the anonymous talmud, that is, the teachers and the students producing the series of refutations and defenses. “The words of talmud” for Friedman are “words,” not people, “words,” not “stamm’aim.” Without this tripartition (tann’a-amor’a-talmud) everything in a sugya will be one and the same, as he stresses. Per some later commentators, he contends that the distinction between dicta and talmud is not merely formal but also hierarchical in terms of authority. Dicta have more authority than Talmud.20 Friedman proposes: Whoever wants to inquire about the composition of the sugya is to use the same formal distinction [between apocrypha, dicta, and talmud—S.D.] in analyzing literary composition of the sugya and the relationships between its parts— because a sugya is not one cohesive oration, neither is it a dress cut from one piece of leather, a s i t l o o k s to a n a i v e r e a d e r a t t he f i r s t g l a n c e [spacing is Friedman’s—S.D.]—it is upon that inquirer to isolate these three components with precision, and to figure out their relationships.
In Friedman’s theory, what is important is that the partition of texts into the forms of apocrypha, dicta, and talmud in the New Talmud reflects no more and no less than the present, modern form of the Talmud in which these sources appear, not the form in which these sources might have originally been.
Talmud Criticism, an Analytical Example 281
You should pay attention to this: marking the source in the present edition as an apocryph as opposed to a dictum tells only how the source appears in front of us: in the form of an apocryph or in the form of a dictum, regardless of whether the [form of the] source changed—either a little or a lot—as compared to its earlier forms, because such changes [of form] often affected our sources, along with the opposite phenomenon of keeping the form unchanged.21
The resulting principle, which Friedman also supports with examples, is “the form is one thing, and its historical components are another thing.”22 That principle explains his twofold goal of research: (1) “to mark the form” as it presently stands, and then (2) to discover the process of transformation that led to the present partition of sources into the forms of apocrypha, dicta, and talmud. In turn, working toward this dual goal, leads Friedman to his notion of a prior composition of the Talmud, or the Old Talmud. Now, since new forms of the sources may have replaced the old forms, these old forms might have participated in the “original,” “previous,” or “earlier” Talmud, as these three meanings are collectively implied in his term “old,”23 referring, as they do, to a theoretically possible earlier partition of sources into apocrypha and dicta, which a newly added form, that of the “anonymous talmud,” embraces in a new frame.24
Navigating the Transformations How, then, is a researcher to discern different genetic components in a sugya, obscured by stylistic and other exchanges of the materials/sources between the forms? Friedman claims that the “form-distinction between sources is positive from the start, and thus always helpful.”25 That is to say the formal partition of sources into apocrypha, dicta, and “talmud” is found not only in the final form of the Talmud but also from the start: Any source always had one of these three forms. Therefore, the researcher is to establish philologically the modern form of the Talmud by using medieval manuscripts and the first printed editions first, and then analyze the genesis of that form literarily, again by engaging both literary analysis of the textual transformation and any textual evidence of such transformations in the above manuscripts as well as in the literature of commentary of the medieval period. All this is
282 Appendix
limited to analyzing texts and does not require any commitment to either the metaphysical or the historical existence of the agencies of the transformations, much less the agencies behind a given form. Unlike Halivni, Friedman does not consider “talmud,” or “anonymous talmud,” or stam-talmud either an agency (stam-of) or an agent/person (stam-in) but rather one form a source can assume in the composition of a sugya, as opposed to two other forms— apocryph and dicta. In three sentences, unlike Halivni, Friedman does not personify the form “stam” as the agency of “stamm’aim,” but instead talks about stam talmud, the anonymous talmud, as a form along with the other two forms of presenting sources in the sugya, the apocrypha and dicta. He holds that any source had always assumed one such form, even if on the way to final composition, and that form might have changed to another form.26 His refusal to personify the form “stam” as the agency of the “stamm’aim” clears the room for his rigorous literary analysis of the genesis of the composition without predicating it on either the metaphysical or the historical existence of the composer.27
Thus, Who Remembers? “The Master” between Two Times In the works of Halivni and Friedman, neither metaphysics nor discussion of its limitations in application to the critical theory of the Talmud became a theme for analysis, nor did philosophical elaboration emerge in the thought of these scholars. Yet metaphysics entered their work through the back door, shaping, as it did, the foundational philosophical constructs in their respective theories of either historical redaction or literary composition of the Talmud. Explicating these constructs yields important heuristic resources for answering the question of who remembers in the Talmud, even if these scholars never addressed or even implied that question. In addition to the already demonstrated Cartesian ways of thinking in Halivni, philosophical constructs permeate the theories of both scholars in the form of distinctions they drew between being, seeming, and a persistent illusion. For Halivni, the redactors must exist in historical reality, not only in literature (being); and for both scholars the Talmud only seems to be a series of either synchronic or diachronic conversations between amor’aim and their students, but in reality it has gone through the process of either redaction of amor’aim’s teachings (Halivni) or transformation of texts, as they
Talmud Criticism, an Analytical Example 283
migrate between forms (Friedman). Furthermore, synchronic and/or diachronic conversations in the rabbinic academy is not simply an illusion to be explained and thereby removed, but rather it is an illusion that does not go away even when its genesis gets critically explained (persistent illusion). Both scholars, that is to say, move away from the illusory reality of synchronic and/or diachronic conversations in the Talmud. Yet where do they go from there? And can one end up in a different place? Following a model of Platonism, both scholars find behind the illusion of conversations another reality, that of genesis. Albeit the new reality has features similar to the first one. It is also populated by thinking subjects, who, once again, live in time. The redactors live in what for Halivni is historically real time; and for Friedman, composers live in, or indeed entail, the time—and in that sense the reality—of the transformations. In both cases, contrary to models of Platonism, the new reality happens in time, which suggests the question: What guarantees that these scholars’ moving from the illusion does not lead to yet another illusion? And if not toward another world, which again exists in time, where else might their way lead from the illusory reality of either synchronic or diachronic conversations? An answer to that two-part question lies in details. Unlike Halivni, Friedman refuses to explain literary forms by attributing them directly to historically existing personae, individual or collective. Instead, he considers any persona associated with these forms as a part, in particular a counterpart, of these forms—in the same way in which “The Author” is both a part and the counterpart of a novel. Accordingly, Friedman’s refusal to attribute literary forms to historically concrete people applies not only to the named and unnamed characters mentioned or acting in Talmudic discussions, but also, and indeed first of all, it concerns the figure of the composer of the “new talmud,” “the master of the sugya.” To the latter, Friedman does not necessarily grant historical concrete flesh and blood—in part because that composer is both an effect of the resulting composition and, at the same time, a personification of the cause of the transformations or migration of texts between forms. Emphasizing forms and transformations rather than personae, Friedman takes the position of pure formalism. In this position, forms of the text and characters they produce take precedence over any historically real personae the texts might or might not imply. Formalism further leads Friedman to his positive two-step program of analyzing the genesis of the forms. To maintain
284 Appendix
this program he has to draw the already mentioned metaphysical distinction between (1) what seems to be the case in what he calls “The New Talmud” (talmud shelefaneinu), and (2) how the things truly are in genesis, on the way from “the old” to “new.” Along those lines, in a concise formulation of his method, he states: In the light of the phenomena mentioned above [migrations of texts between forms—S.D.] and similar to them—changes in language and style in apocrypha and dicta; traditions of different interpretations; use of earlier traditions of talmud in place, as well as other markers—it becomes clear that the research of the three layers in the Babylonian Talmud does not stop with [the distinctive] marking [of the forms of apocrypha, dicta, and talmud] but rather starts with it.28
This means his research program does not at all stop with marking out where exactly the apocrypha, dicta, and the anonymous talmud are in the present form of the text. Rather everything only starts there, and continues to isolate anterior, earlier, and original, or in one word, “old,” forms of the apocrypha, dicta, and Talmud. Friedman’s construing the “original” as “old” is philosophically complex, for it involves more than one line of time. Indeed, to which time does the “old” belong? His answer necessarily involves and combines two senses of time, or, if one prefers, two times: the time of montage and the time of diegesis. One is the time of transformations or migrations of texts between forms of apocrypha, dicta, and “the anonymous talmud.” In that time, texts move from one form to another in any direction, which gives every text an anterior and posterior form; the latter is “new” and the former is “old.” In that sense, I render it the time of the texts, or, applying film theory, the time of montage. The second sense of time is, respectively, the in-text or diegetic time of discussion in the illusory “here and now” of the conversations. Friedman sees that latter time as a literary fiction, that is, the illusion of conversations, again either directly recorded and synchronic, or happening over generations, that is, diachronic. He strategically resolves to go away from that literary fiction to what for him would be the real time, from diegesis to montage. The two lines of time of “illusory” conversations versus “real” transformations are not only parallel but also incompatible in direction. In the diegetical time of the resulting composition of conversations, the anonymous
Talmud Criticism, an Analytical Example 285
talmud comes first and the dicta and/or apocrypha come in response. In the time of montage, however, the allegedly anterior forms of apocrypha, dicta, and the anonymous talmud go through transformations from one another in any direction.
the two times
The above differentiation between two lines of time is thus based on Friedman’s idea of the time of the literary history of the text, which would be completely heterogeneous to the literary time in the text. Among other places, he outlines this idea in the “Summary.” Differentiating the layers [of both content migration and style borrowings between the forms of apocrypha, dicta, and “talmud”—S.D.] makes visible the anterior order of the Talmud, in which apocrypha and dicta are separated from additions of give-and-take of the words of the stam-talmud, which are posterior in their literary establishing, even if stam-talmud can contain some markers of anterior matters and styles. On the other hand, there is no way to claim that fixing the words of stam-talmud in their current position belongs to the latest phase, to the very last moment hit of the hummer in [the formation of] the Babylonian Talmud in its entirety. [Rather,] of course, the latest were the acts of ordering and redacting; in these acts the whole sugyot were installed in their new places, after the languages of give-and-take of stam-talmud were already entirely set.29
In the “anterior order of the Talmud,” Friedman finds apocrypha and dicta, before any additions of give-and-take infused/borrowed from the words of the stam-talmud. I might to highlight: the stam-talmud gets added later, either in the form of direct anonymous discussions, or through infusing such discussion in the dicta. This establishes the time of transformations: from apocrypha and dicta to later recontextualization and/or infusions of arguments into them. In turn, the form of stam-talmud has the internal time of transforming. Shaping the words of the stam-talmud should not be seen as one single act. In several sugyot, we have seen stages of development in the language of talmud
286 Appendix that look and plausibly seem to be the acts of two separate literary composers of the text of the Talmud (baal ha-sugya), when a later composer attacks the sugya produced by the preceding composer, and continues that sugya.30
The literary form of stam-talmud also has its time of transformation, so that later “layers” both renegotiate and continue the earlier ones. I should add in brackets that, in Friedman, promoting “attacks” and defenses is the moving force making the transformations progress. In the larger scheme of things, this force takes us from the anterior forms to their version in the final partition of the Talmud into three forms—apocrypha, dicta, and stam-talmud. Having outlined the positive thesis about the “real” time of transformation as opposed to the in-time of Talmudic conversations, the “Summary” continues with the negative thesis or explains the polemical agenda involved and polemical situation with which it deals. In this context Friedman clarifies how his conception, which I rendered as the conception of persistent illusion, helps him establish himself in the polemics, and to contribute to the side, which historically puts dicta first and anonymous talmud second, without compromising another side, which literarily puts talmud first, and dicta second. The notion of persistent illusion helps Friedman mediate between the historical reality of genesis and the literary illusion of presentation in favor of the first, but without canceling the efficacy of the second. Dicta of amor’aim historically came first, but in the world of literary fiction, these dicta come second, after the anonymous talmud. However, the author of Dorot Ha-Rishonim31 [1901—S.D.] and his cohort increasingly felt inclined to an idea of “anterior stams,” to the extent that some wanted to consider the anonymous texts (stamot) as the fundamental framework, to which the dicta had been added and interlaced at a later time! At the start of the [twentieth] century, researchers were hitting each other [oscillating] between two approaches—the anonymous texts preceded the dicta or dicta preceded them, when the second approach prevailed. In our opinion, the decisive factor in this debate can come from a stronger differentiation between frameworks. That leads to the conclusion that it only seems that amor’aim responded to, and reacted on, the words of talmud, or negotiated [with] them.”32
In the final move in the “Summary,” Friedman shows the continuity of his position with the traditional medieval methodologies of the Talmud.
Talmud Criticism, an Analytical Example 287
Friedman is right about this continuity more then he directly admits: He and other scholars appropriated not only the specific concern outlined in the next paragraph but also a more general take on the Talmud as primarily a rhetorical exercise in refuting and defending or “give-and-take,” as opposed to any other possible understanding of the final composition. However, his direct recognition of the continuity with the tradition concerns only questions about the order of the dicta and the talmud. As he highlights, “Already the traditional methodologists of the Talmud wondered ‘how can we say that amor’a responds to stam gemarr’a?’!”33 That question already suggest, as it were, or at least foreshadows, Friedman’s answer in his theory of persistent illusion: The words of stam-talmud are in most cases not a part of the primary framework in which the words of amor’aim first appeared. Rather, stam-talmud is a secondary, interpretive, framework in relation to the words of the amor’aim. In the conclusion of his argument, when Friedman has fully established the main principle of genetic anteriority and literary posteriority of the dicta, as well as introduced his main methodological solution, that is, of a sharp separation between the forms of material along with mapping any transformations leading to the final partition of the Talmud into the three forms, Friedman addresses exceptional or difficult cases, as well as procedures for deciding them. His language here speaks for itself: Yet, this in no way suggests making “most cases” an absolute rule assumed from the start, or the abstract principle allowing no exceptions. In such a massive literary phenomena, with which we deal, it is quite usual that processes are not totally uniform, and there still is more work to do, and for new elucidations in all these areas. Anyhow, in the majority of cases, borders of dicta and their delimitations are totally clear. In a certain number of places, in which an [additional] clarification is needed, [researchers] should use as many criteria as possible in order to come to a differentiation between components of the sugya in as plausibly as possible. [To that end] one is to conduct a fundamental investigation of the languages used for any given matter [across the board—S.D.]. A general characterization of the language in the Talmud can help do this—that is to say, significant linguistic uniformity in Talmudic terminology—as well as linguistic cliché in dicta as opposed to cliché in Talmud—[make such a fundamental investigation possible]. In a small number of cases, there still are severe difficulties in securing a reliable
288 Appendix differentiation [between apocrypha, dicta, and talmud], and there remain divisions based only on conjectures in marking out the dicta, when the latter stand in the full context in the language of a sugya, which allows readers to investigate the matter independently and to consider all possibilities of a solution.34
The researcher of the Talmud in Friedman thus must separate the forms of the material and map any transformations or migration of the material between the forms. The result allows both to explain why, in the final literary setup of the Talmud, the form of talmud provides the primary framework of the composition, while in the literary-historical genesis, the forms of apocrypha and dicta come first, featuring the anterior Talmud from which “the master,” the designer of Talmudic discussion, arrives at the final form, as he moves, along with Friedman in the genetic time of transformations, as opposed to the in-text time of resulting literary composition. Drawing this distinction also explains Friedman’s careful conceptual separation between anonymous composer(s) of the Talmud (“the master”) and the anonymous talmud (stam-talmud, either the nameless participants of the give-and-take conversations, or those who covertly invade the dicta). “The master” is responsible for the process of transformations; the stam-talmud is only one of the forms established in the result. What, however, remains underemphasized, albeit implied, is that the anonymous composer of the conversations must also go through the stages of development, while, as is the case with every composer of every composition, the composer remains a function of the composition, just as, unlike to what stably seems, the author is and will always be a function of his or her work, rather than what he or she seems to be—a “natural”-historical being that “precedes” and “creates” the work. Since being an author, or a composer, is no less performative in nature than is the composition, the former must evolve with the latter. Importantly, that renders the time of transformations a literary time of the second order, rather than an absolute chronological time of history. Nevertheless, this literary rather than historical nature of the history of the Talmud’s composers stands not at the center in my juxtaposition of Friedman and Halivni, because Halivni does not entertain the second order of time in literary-historical genesis, but rather intertwines the in-text time of composition and the time of composing, in two respective senses of stam (stam-in and stam-of) as closely as possible. And yet, and still not all the cases allow
Talmud Criticism, an Analytical Example 289
that conflation of times and anonymous agencies. As I will soon show, at least one such exceptional case comes into the crossing light of Halivni’s and Friedman’s views, and will serve as a point of both convergence and divergence of their approaches.
do parallel times intersect?
Yet, the relationships between the diegetic time of illusory conversations and the transformative time of montage is even more complex than that of two parallel lines. This is because there is yet a third dimension, in which the otherwise parallel lines of time of montage and diegesis intersect. That dimension has to do with the figure of “the master” who paradoxically belongs to both. “The master” is both a literary effect of the resulting composition, and if not the cause, than at least a personification of transformations leading to that composition. Virtual as it is,“the master” therefore belongs to both times, and to none. Understanding the complexity of “the master” positioning not only allows, but also requires, seeing the master as residing in, and entailing, the third time-dimension; call it virtual time.35 What exactly makes Friedman’s “the master” a philosophical construct? Because “the master” resides in the virtual, and in particular in virtual time, “the master” would be most readily a version of a virtual agency. Yet, can that virtual agency be a thinking subject, as Friedman seems to assume? Is “the master” a thinking subject remaining, as it therefore must, the same in both the time of montage and in the time of diegesis, as the notion of the thinking subject presumes? The answer is negative. Because the lines of time do not intersect in any actual dimension, the thinking subject cannot belong to both of them. There are at least two additional reasons for that negative answer. First, just as, per the analysis in Chapter 4 above, Samuel Clemens and Mark Twain are not exactly the same, so also “the master,” in the resulting composition, or any alleged composer behind the transformations, is not the same either. “The master” as “The Author” of the resulting composition cannot be the same as “The Master” of the transformations. Second, similar to the above analysis of Halivni’s theory of “redaction,” just as heterogeneous acts of redaction do not require positing a redactor as their cause, even if they allow for the redactor as their effect, so also heterogeneous migrations of texts
290 Appendix
between forms do not require a “master” as their cause, even if such a master may personify these transformations as their effect. Yet, if not a thinking subject, what else can Friedman’s theory of textual transformations entail? At stake is Friedman’s dual treatment of the illusory nature of the conversations in the Talmud. As I will immediately show, in criticizing Bacher’s position introducing textual transformations, Friedman explains the illusory nature of either synchronic or diachronic conversations and admits that despite that critical explanation, the illusion does not go away; he fails to explain, or even to address, the question of why it does not. Friedman’s theory thus philosophically implies a triple time structure: (1) a linear one-directional time of the illusory “real” conversations; (2) a parallel two-directional time of historically “real” transformations; and (3) the virtual time of the “master of the sugya,” where the otherwise parallel linear times paradoxically intersect. Friedman does not compromise this complexity. Instead of abandoning the synchronic/diachronic literary illusory time of the Talmud altogether, Friedman departs from the illusion without abandoning that illusion. Criticizing Bacher, who replaced the “naive” theory of the Talmud as a direct record of conversations between amor’aim and their students with a theory of three hundred years of shaping these conversations, Friedman calls Bacher’s diachronic explanation of the conversations “historization of the naive.” In Bacher’s words, The real framework of the Talmud, however, on which the entire structure was built, was provided by the questions, comments, and discussions which are based on individual paragraphs of the Mishnah, and which are anonymous, or not ascribed to any author. Appended to these passages and interspersed among them are sayings whose authors are named: and this class frequently preponderates greatly. The anonymous framework of the Talmud may be regarded as the warp resulting from the united activity of the members of the academy, and upon which the woof of the Talmud was interwoven and developed [italics in Friedman’s text—S.D.] during three centuries, until its final redaction gave it definitive form.36
In Friedman’s interpretation of Bacher, regardless of when (or, for that matter, regardless of whether or not) the historical amor’aim existed, teachings attributed to their names had been “interwoven and developed” “during three centuries.” Friedman admits that this critical reconstruction is of
Talmud Criticism, an Analytical Example 291
course different from a more traditional and naive belief that the Talmud directly and realistically records clever students offering their clever questions about the Mishnah, with amor’aim responding to them, as if all were happening at once, in one room, within no time, or, which is the same, in eternity, so that amor’aim started answering the moment the student stopped asking. The difference between this naive view and Bacher’s is that Bacher takes a historical-genetic approach. and allows three centuries for the genesis of what tradition considered to be the eternal “here and now” of the academy. Yet Bacher only historicized the naive, Friedman argues. Bacher’s historical view is a diachronic projection of the “traditional” or naively realist view, in which anonymous students came first and amor’aim responded to them “in class.” In terms of a synchronic reading versus a diachronic, Bacher’s historical diachronic reading of the genesis of the Talmud is only a chronological projection of his (inherited) synchronic reading of the literary composition of the Talmud. For Bacher, if in the order of composition amor’aim respond, then in the order of history they must have been “interwoven and developed” in the text later. Opposing this view, Friedman introduces a (de facto Marxist-Hegelian) notion of a persistent illusion, or an illusion that cannot be undone practically, but must be overturned theoretically or “scientifically.” As Friedman states above, “The decisive factor in this debate can come from a stronger differentiation between frameworks. That leads to the conclusion that it only seems that amor’aim responded to, and reacted on, the words of talmud or negotiated [with] them.”37 Holding the in-text time as persistent illusion, and thereby opposing Bacher’s projection of that illusion into the literary history of the Talmud, Friedman reads the literary in-text time as ultimately illusory, even and precisely because that illusion does not go away when explained. For him, we cannot correct it, even if we know it is scientifically wrong. Missing from his analysis is an answer to the question of why the illusion persists even after it has been explained as an illusion. However, even without asking about the source of stability of the illusion, Friedman’s departure from Bacher is revolutionary both in its scholarly efficacy and in its theoretical subtlety. Most valuably, Friedman shows that neither a synchronic “naive” reading of the Talmud as a record of “in class” conversations between amor’aim and their students nor a diachronic “historicization of the naive” is satisfactory. On the ruins of either synchrony or diachrony as hitherto predominant ways of
292 Appendix
interpreting the “here and now” of the Talmud, Friedman erects his theory of transformation, which rejects the illusion, recognizes its persistent characters, but fails to explain, or for that matter even to ask, why the illusion does not go away when critically analyzed. Why?
In No Time: Rereading the Surface The root of the illusion of a synchronic conversation in the Talmud might be in a certain notion of time that, as critically demonstrated in phenomenology, as well as in Bergson’s philosophy and thereafter, approaches time in terms of space, reducing time to a line on either a flat or, at best, a curved surface. This linear approach is at odds, for example, with Bergson’s understanding of time as duration, or Levinas’s understanding of time as imminence, or Hermann Cohen’s understanding of time as a function of an anticipated but unpredictable future. Unlike these thinkers, both Friedman and those whom he criticized continue using a linear understanding of time, associating the “here and now” of a Talmudic conversation (a) with a certain point in linear time, and (b) with a certain character, either onstage (amor’aim and their anonymous students) or offstage (“the master”). The juxtaposition of different understandings of time explains both why the illusion of synchronic/diachronic conversation is uncritically accepted by “naive” readers, Bacher included, and why the illusion does not go away for Friedman. All these thinkers assume linear time, and a linear “here and now” in the Talmud, thus sustaining the illusion. Can one get rid of that assumption of linear time in reading Talmudic texts? As shown above, the surface of the Talmud can be read as a series of ordered refutations and counterrefutations. This order might but does not have to suggest linear time. In the genesis, that order is not necessarily chronological. In the genesis, including migrations between the forms, counterrefutations may come before and refutation after, and vice versa. What Friedman’s theory adds to this picture is that refutations and counterrefutations deliver a sense of a certain character, who is not explicitly onstage, or implicitly offstage, but who, nevertheless is intrinsically involved in the process of refuting. Friedman’s complex notion of “the master” can help us discern that character. However, to do so, that notion needs to be carefully separated from the assumption of linear time Friedman still has.
Talmud Criticism, an Analytical Example 293
If divorced from residual constructs of linear time, “the master” would no longer reside in a linear “here and now” of the Talmudic discussion, whether it is an illusory “here and now” or not; rather, as virtual as it is, “the master” would represent only a no longer necessary personification of memory, oriented not to the past as a mode of linear time (in which one remembers what is, and in particular what was) but first and foremost to the task of remembering the authoritative tradition better, thereby also reconfirming the authority of that tradition. Dissociated from the linear “here and now,” “the master” would then become an again no longer necessary personification of that a gaze of memory, and would therefore be no longer connected with any particular character or with any singular moment in present. Instead, it would deliver an orientation, a vector directed toward a point, indeed a horizon in the past, which is no longer a part of linear time. Friedman thus helps establish a view in which the Talmudic conversations are not about any linear time at all, but only about a series of refuting and defending and inventing, all serving the purpose of better remembering the Mishnah or other traditions of the past, no matter where and when exactly that act of memory is taking place. The root of the illusion is in the notion of the linear time, or times, and the text of the Talmud should be reread as a pure surface of refutations and defenses, that is, along the lines of what I have done in the beginning of this appendix, in which case neither illusion of synchrony/diachrony of a conversation nor its critical analysis in the theory of transformation is no longer as unavoidable as they sound in Friedman’s reading.
In the Light of Friedman’s and Halivni’s Views By way of a summary and a theoretical conclusion of what has been an admittedly technical comparison, let me highlight the main points these comparisons yield. Comparing reveals not only explicit differences in critical readings of the Talmud but also the implicit but crucial similarities in how the “naive” or precritical reading of the text of the Talmud is construed. To illuminate that complex dynamics, I proceed with a comparative chart of their theoretical approaches to the precritical object of critical analysis of the Talmud.38
294
Appendix
A Comparative Chart of Two Critical Approaches to the Talmud Table 1. Comparing Two Text-Critical Approaches to the Talmud Friedman (in the formalist approach) The text of the Talmud in follows the pattern of its modern form ( קשיא ופירוקMaimonides) understood as a series of discussions between anonymous rabbis and students (“stam-talmud”) in the form of heuristic refutations of the instructions of the sages of the Mishnah and/or of name-attributed dicta of post-Mishnaic authorities, followed by apologetic deflections of these refutations, often with the help of other such instructions or dicta; constructs literary discussions that proceed in the form of anonymous logical/rhetorical reasoning alone; or else they are conducted by anonymous students, who, by way of argument, might both quote and interpret instructions and dicta of authorities of the historical past;
fup-dolgopolski-12app.indd 294
Halivni (in the realist approach) follows the pattern of ( קשיא ופירוקMaimonides) understood as a series of discussions between named rabbis and their anonymous contemporaries (students) in the form of heuristic refutation of instructions of the sages of the Mishnah and/or of name-attributed dicta of post-Mishnaic authorities, followed by apologetic deflections of these refutations, often with the help of other such instructions or dicta; directly represents discussions that were either anonymous, that is, based on logical/rhetorical reasoning alone (including reasoning about disagreements between named authorities), or else were conducted by anonymous students presenting their questions to the named authorities of historical past, in respective chronological times of these authorities;
10/9/12 12:11 PM
Talmud Criticism, an Analytical Example 295 Friedman (in the formalist approach)
Halivni (in the realist approach)
creates an illusion of representing a live discussion between students and teachers in the academy;
provides a realistic representation of discussions in the academy, even if some elements of that reality are hidden and need to be brought to light through critical analysis of the text;
results from intelligent composing, and thereby implies a composer;
results from chronologically later, but historically real discussions in the academy; these discussions implemented an apologetic program of (re)designing early discourses of refuting and defending [which presumably always attempted but did not always succeed in following the principle of the infallibility of the arguments of earlier authorities];
keeps making the impression of reality, even after the genesis of illusion has been explained.
reemerges in its true light after critical explanation of the genesis of the Talmudic discussions; that analysis discovers the role of hidden authorities in discussions.
