סיג לחכמה, שׁתיקה
(m. Avot 3.13)

Rabbi said: One should never have a multitude of friends in his house, for it is written, A man that hath many friends brings evil upon himself.
(b. Sanhedrin 100b)

Aristeas and the Greek library

1In Aristeas’s epistle to his brother Philocrates, he recounts the story of the acquisition of the Torah for the library of Alexandria. Demetrius of Phaleron’s desire for the sacred Jewish books is told both as part of a comprehensive library project (he wants all the books in the world, §9), and a project of the curation of wisdom (the books of the Jews are deemed "worthy1 of transcription and inclusion" [§10; καὶ τῶν Ἰουδαίων νόμιμα μεταγραφῆς ἄξια καὶ τῆς παρὰ σοὶ βιβλιοθήκης εἶναι]). The Ptolemaic library project is manifestly one of nation building, meant to translocate the young kingdom from eastern periphery to the cultural and political center of the Mediterranean. That the project was successful, at least in some quarters, is evidenced by Aristeas’s letter.

2This Jewish tale claims its fealty to this Greek national center not only through its contents, but also by mimetic absorption of the language of Ptolemaic literary culture. Its longing for inclusion is felt throughout the work – a Greek literary aspiration visible in the epistolary frames, the tour of the land of Judea that transforms the sacrificing priests in Jerusalem into a spectacle of muscled athletes tearing oxen in half by hand (§92-95), and its treatment of the holy city as a master class in hydrological engineering (§88-91), in the elaborate ekphrasis of the gifts sent to the Temple by Ptolemy (II Philadelphus; §51-82), the translators’ text critical translation seminars (§301-307), and the fantastically tedious symposium cum peri-basileus2 in which the King recognizes the deep Jewish (read entirely anodyne Greek) wisdom of the Jews (§187-300), one finds a Jew eager to display his Hellenistic learning. The text’s Jewish author has clearly spent time reading in the Greek library, writ large.

3The king purchases a Torah with the release of Jewish slaves, and great gifts of gold. In return the Jews send their Torah and teachers to translate and interpret it. One purpose of the text is to reveal the worth of Jewish wisdom to the Greek king, and this is accomplished through its translation of Judaism into Greek terms, genres, and concepts, thus making it legible to the imagined Greek reader. Moreover, as Sylvie Honigman has argued, the text seems to invest the Greek king with the power to authorize the text itself. Indeed, his commitment to the text, both as attentive and approving audience/consumer, and as actual investor in its purchase and perfection, gives to the Greek Septuagint what Sinai gives to the Hebrew – a sacrality that is fixed in its very letters. Her argument makes abundantly and dramatically clear the impact of imperial reading on the Jewish book.3

4The narrative’s ceremonial potlatch articulates national pride through a profound desire to have Torah catalogued in the Alexandrian library. However, The Letter of Aristeas is a work of cultural translation that does not, and cannot, imagine an analogous Jewish library.4 The manifest narrative is one of Greek thirst for Jewish wisdom and texts, but the tactic narrative tells the opposite story – of Jewish desire for and absorption of Greek wisdom and texts. Here the dynamics swerve off course in what I suggest are imperially determined ways. If we imagine the author as well-read Jew – then where do we imagine he has done his reading? In the figurative Greek library. As such the dynamics here are oddly non-reciprocal. The fantasy is that the Greek library wants the Torah, so too the Greek king desires Jewish knowledge. Since, the thinking seems to go, Torah both embodies and perfects Greek wisdom, the Jews have no need for Greek books, a logic materialized in the cultural exchange of Torah for gold. However one need not step too far beyond the text’s propaganda to discern that the opposite is more accurate. It is because we inhabit the text’s message that scholars have not asked why the Letter of Aristeas does not imagine that Greek books might be found in or desired by the archives of Jerusalem.5 Why, in other words, aren’t the cultures exchanging books? Just as in the banquet scene, where Ptolemy draws Jewish philosophy from the translators but they evidence no curiosity about him, it is nowhere imagined that the Jews might request a copy of Homer in order to have it translated it into "their own characters" (cf. §11).

5The Bible and her attendant Jews here function as sources of wisdom. The seeker after wisdom is someone else. (Note that the curious Aristeas masks himself as a Greek). While this seems to set the Jew higher in the pecking order – teacher is after all superior to student – the trope of the library itself turns the tables. The collector of wisdom is the colonist, and the Jew, no matter his magnificence, a butterfly held by a pin.