Critical analysis of the text explains a constructed and therefore illusory nature of the discussions, but does not dissipate the illusion of reality that construction creates.
replaces an as it were incomplete picture of historically real discussions between the rabbis by a more complete and therefore more “real” historical picture of discussions.
296 Appendix Table 1 (cont’d)
Difference in understanding of the role of refutations-defense pattern in stamm’aim versus baal hasugya vis-à-vis amor’aim.
Friedman (in the formalist approach)
Halivni (in the realist approach)
Baal hasugya (the composer) creates a refutation-defense pattern of anonymous talmud. He inscribes amor’aim into that pattern, without assuming that any differences between amor’aim about interpreting a text had to mean a disagreement between these amor’aim about the text.
Stamm’aim (hidden redactors of amor’aim’s teachings) reframe (redact) the teachings of amor’aim, for which purpose they create new frameworks of anonymous discussions (stam) in the Talmud. The stamm’aim assume the amor’aim were always already in disagreement one with the other. In the process of redaction they attempt to make sure all of the amor’aim have valid arguments against each other, thereby striving to represent “original” disagreements correctly.
invariance in the precritical object
In the approaches delineated in table 1, there is an invariable assumption that in the Talmud the main pattern of discussions is heuristic refuting of a given authoritative or normative text followed by apologetic defense against that refutation. Furthermore, the refuting and defending parties meet each other on the same stage in a shared “here and now.” Some of these features survive in the critical reconstructions as well. Although, in these critical reconstructions, the scholars of the Talmud have chronologically placed the “here and now” of the Talmud at much later time periods, they always targeted it as a mode of face-to-face communication in a “historical present.” That resulted in productive and sometimes mutually exclusive critical readings of Talmudic texts, which, as I will exemplify in this book, have not only reassessed the chronological placement of the “here and now” of Talmudic discussions but
Talmud Criticism, an Analytical Example 297
also revealed the process of their historical and literary genesis, and yet at the same time uncritically approached the target of their analysis, the refutation- defense series in the Talmud, as if that series represented face-to-face interactions of speakers who sat, stood, walked, and talked, refuted and defended on one and the same stage, in a face-to-face setting of either historically real or literarily constructed rabbinic academies. As productive as these scholars’ approaches have been, they still show an uncritical remnant of the traditional normative approach (of Maimonides, his rivals, his followers, and his immediate predecessors) to the Talmud, in which (a) the Talmud is both composed by one of the latest name-identified authorities mentioned in it, Rav Ashe,39 and (b) by the same token, is staged as early as in fifth-century Babylonia, or in respective academies of the rabbis of previous generations. Scholars critically denied part (a) of that traditional theory through showing that Rav Ashe was not the composer of the Talmud. They consequently reconsidered part (b) by arguing for a much later chronological placement of the processes of composing the Talmud. Yet one element in the second part remained intact. It also kept an attached string from part (a). The traditional claim that heuristic refutation of a given authoritative/normative text followed by apologetic defense thereof is the main pattern of Talmudic composition remained a working assumption of the critical scholars, as well as its counterpart, a belief that refuting and defending demands that the refuter and the defender meet face to face, in the same “here and now,” on one and the same stage. As a result, even if Rav Ashe was no longer the composer of the anonymous parts of the refutations and defenses in Talmud, and even if his academy was no longer the stage where these refutations and defenses took place, these refutations and defenses still retained the role of the main pattern of Talmudic discussion. They also still had to be staged in a face-to-face interaction between a refuter and a defender, who were to be equally presented on one and the same stage at one and the same time, in one and the same “here and now.” Even if the chronological placement of that “here and now” changed, what has been newly placed at a later time period still remained what it was at an earlier period: a face-to-face interaction of refuting and defending going on between the rabbis on stage. However, the pattern of refuting followed by defending against it does not intrinsically have chronological anchoring. Instead, it requires a
298 Appendix
rhetorical sequence alone: Refutation comes first and defense second, even if chronologically that defense may come several generations later, or even may precede the refutation in historical time (“I have an answer, give me the question!”).40 What it means, however, is that not only in genesis but also in its outcome, the refuter and the defender no longer have to appear on stage face to face. Instead, on the precritical stage of the Talmudic “here and now” there might be either named or anonymous speaker(s) who propose refutations and defenses and recall quotes of the teachings of the earlier masters as a part of the refuting and defending process. More radically, that stage might belong to someone who does not speak at all, but whose only job is to bespeak refutations, defenses, and in that way recall the dicta of past masters as well. For example, a Talmudic formula “Said Rabbi X, said Rabbi Y that Rabbi Z ruled such and such” would not require Rabbi X (much less Rabbi Z) be on stage but rather indicate no more and no less than the presence of a speaker on stage, who makes a point in either refuting or defending by recalling what Rabbi X said. The questions then would be, What can scholars say or know about that speaker? and How should they inquire about him, her, or them: Should they consider him, her, or them as historical figure(s)? Should alternatively they treat that speaker, who says, “Said Rabbi X etc.” as a bare being there, produced as a purely literary effect, a pure human presence without any further emanation of flesh, with refutations and defenses unfolding in front of that presence as it remembers, recalls, and recollects—in the senses which I will momentarily explain—the ideational content of a given text? Should that pure human being there be conceived as literary function of texts, for example, as authors either speaking in these texts or making other characters speak, or as a biographical and historical entity? Approaching these questions may benefit from an even more careful disentanglement of rhetorical and historical-chronological orders. In a purely rhetorical order, a refutation comes first and the defense must come second. However, in historical order of genesis, one may already have the defense, which can be consequently preceded by a refutation, to which the defense would “respond.” More important, navigating these questions requires the careful work of reinscribing the Talmudic pattern of refuting and defending into a larger context of various traditions of the art of memory and/or intellectual practices of memory either in late antiquity or at later periods. Attending to the purely rhetorical rather than the chronological order of refuting and defending allows disentangling the refutation-defense series in
Talmud Criticism, an Analytical Example 299
the Talmud from their hitherto predominant interpretation in terms of “historical present” or face-to-face communication between the parties. Once the refutation-defense paradigm of the Talmud is freed from the demand of synchronic face-to-face interaction, other possibilities of interpreting Talmudic composition open up. Canceling the demand to have all refuters and defenders on stage creates new perspectives for understanding the here and now of the Talmud rhetorically, rather than only in literary-formalist terms. In that context, refuting and defending in the Talmudic discussion reemerge as part of a classical rhetorical technique of “memory for things,” which includes thoughts (as opposed to “memory for words”), in which elaboration on ideas instead of memorizing their verbal presentations becomes an intrinsic part of the art of memory, in which to remember means to understand and in turn to understand means to remember. Needless to say, this context represents a conflation and thus a place of competition between Platonic views of memory as a fundamental feature of knowledge as memory of ideal forms with its rival view of memory as mnemonics or a technique of recollection; these two views were conflated, and competed with each other in Late Ancient traditions of rhetoric. In this perspective, refuting and defending became tools of the art of memory— specifically of the “memory for things.” As a result, the Talmud’s affording direct speech to the teachers of previous generations no longer means that refuting and defending belongs to that generation. Instead, approaching a given tradition by means of heuristically refuting and than apologetically defending that tradition is an act of the “memory for things” which, in Platonic interpretation of it is an attempt to elicit, to remember and thereby to understand things (including thoughts) concealed in the otherwise mechanically memorized words of a given tradition. This has an important implication for understanding the “here and now” of the process of the talmud. The “here and now” of this process is the stage of remembering, on which the teachers of the past are recalled by way of memory, not by way of getting these teachers (much less their ghosts) on the stage in a tête-à-tête. Memory, of course, means here both an “art of memory” in the sense of techniques of effective recalling and memory in the Platonic sense of remembering the forgotten knowledge the soul already had. These two meanings come from the traditions of philosophy and rhetoric.
300 Appendix
Friedman’s analysis thus invites a reconsideration in the contexts of philosophical and rhetorical traditions, where memory, as we have seen, has two generally completing approaches, which, in the case of the Talmud, are conflated rather than compete with one another. One approach is mnemonics, or the art of managed recalling, also called artificial memory. The other is Platonic view of memory as the ontological and epistemological condition of the soul, which sees, forgets, and than recalls ideal forms of the things in the world, resulting in what is otherwise known as knowledge or understanding. In the Talmud the two approaches to memory are present, and in some instances of Talmudic discussion conflate one with the other. A dominant place of their conflation is in the refutation and defense process, in which heuristic refutation of a normative text followed by apologetic defense against that refutation serves the dual purpose of understanding and thereby better remembering the normative text and of confirming that the wording of this text is remembered correctly as well.
The Example of bBM76b Having expounded details of Friedman’s theory in juxtaposition with Halivni’s, I conclude with an example of how the difference translates into the practice of reading. I also propose a possibility of a rereading that text critically in response to irresolvable differences between the two scholars, the outcomes of the analyses the two scholars, which arise, despite their shared view of the precritical object. We have already seen how Halivni read the text in question, and will now see how Friedman’s reading differs. I now map the tapestry of the same text according to Friedman. For convenience I index the text with numbers; and add sections (7) and (8) extending to the next folio of the Talmud, which Halivni did not address in his analysis of the genesis. I again boldface what Friedman identified as text that migrated from (7).
Talmud Criticism, an Analytical Example 301
(1) —It [the baraitha] has been recited in presence of Rav, “[The master] gives them their full wages.” Rav said to him [to the reciter—S.D.] Habibi [i.e., my uncle, Rav Chiya—Rashi] used to say, “if I were there, I would not pay but a wage of an idling worker,” and you recite, “[the owner] is to give to them their full wage!” —But did not the baraitha [also] stipulate, “But a worker coming with full load is not the same as coming empty, or doing the work is not the same as sitting idle”?
(2)
—They did not finish [reciting] it in front of him.
(3)
—In another version, they did finish reciting it, but Rav said Habibi [i.e., my (4) uncle, Rav Chiya—Rashi] used to say, “If I were there, I would not pay anything,” and you recite, “[the owner] is to give to them as to an idle worker!” —This time [Rav’s position indeed] refutes [the baraitha]!
(5)
—No, it does not! This [i.e., the exclamation of Habibi which Rav quotes—- (6) S.D.] might have referred to [workers] who did see the field [was wet and thus unsuitable to work—Rashi] the evening before, but that [i.e., the baraitha] to [workers] who did not see the field wet. —That goes in accord to what Rabba said: [If] one hired workers to dig [ditches] and then it started raining and it became full of water, [then] if he [and the workers] saw the field [wet] the evening before, [then] this is [77a] the loss of the workers. However, if he [and the workers] saw the field in the evening, the loss is on the owner, and he has to give to them as to an idle worker.
(7)
—Also, Rabba said: If one hired workers to draw water [from the river to irrigate the field] and it started raining, [then] this is the loss of the owner; however if the river stopped, this is the loss of the workers. Also, Rabba said . . . [etc.]
(8)
Characters participating in these exchanges are anonymous. The first character (call it “student”) cites (1) in an effort to attack an apocryphal text or baraitha, to which this student responds. Another anonymous character (call it “teacher”) defends the baraitha or rather undermines the attack by arguing (2). In (3) the student is no longer attacking the baraitha but only defends his report in (1). In (4) the student opens a second round. He has an adjusted version of (1). In (5) the teacher validates the attack, and in (6) the student defends the baraitha against it, thereby completing the exercise of attack-defense. At face value (2), (3), (5) and (6) are purely anonymous arguments. They are made by only anonymous characters, and they do not
302 Appendix
mention any names of authorities. However, as we have seen, Halivni’s critical reconstruction of the genesis of that text shows the movement by which the “source,” the words of Habibi, “If I were there, I would not pay him” in (1) and (4) is transformed into anonymous arguments. In turn, Friedman offers a different reconstruction of the genesis. He explains the sugya as a work of a composer who artificially created the sugya to connect the apocrypha with the consequent file of teachings of Rabba, which in “the old Talmud” was immediately preceded by the apocrypha without any link whatsoever. As a result, what Halivni considered part of the source—Habibi’s words on the workers who did see the field the night before in (6)—becomes the composer’s artificial borrowing from a different source, that is from the teachings of Rabba. What in Halivni was already in the source, in Friedman became a part of the latest transformation of the text. Let me compare Halivni’s and Friedman’s readings of this text in the form of a chart (following pages). Despite the irresolvable differences in their critical readings of the text, in both Friedman’s and Halivni’s readings, this text fictionally represents refutations and defenses, which, as intrinsically indifferent to the order of time as they are, emerge, nevertheless, in the form of a synchronic conversation staged in a rabbinic academy. Both scholars recognize that that representation is an illusion that can be explained but cannot be canceled. They thereby allow seeing the power of the illusion; however, they miss the virtual as a register in which that power is formed. Instead, they find a stabilizing factor of that illusion in the “real” bodies of either “anonymous redactors” or a “designer of the discussion.” Taking their approaches back to the context of the practices of memory, I suggest that both “anonymous redactors” and the “designer of the discussion” are imageless, and in that sense virtual, bodies that facilitate and manage the process of remembrance while these bodies are a part but by no means a center of that process, which might not have any center at all. What that means, however, is that the “here and now” on the stage is no longer a rabbinic academy of a certain place and time but rather an imaginary palace of memory in which nothing happens until the visitor enters and starts the work of remembering through refuting and defending, without any strong commitment to either chronological or genealogical order, but rather with a very strong commitment to the task and order of memory.
Symptom
Anonymous defense in (2) against Rav’s attack on the tann’a in (1) is artificial, because it responds to the question that should not have been asked in the first place. Indeed, we do not need to solve the contradiction between Habibi and tann’a, because Rav was reproaching that tann’a for the alleged divergence of his position from that of Rav Chiya = Habibi. Thus Rav insists on the divergence of their positions, not on their contradiction. However, if we approach it as the contradiction between the tann’a and Habibi, and solve it, as in (3), then by the same token we attack Rav by making his reference to Habibi nonsensical.
Friedman
Table 2. Comparative Critical Readings of bBM76b
Comparative Critical Readings of bBM76b Traditional (illusory) reading Anonymous disputants attack Rav’s dictum based on the apocryphal text (or the apocryphal text based on Rav’s dictum).
Halivni Why do the discussants in the Talmud attack Rav based on the apocrypha instead of assuming that Rav and the apocryphal text represented divergent points of view of equal authority?
Talmud Criticism, an Analytical Example 303
It seems that the composer of the sugya was primarily targeting the first teaching of Rabba in (7) and imported it from a different place by way of reinforcing the defense in (6), and by the way kept all the attached teachings of Rabba in (8) as well. However, according to the genesis of the sugya, it appears more plausible that all the file of teachings of Rabba was there from the start, but the composer of the sugya used the very first teaching in the file to find an argument connecting the file with the discussion of the apocryphal text, the baraitha that preceded it.
What seemed to be an anonymous argument in the Talmud is in fact a part taken from Rabba’s original dictum.
Solution
Nature of the solution
Friedman There is no symptom.
What seemed to be an anonymous argument in the Talmud is in fact a part taken from Rav’s original dictum.
Critical etiology of cutting and interpreting the words of Rav (on Habibi) [to create a refutation- defense pattern instead of simple divergence in the positions of sages and/or early amor’aim].
Traditional (illusory) reading
Halivni
Table 2. Comparative Critical Readings of bBM76b (continued)
304 Appendix
What seems is not what was; but when explained, illusions disappear and the true picture of the Talmud emerges (with the stamm’aim as a new, but previously overlooked element in it). Anonymous redactor tend to transform divergences between the sages of the Mishnah and the earliest amor’aim into the pattern of refuting and defending promoted by anonymous disputants in the Talmud.
What seems is not what was, but what seems does not disappear when what was is discovered.
Composer of the discussion promotes compositional integrity of the composition through introducing anonymous disputants’ presence and style (the pattern of refuting and counterrefuting) to create a coherent composition from what was a series of loosely thematically or associatively connected dicta.
Underlying principle of the solution
What seemed to be anonymous words of discutants is reascribed to Rav (locally).
What seemed to be anonymous words of the discutants are reascribed to Rabba (forward).
Halivni
Textual component of the solution
Friedman
Table 2. Comparative Critical Readings of bBM76b (continued) Traditional (illusory) reading
Talmud Criticism, an Analytical Example 305
notes
preface 1. See Ned Markosian, “The Open Past,” Philosophical Studies 79, no. 1 (1995): 95–105.
introduction 1. Alain de Libera offers a detailed analysis of the birth of the idiom of the “thinking subject” in his Archéologie du sujet I. Naissance du sujet (Paris, 2007). 2. Sometimes such technological enhancement of our powers of memory happens at the expense of narrowing the role of memory in thinking, as well as of thinking in remembering. 3. Here and below, the term “composer” renders Shamma Friedman’s term “baal ha-sugya,” literally, “the master of a sugya” or “the master of a unit of Talmudic discussion.” For the convenience of usage in the context of more detailed analyses below, I also render the term as either “designer,” or “composer” or “master.” 4. In this construction, the storytelling parts are singled out and thereby marginalized as literary, rather than rhetorical parts of the text that call for philosophical interpretation in their own right. 5. So it is in the introduction to his Mishneh Torah, as this work will show in more detail on several occasions. 6. Libera, Archéologie du sujet I. 307
308 Note to page 4
7. I particularly refer to nous and related terms, as they appear in Protagoras, Parmenides, Anaxagoras, Plato, and Aristotle. How exactly can the term nous refer to an apersonal process? An example is in the all-important fragments II and III in Parmenides, On Nature. “[II . . . ] Come now, I will tell thee—and do thou hearken to my saying and carry it away—the only two ways of search[ing] that can be thought of. The first, namely, that it is, and that it is impossible for anything not to be, is the way of conviction, [5] for truth is its companion. The other, namely, that it is not, and that something must needs not be,—that, I tell thee, is a wholly untrustworthy path. For you cannot know what is not—that is impossible—nor utter it; [III] For it is the same thing that can be thought and that can be (to gar auto noein estin te kai einai.).” (John Burnet’s translation.) In this example, the question of whether what-is-thought-of must be or could also not be invites two approaches. In the “trustworthy” approach a person relies on the fact of his or her thinking, which renders thinking independent from the person, and in that sense, a-personal. Of course, choosing another approach, the person takes control over thinking to decide whether or not what is thought of is or is not, which the goddess also allows for, even if she considers it the way of confusion. From a different standpoint Alexey Losev, a twentieth-century eastern Christian orthodox philosopher from Russia, privileges apersonal over either personal or impersonal in Parmenides. He does it, of course, from a standpoint of trinitarianism, where true personality emerges not before the trinitarian doctrine is formed. As he has it in his History of Ancient Esthetics: Early Classics,“What is, in Eleatic [philosophers], is absolutely objective. It serves them as a principle shaping the sensual flux, as well as the principle of knowing the material world as an organized, lawful one, and therefore as cognized in a certain way. It is telling that the same fragment (Parmenides, A22) addresses at once falsity of sensations, their connection to the world of becoming, eternity and immobility of the cosmos, as well as the genesis of earth from thick air. There is no contradiction here but only a coherent conception having nothing to do with any subjectivism or any dualism” (Alexey Losev, History of Ancient Esthetics: Early Classics [Moscow, 1963]; in Russian (my translation). On that reading, Parmenides celebrates the discovery of being in cognition, through making this being the organizing principle of everything, including individual perception, but he does not see any sharp contrast between being and becoming. Even if that changes in later Greek philosophies, the apersonal aspect of thinking and contemplating being remains. For example, for impersonal aspects of the notion of nous in Aristotle, see Brentano’s interpretation of Aristotle. In that interpretation personality, if any, would be secondary to the nous. More specifically, nous poietikon involves components of the fundamental act of being “thought of” as opposed to a naturalistic/psychologist approach to “thinking
Notes to pages 4–8 309
of” as a function of a preexisting person addressing a preexisting object. See Franz Brentano, Die Psychologie des Aristoteles, insbesondere seine Lehre vom Nous Poietikos (Mainz: Verlag von Franz Kirchheim, 1867). 8. One might compare the task of “remembering authoritative texts of the past” with a category of basic cultural process in oral cultures—myth, ritual, and so on. Within that comparison, however, there is an important difference to highlight: Along with the “myth,” “logos” is also, indeed very intensely, involved in Talmudic remembering. What is remembered becomes, therefore, not a “myth,” but rather an archive, constantly revised and reinvented through logos, which makes this archive a living archive. 9. Sergey Dolgopolski, What Is Talmud?: The Art of Disagreement (New York: Fordham University Press, 2009). 10. Of course, the primacy of memory over cognition on one’s way to being, as well as the creation of a discipline or practice within which to embody this primacy, is arguably a key goal of Plato’s dialogue Laws. The primacy of memory is also one of the central concerns in Leo Strauss’s Philosophy and Law (Leo Strauss, Philosophy and Law: Essays toward the Understanding of Maimonides and His Predecessors, translated from the German by Fred Baumann, foreword by Ralph Lerner [Philadelphia: Jewish Publication Society, 1987]) or in Heidegger’s task of remembering the forgetting of being. Even broader, “return to memory” is a basic concern of post–World War I modernity. As connected to that move, as the understanding of the Talmudic practice of memory might be, in the Talmud, memory is not as strongly oriented to being as it is in other figures of privileging memory over thinking, wherein addressing what is/was still remains the main stake. It is characteristic that Leo Strauss, following Maimonidean philosophical reservations against rhetorical techniques of memory, or against rhetoric at large, dismisses both Mishnah and the Talmud in his project. 11. I mean here an “oral” or more precisely “recorded” text. I differentiate between recording and writing and refer here to a recorded text, which does not necessarily have the status of officially “written.” Unlike graphical writings, such as, for example, Scriptures or Torah Scrolls in Jewish tradition, a recording can but does not have to be graphical. It may use other media, such as mechanical memorization and/or chanting. What is more, if a graphical media is used, the graphicality of letters, words, or phrases does not serve as a source of meaning in a recording, whereas it does in Scriptures. 12. This has implications for canonicity. The drive to fix formally “written” tradition as opposed to what I am describing through the art of a scribe or a philologist does not fully apply to the Mishnah and the Talmud, because the Mishnah in the Talmud is not canonical in the same sense in which the scrolls of the Torah are. Rather, there is a dialectical interplay between
310 Notes to pages 8–12
memory-in-the-service-of-one-text and memory-in-the-service-of-infinite-texts, which deserves further articulation. 13. I refer here to textual traditions in the Talmud rather then to the Talmud as an object that comes in the form of manuscripts from the Middle Ages and on. Textual tradition quoted and discussed in the Talmud includes, for example, a rendition of the Mishnah, or an apocryphal tradition attributed to a Mishnaic sage, which in Talmudic parlance would be a baraitha; or else “textual tradition” refers to a tradition attributed to a post-Mishnaic authority, an amora, which would then be called in Talmudic parlance, memra or dictum. Of course, the Talmud in its entirety is also a “textual tradition” that reaches us in the form of medieval manuscript with a significant number of textual variations and in different versions. That latter meaning of “textual tradition” has already received scholarly attention elsewhere and, as a form, pertains to the Talmud as a source of medieval and modern authority, rather than to the Talmud as a way of dealing with textual traditions of the past in the sense indicated above. 14. Sacha Stern, Time and Process in Ancient Judaism (Oxford, UK; Portland, Ore.: Littman Library of Jewish Civilization, 2003). 15. Because the gaze is independent of the content it grasps, it could later lead to either selecting frames with a more certain posture in their content or to letting them be a “snapshot” or a “whatever-frame.” 16. The word “descriptive” here means “is described in the text” rather than “can be described in the text.” “Descriptive” means here that a character is mentioned or implied in the text as either present or absent, that is, as acting either on stage or off stage, in a world, which the stage represents. In contrast, a purely performative character, such as an implied author and/or narrator, is neither absent nor present, but is performed as a necessary position for the text to work. Importantly, either described or performed or both, human existence in the text remains predicative, that is, entails a property (humanness) to whoever is described and/or performed in the text. Thus in Platonic dialogues, not only Socrates but also Plato are human: Even if Plato does not necessarily act in the dialogue, he is performed therein as the narrator, who, of course, is human. 17. Bracketed must remain the question whether the gaze assumes intentionality of choosing the content. Whether or not it is a snapshot that one sometimes does not even want to keep or a carefully constructed family portrait, the gaze remains independent of these variations of content. 18. The gaze requires presence, and in particular the absence, of the operator. Therefore, the gaze is not reducible to an automatism of the camera. The latter can photograph but cannot see, can take what-so-ever frames, but cannot either selet the frames, or create a representation out of them. The Talmud therefore should not even begin to seem like a sort of automatism. The
Notes to pages 12–22 311
apersonal and nameless character of the gaze in no way means an automatic “look” of the camera. 19. Such a performative act comes before any subject, which consequently becomes attached to it. For the discussion of the complexity of the notion of the subject involved in this discussion see Chapters 7 and 8. 20. Action can be understood as that of a subject, or an intentional being. Otherwise machines could act in thinking as well, which affords yet another possibility of perceiving the anonymous voices in the Talmud as programmed devices. That would mean embracing de Manian nihilism, based, as it is, on at least heuristic overemphasis on the role of rhetorical devices in thinking (cf., inter alia, Paul de Man, Blindness and Insight: Essays in the Rhetoric of Contemporary Criticism [New York: Oxford University Press, 1971]). I instead do not consider reducing philosophical, Talmudic, or rhetorical approaches to thinking to either one of the three, and, at least for that reason, de Man’s approach to the relationship between rhetoric and philosophy conceived starting from his Allegories of Reading: Figural Language in Rousseau, Nietzsche, Rilke, and Proust (New Haven: Yale University Press, 1979) remains outside of my project. Even if the gaze and the existential I describe involve some machine element, some relationship to human intentionality remains in place. 21. Studies in the philosophy of thinking in the theory of new media (both historically new media, such as writing or portable print, and contemporary digital media, such as computing), and in the Talmud have been artificially detached due to a lack of an overarching interdisciplinary framework of analysis. Had this detachment not happened, questions such as “Who thinks in the Talmud?” that this book proposes might have been of explicit interest to scholars in philosophy and rhetoric in the context I have outlined, as well as in the theory of new media and to scholars of the Talmud. Although the Talmudists have asked this question, they have not had the philosophical tools to understand properly the question that they posed. Nor did they critically ask about the relationship between thinking and remembering in the Talmud.
1. what happens to thinking? 1. Here, my (dis)engagement with Heidegger and his construing of state, technology, and Jews is inevitably limited to setting the broader context of my argument, although of course, closer textual readings of Heidegger’s texts, in particular in his unpublished seminars of the mid-thirties introduced in Faye’s work and in consequent discussions are worth pursuing in the light of the question of the role of Jews in Heidegger’s understanding of the political
312 Notes to pages 22–29
(Emmanuel Faye, Heidegger, the Introduction of Nazism into Philosophy in Light of the Unpublished Seminars of 1933–1935, trans. Michael B. Smith, foreword by Tom Rockmore [New Haven: Yale University Press, 2009]). To me, this engagement should be done only in the scope of the question of the political in the light of, and after having done, the analysis of the positioning the late ancient text of the Talmud vis-à-vis the tradition of ontology, which I address in the present book. 2. In that respect, phenomenological reduction or bracketing of the question of what exists and what does not is always already automatically done in consumption. 3. Cf. Martin Heidegger, translated, and with an introduction by William Lovitt, The Question concerning Technology, and Other Essays (New York: Harper and Row, 1977); Jacques Derrida, Of Grammatology, trans. Gayatri Chakravorty Spivak (Baltimore: Johns Hopkins University Press, 1998). 4. Connecting, as he does, Being with Time, Heidegger would agree that Being is only a possibility and thus, in the course of time, might imminently reverse to another possibility, that of nonbeing. However, he firmly assumes it has not yet. 5. That, of course, does not exclude a possibility of a work of poetry that both expresses or lays bare not only the forgetting of Being, but also the forgotten Being, if such work is read by a philosopher, similarly to how Heidegger reads Trakl and Hölderlin. In that interpretation of Heidegger, memory does not serve thinking but makes memory of Being and thinking of Being the same. That reading accounts for aletheia, truth, as a-letheh, not-forgetting, the suggestion that truth equals remembering or letting Being shine precisely through revealing of its occlusion. In a very fine differentiation from this ontology, the Talmud might also invite reading it as poetry, revealing, expressing, or laying bare, as it would thus do, if not the Being, then memory, and thinking as the way to remember, in its complex relation to forgetting. 6. As Emmanuel Faye shows in Heidegger: The Introduction of Nazism into Philosophy in Light of the Unpublished Seminars of 1933–1935, in an unpublished seminar on Hegel and State in 1934–35, as well as in Heidegger’s late work in Beitraege zur Philosophie, Heidegger adopts Carl Schmitt’s definition of the political as an absolute division between friend and enemy to make that division a result of a more fundamental act of self-affirmation of the subject in its existence. The self-affirming subject here means not an individual “I” but rather a state. If it has a sovereign leader, who proclaims, as he must be able to do, “the state of exception” from that law without undermining that law, such a state can achieve full self-affirmation through separating between friends and enemies. The self-affirmation of subject’s being is thus the foundation of the political, in
Notes to page 29 313
Heidegger, and separation between friend and enemy comes as a result thereof. On an individual level, it means replacing the old notion of human being as opposed to animal with a new definition of human being as Dasein, or in Heidegger’s words “an essential transformation [Wessenswandel] of man from rational animal to Da-sein” (cited in Faye, Heidegger, 274. In that approach, Jews might have been human according to the old definition of rational animal but are no longer according to the new one. They do not make it to new essence of the human not only because they threaten to undermine the Schmittean political dichotomy of friends and enemies by being neither, but passing for either, but rather for an even more important reason. This reason singles Jews out. Christians, in their conception of death on the Cross, appreciate the Messiah’s will to recognize the possibility of death as nonbeing, and thus his appreciation of life as being. In contrast, Jews, in their concept of resurrection of the dead, are striving to overcome death. They are thereby willingly denying nonbeing, which means they have no will to be either. Their life and their being do not come together. This means, in the logic of Heidegger’s ontology of the political, that the Jews are striving for life without being, and cannot be Dasein. However, the discussion of the relation between Heidegger’s philosophy and Nazism is beyond the scope of present work, and I therefore limit myself to this short comment only. 7. Jews as a construct in Heidegger are, of course, not the same as Jews in their historical relationship to the Talmud. Jews, as Heidegger’s view implies, deny what is the most important element of being human: the finitude or “the being to death.” These implied “Jews” are not willing to confront death as such. If Christians, in one way or another, recognize the willingness of their messiah to die, or at least the redemptive importance of that death, Jewish beliefs in resurrection of the body, and more importantly in the eternal life after death in the world to come mean, in Heidegger’s view, that they consciously deny “the will to death,” or refuse to recognize that the world might not be. They therefore cannot appreciate the world’s being either. Tied to the miracle of creation, this belief denies that the world, by itself, could just not be. Heidegger and the German poets, however, see the world as always capable of becoming nothing; if not for their thinking, that would sink into oblivion. The intimacy of the historical relationship between Jews and the Talmud is, however, a completely different issue, having nothing to do with Heidegger’s construct of either Jews or the Talmud, except for the fact of a strong connection between them, which holds both in, beyond, and, as Levinas would suggest, despite Heidegger’s views. 8. “Ackerbau ist jetzt motorisierte Ernährungsindustrie, im Wesen das Selbe wie die Fabrikation von Leichen in Gaskammern und Vernichtungslagern,
314 Notes to pages 29–30
das Selbe wie die Blockade und Aushungerung von Ländern, das Selbe wie die Fabrikation von Wasserstoffbomben.” Martin Heidegger, “Einblick in das was ist. Bremer Vorträge 1949,” in Martin Heidegger, Bremer und Freiburger Vorträge: Gesamtausgabe Bd. 79, III. Abteilung: Unveröffentlichte Abhandlungen –Vorträge—Gedachtes (Frankfurt am Main: Vittorio Klostermann, 1976–2011), 3–77, 27; translated as “Agriculture is now a motorized food industry, in essence the same as the manufacturing of corpses in gas chambers and extermination camps, the same as the blockade and starvation of the countryside, the same as the production of hydrogen bombs.” Heidegger thus inevitably (dis)misses what using his language may be called the “essence” of the “gas chambers.” That “essence” is extermination of what-is-not-but- only-seems-to-be, for example, of the human appearance of the “Jews,” which therefore, on its own “logic” would not be killing. Instead, and therefore, he reduces that “essence” to that of a “motorized food industry.” Heidegger does not necessarily endorse the “essence” of “motorized food industry,” but he does tacitly reduce “gas chambers” to that, thereby masking the other “essence,” the annihilation of what-seems-to-be. However, that necessary and (dis)missed “essence” relates to none of his other examples, not even to “hydrogen bomb,” which only has a power to destroy a being, but not to annihilate a nonbeing. 9. Levinas turns to the Talmud in search of a model of philosophical thinking that successfully resists Heidegger’s philosophy. His move exemplifies and helps illustrate the connection to my own interest in “Talmudic thinking.” I engage with that thinking in the intertwining contexts of philosophical and rhetorical traditions of thinking and in its own right. This relationship is highly nuanced and different from Heidegger’s relationship to “Jews” or “Talmud,” according to the account of a National-Socialist psychologist Dr. Erich Jaensch. For Jaensch, Heidegger’s philosophy is “proper to the Jewish spirit.” As he says, Heidegger has a “manner of thought” characterized as “exactly that kind of talmudic chicanery, which has always been resented as something particularly foreign to the German spirit” (quoted in Rüdiger Safranski, Martin Heidegger: Between Good and Evil, trans. Ewald Osers [Cambridge, Mass.: Harvard University Press, 1998]). I thank Krzysztof Ziarek for pointing me to this text. Jaensch might have complicated the simple binary opposition between Heidegger and the Talmudic “manner of thinking” as construed by the sensitivities of National Socialism, or even help draw a distinction between the “Jews” implied in Heidegger, and the “Jews” implied in National Socialist ideology. However, Jaensch does not relate to Talmudic thought in its intrinsic connection with, and irreducibility to, philosophy and rhetoric as traditions of thought. I am interested in the latter.
Notes to pages 30–42 315
10. See: Emmanuel Levinas, “The Strings and the Wood: On the Jewish Reading of the Bible,” in Emmanuel Levinas, Outside the Subject, trans. Michael B. Smith (Stanford, Calif.: Stanford University Press, 1994), 126–34. 11. Ibid. 12. That happens in particular in his writings contemporaneous with his first encounters with Talmud study, in particular in Time and the Other. See Emmanuel Levinas, Time and the Other and Additional Essays, trans. Richard A. Cohen (Pittsburgh: Duquesne University Press, 1987). This concern continues in his later work as well.
2. ego cogito, ego meminí: i think, therefore i remember 1. This four-faceted line of inquiry is intimated by the difference between the explicit question “Who redacted the Talmud?” and the question “Who thinks in the Talmud?” At the same, in the text-critical approach to the Talmud (for brevity, I will call it Talmud criticism) that difference is occluded by a binary opposition of form (the text) and content (the plot, subject, or content of the argument). Talmud criticism prescribes that the text (form) must be philologically established before drawing any conclusions about the content. At the same time, Talmud criticism posits both the existence and the activity of a thinking (“redacting,” “designing”) subject who is not a part of the content of the text (not directly represented therein) and is not a part of the form of the text, either, because this subject produces (“modifies”) the text. As a result, Talmud criticism cannot directly address the thinking subject (much less subjectless thinking) in the Talmud, and instead both introduces and occludes it under the names of “redactors” or “designers” in/of the Talmudic discussions. 2. See Gilles Deleuze, The Logic of Sense, ed. Constantin V. Boundas, trans. Mark Lester and Charles Stivale (New York: Columbia University Press, 1990). 3. This, of course, is not to deny interiority, but it is to affirm that interiority is available only through an expression thereof in the act of speaking, in which the “interior” mimics itself in expression as the only way to become accessible. 4. See Jeffrey L. Rubenstein, The Culture of the Babylonian Talmud (Baltimore: Johns Hopkins University Press, 2003). 5. This includes all known sayings of the sages of the Mishnah, both included in the Mishnah and preserved in the forms of apocryphal sayings or in the form of a slightly later code, the Tosefta. 6. Heinrich Heine, in his Zur Geschichte der Religion und Philosophie, writes in the third book in a paragraph that begins “Das Fichtesche Ich ist aber
316 Notes to pages 42–46
kein individuelles Ich, sondern das zum Bewußtsein gekommene allgemeine Welt-Ich” (“Fichte’s ‘I’ is, however, not an individual ‘I,’ but rather a universal World-I which arrives to consciousness”). Heine continues, “So wie man sagt: es regnet, es blitzt, usw., so solle auch Fichte nicht sagen ‘ich denke,’ sondern: ‘es denkt,’ ‘das allgemeine Weltdenken denkt in mir’ ” (“Just as one says ‘it rains,’ ‘it shines,’ and so on, so also Fichte should not have said ‘I think,” but rather ‘it thinks’—‘the universal world-thinking thinks in me’ ”). See Heinrich Heine, Heines Werke in Fünf Bänden (Weimar: Aublau Verlag, 1981), 5:113. I thank Annika Thiem for a very productive discussion of that text. For now, I leave aside any implications that this approach to “I think” may have for gender theory, and I do not interrogate further whether Fichte in Heine’s reading is different from Descartes in the same way in which a romantic thinker might be different from a classical one. It is, however, important to highlight a universal and at the same time singular nature of “I” on either reading: Lonely as it is, it requires no other human “I” to establish the certainty of “I think.”
3. through talmud criticism to the talmud as thought and memory 1. An earlier version of the argument in this chapter appeared as a “Methodological Postscript” to my article “Sense in Making: Hermeneutical Practices of the Babylonian Talmud against the Background of Medieval and Contemporary Views,” in Judaic Logic, ed. Andrew Schumann (Piscataway, N.J.: Gorgias Press, 2010), 227–37. 2. Disentangling the notion of the agent serving the process of thinking from that of the subject substantiating that process also dissociates the task of critical interpretation of the thinking practices in the Talmud from finding their chronological-historical explanation. Chronology serves the purpose of historical explanation, which, as Leo Strauss suggests, should come only after having interpreted the thinking practices in the Talmud on their own terms. By the same token, external chronology no longer mingles with hermeneutical analysis of the thinking practices in the text. Instead, chronological contextualization serves the purpose of explanation rather than interpretation. (See Leo Strauss, “How to Study Spinoza’s Theologico-Political Treatise,” in Jewish Philosophy and the Crisis of Modernity: Essays and Lectures in Modern Jewish Thought, edited with an introduction by Kenneth Hart Green [Albany: State University of New York Press, 1997], 181.) Austin draws an almost verbatim similar distinction between interpretation and explanation. See J. L. (John Langshaw) Austin, Philosophical Papers, ed. J. O. Urmson and G. J. Warnock (Oxford: Clarendon Press, 1970), 181–82.
Notes to pages 48–59 317
3. David W. Halivni, Sources and Traditions: A Source Critical Commentary on the Talmud: Tractate Baba Metzi‘a (Jerusalem: Hotsa’at ha-sefarim ‘a. sh. Magnes, Universitah ha-‘Ivrit, 2003). 4. See Alain de Libera, Archéologie du sujet I. Naissance du sujet (Paris, 2007). 5. See Jeffrey L. Rubenstein, The Culture of the Babylonian Talmud (Baltimore: Johns Hopkins University Press, 2003), and Devora Steinmetz, “Agada Unbound: Inter-Agadic Characterization of Sages in the Bavli and Implications for Reading Agada,” in Creation and Composition: The Contribution of the Bavli Redactors [Stammaim] to the Aggada (Tübingen: Mohr Siebeck, 2005), 293–337. These approaches allowed researchers to address reality or “culture” either implied in the Talmudic narratives or hypothecated as a condition leading to the emergence of these narratives. An analysis of relationships between literary and historical approaches to reality in Rubenstein’s notion of the culture of the stamm’aim is yet to be done.
4. thought and memory in the talmud: the ambiguous status of “the author”—and beyond 1. I use The Adventures of Tom Sawyer as no more and no less than an example. I follow Brian Massumi’s take on the strategy of example, which he finds in Agamben’s Coming Community (Giorgio Agamben, The Coming Community, trans. Michael Hardt, vol. 1 [Minneapolis: University of Minnesota Press, 1993], 9–10). Massumi explains, “Logically, the example is an odd beast. ‘It holds for all cases of the same type,’ Giorgio Agamben writes, ‘and, at the same time, is included in these. It is one singularity among others, which, however, stands for each of them and serves for all.’ An example is neither general (as is a system of concepts) nor particular (as is the material to which a system is applied). It is ‘singular.’ It is defined by a disjunctive self-inclusion: a belonging to itself that is simultaneously an extendibility to everything else with which it might be connected (one for all, and all in itself). In short, exemplification is the logical category corresponding to self-relation.” Brian Massumi, Parables for the Virtual: Movement, Affect, Sensation (Durham, N.C.: Duke University Press, 2002), 18. In that strategy example sharply differs from application. One might approach The Adventures as a place in which specialized theories on Mark Twain would be applied and renegotiated. In contrast, the strategy of example allows me to think about both The Adventures and the Talmud as comparative examples of different types. Here I draw very close to, but not quite enter in to, Massumi’s field of cultural studies. Rooted as he is in that field, Massumi
318 Notes to page 59–60
heuristically uses Deleuze’s notion of the virtual as the “empirical transcendental” to reclaim an otherwise bracketed linkage between movement and sensation in the body, which he opposes to connecting “The Subject” and “The Body” through either static or dynamic positioning of the subject. Understood in that way, the notion of the “subject position” dominated the field of cultural studies, Massumi claims, as he attempts to undo that domination by reintroducing movement linked to sensation in the body. With respect to movement of the body, I, however, ally with late Husserl’s view of the body as the foundation of experience. As such, the body neither moves nor stays at rest but rather grounds our experiences of movement, rest, and other transformations in our experiencing of the world (Edmund Husserl, “Grundlegende Untersuchungen zum Phaenomenologischen Ursprung der Raumlichkeit der Natur,” in Philosophical Essays in Memory of Edmund Husserl [Cambridge, Mass.: Harvard University Press, 1940], 307–25). This alleviates the opposition between positionality of the body and its movement, and translates into a less “cultural” and more philosophical and/or, as it will become clear, Talmudic take on the virtual, which I explore. 2. A text-critical reading suggests that “it” may or may not originally refer to this specific fragment of law. In fact, the attack on Rav Yehudah said Rav said in the second refutation strongly suggests the “it” does not originally refer to this fragment of the recited law but rather was used in the first defense by mistake. 3. It is more certain that Rabbi Yehudah is not a participant in this conversation but rather someone recalled in it. However, the text does not explicitly say that, either. Because of the lack of any indication of what is the speech of the narrator (if any) in the text and what is that of a character, one might attempt to read at least the first round of refutation and defense as if a student is attacking the recitant and Rav defends him or her. In this case, “Rav Yehudah said” would be the words of the narrator, and “Rav said” and so on direct speech. Yet such a reading would mean attacking Rav using another teaching by Rav himself, which is a contradiction. Yet, if we still explore the possibility that “Rav Yehudah said” are the words of the narrator, we would rather have to assume two narrators instead of one. The earlier narrator is attacked by a later student or teacher who proposes a better version of the same narrative. In that reading, however, the second narrator has no words to say and would have to remain silent. Introducing the silent narrator means that the second time that “Rav Yehudah said” appears, these are not the words of the first narrator, and thereby helps escape the contradiction. Yet if there is a voice of a narrator at all, and if there must be two of them, and if the second is to be silent, the first narrator could be silent, as well. In this case, both the first and second instances
Notes to pages 60–66 319
of “Rav Yehudah said” would be the direct speech of a character introduced by the silent narrator. This means we are back where we started: The voice of the narrator remains silent in all instances, but Rav Yehudah does not participate in the conversation either. Instead, he is quoted in both rounds: in the first by the student and in the second by the teacher. There is yet another, syncretic reading of this text to exclude. In that reading, in the first round of refuting and defending, Rav Yehudah responds directly to the student, which implies that Rav already directly has responded in the same way to another student. In the second round, another student attacks Rav’s response, and Rav Yehudah again responds by way of repeating what Rav had already said to that objection before. On this reading we have two direct conversations between a student and Rav that are repeated by Rav Yehudah in response to another student. Instead of duplicating narrators, this reading duplicates students. But even if this would be acceptable, this reading still does not explain why “Rav Yehudah said” does not frame each round in full, but only appears by way of response. 4. It is perhaps even more ironical that because the teacher’s citation is thematically much more remotely connected to the text of the law, using that citation from the start would look much less persuasive than it does now, after failing the thematically more obvious citation. 5. Unlike Clemens’s text, in the text of the Talmud, the voice of the narrator is remarkably silent. Nobody says anything comparable to “No answer!” 6. Mishnah and other rabbinical traditions are not Scriptures. They therefore are only recorded on paper or in mechanical memory of reciters, but they are not officially “written” by the scribes the way Torah is. 7. The scholars of the Talmud approach this question from a variety of theoretical premises that I will discuss later in the book. They phrase the question in terms of who the “designer of the sugya” is or who “redactor in/of the Talmud” is. 8. See J. N. (Jacob Nahum) Epstein, Introduction to Tannaitic Literature: Mishna, Tosephta and Halakhic Midrashim (Jerusalem: Y. L. Magnes; Tel Aviv: Dvir, 1957). 9. There is an important theoretical distinction to make in analyzing a text by differentiating between the text as a form of a representational content and as a form of expression. In the interest of what will become important at a much later stage of analysis, it is worth highlighting that text-critical approaches in general do not differentiate between these two concepts of text as form. Seeing texts as a form of representation of a certain reality or “content” only but failing to see that the form has the power to express something beyond the a specific content it represents, these scholars nevertheless find evidence
320 Notes to page 66–68
of the process of thinking expressed in the shaping of these texts. Lacking the notion of a form of expression, these scholars axiomatically associate expressed thinking with different, but always personified, figures of “The Author,” for example with character(s) (such as hidden redactors) in the reality represented by text or with hidden designers whom these scholars positioned as authors who do not have a place in the represented reality but still exist in the represented reality as producers of that representation. 10. The open character of the reference of “it” is sufficiently demonstrated by the fact that in the different folio of the same tractate, 34 a, a sugya that almost verbatim repeats our example is used to discuss a similar, but distinct law about “four portions in phylacteries,” instead of “two portions in the mezuzah.” In addition, in folio 29b, Rav Yosef (genealogically a student of Rav Yehudah) is reported to suggest that the dictum of Rav refers to the parchments of the Torah scrolls, rather than to either phylacteries or mezuzahs. This suggestion only reinforces the relativity of the reference in Rav’s dictum and thus of its original context, which might have been not polemic at all. Since genealogically Rav Yosef is is a student of Rav Yehudah, he is not disagreeing with either Rav Yehudah or Rav, but only clarifies that Rav’s reference does not have to do with the laws of either mezuzahs or phylacteries. The redactors of the sugya there (29b) make the teachers or students in the academy counter Rav Yosef’s clarification of Rav’s dictum by an apocryphal teaching of the sages of the Mishnah, which other students or teachers immediately relativize (and weaken) by citing a contrary apocryphal teaching that speaks on the same theme. According to literary-historical approach, this all might mean that, in the original, the redactors had to deal with Rav Yehudah’s report of Rav’s dictum on the “tittle of the letter yod” in reference to the Torah scrolls per Rav Yosef’s explanation thereof. These redactors also had to deal with two mutually exclusive apocryphal traditions on the same theme (scrolls) from the sages of the Mishnah, each of which both exceeded Rav in authority and was contrary to Rav’s dictum as explained by his student, Rav Yosef. To secure the failing status of Rav’s dictum, the redactors reassigned its reference to mezuzahs (in b Menachot 29a) and to phylacteries (in b Menachot 34a) and also created the sugya, in which Rav’s dictum sounds both necessary and ultimately either surpassed or respectfully disagreed with, but in no way dismissed. An alternative text-critical literary-historical explanation might suggest that, originally, Rav’s dictum clarified one or another or both of the apocryphal traditions of the sages of the Mishnah, yet because these traditions were contrary to each other, his dictum was lacking solid grounds, and the redactors decided to give it a more reliable reference and a stronger context, as they did in 29a and 34a.
Notes to pages 68–73 321
Both explanations suggest that the redactors were simultaneously invisible participants in Talmudic discussions (by acting in them as anonymous teachers and students) and the historical creators of these discussions, who implemented their apologetic program of making the amor’aim effective in polemics (rather than in the mere elucidation or interpretation of teachings of the Mishnah) by making these anonymous teachers and students act as they do. 11. In a more detailed analysis of real examples in the following chapters, I will show where this program fails to conceal the heterogeneity of the notion of the redactors as both historical-biographical producers of the Talmud and characters (doubly) disguised as the anonymous teachers and students in the Talmud. 12. The term “apocrypha” refers here to teachings of the sages of the Mishnah quoted in Talmudic discussions. “Dicta” refers to pronouncements attributed in the Talmud to the names of the masters of the post-Mishnaic period. “Discourses” refers to arguments, discussions, or exchanges in the Talmud whith no explicit attribution, but implicitly attributed to rabbis, students, or teachers in a rabbinic academy, by default in Babylon. 13. As I will argue in later chapters, this conclusion remains open to logical criticism. 14. The divide between literary-realist and literary-formalist groups of text-critical scholars of the Talmud is empirically observable. The distinction between “realist” and “formalist” types of scholar is abstract and theoretical. In practice, both groups of text-critical scholars show both “realist” and “formalist” elements. This is more obvious in the case of literary-realist scholars, for they insist on both the “literary” and the “real” existence of the redactors. However, the literary-formalist approach is also confined to the dynamics of “literalism” versus “realism.” In this approach, the designer or master of a Talmudic composition (baal hasugya) is a literary figure that has no place (direct or indirect) in the literary reality represented by the composition, yet is still implied as a producer of that composition. As such, such masters or composers have an element of literary reality that transgresses the reality represented in the literature, which welcomes (even if does not require) the “realists’ ” claim that the designers of the compositions not only exist in literature but also have existed in history. 15. I will return to the issue of experiential models in chapter 6. 16. The play between “composite” and “composition” cannot be overemphasized in the present context. Note, also, that this is a somewhat complex phrase, both in terms of the relation of characters to reality and in terms of “The Author’s” own relationship with himself. As noted in the original quote, the character of Tom is a composite of three real people. But “composite
322 Notes to page 73
architecture” here is also a self-deprecating (or ironically self-deprecating and hence not self-deprecating at all) joke: The “composite order” in Roman architecture mixed elements of the Ionic and Corinthian, but in nineteenth-century America, it tended to refer to any mishmash of architectural elements (think of a Greek temple with a Mansard roof and gargoyles) and hence offered material for jokes about the bad-taste catachresis of elements—”composite” involves composing and thus applies not only to the character of Tom Sawyer but also to that of “The Author,” as such a “composite order” of Samuel Clemens, Mark Twain, and the figure of romantic author in general. I thank Bud Bynack for pointing out that connection. 17. Of course, in either approach, the question of “The Author” involves a much broader problem of authority, and therefore of personality. That broader problem includes questions about authorship and authority in the Talmud. Who might be the authority in the Talmud? Is the authority in the Talmud associated with a person or a group of people, or is it instead impersonal? Does “The Author,” however understood, have authority? Is authority always a person, even if a fictional person, like an institution or a corporation? Of course, authority (auctoritas) of the law requires a commanding voice, not just a legal argument that is impersonally written down. Otherwise why obey it or consider it binding? Therefore, the quest of “who speaks in the Talmud?” was always a quest for an authority to ground the authority of the text, and the Talmud, speaking in broader terms, seeks to establish the authority of the Mishnah. I address only one, much narrower aspect of the problem of authorship and authority: the way in which the characters in the Talmud verify the authority of the record through remembering that record better. In this aspect, obeying the authority of any personae mentioned in the record (for example, in the Mishnah) depends on the accuracy of the record. In this case the authority is located on the site of the characters who analyze the record, even if it comes from the personal names in the texts analyzed. In that sense, authorship and authority become separated, as the analysis in this chapter further exemplifies. Of course, this complex structure of rabbinic authority provoked much criticism, for example from the Karaite community. (Chapter 10 addresses the notion of authority in that particular example further.) Needless to say, this structure of authority resembles Foucault’s famous essay “What Is an Author?” (Michel Foucault, Language, Counter-Memory, Practice: Selected Essays and Interviews, translated from the French by Donald F. Bouchard and Sherry Simon [Ithaca, N.Y.: Cornell University Press, 1977], 113–39). Foucault offers “discourse” or “episteme” as substitutes for “author,” and “will to power” coming to expression in an episteme as an authority. On my reading, in application to Talmudic texts, the figure of “The Author” is one among other characters working to
Notes to pages 73–80 323
establish the authority of the records of the past. It thus functions in a much more specific way than Foucault’s more general notion of the episteme as the author addresses. Of course, Foucault’s general characterization of authority as an expression of the will to power in the “discourse” applies. However, what interests me here precisely is the structure of this expression, based, as it is, on separating authorship from authority, rather than on any direct expression of authority in the author, or of the will to power in the discourse. The Talmudic separation of the authorship and authority, along with other late ancient versions of such separation, might still be a particular case of what Foucault describes, yet understanding that separation goes beyond any direct application of Foucault’s more general model. 18. Not only are there female characters in the Talmud, Jewish and non- Jewish alike, but also, contrary to prevailing cultural convention, the anonymous speakers in the Talmud are not necessarity gender-specific, so that in that respect “her” is as good as “his.” 19. See Irenaeus, Against Heresies, preface to bk. 1, The Ante-Nicene Fathers: Translations of the Writings of the Fathers Down to A.D. 325 (New York: C. Scribner’s Sons, 1899–1900), 315–16. 20. I thank Bud Bynack for suggesting that reference, as well as for his extremely helpful critique of the first draft of this chapter. 21. Unlike Hegel or Very, according to Irenaeus’s approach, in G-d, authorship and authority cannot coincide, and therefore G-d cannot be simultaneously the author of and the authority behind Irenaeus’s thinking, because that would render G-d a heretic. Irenaeus’s alternative theological model of the relationship between authorship and authority in the divine insists on differentiating between the instantiations of authorship and authority in the divine, thus leading to a complex dialectics of unity and plurality in the figure of G-d. This approach finds a continuation in Trinitarian doctrines of G-d, in which to think and to be a subject of thinking are far from being the same thing.