Josephus and elite Roman readers

6One empire later, Josephus embarks on his literary career. One finds occasion to think again about Jewish book production in relation to imperial libraries.

7Embedded in the Josephan corpus we find many of the markers of ancient ethnographical traditions. His work might be loosely described as a sort of apologetic autoethnography.6 In his book on the topic of ancient ethnography, Greg Woolf includes Josephus as an aside, among a list of home grown autoethnographers – among them Trogus, Berossus, Juba II, and others7 – whose Hellenization gave them the tools of participation in the dominant literary culture. The minimum of such inclusion, Woolf avers, is an awareness of the conventions of ethnographic writing and knowledge derived from classical sources. These two things Josephus has in spades; he is clearly aware of these conventions. Given the Torah’s own emphasis on genealogy, its broad scope of universal history, focus on nomoi, and toponyms and to a certain extent geography, Josephus’s object seems ready made to blend with Hellenistic ethnographic tropes, and indeed Josephus reshapes them into a Roman idiom in many places.8

8But reading Josephus’s self-study, the disanalogies with certain elements of ethnographic convention are striking – and to illuminate the divergence I would like, heuristically, to contrast Josephus to another ethnographer, writing from a different social location: Tacitus, mutatis mutandis. Think of the colorful barbarians described by Tacitus – the honorable Britons, savagely handsome Germans, oversexed Jews. Through these depictions, he populates the Roman boundaries with barbarians that do a great deal of work for him. Josephus does not take the opportunity his narratives offer to do the same. Take, for example, the primordial history from Genesis in the first book/s of his Antiquities, where Josephus recounts Jewish history from the dawn of time. Paraphrase is a powerful tool, and Josephus, despite his respect for the Bible’s authority, shows no fear in reworking or interrupting it to tell his own story. Given the ambitious contours of what seems a universal history form9 (in Genesis, after all, God creates, subdues and peoples the cosmos, then creates the architecture of unfolding human history10), the opening chapters of the Antiquities are oddly provincial. Josephus communicates no sense of ownership over the peoples his own God has spawned, and is not tempted by the easy prompts offered by primordial enemies or kin to inhabit a conqueror’s stance in relationship to traditional rivals such as Egyptians, Canaanites, or Assyrians, even when they are tucked safely in the mythic past.11

9Moreover, he has no real taste for the noble savage or the irreducible barbarian. Josephus has ample opportunity for paradoxography, and I am not looking for a long essay – a lot can be done with an aside (i.e., "The Hivvites, who have feet facing backwards and eat a foul smelling porridge thrice daily,…") – it does not require straying far from the biblical frame, certainly no farther than he regularly does to serve other agendas.12

10For Tacitus, additionally, ethnography is a mode of self-critique; "the fringe", writes Woolf, "as often incriminates as elevates the center".13 A simple dichotomy, good Roman vs bad barbarian hardly describes the sources. Tacitus’s Britons are at times more honorable than the Romans, his Germans are braver. In his turn, Josephus might have populated a Jewish-centered world which gazes, with condescension, disgust or even with admiration, on lesser souls that populate his primordial epic; he does not.14

11While ethnography may not perhaps be reducible to imperial ends, the link between them gives us a view into the ways that certain provincials both could and could not see themselves and others. Indeed Josephus’s finest field work he does among the Jews – exploring the Pharisees, Essenes15, and Sadducees (and there is more to be said about Josephus’s treatment of the Romans in the War.) In this he is a culture broker for Roman eyes, operating in a middle zone, more an interpreter for a Roman anthropologist than one in his own right. In short, Josephus adopts the tools of the imperial ethnographer, but does not adopt a central position for himself and his ethnos – he looks at Jews from the imperial center, that is, but does not make himself the middle of a Jewish universe.16 We neither have nor imagine finding a work by Josephus called "On the Nabateans, " or "On the Syrians. " As with Aristeas, this simple lacuna invites reflection.

12Like the author of Aristeas, Josephus is stung by the lack of mention of the Jews in early Greek sources, and perturbed by the lies preserved in later ones.17 Yet in composing his Antiquities, as in answering Apion – projects that digest many of the same sources that inform Tacitus’s treatment of the Jews in book 5 of the Histories – one has the sense that he seeks less to be a peer of Tacitus, and his ilk, than to be the primary source of record in the Roman library. Like Aristeas, he seeks to "be worthy" in Josephus’s own words "of study for the Greeks" (τοῖς Ἕλλησιν ἀξίαν σπουδῆς).18 As such, his rival is not Tacitus himself, but Manetho. How galling in the end that in his work on the Jewish War, Tacitus ignores Josephus completely.