5. human existence in the talmud: thinking as multiplicity and heterogeneity 1. A. F. Losev, Istoriia Antichnoi Estetiki: Poslednie Veka (Moscow: “Iskusstvo,” 1988), 7. 2. This philosophical theory was at the same time and by the same token a practical performance of living the life of thinking. The compound structure of thinking that Amelius describes for Plotinus corresponds to Plotinus’s
324 Notes to pages 80–81
practice of philosophy as the way of life, a practice in which the philosopher refrained from mirror images of himself in order to not complicate what Pierre Hadot has described as “simplicity of vision.” See Pierre Hadot, Plotinus, or, The Simplicity of Vision, trans. Michael Chase, with an introduction by Arnold I. Davidson (Chicago: University of Chicago Press, 1993). 3. This difference, as I will argue in this book, has to do with the relationship between thinking and memory. 4. Below in this and following chapters, I will complicate that model, and the notion of virtual it implies, by eliminating the linearity of thinking moving back and forth between what is expressed, what is being expressed, and expression. Instead, when loosened from its initially “false” (by De Libera) connection with thinking subjects, thinking, and thereby expression, in the Talmud happens in the virtual as a site that orients thinking, and again expressing toward the goal of remembering. Unlike modeling the process of expression on thinking alone, that orientation does not define its goal well enough to create the sense of a unity, much less ontological identity of what is being remembered, or what is actually remembered, or of the difference between the two. Instead, the task of remembering takes over and tames an urge to decide what truly is (or was) in one single way. The orientation of thinking to memory, and the virtual as a site in which the orientation is performed entails a multiplicity, which embraces not only multiple versions of what is expressed, but also of what is being expressed. This multiplicity is different from the multiple in the Canpantonian sense of refuted opinions as foundations needed to shape the meaning of the text in one’s mind. A single mind stands not at the center of thinking or for that matter of remembering, and the single mind is not the only, and certainly not the first, to populate the virtual as the site where memory is performed. Even if that site of thinking where remembering is performed is not reducible to any set of either homologous or heteronomous agencies that may come to populate it, these agencies are closer to the “hand of heart” (see Chapter 7) in the palace of memory than to the “eye of mind” in the theater of expression. 5. The question of how to interpret these thinking forces in philosophical terms calls for a separate analysis, as well as the question of how that analysis helps explain the specific work of these forces. The most immediately available starting point is, once again, Foucault’s account of what comes to expression in a discourse or episteme: will to power. This is the will to overcome any obstacle, to never rest with a homeostatic balance, but to push beyond to test the limits. Ultimately, this is a Nietzschean concept. The way the forces of thinking work in Halivni’s and Friedman’s analysis are more specific that that: It is not only about testing the limits, but even more about creating the limits to test. First, the forces of thinking create an “epistemic” construction, in
Notes to pages 81–88 325
which any differences between earlier authorities (amor’aim) become irresolvable disagreements between them, and second, the forces produce accounts and interpretations of the dicta of these authorities to comply to that standard. Because the constructive element of the forces, Bergson’s notion of élan vital, and a notion of the virtual that it implies can provide a better framework for an initial understanding of these forces. I for now leave open the question of whether the notion of the virtual implied in élan vital best describes the forces of thinking in question. 6. See Hayim Lapin, Jews, Antiquity, and the Nineteenth-Century Imagination (Potomac: University Press of Maryland, 2003). 7. In the Appendix, I address an example illustrating both empirical and theoretical complexity and heterogeneity of the in/of positioning of the allegedly homogeneous agency of the redactor of a sugya in Halivni. 8. See J. N. (Jacob Nahum) Epstein, Introduction to Tannaitic Literature; Mishna, Tosephta and Halakhic Midrashim (Jerusalem: Y. L. Magnes; Tel Aviv, Dvir, 1957). 9. See Jeffrey L. Rubenstein, The Culture of the Babylonian Talmud (Baltimore: Johns Hopkins University Press, 2003). 10. For an analysis of implied and thereby constructed reality in the Talmud see Jil Klein, “Non-Canonical Towns: Representation of Urban Paradigms in Talmudic Understanding of the Jewish City,” Studia Rosenthaliana 40 (2008): 231–36. 11. Answering these questions also involves a more refined determination of status of any agents and agencies involved in thinking in the Talmud as virtual rather than either fictive or real. I will not suggest that simple replacement of the alleged thinking subject with a virtual agent solves the problem. It does not, even if that subject is construed as G-d. But I do suggest multiplicity of virtual agents and agencies in thinking and remembering as coparticipating factors in what is thinking and remembering in the Talmud. 12. By way of contrast, various versions of the figure of “The Author” in modern times shed light on different aspects of late ancient ways to connect “The Author,” thinking, and remembering. In modern views, “The Author” functions as the privileged and sole site of thinking. That position is possible only because “The Author” is imagined as a homogeneous, rather than heterogeneous entity. For example, Mark Twain can be imagined as either as real as Samuel Clemens or as fictional as Tom, but never as a site for these two heterogeneous positions together. 13. “United plurality” means a sum, if not a mob, of individuals all making for one person’s thinking and acting.
326 Notes to pages 89–93
14. I have already highlighted heterogeneity in this concept of Halivni’s under the rubric of the aporia of “in/of” in Chapter 4. I will further show that heterogeneity in an example analyzed in the Appendix 15. That’s in effect what we’ve argued above: Hegel’s way of detaching thinking from the subject is possible only because thinking is reattached to a different subject, “Spirit.” 16. Granted, unlike Clemens’s “The Author,” Heidegger’s Dasein takes another option in responding to being. Dasein philosophizes by deconstructing boredom and anxiety, or openly confronts nothingness through the will to death, and thereby appreciates its own being, and by extension the being behind all other beings. Yet the difference is of implementation, not of structure. Both “The Author” and Dasein happen as multiplicities, in which going one way is predicated on having the other, incompatible, way both available and not taken. 17. Aesthetically, structural similarity between Clemens’s “The Author” and Heidegger’s Dasein becomes even more apparent in The Adventures of Huckleberry Finn, where “The Author” assumes the point of view of another character, Huck, and explores the world through Huck’s eye and mind. 18. Note at least a brief comparison of the model of multiplicity and heterogeneity in the Talmud and Bakhtin’s model of polyphony in the poetics of Dostoyevsky. Polyphony is a form of what Bakhtin would hope can be harmony, in which ideally, every person can understand his or her actions and motives and those of others in an infinity of ways and still be in accord, thereby striving for true justice and responsibility for one another, despite plurality and infinity of possible explanations. Yet, the macrocosm of polyphony begins and ends with a microcosm of a personality among other microcosms, a center among other centers, perhaps even a monad within other monads, which leads to the irreducible multiplicity these monads imply and create. The world of Talmudic memory, in contrast, is not necessary predicated on the microcosm of a person (either romantic, or trinitarian in Augustine’s sense of perichoresis of one essence in three hypostases, or Cartesian). Rather, the world of memory in the Talmud approaches thinking not as a part of an initial microcosm in relationship to others, but as discovering of the other (and oneself) as always already a response (refutation of) to a statement attributed to another person, to whom that statement is not intrinsically connected either, but whose name is associated with that statement. The association of a statement to a name by way of memory works as (re)grouping (often of the same) sets of statements in different sets under different names. Another way to draw this fine line between Bakhtin’s world of polyphony and Talmud’s world of memory has to do with an agreement between persona and other, as hoped to be achieved by Dostoyevsky’s characters, versus the Talmudic hope for a well-structured disagreement as the
Notes to pages 93–105 327
vehicle of remembering the authoritative texts of the past, which the characters in the Talmud are striving to gain. Centered as they are on disagreements between earlier authorities, both named and nameless characters of the rabbis in the Talmud are individuals, who embody certain dicta and/or perform the dance of refutations and counterrefutations. The resulting dance is a macrocosm of well-structured disagreements, inventions, and memoirs. In contrast, in Bakhtin’s model, Dostoyevsky’s characters are personal microcosms centered on the personal psychology of each character, which results in a polyphony of the whole. One ramification of that fine but crucial difference in models is that all characters in Dostoyevsky have names and are persons, whereas in the Talmud some characters do not have names but still act as individuals performing refutations and counterrefutations. In the Talmud, the names of the individual characters serve to mark positions, or versions of positions, in the macrocosm of a well-structured disagreement, yet they do not necessarily represent microcosms of a person striving to go through all ambivalence to attain any kind of polyphonic agreement with herself and other persons in the world. An implication of that difference is that the characters in the Talmud are oriented toward the past beyond their personal life, whereas Dostoyevsky’s characters are oriented on the past, present, and future of their personal existence. 19. I analyze that connection in Chapter 7 in greater detail. 20. William J. Richardson, Heidegger: Through Phenomenology to Thought, preface by Martin Heidegger (The Hague: Martinus Nijhoff, 1974), xii. 21. Quintilian, Institutiones Oratoriae 10.6.6. 22. In speaking about a clavis universalis, Hutton (see next note) refers to Paolo Rossi, Clavis Universalis: Arti Mnemoniche e Logica Combinatoria da Lullo a Leibniz (Milan, 1960). 23. See Patrick H. Hutton, “The Art of Memory Reconceived: From Rhetoric to Psychoanalysis,” Journal of the History of Ideas 48, no. 3 (1987): 372.
6. sense in the making: hermeneutical practices of the babylonian talmud 1. An earlier version of this chapter appeared as a part of my article “Sense in Making: Hermeneutical Practices of the Babylonian Talmud against the Background of Medieval and Contemporary Views,” in Judaic Logic, ed. Andrew Schumann (Piscataway, N.J.: Gorgias Press, 2010), 199–227. 2. I use their conception of making sense of a sentence, utterance, or language as example. As already explained with regard to my use of Mark Twain’s concept of “The Author,” in contrast to applications of a theory to a case,
328 Notes to pages 105–6
examples are definitive of their kinds, thereby allowing for comparison with other examples, which introduces (or does not introduce) other kinds, which of course also means it is not my purpose to renegotiate the results of Russell’s or Austin’s theory of making sense but only to use these results heuristically for introducing another example, that of the Talmudic ways of making sense of a language. Even if I use the concepts of these philosophers as examples, some readers may feel surprised to see an engagement with Russell and Austin in a book engaging the context of “continental” philosophy. That impression of eclecticism may be formed due to the dominant divide of philosophies into “continental” versus analytic, which itself is, it may be argued, analytic. Therefore, following the divide implies a firm subscription to analytic tradition, which, at least in this book, I do not have. However tacit or explicit such subscription may be, it always already takes place when the divide is drawn. More specifically, I justify my choice of Russell and Austin as a comparative example in exploring making sense in the Talmud because turning to their conceptions of making sense is heuristic. I use their conception as a comparative example in the sense above. For me, Russell and Austin are modern thinkers situated most closely to the classical theory of truth as correspondence, which translates into a respective idea of what it takes to make sense of a sentence or claim. Therefore, these philosophers are instrumental for my task of fine-tuning both connections and differences between sense-making based in the correspondence theory and the Talmudic practices of making sense of the Mishnah. These practices do not challenge Austin (and perhaps even do not challenge Russell), but do offer a possibility to inscribe their theories into a broader scope of the ways people and other characters make sense of the words they encounter. 3. Shamma Friedman criticizes an earlier, more naive view of the Talmud in the following words: “One of the conceptualizations of Talmudic literature to which mid-twentieth century scholarship was heir may appear fundamentalist and simplistic today. The Talmudic sugya was viewed as a protocol recording debate in the academy. Statements attributed to ancient sages were accepted at face value as the utterances of these sages, with a tendency to accept the interpretation provided in context, unless demonstrated otherwise. Events described were largely accepted as historic fact.” Shamma Friedman, “A Good Story Deserves Retelling—The Unfolding of the Akiva Legend,” JSIJ 3 (2004): 55. However, if the Talmud is a literary composition, it also has an authorial agency or a literary counterpart, just as a novel has. This literary counterpart is the author or authors, or, as scholars of the Talmud have it, “redactors” of the discourses. Despite their status as literary functions rather than natural essences, these literary agencies have been predominantly understood in essentialist historical terms as historically distinct or “real” groups of people. This means
Notes to pages 106–8 329
that even if scholars no longer believe in the reality described in the Talmud as historical, some continue to believe in the reality of the Talmud’s authors as historical personae, rather then as literary agencies. Discontinuing this belief, this book treats both the Talmud and its authors in terms of their literary reality, focusing in particular on the processes and practices of sense making that the Talmudic discourse follows and thereby demonstrates. 4. In this historical view, the genesis of the Talmud stopped when it reached its goal, the telos of history; in this case, that means that the Talmud as we find it in medieval manuscripts and early modern printed editions is the end result of the historical process of the Talmud’s genesis, beginning in the third century C.E., at the earliest, and ending in the ninth century C.E., at the latest. 5. For a historical account of post-Talmudic scholarship of the medieval period, see Yisrael M. Ta-Shema, Ha-Sifrut Ha-Parshanit la-Talmud be-Eropah Uvi-Tsefon Afrikah: Korot, Ishim Ve-Shiot (Jerusalem: Magnes, 2003 or 2004). 6. As reading Ned Markosian shows, correspondence and correlation theories of truth in the analytic tradition yield a homonymic notion of the open past, which, as mentioned in the Preface, differs from that notion of the open past arising if one, as I do in this chapter, heuristically applies correspondence and correlation truth theories to reading Talmudic discussion, which ultimately takes me beyond the correspondence or correlation to pertinence as a different, and much more open, truth criterion. Needless to say, the open past, in the context of this analysis, is not about a temporary inability to determine which of the two or more versions of what-was is the true one (as Ned Markosian has it in the framework of the tradition of analytical philosophy [Ned Markosian, “The Open Past,” Philosophical Studies 79, no. 1 (1995): 95–105]). The difference is not only that in the discussions in the Talmud there could be more than one legitimate and true version of what-was; and not only, as Deleuze would have it, that what-was is always an event (that is always already or always not yet) and thus is never simply a what-is in the mode of what-was, but rather also, and more important, the difference is that the past is the orientation of thinking toward remembering better. That means the past is much more than a state of affairs to store in a reservoir of memory and to bring to thinking as needed. 7. This might sound rather like Wittgenstein, where philosophical language would be the Mishnah, and philosophy tests varying sentences for when they are used in ordinary language. Thus, for example, “I know I feel pain” works (makes sense) in certain ordinary language scenarios and not others. It shows that “knowledge” (in the language of philosophy) is not a special thing in philosophy different from its use in ordinary language. The task of philosophy is to show how everything in ordinary language makes sense. Indeed, for some readers, that analogy between Mishnah and the Talmud and language of
330 Notes to pages 108–12
philosophy and ordinary language helps explain how close Mishnah and the Talmud are intertwined. However, in these terms, the task of the Talmud is not in applying the Mishnah (that is, putatively, philosophical language) to showing how ordinary language (that is, putatively, situations discussed in the Talmud) makes sense, but rather in inventing a situation to which the language of the Mishnah would apply. Another difference has to do with the notion of refutation as central for making sense of the Mishnah in the Talmud, which is not necessarily the case in Wittgenstein”s model of philosophy of language. 8. See bBM76b, a text that I do not directly address in the present work. 9. See his Introduction to Mishneh Torah in Moses Maimonides, Mishneh Torah, Hu Ha-Yad Ha-Hazakah,’Im Haagot Ha-RABaD, u-Ferush Ha-Magid Mishneh, Ha-Kesef Mishneh [Ve-’od]. Ve-’Atah Hosafnu Hosafot Hadashot.. Teshuvot Ha-RaMBaM, Teshuvot Rabi Avraham Ben Ha-RaMBaM. Be’urim, Hidushim Ve- Hagahot Me-et Hazon ISH (Jerusalem: El ha-mekorot, 1953/54–1956/57). 10. See Daniel Boyarin, Ha-‘Iyun Ha-Sefaradi: Le-Farshanut Ha-Talmud Shel Megorshe Sefarad (Jerusalem: Mekhon Ben-Tsevi le-heker Kehilot Yisra’el ba- Mizrah, 1989); Daniel Boyarin, Socrates and the Fat Rabbis (Chicago: University of Chicago Press, 2009). 11. One reason for that limitation might be that Canpanton shares— specifically, reverses—Maimonides’ view that refutation and defense constitute the main intellectual form of the Talmudic discussion, thereby circumscribing the question of making sense by an admittedly viable, and undeniably strong, but perhaps too strong, an answer to it. Without denying the validity of this answer, my goal here is to pose the question of making sense in the practices of the speakers in the Talmud in its own right first, independent of any specific answer to it, thereby making other possible answers staying in dialogue with the answer given by Canpanton. His model of invention as both sense-and truth-making protocol remains useful. It helps explain one way in which a student can discern the sense of the Talmud. Yet his model generates no interest in the question of how sense is made in the Talmudic discussion in the first place. More generally, owing to their interest in what the sense was, rather than in how it was made or produced, the medieval and early modern theorists (as well as their contemporary followers in scholarship) paid little to no attention to the importance of the criterion of referential pertinence, much less to the generative power of that criterion in the rabbinic academy practice of sense making. 12. For a critical edition of the text, see Shamma Friedman, Talmud Arukh: Perek Ha-Sokher et Ha-Umanin: Bavli Bava Metsia Perek Shishi: Mahadurah al Derekh Ha Mehkar Im Perush Ha-Sugyot (Bet ha-midrash le-rabanim ba- Amerikah, 1990). 13. Quoted from MS Parma, Biblioteca Palatina, 3173.
Notes to pages 113–36 331
14. Since the Mishnah is a normative discourse, rather than a descriptive one, its references are prescriptive, rather than descriptive. Yet this in no way excludes the Mishnah or any other normative text from the realm of the referential theory of truth, as far as the procedures of making sense are concerned. 15. bBM76a, MS Firenze, Biblioteca Nazionale Centrale, II.1.8–9. 16. bBM76a. Translation is mine. 17. To avoid a possible confusion between what is type and what is situation in each instance is to remember that a certain understanding of the language of the Mishnah is a type, whereas cases to which the Mishnah thus understood would apply are situations. 18. bBM76a; my translation. 19. Ibid. 20. Or, in David Constantine’s translation, “Where there is danger some Salvation grows there too.” Hölderlin, Patmos, in Friedrich Hölderlin, Selected Poems, trans. David Constantine (Newcastle upon Tyne: Bloodaxe Books, 1990), 39. 21. bBM76a. 22. Ibid. 23. Ibid. 24. Ibid. 25. Ibid. 26. Ibid.
7. who thinks in the talmud? 1. An earlier version of that chapter appeared as an article “Who Thinks in the Talmud?” Journal of Jewish Thought and Philosophy 20, no. 1 (2012): 1–34. 2. Alain de Libera, Archéologie du sujet, vol. 1: Naissance du sujet (Paris, 2007). 3. By way of synecdoche, let me mention three names and chart three main lines of scholarship I engage with here by mentioning just three book titles. These three books are from three different fields: rabbinics, studies in gender and sexuality, and critical theory of thinking processes. One is Carnal Israel: Reading Sex in Talmudic Culture (Berkeley: University of California Press, 1993). This work by Daniel Boyarin introduces us to the area of sexuality in rabbinic writings. Another is “About the Beginning of the Hermeneutics of the Self,” by Michel Foucault (in Religion and Culture, selected and edited by Jeremy R. Carrette [New York: Routledge, 1999], 158–82). In this late work, focused primarily on Christian and pagan texts of confession in late antiquity, Foucault explains the same theoretical object, sexuality, using what he calls techniques of the hermeneutics of the self. In the essay, he studies the
332 Notes to pages 136–41
ways—the techniques and practices—by which the self accesses itself through experience, in particular the monastic practices of interpreting the self, that is, finding in the self something hermetic or obscure to be clarified via techniques of “eximologesis.” As he submits, without these techniques, there simply is no place for sexuality, as opposed to sexual differences. Without a dark, elusive, and for that matter sinful element in the self and its thoughts, there would be no need for hermeneutics. We would need only to regulate the transparent thoughts and deeds of the self in a fully controlled effort in which a person conceived as a thinking and willing self avoids mistakes that are committed in the past, discovered in the present, and thereby made avoidable in the future. As a consequence, a well-administered and transparent willing and thinking self allows only sexual relationships, not sexuality. In contrast, the presence of an obscure, elusive element in the self requires hermeneutical techniques in order to read it. This linkage between sexuality and the hermeneutics of the self sufficiently explains the transition I make from reading sex to reading self. However, this leaves unexplained another transition that I make here: from culture to literature. A recent critique of Foucault by Alain de Libera helps navigate that transition. De Libera’s Archéologie du sujet I is the third book I will mention here. The underlying assumption of Foucault, as well as of his teacher, Heidegger, was the axiomatically assumed notion of the self as a thinking subject. After de Libera’s analysis of the birth of that seemingly obvious but in fact contradictory philosophical concept, the concept of the self as a thinking subject can no longer be an axiom. As a result, the primary object of hermeneutics, a thinking self, can no longer be taken for granted, at least not in the late antique text of the Talmud. Therefore, an inquiry into Talmudic texts that hitherto axiomatically assumed either a dark side in the thinking self and hence a sexuality therein needs to be started anew. The self as a thinking subject is no longer axiomatically assumed. Rather, it becomes what it always was—the indication of a question: Who thinks in the Talmud? 4. Martin Heidegger, Being and Time: A Translation of Sein und Zeit, trans. Joan Stambaugh, SUNY Series in Contemporary Continental Philosophy (Albany: State University of New York Press, 1996). 5. Foucault, “About the Beginning of the Hermeneutics of the Self,” 158–82. 6. This concerns only one of the three following aspects of Aristotle’s model of thinking as a faculty (dunamis) as opposed to an action and/or actuality (energeia). The model includes (1) the potentiality (ability, and thus faculty) to learn a subject (math, for example); (2) the actuality of knowing (math, for example), but not using it (say, when sleeping); and (3) the ability of using the
Notes to pages 141–42 333
knowledge, in actualizing the sleeping potential. In (3), one can argue that for Aristotle the “thinker” is not individuated as a specific flesh-and-blood human as in (1) and (2), because thinking “2 + 2 = 4” is not achievable in space-time. In addition, thinking “2 + 2 = 4” is also an action (dunamis), not a mere faculty (energeia). Of course, as de Libera emphasizes, these notions in Aristotle have nothing to do with the notion of hupokeimenon (subject in the sense of substratum), even if, for him, energeia, dunamis, and hupokeimenon paradoxically co-participate in the modern notion of the thinking subject. Still, the thinking involved in the notion of the thinking subject has to do with only (3) above. The first two aspects are spatial and temporal and could be illustrated with an example: When Crito knows the alphabet and uses it, then (maybe) Crito is only “substrate” (in the sense of a faculty or potential) but not an agent. When Crito is learning the alphabet, we can say that it is Crito, who is actualizing his potential. The (3), however, relates to an ability and action of thinking (or cognition, broadly construed) which is not located in space-time. According to de Libera, if the distinction between energeia and dunamis happens to disappear in certain understandings of thinking, and if action gets attributed to subject both in the aspect of subjectness (to be a subject of something) and “subjectity” (to be subject to something), the modern notion of the “thinking subject” emerges. In contrast, for Aristotle, any conflation of the notions of dunamis, energeia, and hupokeimenon (subject) would only lead to confusion in the understanding of thinking. Therefore, the participation of Crito in thinking, teaching, and learning is not the same as Crito’s being both subject to, and subject of, either of these processes, which would have rendered him a thinking subject. Yet even if not a thinking subject, Crito as a person in some ways participates in teaching, learning, and more broadly, thinking; albeit he always stands at the center of each of these processes, being its alpha and omega. The Talmud however demonstrates a completely apersonal and not-person-centered engagement in memory. Thinking, as well as learning and teaching, work as parts of the Talmudic process of memory, facilitated, as it is, by names and places but not by personae, and certainly not by an individual at the center. (This applies to Heidegger’s construct of Dasein as well, which still is a version of the thinking subject: truth as memory of being in Heidegger is still centered on Dasein as lonely and as central as that version of persona can be.) 7. In that respect, a sage (or more precisely a student of the sage) in the Talmud is comparable to a similar category in Aristotle, that of bios theoretikos, contemplative life. It is of course much too strong a claim to argue that Aristotle did not think that Crito had thoughts or was a spatiotemporal locus of thinking. If Crito thinks about what to do in a particular circumstance, or how to generalize from a certain perception to a rule, Crito is thinking in the sense
334 Notes to pages 142–43
that he participates in the thought process, as connected as it may be to his particular spatiotemporal situation. Crito thus is participating in thinking qua specific person in a specific situation, even if not necessarily as a Cartesian subject of that thinking: To participate in and to be the subject who thinks are not the same. An approximate (and short) version of that difference is that Crito participates in thinking defined by his locality, without becoming a thinking subject—a subject of thinking who is also subject to that thinking. Now when the “sage” (in the sense the term is used in philosophy) contemplates the nature or meaning of Being (to on) she of course is not thinking qua specific person nor is the locality important. Even still that contemplator of Being can be said to have participated in the thinking of Being. Lonely as it is, in this respect, the sages’s thinking is not “thinking” simpliciter but rather “contemplating” (theorein). The position of a Talmudic sage is somewhere in between these two extremes, between engagement in a locality of thinking and disengagement in contemplation. The thinking of the Talmudic sage, as participatory as all modes of Crito’s thinking are, always presumes a spatiotemporal context, or a situation, which the sage addresses, when and as long as this sage contemplates (in the sense of theorein) another sage. This other sage, and in particular his words and his contemplated locality, take the place of Being. The contemplation of the second sage addresses the spatiotemporal situation in the thinking simpliciter of the first sage contemplated by the second. 8. See the analysis of that Midrash in Emmanuel Levinas, Difficile liberté: Essais sur le Judaïsme, Présences du Judaïsme (Paris: A. Michel, 1976), 47–50. 9. In scholarship on late ancient Jewish texts, the once-predominant direction of research from Jewish to Christian texts has been reversed relatively recently. Daniel Boyarin, Azan Yadin, and Ishai Rozen-Zvi (see Bibliography) have ventured to use Christian texts and theology to understand Jewish texts and Jewish theology. Most recently, that interest was significantly expanded to embrace not only Christian theology but also pagan philosophy, either in terms of period specific linkages during a particular period or in the broader scope of the philosophical tradition of the West. This line of scholarship reached the point of drawing a complex intellectual map of late antiquity in which the divisions between Jews and Christians, and Judaisms and Christianities, were far from definitive in terms of intellectual movements. In this connection, a parallel interest in illuminating Jewish texts anew in their context in pagan thought and philosophy has recently strongly emerged. See, for example, Naomi Janowitz, Magic in the Roman World: Pagan, Jews, and Christians (London: Routledge, 2001). At the same time, in this line of scholarship, the questions go beyond historical recontextualization per se to establish intellectual linkages between the Jewish and pagan traditions of
Notes to pages 143–49 335
thought and philosophy in their intrinsic scopes, rather than under the aegis of a totalizing historical-chronological view, for example, that of intellectual history or even of a history of concepts. 10. See Mosheh Chayim Lutsatto, Sefer Derekh Tevunot: Kolel Bo Kol Darkhe Ha-Hitbonenut Veha-Haskalah le-Havin be-Nakel Darkhe Ha-Talmud Vi-Yesodot Ha-Pilpul be-Derekh Ketsarah (Jerusalem: Yeshivat ha-tefutsot: Feldhaim, 1988 or 1989). 11. See inter alia Chanoch Albeck, Untersuchungen über die Halakischen Midraschim, von Chanoch Albeck (Berlin: Akademie-Verlag, 1927), Chanoch Albeck, Untersuchungen über die Redaktion der Mischna (Berlin: C. A. Schwetschke, 1923); Chanoch Albeck, Mavo la-Mishnah (1966 or 1967); Chanoch Albeck, Einführung in die Mischna. [Aus d. Hebr. übers. von Tamar u. Pessach Galewski], Studia Judaica, vol. 6 (Berlin: De Gruyter, 1971); Chanoch Albeck, Shishah Sidre Mishnah (Jerusalem: Mosad Bi’alik., 1969). 12. Maimonides, Mishne Torah, Introduction. Here and below, with no academically published translation of this text in English available, I have used a translation from http://www.mechon-mamre.org/e/e0000.htm, accessed 10/26/08. I have slightly modified the terminology of the translation to make it perhaps less idiomatic but more literal in the places essential for the current discussion. I have also modified the transliteration of names to maintain the coherence of reference. 13. Maimonides, Mishne Torah, Introduction. 14. Lutsatto, Sefer Derekh Tevunot, 11. 15. I am presenting the Mishnah and the Talmud on it in an English translation to accord with David Weiss Halivni’s interpretation of it. 16. The soliloquist needs to do one more thing: answer what exactly would in the amor’aim’s perspective be the import of the Mishnaic instruction in practical terms. The answer that he contemplates is: In mentioning, “It is all mine!” the instruction refers to the case of two litigants each claiming that the cloak was their purchase made in a store. Of course, the court might just ask the merchant and decide based on that; yet, for that, no special instruction is needed. Therefore, the real point of the instruction is whether the merchant took money from both of the litigants. In this case, the Mishnah prescribes that the court choose to have the litigants swear, rather than leave the case unresolved. In the soliloquist’s own words: “One still might just ask from whom the merchant took money; yet, if the merchant took money from both of them, the instruction to swear is necessary: It teaches that a possession in doubt is to be split [by that swearing], etc.” The first-person inquirer rhetorically addressed himself or herself, or what was the same thing, the audience, as he or she made an attempt to solve the concern about the Mishnah using his or her own
336 Notes to pages 149–55
power of judgment, but ended up eliciting an answer from the teaching of the amor’aim, followed by making a viable argument for the answer that he or she thus found. In sum, stylistically, we have just observed a soliloquy refuting the Mishnah, followed by a defense that recalls a tradition, which, regardless of its original purpose, helps respond to the attack and thus provides the counterrefutation required. 17. This is, of course, not to say that Crito does not participate in thinking. Crito learns the alphabet, and learning is actuality of a potential ability to know the alphabet. However, the teacher of any knowledge is not Crito, that is, a specific person. The knowledge itself actualizes the potential in Crito. So teaching and learning in Aristotle are not exactly personal, but that does not apply to all thinking. The Talmud is thinking by way of teaching. Moreover, for Talmud criticism, and other conceptions based on the notion of a thinking subject, the Talmud is thinking in the more general sense of Aristotelian contemplation. And yet, again, the analogy with Aristotle is not complete, because as a Talmudic character approaches a sage in the Mishnah or another sage of an earlier generation, and/or of greater authority, teaching remains an impersonal process but involves a contemplation of the spatiotemporal specificity of the sage to whom what is taught is attributed by the one who learns it. 18. It might seem that strategically, the soliloquist had the amor’aim in mind from the outset and does nothing but ask questions that prepare for their emergence on the stage. In that sense, the soliloquist is a direct adept of, and an apologist for, the amor’aim. But Halivni’s point is that there also is an important difference between the amor’aim and the soliloquist: The soliloquist is a rhetorical position created by the redactors, not the redactors “themselves.” 19. For his biography and bibliography on his work, see Yitzhak Barzilay, Manasseh of Ilya: Precursor of Modernity among the Jews of Eastern Europe (Jerusalem: Magnes Press, Hebrew University, 1999). 20. David W. Halivni, Sources and Traditions: A Source Critical Commentary on the Talmud: Tractate Baba Metzi‘a (Jerusalem: Hotsa’at ha-sefarim ‘a. sh. Magnes, Universitah ha-‘Ivrit, 2003), 2. 21. Halivni’s analysis calls for the following questions. Can we be sure that the quote from the amor’aim that he gives relates to the Mishnah at hand exactly in the way the soliloquist decided it does? When Halivni says, “ ‘Said Rav Pappa, or according to some, Rav Simi the son of Ashe means that the opening of the Mishnaic instruction refers to finding a thing, but the continuation of it to buying and selling it, because and only because of the power of the question that the soliloquist proposed” (Halivni, Sources and Traditions, Bava Metzi‘a, 2), this suggests connecting the saying of the amor’aim to the Mishnah at hand in this specific way was a silent decision of the amor’aim based on their
Notes to pages 155–57 337
dismissal of another option, which Halivni does not seem to feel any need to mention, but which might relate to partitioning of the Mishnah into different beginning and ending parts. Thus, for example, since, as Halivni says, according to Rabbi Manasseh, the Mishnah itself does not give any direct or indirect reference to buying, but can be applied to buying only through an inference or by implication, “the Mishnah itself does not give any hint about buying or selling” (Halivni, Sources and Traditions, 2)—the amora’ims’ term “ending part” might refer to the next part of the Mishnah talking about “two people riding a donkey or one riding and one leading, and each says ‘it is all mine’ without saying I have found it” (!). In fact, Halivni gives another, more compelling answer to that question: By the beginning and by the end of the concluding phrase of the Mishnah, the amor’aim might have meant the first and second topic in the Mishnah’s last phrase: “At the time when the litigants agree, or have witnesses, it should be split without any swearing involved.” The “beginning” then is “when litigants agree,” and, according to this reading of the amor’aim, the Mishnah deals with the things found (for in the case of the things bought, there is no point even to mention that swearing is not needed). The “ending” is “when they have witnesses” and refers to litigation on buying and selling, which in the case of conflicting witnesses should not be resolved with an oath but rather should be left unresolved. Also, can we be sure that the disagreement between the amor’aim on the issue of acquiring an object for the other without acquiring part it for yourself is not a retroactive construct? 22. In Halivni’s words, “Rabbi Manasseh did not differentiate between stamm’a de gemarr’a and the amor’aim . . . . However, we do draw a distinction between stamm’a de gemarr’a and the amor’aim. In many cases in the Talmud, other arguments made by the stamm’a de gemarr’a are assumed to belong to the words of the amor’aim. Yet we also are trying very hard to explain why [Halivni’s emphasis] stamm’a de gemarr’a neglects to explain [the Mishnah] our way [i.e., Rabbi Manasseh’s and Halivni’s way]. Only we do not satisfy ourselves with an explanation that it simply did not occur to them to think that way. In this case, too, we will be looking for an explanation: Why was stamm’a de gemarr’a not able to accept Rabbi Manasseh’s way of reading the Mishnah?” Halivni, Sources and Traditions, Baba Metzi‘a, 2–3. 23. I will address the attributivism principle later in this argument, following de Libera’s treatment of it. 24. Aristotle, De Anima 2.415b1–15. 25. Once again, this is not to say that in Aristotle, a person does not participate in the thinking, but it is to say that the subject (hupokeimenon) is not a person who thinks, nor is it a thinking substance. Aristotle does not deny that
338 Notes to pages 157–66
learning the alphabet alters the learner qua potential knower of the alphabet. Teaching the alphabet does not alter the teacher, however. “Alter” means going from privation to presence of something (not knowing the alphabet to knowing it), but teaching only actualizes a present dunamis. Thus “Crito is learning the alphabet” is true, but “Socrates teaches the alphabet” is, strictly speaking, false. In de Libera’s view, teaching and thinking in general in the sense of contemplating got confused in the modern notion of the thinking subject. For Aristotle, however, hupokeimenon can neither think in general nor can it teach or learn anything, including the alphabet, nor can it contemplate. 26. That leads to yet another conceptual question to ask about Halivni’s work and perhaps of the larger tradition of harmonizing the Talmudic arguments in one coherent piece of intellectual production rather than as local refutations and counterrefutations. Is such harmonization possible only if the statements of the amor’aim are taken as apodictic claims, rather than as products of a rhetorical-probabilistic give and take and intrinsically bound to the local argument in which they occur? At this point, it would also be important to ask if Maimonides (typologically, if not also historically) is responsible for the view of the Talmudic local rhetorical lemmata as universally apodictic claims. The early Talmudic commentators, with their integrative approach to the Talmud, have already identified that problem, but without such an integrative approach, the Talmud only does what it does—it locally defends the Mishnah. This question, however, is beyond the scope of my current analysis. What remains important in the context of my present inquiry is that without that integrative approach, claiming the existence of the stamm’aim, or anonymous redactors, is not possible. Conceptually, they are a projection of that integrative view of the Talmud as a harmonic whole with one, uniting thinking subject at work amid and despite the local but always required format of refutations and counterrefutations.