Rabbis and the imagined Roman reader

13In Aristeas and Josephus, the idea of the "library" – whether a brick and mortal establishment as in Alexandria, or signifying the standard works consulted by an aristocratic Roman historian like Tacitus – stands in for both the temporal and cultural power of empire. Empire, like a librarian, exercises control over selection for inclusion, and sets terms for access to and use of its resources. Decisions concerning what to catalog, copy, read, and cite, manifest the imperial gaze, which would have been felt especially keenly by learned provincial elites. Josephus’s and Aristeas’s literary styles display their classical educations, but the trope of the library allows us to see beyond their mimetic self-styling; they wish not only to be recognized, and treated, as Roman or Greek author, but as a Roman or Greek book – each promoting the Jewish book as worthy of study and a prerequisite to enfranchisement. In other words, vis-à-vis the classical library, each may have a library card, but would prefer to be in the card catalog. Moreover, this desire to be read by the imperial reader, has an impact on how they see and depict themselves and the world around them. The tacit curatorial/ethnographic perspective of empire directs their own focus back upon themselves, and appears even on some level to thwart their ability to look out.

14For the remainder of this essay I turn to a very different Jewish Roman corpus, that of the early rabbis. I’ll posit that the idea of the imperial library, so figuratively understood, is a factor as well in rabbinic literary development, albeit with wildly different results. Roman book culture and the silent but pervasive notion of the imperial reader is a frame that allows us better to see the rabbis as provincial exemplar and not cultural exception or outlier, tying rabbinic literature into the thick web of cultural relationships built around reading and writing that includes both the imperial center and the Jewish authors writing in Greek above.

15If the aim in this volume is to reflect on how Rome was perceived by those under its power, the rabbis are distinctive in their strenuous effort to stay mum on the topic, to avert their eyes. The earliest rabbinic literature was composed at the height of empire. It circulated officially only orally among literate Jews, conservative scholastics of a sort who knew Greek, walked on Roman roads, and moved through cities plastered with Roman decrees.19 Numerous scholars have charted the deep incursion of Greek and Roman culture into rabbinic texts and lives, and the rabbis’ awareness of and reaction to external perceptions. Roman/Hellenistic influence issuing from law and philosophy, literature and rhetoric, can be traced in rabbinic thought.20 Still, it is worth underscoring that the rabbis – perhaps more-determinedly than other provincials – are uneager to help the scholar in search of Roman, or for that matter any, historical context. One struggles to find an identifiable date, an explanatory prologue, an historical emperor or transcribed rescript, a map of a military base, or even a Christian. Typology replaces specificity, names are masks that when removed reveal only a character serving an ad hoc homiletic or halakhic need. That the rabbis don’t do history is a theme about which much as been written, in what way rabbinic literature can be used to write history is a problem that must be addressed in some way at every scholarly encounter.

16The provincial has a voice in how he relates to power. However, the raw materials for his self-fashioning are often provided by that same power, and the empire, like water to the fish, can be difficult to see and therefore to assess. Rabbinic literary culture is markedly discontinuous with Jewish precedent. We should be curious as to why this is; I will argue below that seeing rabbinic choices in relationship to certain aspects of Roman book culture helps explain these innovations. Juxtaposition with larger Roman trends has the potential to reorient standard takes on familiar rabbinic phenomena concerning attribution, composition, authorization, censorship, and citation, and will, I hope, suggest something new about both Romanization and the rabbinic strategic reaction to their location as provincials.

17The rabbis force the scholar to be a close reader of aporiae. I want to focus on a pair of these in particular – two activities that the rabbis seem rather selectively, yet strenuously, to avoid depicting: namely, reading and writing.21 These lacunae can be interpreted as part of a rabbinic strategy to opt out of the intellectual culture surrounding them, representing among other things, a refusal to be consumed by imperial readers. Seen from this angle, the amplification of the written bookishness of Scripture, coupled with the pointed quiescence of other sites of reading and writing, collectively represent a reaction to the same determinate reality as do the works of Josephus and Aristeas. The latters’ desire to be read is met in the rabbis by an equal and opposite desire not to. (It matters little, of course, whether the Romans sought rabbinic books or knowledge, as it matters little whether Demetrius wanted a Torah, or Tacitus ever consulted the Antiquities. It is the desire here that matters.)