8. the hand of augustine: thought, memory, and performative existence in the talmud 1. Augustine, Confessions, ed. George Stade (New York: Barnes and Noble Classics, 2007), 10.8. 2. See the anonymous Rhetorica ad Herennium, Cicero’s De Oratore, and Quintilian’s Institutiones Oratoriae. The palace of memory or using places to remember images and their order is introduced, for example, in Rhetorica ad Herennium (4:16) as a house, an intercolumnar space, a recess, or an arch: “The
Notes to pages 166–70 339
artificial memory includes backgrounds and images. By backgrounds I mean such scenes as are naturally or artificially set off on a small scale, complete and conspicuous, so that we can grasp and embrace them easily by the natural memory—for example, a house, an intercolumnar space, a recess, an arch, or the like. An image is, as it were, a figure, mark, or portrait of the object we wish to remember; for example, if we wish to recall a horse, a lion, or an eagle, we must place its image in a definite background” (Marcus Tullius Cicero, Rhetorica ad Herennium, trans. Harry Caplan [Cambridge, Mass.: Harvard University Press, 1999], 207). For critical analysis see Frances A. Yates, The Art of Memory (Chicago: University of Chicago Press, 2001). 3. See Preface to Mishneh Torah, and my reading of it in Chapter 7 and the reading below. 4. Italics are mine—S.D. 5. Such is Maimonides’ view thereof. 6. Biblical passages are quoted here according to JPS 1917 Edition. 7. Timewise, in this apologetic exercise, ideas come from the interpreter, and go backward to their place in the genealogy of authorities, which means that “clarification” does not have to do with historical time, but only with the genealogical rank of the authority for any given idea, which may but does not have to translate into history. 8. Translation of Mechon Mamre (http://www.mechon-mamre.org/e/e0000 .htm; accessed August 7, 2009), heavily emended. Italics are mine—S.D. 9. Another formative notion of the precritical reading of the Talmud, as formulated by, among others, Maimonides, suggests that the “here and now” of the Talmud is fifth-century Babylonia, the Talmudic schools of the amor’aim Ravina and Rav Ashe, who genealogically are the latest listed authorities mentioned in the Talmud by name. Yet another component of that reading is that the genealogical order of the generations of the masters of the Talmud or amor’aim coincides with the chronological order of their lives, as well as with the descending ranks of the authority of the teachings ascribed to each master. Since the ranks of genealogy, chronology, and authority coincide, I call this the triple principle of coincidence of chronology, genealogy, and authority. This principle regulates the dynamics of refutations and defenses, as well as divergences in positions between different masters (amor’aim) in their arguments with each other as well as with the sages. One practical implementation of the triple principle of coincidence is that a master of the Talmud cannot diverge from the teaching of a sage unless this master identifies with the position that he ascribes to the authority of another sage. As a result, not only people named in the Talmud, but also their ideas and teachings follow the principle of
340 Notes to pages 170–75
triple coincidence, and the pattern of refuting, defending, and disagreeing also follows that principle. 10. As a result, a modern researcher will no longer be able to occupy this position. 11. The Talmudic legalistic concept of “four cubits” or of immediate private space always possessed by a person, which always “travels” with any human being, might in turn serve as an analogy illustrating Husserl’s concept of earth as perceptual ground as opposed to earth as a planet. One never departs from the ground, but can leave the planet of earth for an adventure in space. 12. See Edmund Husserl, “Grundlegende Untersuchungen zum phaenomenologischen Ursprung der Raumlichkeit der Natur,” in Philosophical Essays in Memory of Edmund Husserl (Cambridge, Mass.: Harvard University Press, 1940), 307–25. 13. Although the anonymous speakers in the rabbinic academy are identified neither by names nor by any other markers, only by their statements, they, too, have bodies in this sense. Their bodily differences guarantee that they follow turns in the well-orchestrated dance of their collective argument. Their nameless, but distinct bodies are what guarantee the coherent differences between them, providing distinctions between their faceless voices. Are these bodies intrinsically gendered? The answer has to be “yes,” because these bodies most plausibly assume certain social roles—that of a student or teacher in a rabbinic academy or a judge in a rabbinical court, and these social roles, in all rabbinic discourses, seem to be gendered—provided that, as gender theory tells us, “gendered” no longer automatically means “feminine” and “masculine” no longer means “disgendered,” that is, that being a woman is no longer considered the default state of being “gendered” and that being a man is no longer considered the default state of being a human being without gender. In that case, we can safely say that the bodies in the rabbinic academy are gendered. However, the question of the gender definitions of teachers and students in the academy as well as of their relationship to more senior authorities, such as the sages of the Mishnah, and to amor’aim, or to more junior authorities, such as the reciters of the Mishnah, or tann’aim, remains outside of the scope of this book. 14. Louis Althusser has theorized that situation in his method of emphatic reading in Reading Capital, a work that pays attention to answers first and proceeds to the questions afterward. See Louis Althusser and Etienne Balibar, Reading Capital (London: Verso, 2009). 15. Is this bare being-there of memory the one who historically constructs the Talmudic discussion? It clearly is not. Instead, this being there is a resulting literary construct of that very discussion, its literary effect, which, unlike to
Notes to pages 175–85 341
what scholars hitherto have assumed, should not be confused with any possible historical subject(s) involved in the genesis of the Talmud. 16. Anticipating another possible way of looking at the composition of the Talmudic text, let me first ask the following question: Could the designer of Talmudic discussions have a slightly different apologetic goal—to orchestrate what German philosophy calls Erinnerung (or intellectual memory) by rhetorically inventing the truth of the Mishnah? By way of anticipation of a more detailed and more contextualized analysis, let me note that if the answer is “Yes, the designer could!” then the designer of Talmudic discussions may be seen as choreographing the conversations as a way to advance his memory (precisely not as a thinking subject, but as the bodily location of “memory for things,” which thus is not a subject-centered recollection, but instead is a rather pure— indeed bodiless, or, more precisely, bodily but organless—work of memory, both artificial and Platonic at once in a way that, as a detailed analysis shows, is not fully unlike that of Augustine’s searching for G-d in his texts on memory in the Confessions. 17. In the philosophical terminology of modern times, acts and affects are mutually exclusive, because acts come from acting subjects and affects from the things acting upon these subjects. Accordingly, for a philosophical rationalist such as Descartes, memory would be an affect, but thinking would be an act, because memory does not have a clearly acting subject. However, as we have seen, memory does not have to be divided in this way. Instead, that division might be construed as extraneous to memory, although perhaps relevant to the process of recollection. 18. Although, of course, this “entity” is accessible to the reader of the Talmud through the experience of reading: Attuned to the text, the reader might position herself as just that: the bare being-there of memory. 19. As per the interpretation of Pierre Hadot in Pierre Hadot, Plotinus, or, The Simplicity of Vision, trans. Michael Chase, with an introduction by Arnold I. Davidson (Chicago: University of Chicago Press, 1993).
9. what is the sophist? who is the rabbi? the virtual of thinking 1. For the analysis of meontology as a theory of nonbeing in connection with messianism see Martin Kavka, Jewish Messianism and the History of Philosophy (Cambridge: Cambridge University Press, 2004). 2. An earlier version of this chapter was published in New Directions in Jewish Philosophy, ed. Aaron W. Hughes and Elliot R. Wolfson (Bloomington: Indiana University Press, 2010), 253–85.
342
Notes to pages 186–212
3. The term translates Plato’s philodoxos, meaning “lover of opinion.” In Plato it has a negative connotation: To love opinions is not as good as to love truth, and it is opposite to philosophy or love of wisdom. See Plato, Republic 5.480a. 4. “ . . . and make plain by argument who he [the sophist] is.” Plato, The Sophist 218c. 5. מאן תנאBabylonian Talmud (BT) Ps. 115a, Yom 39a, Qid. 15a. 6. Should asking these questions still risk an allegation of anachronism, I will ally myself with Elliot Wolfson’s response to a similar charge. In his prologue to Language, Eros, Being (New York: Fordham University Press, 2005), Wolfson defends using recent intellectual achievements in both philosophy and science to interpret the earlier masters of mysticism. Reduced, as it was, to empirical philology (glossed as a discipline studying the authorial intent), hermeneutics inevitably has subscribed to a linear conception of time, having just three dimensions, past, present, and future. Wolfson firmly opposes such a reduction. In his rereading of Merleau-Ponty’s philosophy of perception in the context of both philosophical and scientific notions of the reversibility of time, he proves the linear dimensions of time insufficient to grasp the workings of time in full. Itself never timeless, time always involves a perceptual and therefore always partial exposition of either what is perceived or of the perceiving being. Just as the three dimensions of space—length, width, and height—do not allow us to perceive any object in full, but only in one aspect at a time, and yet we still perceive the object, not the aspects of it, so also the three dimensions of time do not allow us to get a correct grasp of time, forcefully reducing it to a line—direct, inverted, or curved. In contrast, a correct grasp of time must allow for what cannot be fully perceived, not even from a hypothetically eternal point of view. If so, the allegation of anachronism concerns only the first three dimensions of time, while hermeneutics has to do with the fourth. 7. See Irenaeus, Against Heresies (Oxford: Parker, 1872). 8. See Michael A. Williams, Rethinking “Gnosticism”: An Argument for Dismantling a Dubious Category (Princeton: Princeton University Press, 1996). 9. Translated from Talmud Yerushalmi according to Ms. Or. 4720 (Scal. 3) of the Leiden University Library with Restorations and Corrections. Introduction by Yaacov Sussman: Berachot 12:4, 9:1. P. 65.
10. the talmud as film 1. Examples include Merleau-Ponty’s theory of body as not only an image but also a foundation of experience, or a similar theory of earth as the never
fup-dolgopolski-13n.indd 342
10/9/12 12:13 PM
Notes to pages 212–14 343
moving and never resting ground of experiencing the moving or resting objects in the world, developed in late Husserl, or Levinas’s vision of the body’s uncontrolled behaviors as always already a site of encountering the other. 2. Shamma Friedman, Talmud Arukh: Perek Ha-Sokher et Ha-Umanin: Bavli Bava Metsia Perek Shishi: Mahadurah al Derekh Ha Mehkar Im Perush Ha-Sugyot (Bet ha-midrash le-rabanim ba-Amerikah, 1990), 1:9. 3. Chanoch Albeck, Introduction to the Talmud, Bavli and Yerushalmi (Tel Aviv: Dvir, 1969); Chanoch Albeck, Untersuchungen über die Redaktion der Mischna (Berlin: C. A. Schwetschke, 1923); Chanoch Albeck, Mavo la-Mishnah (Jerusalem: Bialik Institute, 1966 or 1967). 4. David W. Halivni, Sources and Traditions: A Source Critical Commentary on the Talmud: Tractate Baba Metzi‘a (Jerusalem: Hotsa’at ha-sefarim ‘a. sh. Magnes, Universitah ha-‘Ivrit, 2003); David Halivni, Mekorot u-Masorot: Be’urim Ba- Talmud: Masekhet Shabat (Jerusalem: Bet ha-midrash le-rabanim ba-Amerikah, 742, 1982); David W. Halivni, Mekorot u-Masorot: Be’urim Ba-Talmud: Masekhet Bava Kama (880–03 Jerusalem: Hotsa’at sefarin ‘a. sh. Y. L. Magnes, ha- Universitah ha-‘Ivrit bi-Yerushalayim, 1993); David Halivni, Mekorot u-Masorot: Be’urim Ba-Talmud: Masekhtot ‘Eruvin u-Fesahim (Jerusalem: Bet ha-midrash le- rabanim ba-Amerikah, 1981 or 1982); David W. Halivni, Sources and Traditions: A Source Critical Commentary on the Talmud Seder Moed from Yoma to Hagiga (Jerusalem: Jewish Theological Seminary, 1973). 5. Needless to say, as a result of intermingling and mutual fertilization of discourses, there are, of course, Jewish chronology and Christian genealogy. 6. Read through the lens of the film theory of montage as genealogical rather than chronological in nature, Talmudic arguments cannot help but provide an additional, but quite particular, illustration of the more general argument of Sacha Stern’s about the lack of “time” and the presence of “process” in ancient Judaism. See Sacha Stern, Time and Process in Ancient Judaism (Oxford: Littman Library of Jewish Civilisation, 2003). In this specific case, I would rather talk more specifically about the lack of chronological time and the presence of the genealogical process of montage in the Babylonian Talmud. Sacha Stern would argue, persuasively and also more generally, that there is no time in the Talmud. If this argument is taken in the light of Mary Ann Doane’s views in her The Emergence of Cinematic Time: Modernity, Contingency, the Archive [Cambridge, Mass.: Harvard University Press, 2002], the result becomes even more intriguing and promising. Doane sees cinematic time and subjectivity as effects of the cinematic time representations of the industrial age, that is, by necessity both rationalizing and making things contingent. I thus might conclude that Stern proves the lack of industrially rational time in the Talmud. He dismisses either contingency as an element of time and even Bergson’s element
344 Notes to page 214
of duration and memory, for Bergson pointedly separates duration and memory from cinematically rationalized time. Could we then, in an interrogation that follows Stern’s steps, ask about time in the aspect of duration and memory in the Talmud in the Bergsonian sense, as opposed to the allegedly industrial time of cinematography? What would such an interrogation would mean for Stern’s argument? I still do not think Sacha Stern’s analysis necessarily depends on modern dualisms about time (of which time versus “process” could be an example), but this is not the place to discuss this concern any further. 7. Although I do not interrogate the following three authors and/or books in any explicit way, implicitly, they have been of a particular importance in shaping my inquiry. Yosef Hayim Yerushalmi, Zakhor, Jewish History and Jewish Memory (Seattle: University of Washington Press, 1982) delineates a dramatic disconnect between the scientific, chronological view of time in the discipline of history and time in what he calls “Jewish memory.” Kathleen Biddick in her The Typological Imaginary: Circumcision, Technology, History (Philadelphia: University of Pennsylvania Press, 2003) explores how a seemingly neutral chronological way of thinking of time as a graphical line and empty form or type of progression always already embeds, against any obvious logic or law, the strong anti-Jewish bias that Christian thinkers had in their efforts to disconnect (and sometimes also therapeutically to reconnect) the children of Israel in the Old Testament and contemporary Jews at and after the time of the New Testament. On yet a different level, Philippe Lacoue-Labarthe, in Typography: Mimesis, Philosophy, Politics (Stanford, Calif.: Stanford University Press, 1988), discusses both the disconnect and the inevitable, almost delirious connectedness between typological thinking and the graphical technology of representation that supplies representations of both the self and the time in which the self occurs or even with which it coincides. The motifs of these three books and their respective intellectual times and places are in play here. 8. Gaon Sherira ben Hanina and Benjamin Manasseh Lewin, Igeret Rav Sherira Gaon: Mesuderet Be-Shene Nus’ha’ot, Nusa’h Sefarad ‘Ve-Nusa’h Tsarfat, ‘Im ‘Hilufe Ha-Girsa’ot (Jerusalem: Makor, 1971). 9. The word “talmud,” talmud’a, in late ancient Aramaic texts ranges in meaning from “injunction” of a teaching from the Torah, “teaching,” “learning,” “midrash,” “amoraic discussion”; none of these references strongly implied “a finished text,” much less a “finished book.” See Michael Sokoloff, A Dictionary of Jewish Babylonian Aramaic of the Talmudic and Geonic Periods (Ramat Gan, Israel: Bar Ilan University; Baltimore: Johns Hopkins University Press, 2002), 1209–10.
Notes to pages 215–19 345
10. See, for example, Azzan Yadin, Scripture as Logos: Rabbi Ishmael and the Origins of Midrash, Divinations (Philadelphia: University of Pennsylvania Press, 2004). 11. Jean-François Lyotard, Discourse, Figure (Minneapolis: University of Minnesota Press, 1971). 12. For a case study why the romance between scientific history and rabbinic (Talmudic and post-Talmudic) genealogies is not simple, see David Nirenberg, “Mass Conversion and Genealogical Mentalities: Jews and Christians in Fifteenth-Century Spain,” Past and Present 174, no. 1 (2002). Nirenberg shows how problematic any historian is whose work tends not only to misrepresent what is full of breaks and ruptures as a smooth and linear genealogy, but who also keeps confusing the intellectual coherence of his own soul with the genealogical coherence of the historical period that he or she studied (40–41). 13. Hayim Lapin, and Dale B. Martin, eds., Jews, Antiquity, and the Nineteenth-Century Imagination (Potomac: University Press of Maryland, 2003). 14. Moshe Lutsatto, The Ways of Reason (Jerusalem: Diaspora Yeshiva, Feldheim, 1989). 15. For a historical account of how, by the nineteenth century, chronological time came to be the main and mandatory dimension of history, see, for example, Reinhart Koselleck, Futures Past: On the Semantics of Historical Time, trans. and with an introduction by Keith Tribe (New York: Columbia University Press, 2004). For a book-length philosophical-genealogical account of the transition from restoring the logical foundations of society in the eighteenth century to a linear chronological-historical interpretation of these foundations in the nineteenth century, see Jacques Derrida, Of Grammatology, trans. Gayatri Chakravorty Spivak (Baltimore: Johns Hopkins University Press, 1998). 16. Founded earlier in the twentieth century in the works of such intellectual giants as Yaakov N. Epstein, Chanokh Albeck, Saul Lieberman, and others, the scientific theory of the Talmud’s redaction grew into a highly developed trend of philological, historical, literary, linguistic, and more recently, comparative study of the rabbinic and cognate literatures. These modes of study, however diverse or even controversial they might be, belonged to one and the same intellectual continuum, dominated by the idea of the scientific strictness of textual hermeneutics, as opposed, perhaps misleadingly, to the putatively more lax tradition of hermeneutics that philosophy or metaphysics had to offer. 17. Izh.ak Canpanton, Shemu’el Valenci, and Isaak S. Lange, Darkhe ha-Talmud [The Ways of Talmud] (Jerusalem, 1980 or 1981). For an academic bibliography on this text, as well as for a close reading and analysis in English see Sergey Dolgopolski, What Is Talmud?: The Art of Disagreement (New York: Fordham University Press, 2009).
346 Notes to pages 219–23
18. For an analysis of that context and its extensions, see Jean Antoine- Dunne and Paula Quigley, eds., The Montage Principle: Eisenstein in New Cultural and Critical Contexts, Critical Studies (Amsterdam, N.Y.: Rodopi, 2004). 19. Jean-François Lyotard, “Jewish Oedipus,” in Toward the Postmodern, ed. Robert Harvey and Mark S. Roberts (Atlantic Highlands, N.J.: Humanities Press, 1993). 20. Even if, as in suspenseful scenes in Alfred Hitchcock’s films, the event (for example, murder) happens more in the heads of the implied viewer, that is, by way of inference that the viewer makes from what is displayed on the screen, whereas no direct depiction of the act of killing is given, this is still a singular event in a frame rather than an effect of cross-cutting, which creates thoughts and feelings in the viewer, which do not relate to any direct act on the screen, but rather arise from the sequence of acts. 21. In his “Montage” (“Montaz”), 1937 Eisenstein writes: “By playing with cuts of film, [the leftists of montage (Eisenstein refers to leftwing filmmakers— S.D.)] discovered a quality that surprised them very much. This quality was that putting together two random cuts of film inevitably unites them in a new representation emerging from them as a new quality” (Eisenstein’s italics—S.D.). There is nothing particularly cinematographic about that, as Eisenstein explains. What really is cinematographic for him is using montage for what for him is the main task of a film: “The task of our [“Soviet”—S.D.] films is not only a logically connected narration, but also a maximally emotionally exciting one.” In his example, “A [face of a] woman is an image, a black closing [black dress—S.D.] on that woman is another image, and both are concretely representable [on the screen]. However, ‘widow’ emerging from juxtaposing these two images is no longer concretely representable; this is a new representation, new concept, and new image.” The “widow” is both the emotion and the thought of the implied viewer created by juxtaposing the images of a “woman” and “black closing” on the screen. This is truly cinematographic,” Eisenstein claims. See Sergei Eisenstein, Montage (Moscow: Isskusstvo, 1938); accessed at http://lib.ru/CINEMA/ kinolit/EJZENSHTEJN/s_montazh_1938.txt on June 11, 2012). Translation is mine—S.D. 22. It will suffice here to mention the names of several “pioneers” and the fields of the humanities they sought to reinvent along the lines of strict and objective science: Husserl for phenomenology, Freud for psychoanalysis, Windelband and logical positivism for philosophy, and Saussure for linguistics or semiology. 23. For an example of this axiom at work in a traditional setting and compared to the scientific-historical discourse, see Jonathan Boyarin, “Voices around the Text: The Ethnography of Reading at Mesivta Tifereth Jerusalem,”
Notes to pages 223–29 347
in The Ethnography of Reading, ed. Jonathan Boyarin (Berkeley: University of California Press, 1993). 24. See Maurice Merleau-Ponty, The Incarnate Subject: Malebranche, Biran, and Bergson on the Union of Body and Soul, ed. Andrew G. Bjelland and Patrick Burke, preface by Jacques Taminiaux, trans. Paul B. Milan (Amherst, N.Y.: Humanity Books, 2001). 25. Just as Karl Popper’s idea of “falsifiability” establishes the protocol of a positive scientific truth as opposed to a philosophical or metaphysical “truth” in a sense that any “metaphysical truth” can neither be verified or even falsified, the Talmudic protocol of “refutability” is a diegetic protocol of the Talmud that establishes not the truth of the tannaitic teachings on the record, but rather, through the process of refutation followed by a successful defense, establishes the reliability of the record itself. 26. See Koselleck, Futures Past. For an analysis of the connection between textuality and historical consciousness, see Mary Carruthers, The Book of Memory: A Study of Memory in Medieval Culture (Cambridge: Cambridge University Press, 1990). 27. These axioms relate to linear time and linear signification as they progress respectively from past to present and from oral to written. Of course, the order of analysis could reverse the directions and go from present to past and from written to oral, yet this always will be a reversal of what the axioms posit as the original direction. 28. Halivni, Sources and Traditions: A Source Critical Commentary on the Talmud: Tractate Baba Metzi‘a. In my reading, I suspend any judgment of whether the new scientific concept of stamm’aim fits well enough with the old traditional concepts of tann’aim, amor’aim, and savor’aim. Instead of making such a judgment, I heuristically return Halivni’s theory of the stamm’aim to the literary and ontological framework to which it initially intellectually belongs. The framework in question is not only ontological, but also literal. The notions it conveys were first conceived in film theory in the first half of the twentieth century and then were extended beyond film theory to both literary theory at large and its ontological extension, known as political ontology, that analyzes human being not as an object of science, but rather in the aspect of its being- there, seen as a result of a process similar to that of filmic montage. 29. On a different note, there are of course big differences between the classical rhetorical category of invention and the Talmudic category of h.iddush (which literary also means “invention or novelty”). Among other differences that Canpanton taught, there is one that is perhaps the most important. For him, an invention made in Talmudic contemplation is different from the classical rhetorical invention because the latter is going forward in time, from
348 Notes to pages 229–31
no invention to making one, while the former is going backward in time, from invention having been made to discovering the intellectual situation as it “must have been” before the invention took place. In such a process we recognize that something that now seems obvious might not have been obvious earlier and therefore is an invention, and not something already known. However, the difference is not only in the direction of time. It also has to do with the quality of its flow. Classic rhetorical invention is a smooth or at least a welcoming transition from what was before to what is invented now. Canpanton’s contemplator of the Talmud has to deal with a rupture in time—he or she needs to reveal the world of mind as it existed before the invention took place and thus needs not only to hold the invention in mind, but also to be able to discover a world in which the invention was not yet there, not even in a latent form. See Daniel Boyarin, Ha-‘Iyun Ha-Sefaradi: Le-Farshanut Ha-Talmud Shel Megorshe Sefarad (Jerusalem: Mekhon Ben-Tsevi le-heker Kehilot Yisra’el ba- Mizrah, 1989). 30. My decision to proceed to a literary analysis of the Talmudic montage first and only then to continue with the ontological analysis thereof is less self-evident than it seems. Although that decision sounds quite natural, it still is not completely obvious what kind of analysis, literary or ontological, should be done first. Since, as already mentioned, political ontology shows the literary analysis of cutting devices and their representational effects to be only a partial case of a larger analysis of political-ontological constructions of reality, I have an equally good reason to begin either from the literary analysis or from the ontological assumptions that underlie a specific version of it. However, in the order of that presentation, I will proceed in the literary mode of analysis first, keeping in mind that ontological analysis should always go in parallel with it, not after it. 31. David Weiss Halivni, Midrash, Mishnah, and Gemara: The Jewish Predilection for Justified Law (Cambridge, Mass.: Harvard University Press, 1986), 78–79, cited in Halivni, Sources and Traditions: A Source Critical Commentary on the Talmud Tractate Baba Kama, 479. 32. My argument does not, however, make any direct suggestion as to the typological place that the Talmud Yerushalmi, the Palestinian Talmud, may have on a map defined by differences between the Oral Torah as Mishnah, midrash as expounding the Written Torah, and Yerushalmi itself as a place where the competing projects of the Mishnah and Midrash come together in a nascent intellectual form of the Talmud. For discussion of that place see Slomo Naeh, “Mivneho Ve-Halukoto Shel Midrash Torat Cohanin. 1. Megilot (Le Kodikologia Ha Talmudit Ha Kduma); 2. Parashot, Prakim, Halakhot,” Tarbits (1997).
Notes to pages 231–35 349
33. As Halivni specifies in his introduction to the volume on Baba Metzi‘a, in his previous work he talked extensively about the activity of the redactors and divided that activity into three kinds: extending, adding to, and filling in the gaps in [the texts.] In the introduction to his commentary to Baba Metzi‘a he sets out to reinforce his earlier words by means of defining the above kinds further, and by adding the fourth kind, changing [the text]. as well as by looking at the relationship between them. See Halivni, Introduction, to Sources and Traditions, Baba Metzi‘a, 11–26. 34. See David Weiss Halivni, Introduction to Sources and Traditions: A Source Critical Commentary on the Talmud Tractate Baba Kama (Jerusalem, 1993). 35. See, for example, Chanoch Albeck, Mavo’ la-Talmudim (Tel Aviv: Devir, 1969). 36. Halivni, Sources and Traditions: A Source Critical Commentary on the Talmud Tractate Baba Kama, 9–10. 37. Ibid., 9. 38. Ibid. 39. Sharing the ideas of montage in the sense of putting fragments together, Halivni still implies in this theory the necessity of the historical existence of a director, a cinematographer, and a cutter-editor, or their equivalents, for the Talmud, rather than conceiving these figures as part and parcel of montage, of which these figures are functions in the same way in which “The Author” is a function of the novel. As discussed above, this is not to deny the existence of Samuel Clemens, but it is to resist proving the necessity of the historical existence of Mark Twain based on his status as “The Author.” What is more, montage as putting the cuts of the film together, produces an implied viewer with thoughts and emotions that are not on the screen, regardless of whether these thoughts and emotions are preconceived by the cutter. In fact, even if they are, the subject position from which the cutter operates does not exist outside of that of the implied viewer. Even if the two positions do not fully coincide, they can not be fully separate either. Rather than seeing them as historically actual, the positions of implied viewer and of the cutter should be conceived as virtual. 40. A recent line of research initiated by Jeffrey Rubenstein and Daniel Boyarin presents a still-historical, but no longer necessarily historicist, exploration of Talmudic diegesis. It takes a radically new direction. These researchers abandoned the earlier external historicist-chronological approach to the analysis of the Talmud and initiated a new project, now explicitly in the register of diegesis. They addressed the activity of the stamm’aim from within the diegesis of the Talmud, not from the point of view of a historicist chronology thereof. As a result, we gain not only more knowledge about the diegetic background in which the Talmudic action occurs (Rubenstein), but also a clearer sense of
350 Notes to pages 235–37
the diegetic nature of the past that the Talmudic stamm’a entails. The question of the residual role of the historicist concept of stamm’aim as opposed to stamm’a, plus the questions of how the diegetic past (Yavneh) relates to the time/temporality of the Talmud’s montage and what the former can elucidate about the latter, are yet to be explored. Moreover, Boyarin’s historical analysis of the tann’aim and the amor’aim as more ordinary “Greeks” and the stamm’aim as inventors of midrashic/Talmudic antiepistemology, as opposed to stamm’a of the Talmud, can speak to the time and indeed the temporality of the Talmudic montage, which in itself does not have to be timeless and may very well have its own dynamics of change. See Jeffrey L. Rubenstein, ed., The Culture of the Babylonian Talmud (Baltimore: Johns Hopkins University Press, 2003), and Daniel Boyarin, “The Yavneh-Cycle of the Stammaim and the Invention of the Rabbis,” in Creation and Composition: The Contribution of the Bavli Redactors (Stammaim) to the Aggada, ed. J. L. Rubenstein (Tübingen: Mohr/Siebeck, 2005), 237–89. 41. This is not to exclude that, on Halivni’s reading, the redactors may create anonymous questions and answers to set the stage for a dictum they ascribe to an amor’a. 42. The result of the analysis takes us beyond either answering who “edited” the film or claiming that no one did. Instead, it implies figures produced by montage, which is, in Eisenstein’s terms, ustroistvo vsech vechei (meaning, as already discussed, “the setup of all things,” or “creation of all things”). Even Eisenstein positions himself as a figure participating in the work of montage rather than as someone standing in a hypothetical, but impossible historical outside. That, of course, does not set researchers free from continuing to analyze the characters, both diegetic and extradiegetic, produced, as they are, in the film called The Talmud. 43. “When was the decision [to collect everything that remained from the previous give-and-take] made? The dominant view of today (according to Ch. Albeck) is that the decision is to be related to the beginning of the time of the amor’aim, because the stammaitic give-and-take appearing in the Talmud after the Mishnah and Baraitha relates to that time. From that time and on, the give-and-take is subdivided according to the generations and times of the amor’aim, whom it [the process of give-and-take] interpreted. In each generation, the stamm’aim explained and argued for the sayings of the amor’aim of the same generation (with their permission and under their supervision?) mostly in the form of dialectics. [Approvingly] the amor’aim attached it [i.e., the dialectics—S.D] to their own sayings and transmitted [all of] that together to future generations. Hence, the time of anonymous give-and-takes is the time of the amor’aim whom the give-and-takes interpreted, and only the time of the
Notes to pages 237–39 351
anonymous give-and-take [found] in the Mishnah and Baraitha belongs to the beginning of the time of amor’aim.” Halivni, Mekorot U-Masorot: Beurim Ba- Talmud: Masekhet Bava Kama, 8. (Translation is mine—S.D.) 44. They are fully present, if anywhere at all, only in the “external social history” of the Talmud, a direction recently fruitfully supported, amplified, and extended by Rubenstein in The Culture of the Babylonian Talmud. While remaining committed to history and chronology, Rubenstein is in fact carefully and I think very fruitfully reconstructing the diegetic intra-Talmudic backgrounds of the Talmud. 45. It is worth noting that alongside Halivni’s project of identifying sources and traditions, which focuses on the Talmud “itself” and on redactors as the intrinsic part of the Talmud (historic or not), there also exists a less literary and more philologically oriented approach in which the focus is not on what was going on in the Talmud, but rather on what is going on with the Talmud in the long chain of the transmission of its manuscripts. Such an approach, of course, not only applies to the post-Talmudic transmission of the Talmud but also extends itself to the making of the Talmud, that is, to the historical period “before” the Talmud’s completion. The striking difference between that philological approach and Halivni’s literary one can be better seen in answering the basic question of what the Talmud fundamentally is. For a philologist, it is always book, which even in status nascendi was constantly edited as such. For a literary analyst such as Halivni, the Talmud is, first of all, a literary form and only afterward a book. In philological research on the transmission of the Talmudic manuscripts, an anonymous, dynamic, and perhaps heterogeneous agency has, again positively, been identified as “the redactors of the Talmud,” not as the redactors in the Talmud, as if the Talmud has always been a written book to be scientifically edited. Unlike the philological analysts of the editing of the Talmud across the chain of its transmission in manuscripts, Halivni focuses on the intra-Talmudic analysis of the Talmud while putting the “post- Talmudic” transmission of the Talmud aside. 46. Sigmund Freud, Moses and Monotheism, trans. Katherine Jones (New York: Knopf, 1939). 47. There are also other ontological assumptions at work here, but this is not the place to elaborate on them, I therefore limit myself to a brief list of them. In addition to the historical existence of the stamm’aim as self-conscious individuals with conscious and unconscious actions that moved them to act, Halivni thinks along the lines of the Platonist assumptions of the idea (both original and origin) versus its copy—oral signs/origins versus their ideally, but not always really accurate recordings. I refer to the Platonist assumption of an idea equated to both an original and origin, which then gets copied and
352 Notes to pages 239–43
distorted and therefore needs to be restored. That assumption underlies the basic conceptual distinction of Halivni’s project: original sources versus their secondary transmissions, mekorot versus masorot. The Platonist nature of his concepts of makor and masora reveals itself not only in the quite obvious borrowing of the term: Makor is modern Hebrew for always Platonically colored terms of origin or source. As an important addition, Halivni’s theory embeds an automatic preference of qualifying the stammaitic arguments as dialectical versus rhetorical. Unlike rhetorical arguments that change as quickly as the situations in which they arise and therefore do not really welcome mechanical recording or memorization, dialectical arguments always welcome the monumental recording of dialectical victories. The “filmic” analysis of the stammaitic arguments should therefore take into account Halivni’s literary analysis of them and not only clarify their ontological grounds but also revisit those grounds, enabling thereby a more critical application of “filmic,” that is, literary and ontological, analysis to the stammaitic arguments. 48. See Halivni’s own words on the stammaitic arguments and his own conjecture in his Midrash, Mishnah, and Gemara: The Jewish Predilection for Justified Law (Cambridge, Mass: Harvard University Press, 1986), 79. 49. Whether understood as a historical layer of the Talmud, late but formative or expressed in terms of film theory, as a strategy of the Talmud’s cutting, stammaitic discussion has always had a recognized form of dialectical give-and- take. It might seem even possible to think of the structure of at least some units of the Talmudic discourse as being like successive “takes” of a scene in a film. However, such a direct comparison between the Talmud and film, rather than between the theory of the Talmud and film theory, would be far beyond the task of the present analysis. In fact, that comparison may well turn out to be too naturalistic, since even if montage creates a preemptive place for a “viewer” in both the Talmud and a film, in the case of the Talmud, that place is occupied by a “learner,” and because the Talmud comes to him or her from a distant past, the learner is supposed to bridge that distance and to be a much more active participant than the average movie viewer. 50. Halivni, Mekorot U-Masorot: Beurim Ba-Talmud: Masekhet Bava Kama, 7. 51. This question has not been of any interest in either the scientific or the traditional study of the Talmud. Since stammaitic arguments have been treated as dialectical, approaching them through only the secondary form of a record was not a problem, because dialectics, despite its allegedly oral nature, never has resisted recording. As proved by Plato’s record of the oral dialogues of Socrates, even despite Socrates’ objection to recording and memorization, Platonic dialectics welcomes recording, not least because recording can serve as a monument of one’s victory in the dialectical contest. Therefore, Platonic
Notes to pages 243–61 353
dialectical forms of argument not only survive recording, they welcome it. This, however, is not the case with the art of rhetoric, an enterprise that was always intrinsically oral. In monument-oriented reproductions of dialectical victories, one can only comment on what has already been said. However, imitating the masterpieces of rhetoric was never a mechanical reproduction of a record followed by a comment, nor was it limited to commenting on the speeches of the past. Instead of either memorization or recording, a successful student of rhetoric has to produce a new speech, not simply reproduce the old one. As a result of this difference between the statically reproductive nature of dialectics and the dynamically imitative nature of rhetoric, a rhetorical approach to stammaitic argumentation promises to say more about the oral status of stammaitic argumentation than the hitherto dominant dialectical-logical approach. At the end of the day, what we are left with are the “stammaitic” arguments themselves, no matter if they were initially oral or not and no matter who really made them, the earlier Talmudic masters, amor’aim, or their later redactors, or stamm’aim. That, of course, puts the historical interpretation of the Talmud aside. However, it does not help us understand the inner form of these arguments, and therefore a rhetorical analysis of them must be done in terms of the preemptive space of comprehension that they rhetorically create (in montage) and in terms of the actively oral character of themselves (in diegesis).
appendix: talmud criticism, an analytical example: “composer” versus “redactors”: david halivni’s and shamma friedman’s competing readings of baba metzi‘a 76ab 1. David Halivni, Sources and Traditions: Mekorot u-Masorot: Be’urim Ba- Talmud: Masekhet Bava Metsi’A (Jerusalem: Magnes, Hebrew University of Jerusalem, 2003); Shamma Friedman, Talmud Arukh: Perek Ha-Sokher et Ha- Umanin: Bavli Bava Metsia Perek Shishi: Mahadurah al Derekh Ha Mehkar Im Perush Ha-Sugyot (New York: Jewish Theological Seminary, 1990). 2. For convenience of reference, I will index the texts with bold capital roman letters from A to J. 3. I add interpolations in square brackets. If an interpolation crosses the limits of what is a bare grammatical necessity of English rendition and becomes instead an interpretation of meaning, and if that interpretation is mine, I mark it “S.D.”; if it comes from a commentary, I indicate that. 4. There are at least four different ways to understand the compositional connection between (1) the disputants’ reciting the report of the precedent in
354 Notes to pages 261–66
the Rav’s court and (2) the following rebuttal from Rav Ashe. First is a traditional way, in which Rav Ashe directly responds to the inference the disputants in the academy made from the precedent. Second is a critical scholarly reading, in which the first inference belongs not to the disputants, but to one of the transmitters of the tradition about the precedent, to whom Rav Ashe responds, i.e., to Rav Nachman or Rava Bar Abahu. See David W. Halivni, Sources and Traditions: Mekorot u-Masorot: Be’urim Ba-Talmud: Masekhet Bava Metsi’a (Jerusalem: Magnes, Hebrew University of Jerusalem, 2003), 244. Yet another critical reading considers the present form of the composition corrupted, restores the texts, and, as a part of the process of reconstruction, re-ascribes the first inference directly to Rav Ashe, rather than to either the disputants or to the later transmitters of the precedent, thereby establishing coherence of the here and now of the composition. (See Friedman, Talmud Arukh, 22–43.) Finally, there is yet another possibility that fundamentally contradicts none of the critical readings but eliminates part of the concerns they address, and at the same time slightly deviates from the dominant traditional reading. On that fourth reading, the disputants had purposely invoked Rav Ashe’s text, or, if you prefer, rhetorically “recalled” it in the same way in which they recalled the report of the precedent in Rav’s court by way of rhetorical recollection. In this reading, both Rav and Rav Ashe are figures from the past, and thereby offstage. This applies to any inferences the discutants made on their behalf as well, for if they are figures of the past, and are offstage, the difference between the precedent and inference derived there fades, as many other differences do when they belong to the past, which is not considered to have had been present at any time, and surely not in the “here and now” onstage, but are only invoked in the modus of remembering the thoughts, not in the way of remembering the words (or more broadly, the things). 5. And, in larger picture of the sugya in the Talmud, it also leaves the rabbis with a refutation of the Mishnah that was in turn neither finally refuted nor fully accepted. This is, however, beyond the scope of the current analysis. 6. He and other modern researchers of the Talmud borrow the word, but not the concept, from stamm’a de-gemarr’a (“anonymous tradition”), which in this context is a term used in early medieval post-Talmudic literature for unattributed texts in the Talmud. 7. Unlike Halivni’s explicit classificatory divisions of the stamm’aim into kinds found in Halivni and his students, such as “stamm’aim tann’aim”— anonymous reciters of the Mishnah versus “stamm’aim orkhim”—anonymous redactors of the teachings of amor’aim, or “earlier stam” versus “later stam,” etc., the twofold nature of the core concept of the stam is implicit in his argument, rather than explicitly proclaimed.
Notes to pages 266–82
355
8. Some of these attributions or ascriptions are made in reference to savor’aim, or rabbis who run anonymous discourses in the Talmud, while some of them may be known by name, from the post-Talmudic literature of the Geonim. 9. Halivni, Sources and Traditions, 244–45. 10. Ibid., 244. 11. Halivni supports that ascription, using an example from bKethuboth 21b, where in a similar disposition Rav Ashe again seems to be responding to anonymous rabbis in the academy. As Halivni shows, the continuation of the text there makes it clear that the inference is to be ascribed to intermediary amor’aim, and was discussed among them. Ibid. 12. Ibid. 13. “Therefore it seems that originally Rav faced a refutation presented to him directly based on the Mishnah Rosh Ha-Shanah 4:1, rather than based on the inference [from his words.]” Ibid. 14. Ibid. 15. Ibid., 244–45. 16. Ibid., 245. 17. Ibid. 18. Friedman, Talmud Arukh, 7. 19. Ibid. 20. Ibid., 8. 21. Ibid. 22. Ibid., 9. There Friedman gives an “extreme” example of a change in form. In that example, the form “anonymous talmud” (the words “ ”מה טעםetc.) transforms into a part of “apocryph.” This extreme example entailed inserting the least authoritative words, those of the “talmud” to the most authoritative text, that of apocrypha. See ibid., 8–9. 23. Ibid., 21. 24. Historically these prior forms may or may not fashion a stand-alone literary composition. However, in any case, Friedman’s notion of “anterior Talmud” implies a counterpart, which is the notion of the “posterior Talmud,” which is the modern form of the Talmud. The “anterior Talmud” designates alleged prior partitioning of the sources in the forms of apocrypha and dicta as opposed to final partitioning of the sources in apocrypha, dicta, and stamtalmud. The question is, does “posterior Talmud” require not only a relative concept, that of posteriority, as a new stage in the process of becoming, but also a teleological concept of completion or stoppage of that becoming? 25. Friedman, Talmud Arukh, 13. 26. See ibid., 7, 16ff.
fup-dolgopolski-13n.indd 355
10/9/12 12:13 PM
356 Notes to pages 282–98
27. This sets the researcher potentially free from any implications about the existence or nonexistence of the composer of the anonymous talmud, and of the Talmud as well, even if Friedman himself might still entertain such implications. 28. Friedman, Talmud Arukh, 21. 29. See ibid., “Summary,” 7–23. 30. Ibid. 31. Isaak Halevy, Dorot Harischonim (Frankfurt a. M.: Buchdruckerei von M. Slobotzky, 1901). 32. Friedman, Talmud Arukh, 22–23. 33. Ibid. 34. Ibid., 21–23. 35. Even if the textual elements might migrate between forms in either direction, from dicta and apocrypha to the anonymous talmud, or backwards from the anonymous talmud to apocrypha and/or dicta, to access the time of genesis, Friedman, and for that matter any researchers must always begin with the time of the resulting composition. 36. Bacher, The Jewish Encyclopedia, XII, 15, cited in Friedman, Talmud Arukh, 22n93. 37. Similarly, it only seems that sun rotates around the earth, but that seeming has an element of reality to it. That element of reality comes from our real position on earth, neither somewhere else in space nor on the sun. 38. Needless to say, since both scholars analyze the genesis of the modern form of the Talmud, the outcomes of their analysis do not necessarily bear implications for legalistic inferences from Talmudic discussions developed in medieval and modern periods. 39. According to post-Talmudic genealogical tradition, he was a fifth- century master of the Talmud. 40. The phenomenon of creating a refutation to be deflected by an already existing defense is described in Halivni’s analysis of the Talmud. See, for example, Halivni, Sources and Traditions, 1–7. However, despite his etiological diachronic analysis of the genesis of the scene, Halivni’s running assumption would still be that the resulting conversation is staged as a synchronic face-to- face communication between the refuter and the defender, even if it is taking place much later than it seems for a noncritical reader.
bibliography
Abelard, Peter. Ethical Writings: His Ethics or “Know Yourself” and His Dialogue between a Philosopher, a Jew, and a Christian. Translated by Paul Vincent Spade, with an introduction by Marilyn McCord Adams. Indianapolis: Hackett, 1995. Adorno, Theodor W. Aesthetic Theory. Edited by Gretel Adorno and Rolf Tiedemann. Newly translated, edited, and with a translator’s introduction by Robert Hullot-Kentor. Continuum Impacts. London: Continuum, 2004. Agamben, Giorgio. The Coming Community. Translated by Michael Hardt. Theory Out of Bounds, vol. 1. Minneapolis: University of Minnesota Press, 1993. ———. The Open: Man and Animal. Stanford, Calif.: Stanford University Press, 2004. Albeck, Chanoch. Einführung in die Mischna. [Aus d. Hebr. übers. von Tamar u. Pessach Galewski]. Studia Judaica, vol. 6. Berlin: De Gruyter, 1971.
357
358 Bibliography
———. Introduction to the Talmud, Bavli and Yerushalmi. Tel Aviv: Dvir, 1969. ———. Mavo la-Mishnah. Jerusalem: Bialik Institute, 1966 or 1967. ———. Mavo’ la-Talmudim. Tel Aviv: Devir, 1969. ———. Mehkarim be-Varaita Ve-Tosefta. Jerusalem, 1943/44. ———. Shishah Sidre Mishnah. Jerusalem: Mosad Bi’alik., 1969. ———. Untersuchungen über die Halakischen Midraschim, von Chanoch Albeck. Berlin: Akademie-Verlag, 1927. ———. Untersuchungen über die Redaktion der Mischna. Berlin: C. A. Schwetschke, 1923. Althusser, Louis, and Etienne Balibar. Reading Capital: (Radical Thinkers). London: Verso, 2009. The Ante-Nicene Fathers: Translations of the Writings of the Fathers Down to A.D. 325. New York: C. Scribner’s Sons, 1899–1900. Antoine-Dunne, Jean, and Paula Quigley. The Montage Principle: Eisenstein in New Cultural and Critical Context. Critical Studies series. Amsterdam, N.Y.: Rodopi, 2004. Aquinas, Thomas. De Unitate Intellectus contra Averroistas. In French. Translation, introduction, bibliography, chronology, and notes and index by Alain de Libera. Paris: GF-Flammarion, 1994. Arendt, Hannah. The Human Condition. Charles R. Walgreen Foundation Lectures. Chicago: University of Chicago Press, 1958. Aristotle. Basic Works of Aristotle. Translated and edited by Richard McKeon. New York: Modern Library, 2001. Auerbach, Erich. Mimesis. Translated from the German by Willard R. Trask, with a new introduction by Edward W. Said. Princeton, N.J.: Princeton University Press, 2003. Augustine. Confessions. Edited by George Stade. New York: Barnes and Noble Classics, 2007. ———. De Doctrina Christiana (Books 1–3 396/397, Book 4 426; Christian Doctrine). Austin, J. L. (John Langshaw). Philosophical Papers. Edited by J. O. Urmson and G. J. Warnock. Oxford: Clarendon Press, 1970. Badiou, Alain. Being and Event. New York: Continuum, 2006. Barthes, Roland. Camera Lucida: Reflections on Photography. Translated by Richard Howard. New York: Hill and Wang, 1981.