18Given this direction of inquiry, it is also important to note that rabbinic deviation from elite practices of inscription, circulation, and consumption of written texts manifests itself on a field not entirely alien to those same practices. Rabbinic literature of the tannaitic period reveals marked tension between an absorption and rejection of elements familiar from Roman book culture. In order to orient my analysis toward that external context, I will focus on the two pillars of rabbinic literary authority: the book and the sage. In each section below, I will show first, how certain rabbinic literary practices and assumptions are discontinuous with Second Temple precedent; then, that these discontinuities are consonant with their pagan intellectual environment; and finally, explore the rabbis’ inversion of expectations and divergence from both Jewish and pagan antecedents in context.

The book (reading)

19In the centuries surrounding the opening of the Common Era – before the rise of the rabbis – a wide range of Jewish writings was in circulation and production in Hebrew, Greek, and Aramaic in different versions and with varying significance for different Jews. Jewish literary production happened in a number of genres, mostly narrative, which ranged from parabiblical and novelistic prose to historical, liturgical, oracular, legal, philosophical, and didactic forms. Though not at the same scale as Roman or Greek collection, finds at Qumran reveal that elites collected and studied books from a range of sources, and produced them as well.22

20In addition, Second Temple Jewish literature (especially, but not solely in Semitic languages) tells us that beyond these extant texts, circulated a virtual library made up of imagined, "lost", and esoteric books.23 These ghost volumes, which are connected to Enoch and David, among others, punctuate Jubilees, underpin apocalypse, and chronicle monarchy. Such real and imagined works, where mentioned, are said to contain special laws, testaments, technical manuals and primers, priestly knowledge, science and cosmology, angelic gnosis, keys to reading history, and more.24 Their abundance proves that Jewish books, qua books, were significant markers of elite identity.25 Beyond transmitting divine communication, books are given the job of bearing vital tradition from one generation to the next, and are depicted as the key medium for authorizing, protecting and transmitting important knowledge of many sorts.26 In short, the book and its cognates, scribes and the act of writing itself, were prominent Jewish tropes in a range of pre-rabbinic sources.

21Although a scribal elite themselves, the rabbis, oddly enough, reject this sprawling library, selecting and sanctioning from it only a small collection of what they deem divinely inspired volumes.27 The first five – the Torah revealed to Moses at Sinai – had special status for the rabbis no less than for other Jews. Another nineteen books make up what we know now as Prophets and Writings. The rest of the library is suppressed. Censorship is the rabbis’ bluntest instrument – books beyond the twenty-four are prohibited or otherwise marginalized. The rabbis explicitly debate a few liminal books’ inclusion in Scripture (Song of Songs, Ecclesiastes, and Ben Sira28), but functionally silence the remainder of the Jewish bookshelf, both real and fictional tomes.29 For the rabbis the books of Scripture emerge in a distinct key, one familiar to us as a coherent textuality – "the Bible". Our familiarity with a fixed and circumscribed scriptural library, however, can blur its novelty. Even as the rabbis discourage the reading of what they deem to be non-Scriptural books, they reify the material book-ishness of Scripture in novel ways, which in scale and centrality mark departures from antecedent Jewish literatures.30 This can be traced through a range of textual phenomena that arise surrounding and defining this canon.31

22The reading practice perhaps most central to rabbinic scriptural textuality is the development and discursive prominence of what we might call citation genres – literature that incorporates quotations from other works to authorize, serve as examples in, or generate commentary. It is a discursive habit innovative in its prominence and stridency (i.e., midrash).32 Using precise quotes, citation formulae, and a version of Hebrew different from biblical Hebrew, the rabbis separate themselves clearly from the scriptural text and so communicate that they are not writing or competing with it.33 (Eventually Gemara will do the same with Mishnah and other tannaitic sources.) Though standard modus operandi for the rabbis – and the Romans – citation formulae and exact quotations are exceedingly rare in prerabbinic Jewish literature, and signal a new sense of the Bible as finite fixed text. In Second Temple writing, one is far more likely to encounter the Bible, or the "biblical" through paraphrase, allusion, pseudepigraphy, plot, or style – not citation. In prerabbinic Hebrew sources, "Bible" passages (whatever that would have meant) where verbatim are marked as such most often only by modern compilers of critical editions.34 Outside of Philo35 (whose embeddedness in Greek textual criticism tends to underscore the link between citation and Greco-Roman genres), lemmatic commentary in Jewish sources are rare, and more so in Hebrew sources, showing up in only a very few places. The first real signs appear at Qumran, itself, not coincidentally, a library-centered community – and still, though this community was avowedly textual and Torah centered, the practice of citation is far from dominant there, adhering to technical reading contexts such as legal and mantic exegesis.36