Bibliography 359
———. Mythologies. Selected and translated from the French by Annette Lavers. London: J. Cape, 1972. Barzilay, Yitzhak. Manasseh of Ilya: Precursor of Modernity among the Jews of Eastern Europe. Jerusalem: Magnes Press, Hebrew University, 1999. Berkowitz, Beth A. Execution and Invention: Death Penalty Discourse in Early Rabbinic and Christian Cultures. New York: Oxford University Press, 2006. Biddick, Kathleen. The Typological Imaginary: Circumcision, Technology, History. Philadelphia: University of Pennsylvania Press, 2003. Blair, Walter. Mark Twain’s Hannibal, Huck and Tom. Berkeley: University of California Press, 1969. Boyarin, Daniel. Border Lines: The Partition of Judaeo-Christianity. Divinations. Philadelphia: University of Pennsylvania Press, 2004. ———. Carnal Israel: Reading Sex in Talmudic Culture. Berkeley: University of California Press, 1993. ———. Ha-‘Iyun Ha-Sefaradi: Le-Farshanut Ha-Talmud Shel Megorshe Sefarad. Jerusalem: Mekhon Ben-Tsevi le-heker Kehilot Yisra’el ba- Mizrah, 1989. ———. The Jewish Gospels: The Story of the Jewish Christ. New York: New Press, 2011. ———. Socrates and the Fat Rabbis. Chicago: University of Chicago Press, 2009. Boyarin, Jonathan, ed. The Ethnography of Reading. Berkeley: University of California Press, 1993. Brentano, Franz Clemens. Die Psychologie des Aristoteles insbesondere seine Lehre vom Nous Poietikos. [von] Franz Brentano. Nebst einer Beilage über das Wirken des Aristotelischen Gottes. Darmstadt: Wissenschaftliche Buchgesellschaft, 1967. Briel, Holger, and Andreas Kramer, eds. In Practice: Adorno, Critical Theory, and Cultural Studies. German Linguistic and Cultural Studies, vol. 9. Oxford: P. Lang, 2001. Canpanton, Izh.ak [also Kanpanton, Yitshak or Isaac], Shemu’el V.alenci, and Isaak S. Lange [Kanpanton, R. Yitshak, Shemu’el al Valensi, and Yitshak Shimshon Langeh]. Darkhe ha-Talmud [The Ways of Talmud]. Jerusalem: Y. Sh. Langeh, 1980 or 1981.
360 Bibliography
Carruthers, Mary. The Book of Memory: A Study of Memory in Medieval Culture. Cambridge: Cambridge University Press, 1990. Cavell, Stanley. The Claim of Reason: Wittgenstein, Skepticism, Morality, and Tragedy. Oxford: Clarendon Press; New York: Oxford University Press, 1979. Crocker, Stephen. “Into the Interval: On Deleuze’s Reversal of Time and Movement.” Continental Philosophy Review 34, no. 1 (2001): 45–67. http://www.springerlink.com.www2.lib.ku.edu:2048/content/ xl44243x3w462twv/fulltext.pdf. DeLanda, Manuel. Intensive Science and Virtual Philosophy. Transversals. London: Continuum, 2002. Deleuze, Gilles. Cinema. Translated by Hugh Tomlinson and Barbara Habberjam. Minneapolis: University of Minnesota Press, 1986–c. 1989. ———. The Logic of Sense. Edited by Constantin V. Boundas. Translated by Mark Lester and Charles Stivale. European Perspectives. New York: Columbia University Press, 1990. De Man, Paul. Blindness and Insight: Essays in the Rhetoric of Contemporary Criticism. New York: Oxford University Press, 1971. Derrida, Jacques. Of Grammatology. Translated by Gayatri Chakravorty Spivak. Baltimore: Johns Hopkins University Press, 1998. de Vries, Hent, and Lawrence E. Sullivan, eds. Political Theologies: Public Religions in a Post-Secular World. New York: Fordham University Press, 2006. Doane, Mary Ann. The Emergence of Cinematic Time: Modernity, Contingency, the Archive. Cambridge, Mass.: Harvard University Press, 2002. Dolgopolski, Sergey. What Is Talmud? The Art of Disagreement. New York: Fordham University Press, 2009. Ebbessen, Stan. Commentators and Commentaries on Aristotle’s Sophistici Elenchi: A Study of Post-Aristotelian Ancient and Medieval Writings on Fallacies, vol. 2, Corpus Latinum Commentariorum in Aristotelem Graecorum 7. Leiden: Brill, 1981: 27–152. Eisenstein, Sergei. “Montage.” In Russian. In Sergey Eisenstein, Selected Works. vol. 2. Moscow: “Iskusstvo,” 1968. ———. Montage. In Russian. Edited by Naum Kleiman. Moscow: Musei Kino, 2000.
Bibliography 361
Epstein, J. N. (Jacob Nahum). Introduction to Tannaitic Literature: Mishna, Tosephta and Halakhic Midrashim. Jerusalem: Y. L. Magnes; Tel Aviv: Dvir, 1957. Farías, Victor. Heidegger and Nazism. Edited, with a foreword by Joseph Margolis and Tom Rockmore. French materials translated by Paul Burrell and the advice of Dominic Di Bernardi. German materials translated by Gabriel R. Ricci. Philadelphia: Temple University Press, 1989. Faye, Emmanuel. Heidegger: The Introduction of Nazism into Philosophy in Light of the Unpublished Seminars of 1933–1935. Translated by Michael B. Smith, with a foreword by Tom Rockmore. New Haven: Yale University Press, 2009. Fichte, Johann Gottlieb. “The Spirit and the Letter within Philosophy.” In Early Philosophical Writings, edited by Daniel Breazeale, 192–215. Ithaca, N.Y.: Cornell University Press, 1988. Fischel, Henry A. Rabbinic Literature and Greco-Roman Philosophy. Leiden: Brill, 1973. Foucault, Michel. Language, Counter-Memory, Practice: Selected Essays and Interviews. Edited, with an introduction by Donald F. Bouchard. Translated from the French by Donald F. Bouchard and Sherry Simon. Ithaca, N.Y.: Cornell University Press, 1977. ———. L’Archéologie du savoir. Paris: [Gallimard], 1969. ———. Religion and Culture. Selected and edited by Jeremy R. Carrette. New York: Routledge, 1999. Fraade, Steven D. “Rabbinic Polysemy and Pluralism Revisited: Between Praxis and Thematisation.” AJS Review 31, no. 1 (2007): 1–40. Franks, Paul. All or Nothing: Systematicity, Transcendental Arguments, and Skepticism in German Idealism. Cambridge, Mass.: Harvard University Press, 2005. ———. “Serpentine Naturalism and Protean Nihilism: Transcendental Philosophy in Anthropological Neo-Kantianism, German Idealism, and Neo-Kantianism.” In Oxford Handbook of Continental Philosophy, edited by Brian Leiter and Michael Rosen, 243–86. Oxford: Oxford University Press, 2007. Freud, Sigmund. Moses and Monotheism. Translated by Katherine Jones. New York: Knopf, 1939.
362 Bibliography
Friedman, Shamma. “A Good Story Deserves Retelling—The Unfolding of the Akiva Legend.” JSIJ 3 (2004): 55–93. ———. “Perek Ha-Isha Raba BeBavli be Tziruf Mavo Klali al Derekh Heker Hasygia.” In Mekhkarim UMekorot: Sefer Alef, 275–440. New York: Bet ha-midrash le-Rabanim be-Amerikah, 5738. ———. Talmud Arukh: Perek Ha-Sokher et Ha-Umanin: Bavli Bava Metsia Perek Shishi: Mahadurah al Derekh Ha Mehkar Im Perush Ha-Sugyot. New York: Jewish Theological Seminary, 1990. Gibson, William M. Mark Twain’s Mysterious Stranger Manuscripts. Berkeley: University of California Press, 1969. Goldberg, Hillel. “An Early Psychologist of the Unconscious.” Journal of the History of Ideas 43, no. 2 (1982): 269–84. Golinkin, ba-‘arikhat David, et al. 880–01 Torah Li-Shemah: Mehkarim be- Mada‘e Ha-Yahadut Li-Khevod Profesor Shama Yehudah Fridman. 880–02. Jerusalem: Mekhon Shekhter le-mada‘e ha-Yahadut, 2007. Hadot, Pierre. Plotinus, or, The Simplicity of Vision. Translated by Michael Chase, with an introduction by Arnold I. Davidson. Chicago: University of Chicago Press, 1993. Halevy, Isaak. Dorot Harischonim. Frankfurt a. M.: Buchdruckerei von M. Slobotzky, 1901. Halivni, David Weiss. The Book and the Sword: A Life of Learning in the Shadow of Destruction. New York: Farrar, Straus and Giroux, 1996. ———. Breaking the Tablets: Jewish Theology after the Shoah. Edited and introduced by Peter Ochs. Lanham: Rowman and Littlefield, 2007. ———. Mekorot u-Masorot: Be’urim Ba-Talmud: Masekhet Bava Kama. Jerusalem: Jerusalem: Magnes, Hebrew University of Jerusalem, 1993. ———. Mekorot u-Masorot: Beurim Ba-Talmud: Masekhet Bava Metsi’a. Jerusalem: Magnes, Hebrew University of Jerusalem, 2003. ———. Mekorot u-Masorot: Be’urim Ba-Talmud: Masekhtot ‘Eruvin u-Fesahim. Jerusalem: Jewish Theological Seminary, 1981 or 1982. ———. Midrash, Mishnah, and Gemara: The Jewish Predilection for Justified Law. Cambridge, Mass.: Harvard University Press, 1986. ———. Peshat and Derash: Plain and Applied Meaning in Rabbinic Exegesis. New York: Oxford University Press, 1991.
Bibliography 363
———. Sources and Traditions: A Source Critical Commentary on the Talmud Seder Moed from Yoma to Hagiga. Jerusalem: Jewish Theological Seminary, 1973. Heidegger, Martin. Being and Time: A Translation of Sein und Zeit. Translated by Joan Stambaugh. SUNY Series in Contemporary Continental Philosophy. Albany: State University of New York Press, 1996. ———. The Question concerning Technology and Other Essays. Translated and with an introduction by William Lovitt. New York: Harper Colophon Books, 1977. Heine, Heinrich. Heines Werke in Fünf Bänden. Bd. 5. Weimar: Aublau Verlag, 1981. Hollander, Dana. Exemplarity and Chosenness: Rosenzweig and Derrida on the Nation of Philosophy. Cultural Memory in the Present. Stanford, Calif.: Stanford University Press, 2008. Hughes, Aaron W., and Elliot R. Wolfson, eds. New Directions in Jewish Philosophy. Bloomington: Indiana University Press, 2010. Husserl, Edmund. “Grundlegende Untersuchungen Zum Phaenomenologischen Ursprung der Raumlichkeit der Natur.” In Philosophical Essays in Memory of Edmund Husserl, 307–25. Cambridge, Mass.: Harvard University Press, 1940. Hutton, Patrick H. “The Art of Memory Reconceived: From Rhetoric to Psychoanalysis.” Journal of the HIstory of Ideas 48, no. 3 (1987): 372. Irenaeus. Five Books of S. Irenaeus, Bishop of Lyons, Against Heresies. Oxford: Parker, 1872. Jaffee, Martin S. Early Judaism: Religious Worlds of the First Judaic Millennium. Studies and Texts in Jewish History and Culture. Bethesda: University Press of Maryland, 2006. ———. The End of Jewish Radar: Snapshots of a Postethnic American Judaism. Samuel and Althea Stroum Chair in Jewish Studies. New York: Universe, 2009. ———. Inner-Worldly Monasticism: Towards a Model of Rabbinic-Halakhic Spirituality. Etna, Calif.: Center for Traditionalist Orthodox Studies, 2006. ———. Mishnah’s Theology of Tithing: A Study of Tractate Maaserot. Brown Judaic Studies, no. 19. Chico, Calif.: Scholars Press, 1981.
364 Bibliography
———. Torah in the Mouth: Writing and Oral Tradition in Palestinian Judaism, 200 BCE–400 CE. New York: Oxford University Press, 2001. ———. The Wars of Torah: The Sublimation of Violence in Rabbinic Piety. Eugene: University of Oregon Books: University of Oregon Humanities Center, 2006. Janowitz, Naomi. Magic in the Roman World: Pagan, Jews, and Christians. London: Routledge, 2001. Kavka, Martin. Jewish Messianism and the History of Philosophy. Cambridge: Cambridge University Press, 2004. Klein, Gil. “Non-Canonical Towns: Representation of Urban Paradigms in Talmudic Understanding of the Jewish City.” Studia Rosenthaliana 40 (2007): 231–63. Koselleck, Reinhart. Futures Past: On the Semantics of Historical Time. Translated and with an introduction by Keith Tribe. Studies in Contemporary German Social Thought. With a new introduction. New York: Columbia University Press, 2004. Lacoue-Labarthe, Philippe. Typography: Mimesis, Philosophy, Politics. Stanford, Calif.: Stanford University Press, 1988. Lapin, Hayim, and Dale B. Martin, eds. Jews, Antiquity, and the Nineteenth- Century Imagination. Potomac: University Press of Maryland, 2003. Levinas, Emmanuel. Difficile Liberté: Essais sur le Judaïsme. Présences du Judaïsme. Paris: A. Michel, 1976. ———. “Ethics as First Philosophy.” In The Levinas Reader, 75–87. 1989. Blackwell, 1984. ———. Outside the Subject. Translated by Michael B. Smith. Meridian: Crossing Aesthetics; Stanford, Calif.: Stanford University Press, 1994. Libera, Alain de. Albert le Grand et la Philosophie. A la Recherche de la Vrit. Paris: J. Vrin, 1990. ———. Archéologie du sujet. Vol. 1: Naissance du sujet. Paris, 2007. ———. La Querelle Des Universaux: De Platon la Fin Du Moyen Age. Des Travaux. Paris: Editions du Seuil, 1996. ———. Le Probleme de l’Tre Chez Matre Eckhart: Logique et Metaphysique de l’Analogie. Cahiers de la Revue de Thologie et de Philosophie. Lausanne: Revue de thologie et de philosophie, 1980.
Bibliography 365
———. Penser au Moyen Age. Chemins de Pense. Paris: Editions du Seuil, 1991. Libera, Alain de, Abdelali Elamrani-Jamal, and Alain Galonnier, eds. Langages et Philosophie: Hommage Jean Jolivet. Etudes de Philosophie Mdivale. Paris: J. Vrin, 1997. Losev, A. F. Istoriia Antichnoi Estetiki. Poslednie Veka. Moscow: “Iskusstvo,” 1988. Lutsatto, Mosheh Chayim [Luzzato Moshe Chayim]. Sefer Derekh Tevunot: Kolel Bo Kol Darkhe Ha-Hitbonenut Veha-Haskalah le-Havin be-Nakel Darkhe Ha-Talmud Vi-Yesodot Ha-Pilpul be-Derekh Ketsarah. Jerusalem: Yeshivat ha-tefutsot: Feldhaim, 1988 or 1989. ———. The Ways of Reason. Jerusalem: Diaspora Yeshiva, Feldheim, 1989. Lyotard, Jean-François. Discourse, Figure. Minneapolis: University of Minnesota Press, 1971. ———. Toward the Postmodern. Edited by Robert Harvey and Mark S. Roberts. Philosophy and Literary Theory. Atlantic Highlands, N.J.: Humanities Press, 1993. Maccoby, Hyam. The Philosophy of the Talmud. Curzon Jewish Philosophy Series. London: RoutledgeCurzon, 2002. Maimonides, Moses. Mishneh Torah, Hu Ha-Yad Ha-Hazakah, ‘Im Haagot Ha-RABaD, u-Ferush Ha-Magid Mishneh, Ha-Kesef Mishneh [Ve-‘od]. Ve-‘Atah Hosafnu Hosafot Hadashot.. Teshuvot Ha-RaMBaM, Teshuvot Rabi Avraham Ben Ha-RaMBaM.. Be’urim, Hidushim Ve-Hagahot Me-et Hazon ISH. Jerusalem: El ha-mekorot, 1953/54–1956/57. Manovich, Lev. The Language of New Media. Leonardo series. Cambridge, Mass.: MIT Press, 2001. Markosian, Ned. “The Open Past.” Philosophical Studies 79, no. 1 (1995): 95–105. Massumi, Brian. Parables for the Virtual: Movement, Affect, Sensation. Post- Contemporary Interventions. Durham, N.C.: Duke University Press, 2002. Meillassoux, Quentin. After Finitude: An Essay on the Necessity of Contingency. Translated by Ray Brassier. London: Continuum, 2008. Merleau-Ponty, Maurice. The Incarnate Subject: Malebranche, Biran, and Bergson on the Union of Body and Soul. Edited by Andrew G. Bjelland and
366 Bibliography
Patrick Burke. With a preface by Jacques Taminiaux. Translation by Paul B. Milan. Contemporary Studies in Philosophy and the Human Sciences. Amherst, N.Y.: Humanity Books, 2001. Naeh, Slomo. “Mivneho Ve-Halukoto Shel Midrash Torat Cohanin. 1. Megilot (Le Kodikologia Ha Talmudit Ha Kduma); 2. Parashot, Prakim, Halakhot.” Tarbits (1997). Nancy, Jean-Luc. Hegel: The Restlessness of the Negative. Translated by Jason Smith and Steven Miller. Minneapolis: University of Minnesota Press, 2002. Nirenberg, David. “Mass Conversion and Genealogical Mentalities: Jews and Christians in Fifteenth-Century Spain.” Past and Present 174, no. 1 (2002): 3–41. Philo. Philo. Loeb Classical Library. Cambridge, Mass.: Harvard University Press. Richardson, William J. Heidegger: Through Phenomenology to Thought. With a preface by Martin Heidegger. Phaenomenologica. The Hague: Martinus Nijhoff, 1974. Rosenak, Avinoam. “Philosophy and Halakhic Thought: A Reading in Talmudic Lessons by Rab I. D. Soloveichik in Light of Neokantian Models.” In Hebrew. In Faith in Ages of Change, edited by A. Saggi, 275–301. [Tel Aviv:] Alinor [Library] and [Ein Tsurim:] Yaacov Herzog Center Press. Rosen-Zvi, Ishai. Demonic Desires: “Yezer Hara” and the Problem of Evil in Late Antiquity. Philadelphia: University of Pennsylvania Press, 2011. Rosenzweig, Franz. The Star of Redemption. Translated from the second edition of 1930 by William W. Hallo. New York: Holt, Rinehart and Winston, 1971. Rossi, Paolo. Clavis Universalis: Arti Mnemoniche e Logica Combinatoria da Lullo a Leibniz. Milan, 1960. Rubenstein, Jeffrey L., ed. Creation and Composition: The Contribution of the Bavli Redactors (Stammaim) to the Aggada. Texts and Studies in Ancient Judaism. Tübingen: Mohr Siebeck, 2005. ———. The Culture of the Babylonian Talmud. Baltimore: Johns Hopkins University Press, 2003. Russell, Bertrand. The Analysis of Mind. Muirhead Library of Philosophy. London: G. Allen and Unwin; New York: Macmillan, 1951.
Bibliography 367
Safranski, Rüdiger. Martin Heidegger: Between Good and Evil. Translated by Ewald Osers. Cambridge, Mass.: Harvard University Press, 1998. Samely, Alexander. Forms of Rabbinic Literature and Thought: An Introduction. Oxford: Oxford University Press, 2007. ———. Rabbinic Interpretation of Scripture in the Mishnah. New York: Oxford University Press, 2002. Schofer, Jonathan Wyn. Confronting Vulnerability: The Body and the Divine in Rabbinic Ethics. Chicago: University of Chicago Press, 2010. ———. The Making of a Sage: A Study in Rabbinic Ethics. Madison: University of Wisconsin Press, 2004. Schumann, Andrew. “Non-Well-Foundedness in Judaic Logic.” Studies in Logic, Grammar, and Rhetoric 13, no. 26 (2008): 41–59. ———. Sovremennaia Logika. Teoriia i Praktika. Minsk: Ekonompress, 2004. A Select Library of the Nicene and Post-Nicene Fathers of the Christian Church / Edited by Philip Schaff in Connection with a Number of Patristic Scholars of Europe and America. 1989. Edinburgh: T. & T. Clark; Grand Rapids, Mich.: Wm. B. Eerdmans, 1994. Sherira ben Hanina, Gaon, and Benjamin Manasseh Lewin. Igeret Rav Sherira Gaon: Mesuderet Be-Shene Nus’ha’ot, Nusa’h Sefarad ‘Ve-Nusa’h Tsarfat, ‘Im ‘Hilufe Ha-Girsa’ot. Jerusalem: Makor, 1971. Sokoloff, Michael. A Dictionary of Jewish Babylonian Aramaic of the Talmudic and Geonic Periods. Ramat Gan, Israel: Bar Ilan University; Baltimore: John Hopkins University Press, 2002. Soloweyczik, Elias, ed. Moses Mamionides [sic] Yad-Hachasaka, or Mishne Torah Hilchoth Melahim, Containing Laws Concerning Kings, and Their Wars. Tr. from the Hebrew Into English by Several Learned Writers. Ed. and Rev. London: Printed by T. W. Nicholson, 1863. Solowiejchzyk, Josef. Das reine Denken und die Seinskonstituirung bei Herman Cohen. 1930–32. Pruzana (Poland): Friedrich-Wilhelms-Universitaet zu Berlin, Schoen Books Reprint, 2005. Steinmetz, Devora. “Agada Unbound: Inter-Agadic Characterization of Sages in the Bavli and Implications for Reading Agada.” In Creation and Composition: The Contribution of the Bavli Redactors (Stammaim) to the Aggada. Texts and Studies in Ancient Judaism, 293–337. Tübingen: Mohr Siebeck, 2005.
368 Bibliography
Stern, Sacha. Time and Process in Ancient Judaism. Oxford: Littman Library of Jewish Civilisation, 2003. Strauss, Leo. Philosophy and Law: Essays toward the Understanding of Maimonides and His Predecessors. Translated from the German by Fred Baumann, with a foreword by Ralph Lerner. Philadelphia: Jewish Publication Society, 1987. Talmud Yerushalmi according to Ms. Or. 4720 (Scal. 3) of the Leiden Univeristy Library With Restorations and Corrections. Introduction by Yaacov Sussman. Historical Dictionary of the Hebrew Language. Jerusalem: Academy of the Hebrew Language, 2005. Ta-Shema, Yisrael M. Ha-Sifrut Ha-Parshanit la-Talmud be-Eropah Uvi- Tsefon Afrikah: Korot, Ishim Ve-Shiot. Jerusalem: Magnes, 2003 or 2004. Twain, Mark. Mark Twain’s Correspondence with Henry Huttleston Rogers, 1893–1909. Edited by Lewis Leary. Berkeley: University of California Press, 1969. Twersky, Isadore. Introduction to the Code of Maimonides (Mishneh Torah). Yale Judaica Series, vol. 22. New Haven: Yale University Press, 1980. von Maarten, J. F. M. Hoenen, and Alain de Libera, eds. Albertus Magnus und der Albertismus: Deutsche Philosophische Kultur Des Mittelalters. Studien und Texte zur Geistesgeschichte des Mittelalters, vol. 48. Leiden: Brill, 1995. Williams, Michael A. Rethinking “Gnosticism”: An Argument for Dismantling a Dubious Category. Princeton: Princeton University Press, 1996. Wimsatt, W. K. The Verbal Icon: Studies in the Meaning of Poetry. And two preliminary essays written in collaboration with Monroe C. Beardsley. Lexington: University of Kentucky Press, 1954. Wolfson, Elliot. Language, Eros, Being. New York: Fordham University Press, 2005. Yadin, Azzan. Scripture as Logos: Rabbi Ishmael and the Origins of Midrash. Divinations. Philadelphia: University of Pennsylvania Press, 2004. Yates, Frances A. The Art of Memory. Chicago: University of Chicago Press, 2001. Yerushalmi, Yosef Hayim. Zakhor, Jewish History and Jewish Memory. Seattle: University of Washington Press, 1982. Žižek, Slavoj, and Glyn Daly. Conversations with Žižek. Cambridge, UK: Polity; Malden, Mass.: Blackwell, 2004.
acknowledgments
I am thankful for the example and inspiration I find in Daniel Boyarin’s scholarship and intellectual personality. Important for, and divergent from, the argument in this book, Boyarin’s recent research highlights that dialogues both in Plato and in the Babylonian Talmud are far from being as open as they might seem. In the same argument, he reclaims “sophistic” as the forgotten stream of the Greek/Western tradition of philosophy. Both moves prove heuristic and inspiring for my work, even if, and precisely because, I took a different path of inquiry, rooted in the contemporary context in philosophy as well as in philosophical reflection on contemporary Talmud criticism. This path goes through Talmud criticism to exploring Talmud (from its earlier instantiation in the Babylonian Talmud to other instantiations before and after) as an intellectual project coextensive in scope to the projects of philosophy and rhetoric. That move is no less inspired by Daniel Boyarin’s earlier scholarship as by a (post-) Heideggerian understanding of hermeneutics, which I applied, negotiated, and complicated in this book. As a result of, and despite, these influences, I distanced myself from deciding whether or not Talmudic or Platonic dialogues are genuinely open or only seem to be so; instead, I asked about the openness of the past, which the dialogues in the Talmud perform. Attempting to remember the authoritative texts of the past better, these dialogues suspend, I submit, any Platonic framework of inquiry on what truly is or was, in which the question of openness of a dialogue or discussion can only arise. I am deeply intellectually indebted to many friends and colleagues. In friendship and deep appreciation I thank Edouard Nadtochii of the University 369
370 Acknowledgments
of Lausanne, for our intense conversations on de Libera and the nature of philosophical, literary, and critical work in general. I cordially thank Jonathan Boyarin of the University of North Carolina at Chapel Hill for the sense of genuine intellectual and existential engagement I felt taking shape between us over the years of my journey from California to Kansas to New York and at various academic intersections in New York City, London, and beyond. I also deeply thank him for suggestions he offered after reading the manuscript. I gladly recognize my intellectual debt to Bruce Rosenstock of the University of Illinois at Urbana-Champaign. I benefited from his philosophical and especially historical-philosophical genius, which were revealed to me in our always very stimulating and thought-provoking conversations. I thank Lynn Davidmann of the University of Kansas for teaching me the importance of asking who-questions, which she introduced to me from the point of view of a theoretically open and rigorously disciplined sociologist of modern religion. I thank Ilya Dvorkin of the Hebrew University of Jerusalem and all participants of his seminar on Jewish philosophy there, in which I regularly participated online and for which I presented parts of this book project online in 2007–9 and in person in 2010. I am thankful to David Bates, Richard Cohen, Kenneth Dauber, James Diamond, Paul Franks, Rodolphe Gasché, Robert Gibbs, Martin Jay, David Katzman, Cheryl Lester, Hindy Najman, Benjamin Pollock, Dina Stein, Ewa Plonowska Ziarek, Krzysztof Ziarek, and Azzan Yadin for conversations, readings, feedback, and other forms of intellectual support while I worked on different stages of writing this book. I also thank Annette Aronowicz, David Metzger, and Devora Schoenfeld, and my new colleagues at the University at Buffalo, both at the Institute of Jewish Thought and Heritage and in the Department of Comparative Literature, for stimulating conversations that have helped shape various aspects of my argument. I thank Indiana University Press, Gorgias Press, and Brill for graciously permitting me to adapt papers or parts of papers I published with them for arguments in Chapters 3, 4, 6, and 9 of this book. I thank Julian Park Publication Fund of the University at Buffalo SUNY for a generous grant supporting preparation of this book for publication. I dedicate this book to my wife, Lilia Dolgopolskaia, to express my infinite gratitude for the precious gift of life companionship she gave and continues to give to me. Last but not least, I most deeply thank our daughters, Polina and Elen-Sarrah, for their being there for and with us on this journey.
index
abstraction of the thinking subject, 142–6 act of thinking versus faculty, 141–2, 332–3 action, 311; de Man and, 311; subjectity, 140; thought process and, 140–1 addressee of the Talmud, 214–17 The Adventures of Tom Sawyer (Twain), 60–1; example strategy, 317–18 agency of the Talmud: different views, 47; historical approaches, 46–7; refutations and, 159; virtual, 57–8 Albeck, Chanoch, 143 alphabetic writing, 22 amor’aim, 62; dicta, modifications, 170–1; oral argument, 240–6; Rav, 63–4; Rav Yehudah, 63–4; soliloquies, 153–6 anachronism, 135 anonymous redactors, 99, 157; soliloquies, 154–6 antinomies, 61; recall of dicta, 81; text- critical approaches, 79 apocrypha, 321 apology, 136–8 Apology of Socrates (Plato), 136 aporias, 76–9, 82–7 Aquinas, Thomas: Summa, 136 attacking-defending: memory and, 84 Augustine, 162–6 Austin, J. L., 105, 107–9 author. See The Author: virtual, 57–8 authorship: versus authority, 75; Irenaeus, 74–5; Very, Jones, 75
baal hasugya, 172 Baba Metzi’a, 111–16 Babylonian Talmud, x; Aramaic speakers, 114–16; attacks, word choice, 114–15; Baba Metzi’a, 111–12; correlation theory, 117; critical scholarship, 166; interlinguistic clues, 115–16; Nezikin, 111; Rav Ashe, 143–5; sense-making procedures, 117–28; shadow speakers, 122–5; truth criterion, defense successes, 116; virtual argumentation, 202–5; virtuality of rabbi’s being, 200 Babylonian Talmud bBM76ab, 258–65 Babylonian Talmud bBM76b, 300–6 being: exegesis and, 197–200; forgetting and, 27–8; Heidegger, 312; language and, 198–9; nonbeing, seeming to be, 182; versus nonbeing, Plato, 189–93; ontological tripartition, 182 bracketing, 137 Canpanton, Izh.ak, 36; refutations, 330; Talmud as philosophy, 208–9; The Ways of Talmud, 219 Cartesian thinking, 212–13 casuistry: modern scholarship and, 52 characters, 10, 11; The Author as, 71, 90–1; as constructs, 89–90; film montage, 220; human existence and, 99–100; titleless, 64
371
372 Index Christian chronology: Jewish genealogy and, 214 Christian philosophical thought: Jewish texts and thought, 142–3 chronology, 316; Halivni and, 235; versus semiology, 226 Clemens, Samuel, 61; Dasein, 90 cogito, ergo sum, 41–2 cogito sum: Talmudic thinking and, 53 cognition: Talmud and, 6 composers of the Talmud, 100 composite versus composition, 321–2 conceptual anachronism, 135 confession in the Middle Ages. See eximologesis Confessions (Augustine), 162–6 consumption: Heidegger, 22 conversations: illusion of, 171–6; recording, 63 correlation: Babylonian Talmud, 117; dictum with what it refutes, 111; normative texts, 127; truth and, 107–9, 329 correspondence theory of truth, 107–10, 329; normative texts, 127 counterrefutations: refutations and, 146–56 criterion: Mishnah’s instructions and, 108; truth, defense successes, 116 critical approach to text, 159–66 critical conceptual anachronism, 135 critical thinking: occlusion of being, 25–6 cross-cutting in films, 220–1 cultural environment aporia, 85–7 cultural-literary approach. See Rubenstein, Jeffrey: Harry Potter, 85–6 Dasein, 26, 28–9, 326; Clemens, Samuel, 90; Levinas’s alternative, 31; multiplicity, 92; The Author (Heidegger), 89–91 data for thinking and remembering, 94–5 dating of text, 37 de Libera, Alain, 136; philosophical archaeology, 138–42; subjectity, 140–1; thinking subject and, 138–9 de Man, Paul: action and, 311
deconstruction, 137 defenses: memory and, 84, 161; rhetorical, 63; traditional view of the Talmud, 167–70; The Author and, 161; truth criterion and, 116 Deleuze, Gilles: logic of expression, 80–1; human existence and, 101 Descartes, René: Halivni’s thinking, 50–3; thinking, 42 descriptive: definition, 310 descriptive-theoretical approach, 77 designers of the Talmud, 100; baal hasugya, 172 dicta: amor’aim, modifications, 170–1; recall, 81; refutations as purpose, 111 double forgetting, 25; character definition, 26 Ego cogito ergo sum, 150 Eisenstein, Sergei, 220–1, 227–9 emendation versus interpretation aporia, 76, 83–5 environment. See cultural-literary approach: redactors and, 85 environment of emergence of the Talmud, 76 Epstein, Jacob N., 83–4; reductive understanding of logos, 160–1 example strategies, 317–18, 327–8 exegesis: being and, 197–200; Midrash and, 193–4; Philo of Alexandria, 195; Talmud and, 193–4; as the Talmud’s task, 198–9 exegetical reading, 197–200; philosophical tradition and, 198 eximologesis, 139 existentials, 12–13 expression: humans and literary creations, 102–4; logic of, 80–1; theory of, 103 facts: truth and, 107–8 faculty of thinking versus act of thinking, 141–2, 332–3
filmmaking: cross-cutting, 220–1; Eisenstein, Sergei, 227–9; Kuleshov effect, 220–1; montage, 219–20, 227–9; stamm’aim theory, 222–4; timelines, 219–21 forgetting, 23–8 formalist versus realist, 321 Foucault, Michel: archaeology, 138–9; confession in the Middle Ages, 139; technique of the self, 138 Friedman, Shamma: Babylonian Talmud bBM76ab, 258–65; Babylonian Talmud bBM76b, 300–6; composers, 100; designers, 100; illusion, 171–4; literary- formal approach, 45; M (Master) of discourse, 47–8; Master of the Sugya, 277–8; Master of the Sugya (baal hasugya), 47; migrations of text, 69, 279–82; philological versus literary analysis, 213; Talmud criticism, 45; The Author, 66–9; thinking subject, 3 functions of text, 69 gaze, 177, 310–11; performatives and, 163; photographer’s, 10–12 gender, 340 ground, 174–5 haggadah, 36 halakhah, 36 Halivni, David W.: anonymity of redactors, 145–6, 157; anonymous redactors’ soliloquies, 154–6; Babylonian Talmud bBM76ab, 258–65; Babylonian Talmud bBM76b, 300–6; Cartesian thinking, 50–3; conjecture, 219, 226–7; forced interpretations, 67–8; hupokeimenon, 150; in/of aporia, 82–3; literary design, artificial features, 48–9; literary interpretation, 213–14; literary-realist approach, 45; Rabbi Manasseh and, 152–3; redaction, 170–2; redactors, thinking self of, 143; redactors as
Index 373 historical group(s), 51; redactors as subjects, 151; redactors’ disagreement, 88–9; redactors’ role, 82–3; science of the Talmud, 224–5; soliloquies, 150–1; “source” versus the “of,” 276–7; Sources and Traditions, 231, 265–73; stamm’aim, arguments literary analysis, 230–6; stam-of, 273–5; subiectum, 150–1; T (Thinker), 49–51; Talmud criticism, 45; The Author, 66–7; theoretical- rhetorical thinking, 53; thinking subject, 3 Harry Potter: cultural-literary approach and, 85–6 Hegel, Josef: The Author as Spirit, 89; as thinker, versus author, 75 Heidegger, Martin: The Author, 89; Being, 312; consumption, 22; Dasein, 26, 28–9, 89–91, 326; essence of technology, 22, 314; forgetting, double, 25; forgetting, memory and, 28; Jews and, 28–30, 311–13; Phaedrus, 25; Phaedrus (Plato), 23–4; the political and, 312–13; theory of being, 25 heresy versus orthodoxy, 194–7 hermeneutical method, 13–14, 134–5; mutual hermeneutics of philosophy and the Talmud, 188–90 heterogeneity: multiplicity and, 92–3, 326–7 historical reality of Talmud production, 45, 86 historical thinking versus modern, 2, 34 historical-empirical account of the Talmud’s genesis, 47 history: interpretation and, 218–22 History of Ancient Esthetics: Early Classics (Losev), 308 holy text: falsity and, 198 human existence in the Talmud, 87, 98–104 hupokeimenon, 150 Hutton, Patrick H., 95–6
374 Index The Igeret of Rav Sherira Gaon, 214–15 illusion of a record of conversations, 171–6 in/of aporia, 76; Halivni, 82–3 Institutiones Oratoriae (Quintilian), 95 intellect, 2; Plotinus, 79–80 intellectual discipline: Talmud as, 6–7, 219 interdisciplinary argument, 14–15 interpretations: chronology-oriented, 235; versus emendation, 76, 83–5; forced, 67–8; history and, 218–22 intersecting times, 289–92 invention: inventing and remembering, 64–5; as refutation, 111 Irenaeus, 323; heresy versus orthodoxy, 194–7; refutation, 199–200; thinking and personal authorship, 74–5 Jewish genealogy: Christian chronology and, 214 Jewish text: Christian philosophical thought and, 142–3; pagan philosophy and, 142–3; scholarship, 334–5 Jews: Heidegger and, 28–30, 311–3 Karaites, 215–16 Kuleshov effect in filmmaking, 220–1 language: being and nonbeing, 198–9; virtuality and, 197–200 learning: teaching and, 336 legal debate: thinking subject and, 3 Levinas, Emmanuel, 314; ethical responsibility, 31–2; professional Talmudists, 31; study of Talmud, 30–1 Libera, Alain de. See de Libera, Alain literary creations: humans and, 102–4 literary design of the Talmud: artificial features, 48–9; M (Master) of discourse, 47; T (Thinker), 49–51 literary text of the Talmud, 45 literary-formal approach, 321. See also Friedman, Shamma
literary-philosophical approach, 45 literary-realist approach, 321. See also Halivni, David W. logic of expression: Deleuze, Gilles, 80–1; human existence and, 101 logos: reductive understanding, 160–1 Losev, Alexey: History of Ancient Esthetics: Early Classics, 308 Lutsatto, Mosheh Chayim: The Ways of Reason, 143, 218 M (Master) of discourse, 47–8 Maimonides, 36; description of the Talmud, 167–70; Guide, 136; Mishne Torah, 136; Mishneh Torah, rationality, 3; Oral Law, 159; philosophy, 206–8; Rav Ashe, 143–5; refutations, 144–5; thinking subject, 145; Who thinks in the Talmud?, 143 Markosian, Ned, 329 Master of the Sugya (baal hasugya), 47 Master of the Sugya (Friedman), 277–8 medieval Talmudists, 105–6, 110–11 meminí, ergo sum, 41–2 memorizing: forgetting and, 23–4; versus thinking, 23–4 memory: being-there, 175–6; cognition and, 6; data and, 94–5; embodiment and, 178; forgetting and, Heidegger, 28; human existence and, 101; Hutton, Patrick H., 95–6; mnemonics, 96–7, 165; modern thinking and, 34; protocols, 94; Quintilian, 95–7; reasoning as vehicle, 165–6; refutations and, 97–8; remembering versus the remembered, 94; rhetoric and, 5–6; the Talmud, 97; the Talmud as exercises of, 61; techniques, refutation and defense, 161; thinking, disconnection and, 87–8; thinking and, 5; thinking in ancient texts, 88–9; thinking in service to, 88 memory form: the Talmud as, 44 Menachot, 75
metaphysical premises: science of the Talmud and, 223–4; Who remembers?, 282–92 methodologies of Talmudic study, 106 migrations of text, 69, 279–82 Mishnah, 62; criterion for instruction, 108; decision-making source, 137; litigation cases, 151–2; pertinence criterion for truth, 108–9; references and, 108, 126; Talmudic discussions as apology, 136; The Author and, 161; titleless characters and, 64; truths, types and, 107; types, 112–14, 127 mnemonics, 5–6, 96–7; ancient rhetoricians and, 161; Augustine and, 162–6; distrust, 165; Talmudic practices, 165; tann’aim, 165; two parts in the Talmud, 166; verfication and, 165–6 modern thinking: versus historical, 2, 34; memory and, 34 multiplicity: heterogeneity and, 92–3, 326–7 mutual hermeneutics, 188–90 New Talmud, 278 Nezikin, 111 nonbeing: being, seeming to be, 182 normative texts: truth theories, 127 nous, 308–9 occluding one’s being: technology and, 25–6 Old Talmud, 278 ontological analysis of stamm’aim arguments, 238–40 ontological tripartition (being, nonbeing, seeming to be), 182 ontological virtual, 183 open past: versus open future, x–xi; versus virtual, ix open protocol of making sense, 125–6 oral argument: amor’aim and, 240–6 Oral Law, 159
Index 375 pagan philosophy: Jewish texts and thought, 142–3; Trinitarian theology and, 139–40 Palestinian Talmud, 200–2; virtuality of rabbi’s being, 200 performative: definition, 163; gaze and, 163; thinking, the Talmud, 26; virtual as, x performative actions, 12; I will always love you, 37–9 performative existence: gaze, 11–12; Talmudic sages, 10 performative utterance: the Talmud as, 38–9 persona, 142 persuasion: rhetoric and, 5–6 pertinence criterion for truth, 107–9; medieval methodologies and, 110–11; references for Mishnah theme, 126 Phaedrus (Plato), 23–4; double forgetting, 25; Heidegger, 25 Philo of Alexander, 193–4; exegesis, 195, 198 philodoxy, 192–3; virtual arguments, 193 philosophical archaeology, 138–42 philosophical dialogue: memory and, 5 philosophy: Canpanton, 208–9; exegetical tradition and, 198; Maimonides, 206–8; pagan, Trinitarian theology and, 139– 40; rhetoric and, 6; versus sophistry, 186–7; versus Talmud, 9; Talmud and, 5, 9; the Talmud and, 188; truth and, 8; virtual arguments, 193; virtual personalities, 209–10; virtuality in, 187–90 Philosophy and Law: Essays toward the Understanding of Maimonides and His Predecessors (Strauss), 309 philosophy/theology union: erasure of protocols of thinking, 135–6 photographer’s gaze, 10; as performative existence, 11–12 Plato: Apology of Socrates, 136; being versus nonbeing, 187–93, 198–9; Philo of Alexander, 193–4; philodoxy, 192–3;
376 Index Plato (continued) The Sophist, 187–90; sophists, virtual arguments, 191; virtuality in, 189–97 Platonism: thinking and memory, 5 Plotinus: thinking and intellect, 79–80 polemics: redactors and, 88–9 political-ontological analysis: stamm’aim arguments, 238–40 polyphony, 326–7 positive approach to text, 160 praesens historicum, 182 precritical Talmud, 170–8 Proclus, 79–80 production of texts: as representations of content, 70; The Author as producer, 71 questions and soliloquies from author, 247–56 Quintilian: memory, 95–7; rhetorical exercises, 95 Rabbi Manasseh, 153; Halivni and, 152–3 rabbis: authorship versus authority, 75; G-d and, 75–6; Rav Yehudah, 63–4 rational thinking: body and, 52 rationality: Maimonides, 3 Rav, 63–4 Rav Ashe, 143; Babylonian Talmud and, 143–5 Rav Sherira, 214–16 Rav Yehudah, 60, 63–4, 318–19 Rav Yosef, 320 realist versus formalist, 321 reasoning: as vehicle of memory, 165–6 recall of dicta, 81 recorded text, 309 recording of conversations, 62–3 redaction: conjecture, 219; existence, 83; forced, 149–56; Halivni, 170–2; stamm’aim (redactors/designers), 51; the Talmud, 35–6 redactors: anonymous, 99, 145–6, 157; disagreements, 88–9; dissociating
thinking from Author, 78–9; environment and, 85; existence, detection, 82–3; goal of, 67; as historical group(s), 51; literary-realists and, 66–7; polemics and, 88–9; role, Halivni, 82–3; soliloquies, 154–6; as subjects, 151; sugya and, 67; thinking self of, 143 refutations: Canpanton, 330; counterrefutations and, 146–56; dispensability, 159; genesis, 175; human agency and, 159; invention as, 111; Irenaeus and, 199– 200; Maimonides, 144–5; memory and, 97–8; memory and, techniques, 161; Palestinian Talmud, 202–3; purpose of dictum, 111; remembering and, 64–5, 93–4; rhetorical, 63; soliloquies, 147– 52; synchronous, illusion of, 172–6; tann’a, 146–8; The Author and, 161; traditional view of the Talmud, 167–70; truth criterion and, 116 reinvention: remembering and, 64–5 remembering: attacking-defending and, 84; defenses and, 84; human existence and, 101; inventing, 64–5; meminí, ergo sum, 41–2; performative action, 12; Quintilian, 95–7; versus recalling, 87–98; refutation and, 64–5, 93–4; reinvention and, 64–5; versus the remembered, 94; sugya, 65–6; the Talmud and, 43; thinking and, 5; verification and, 64–5; the virtual and, 184. See also Who remembers? rheme: Mishnah, 112 rhetoric: defenses, 63; versus logic/dialectics, 225–6; medieval Talmudists, 105– 6; memory and, 5–6; mnemonics and, 161; persuasion and, 5–6; philosophy and, 6; Quintilian, 95–7; refutations, 63; soliloquies, 147–8; Talmud and, 5, 9 rhetorical character: soliloquist, 156 rhetorical functions of texts, 69 Romantics: authorial intent, 74 Rubenstein, Jeffrey: cultural-literary approach, 53, 85–7; stamm’aim, 53
rupture in continuity of Talmudic practices, 5 Russell, Bertrand: truth theory, 105; matter of facts, 107 sage and thinking process, 142–3, 333–4 scholarship on Jewish texts, 334–5 science, 222–3 science of the Talmud: axioms of signification and of time, 225–6; Halivni and, 224–5; logic/dialectics versus rhetoric, 225–6; metaphysical premises of the Talmud and, 223–4; stamm’aim and, 226–7 self: birth of, 138–42; thoughts of, 139–40 semiology: versus chronology, 226 sense-making procedures: Babylonian Talmud, 117–28 shadow speakers: Babylonian Talmud, 122–5 Socrates: apology and, 136; Phaedrus (Plato), 23–4; technology, 22 soliloquies: amor’aim, 153–5; anonymous redactors, 154–6; Halivni, 150–1; polemical, 62; refutations, 147–52; rhetoric, 147–8; rhetorical character, 156; stamm’a de gemarr’a, 153 The Sophist (Plato), 187–90 sophistry: versus philosophy, 186–7; virtual arguments, 193 sophists: virtual speech, 192 “source”: versus the “of,” 276–7 Sources and Traditions (Halivni), 231, 265–73 speech: falsity and, 198 Spirit: The Author as, Josef Hegel, 89 stam, 265–73 stamm’aim (redactors/designers), 51, 265– 73; arguments, literary analysis, 230–6; arguments, political-ontological analysis, 238–40; cultural-literary approach, 53; filmmaking and, 222–4; metaphysical premises and, 223–4; science of the Talmud and, 226–7
Index 377 stamm’aim de gemarr’a: soliloquies, 153 stam-of, 273–5 stam-talmud, 285–6 storytelling: thinking subject and, 3 strategy examples, 317–18, 327–8 Strauss, Leo: Philosophy and Law: Essays toward the Understanding of Maimonides and His Predecessors, 309 students, 64 subiectum, 150–1 subject: virtual, 131–3 subjectity: Alain de Libera, 140–1 sugya, 63, 320; Master of the Sugya (Friedman), 277–8; Rav, 63–4; Rav Yehudah, 63–4; redactors and, 67; remembering in, 65–6; rhetorical structure, 63–4 Summa (Aquinas), 136 synchronous refuting illusion, 172–6 synchrony versus critical genesis, 265–73 T (Thinker), 49–51 Talmud: eternity and, 8–9; as intellectual discipline, 6–7, 219; as intellectual project, 6; versus philosophy, 9; philosophy and, 5, 9; rhetoric and, 5, 9; versus “the” Talmud, 4–5; truth, recalling, 7–8; truth and, 8–9. See also Talmud, the talmud: meaning in Aramaic texts, 344 Talmud, the: addressee, 214–17; agency of, 46–7; amor’aim, 62; Baba Metzi’a, 111–12; characters, 10, 11; closed archive, 136; composer, 143; critical approach to text, 159–60; dating of, 37; exegesis, philodoxy, 194; exegetical task, 198–9; exercises of memory, 61; explicit engagement, 46; filmic possibility, 230; genesis, 3, 106, 329; historical approaches, agency and, 46–7; historical reality, 45, 86; historical-empirical account, 47; human existence in, 98– 104; illusion of record of conversations, 171–6; implicit application, 46; Levinas, Emmanuel, 30–1; literary design, 47– 51; as literary representation of
378 Index Talmud, the (continued) reality, 70; as literary text, 45; Maimonides’ description, 167–70; Mark Twain of, 65–6; memory, 97; as memory form, 44; Menachot, 75; Mishnah; mnemonics, two parts in, 166; name de-essentialization, 62; New Talmud, 278; Nezikin, 111; as object of critical analysis, 3; Old Talmud, 278; Oral Law, 159; as performance of intellectual activity, 44; performative, 26; as performative utterance, 38–9; philosophy and, 188; positive approach to text, 160; precritical, 170–8; protocols of thinking, erasure, 135–6; Rav Papp’a, 149; Rav Simi, 149; redaction theory, 35–6; redactors, 42; remembering, 43; speakers, identified, 62; speaking beings, 42; versus Talmud, 4–5; thinking, 26, 135; thinking disciplines during, 3; as thought form, 44; traditional view, 167–70; virtual arguments, 193; virtual personalities, 209–10; virtuality and, 189–97; Who performs Talmud?, 36; Who redacted the Talmud?, 35; Who remembers in the Talmud?, 36. See also Babylonian Talmud; Palestinian Talmud Talmud criticism, 44, 315; Friedman, 45; Halivni, 45; the Talmud as literary text, 45; virtual agency and, 58 Talmudic discussions: as apology for Mishnah, 136; translations, 112 Talmudic lines of time, 221–2 Talmudic practices: mnemonics, 165; rupture in continuity, 5; thinking, memory and, 34 Talmudic process: persons in, 4; weblike, 4 Talmudic sages: performative existence, 10 Talmudic study: methodologies, 106 Talmudic thinking: cogito sum and, 53 tann’a, 111–12; refutations and, 146–8 tann’aim: mnemonics and, 165 teachers, 64 teaching: learning and, 336
technique of the self, 138 technology: alphabetic writing, 22; being, occluding, 25–6; Heidegger, 22; Socrates, 22; thinking and, 21–2 text-critical discussions, 319–20; antinomies, 79; comparison chart, 294–6; dating of text, 37; The Author, 70–1; who composed the text, 161 texts: migrations of, 69, 279–82; production, as representations of content, 70 textual analysis. See Talmud criticism textual traditions, 310 The Author, 322–3, 325; ambiguity, 71–2; characters as constructs, 89–90; defenses and, 161; Heidegger, 89; as historical producer, 71; homogeneity of, 89; in/ of aporias, 76; intelligent designer, 69; as literary character, 71, 90–1; literary- formalist approach, 66, 68–9; literary- realist approach, 66–7; Mark Twain, 60; Mishnah and, 161; multiplicity, 90–2; refutations and, 161; Romantics and authorial intent, 74; speaking versus thinking, 72–4; as Spirit (Hegel), 89; the Talmud, 61; text-critical approaches, 66, 70–1; thinking, dissociating from redactors, 78–9; thinking, homogeneous, 93; thinking, role of, 73–4; thinking and, 70–1; Yehuda, Rav, 60 theme of Mishnah: as type, 127 theology/philosophy union: erasure of protocols of thinking, 135–6 theoretical-rhetorical thinking: Halivni and, 53 theory of being: Heidegger, 25 theory of expression, 103. See also logic of expression thinking: agent, 46; cogito, ergo sum, 41–2; complex structure, 80; as a dance, 93–4; data and, 94–5; Descartes, Rene, 42; disconnect from thinking subject, 104; dissociating Author from redactors, 78– 9; double activity, 80; as faculty versus act, 141–2, 332–3; forgetting and, 28;
goals, 27; Hegel, Josef, as thinker versus author, 75; as independent faculty, 88; Irenaeus, 74–5; memorizing and, 23–4; memory and, 5, 87–9; memory and, Talmudic practices, 34; (mis)understanding, 135; mnemonics, 5–6; modern versus historical, 2; open-ended style, 136; participation in, 2; performative action, 12; Plotinus, 79–80; post-Cartesian, 212– 13; rationality, 3; remembering and, 5; the Talmud, protocol erasure, 135–6; Talmudic practices, memory and, 34 The Author and, 70–1, 73–4, 93. See also critical thinking; modern thinking; rational thinking; Who thinks? thinking Author, 72–4 thinking process: action and, 140–1; sage and, 142–3, 333–4; versus thinking subject, 158 thinking subject, 2; abstraction, 142–6; Albeck, Chanoch, 143; ancient practices, 79–80; composition embodies the composer, 143; de Libera, Alain, 138–9; disconnect from thinking, 104; establishment of idea, 135; Friedman, 3; Halivni, David W., 3; legal debate and, 3; Lutsatto, Mosheh Chayim, 143; Maimonides, 145; redactors, 143; redactors’ anonymity, 145–6; storytelling and, 3; subjectity, 141; Talmud’s genesis, 3; versus thinking process, 158 thought form: the Talmud as, 44 thought-forms: the virtual, 182–3 time: axiom of, 225–6 timelines, 339–40; filmmaking, 219–21; Talmudic, 221–2 times, 282–92; intersecting, 289–92 traditional view of the Talmud, 167–70 translation of Talmudic discussions, 112 truth: Babylonian Talmud, 117; correlation, 107–9, 329; correspondence, 107–10, 127, 329; criterion, 107–9, 116; normative texts, 127; philosophy and, 8; recalling, through Talmud,
Index 379 7–8; references, 105–6; sense-making procedures, Babylonian Talmud, 117– 28; Talmud and, 8–9; types, Babylonian Talmud, 112–26 Twain, Mark: relationship to Samuel Clemens, 72; of the Talmud, 65–6. See also The Author type (truth): Babylonian Talmud, 112–26; Mishnah, 107, 109; theme of Mishnah as, 127; truth and, 107–8 undoing, 137–8 united plurality, 325 verification: mnemonics and, 165–6; remembering and, 64–5, 84 Very, Jones, 75 virtual, 324–5; ancient forms, ix–x; conceptualization, ix; historical examples, 1; older forms, 185; versus open past, ix; as performative, x the virtual, 181–4; ontological, 183; remembering and, 184; Talmudic thought-form, 182–3 virtual arguments, 186; Babylonian Talmud, 202–5; philodoxy, 193; philosophy, 193; sophistry, 193 virtual author, 57–8 virtual speech: sophists, 192 virtual subject, 131–3 virtuality: language and, 197–200; in philosophy, 187–90; Plato, 189–97; the Talmud, 189–97 The Ways of Reason (Lutsatto), 143, 218 The Ways of Talmud (Canpanton), 219 Who redacted the Talmud?, 35 Who remembers?, 14, 16–17, 36, 65–6 Who speaks?, 14–15; dating of text, 37 Who thinks?, 14–16, 131–3 Who thinks in the Talmud?: Maimonides and, 143 Wissenschaft des Judentums, 218, 222–3