23The tannaim deliberately deploy citation in their canon making process. In a world of versions, citation serves to fix and stabilize the wording of a fluid biblical base, and marginalize the fringe. By citing, for example, only some of the many psalms in circulation and not others, or by settling on one of the variations of Samuel, citation communicates a privileged version and fixes the form of what may have been more loosely bounded corpuses and traditions.37 Additionally, according to rabbinic rules of engagement, a question about a verse in the Torah is only answerable through linking it to other verses drawn from across the canon. By asserting that the sacred books were a self-glossing universe, the tannaim export the sacrality and cultural heft of the Pentateuch onto texts that were less potent and normative. We can surmise that omitted works such as Ben Sira or the Enoch epic in whatever form, were, for many, more religiously resonant than say, the canonized Joshua or Obadiah, and so we see how this reading practice is polemical as well. While we take as a matter of course that when decoding a passage from Exodus, the rabbis will key its meaning through a florilegium of passages collected from across Prophets and Writings, this constellation of books had nowhere else been deployed this precisely.38 It is clear that citation is canon-making as much as canon using.39

24Thus what is cited (and what is not) reveals the sources counted as most important, but could also exert a determining influence on that list – and this dynamic need not be limited to sacred corpuses. Canon-making is an effect of all citation genres (including modern scholarship), and it is worth underscoring that rabbinic scriptural citation falls in line with its function in parallel classical genres that rely on citations such as grammars, legal compendia, medical texts, and the like, and carries forward into later citation driven compendia, such as Eusebius’s Ecclesiastical History and Justinian’s Digest to name but two. Moreover, a range of forms that build from citation practices is on the ascendant in Roman literary genres in the late republic and early empire (the tannaitic era), just as the library, both public and private becomes a standard part of elite intellectual life.40 In this same period – encyclopedic works miscellanies, and handbook genres make visible their "bibliographic labor", meaning they draw attention to their erudition and mastery of the canon, to the effort of plumbing the great works of the past, as well as to the dynamics of collocation that go into their presentation of knowledge.41 Midrash fits rather comfortably among this loose collection of works and genres that are ostensibly paraliterary, not manifestly committed to literary aesthetics, rely on books, and are "models of active reading".42 What is commonly described as the midrashic atomization of Scripture, shares features with what Formisano sees in a range of later Latin literatures, namely, an idea of knowledge "conceived of as an aggregate of discrete particles that are subjected to a process of analytical ordering".43 The early rabbis likewise present God’s word as coming to its fullness through the proper perception of it as "discrete particles", interpreted by the rabbis through a range of excerpting and compilatory techniques.

25Thus, in respect to the material stability and prominence of God’s word treated as a bounded library, the rabbis are rather bookish in a classical mode44 – even as their particular canon making project (rather unlike those of their pagan neighbors) aims to end or drastically curtail the authority and cultural importance of books. In their self-glossing universe, they acknowledge no textual authority beyond the Bible. Extra-Scriptural tradition or authority is not depicted, preserved, or imagined in book form. However the citation project is a performance with clear didactic and rhetorical ends; it does not mean that the rabbis did not have access to or consult extra-biblical works.45 Indeed they certainly had access to books and read and used them. They discuss the status and permissibility of books of various sorts, canonizing some, outlawing some (m. Sanhedrin 10.1), and permitting others (m. Yadayim 4.6; y. Sanhedrin 10.1; t.Yadayim 2.13).46 They discuss the merits and dangers of written media, of Greek education for their children (m. Sotah 9.14; t. Avodah Zarah 1.20, 461 MS variant sefer yevani, "Greek book"47 [not merely Greek language, yevanit]), they transmit literary motifs from Jewish and gentile sources – echo rhetorical forms (such as chreia48), or allude to philosophical and (ironically) text critical concepts,49 and more. However, whatever the origin of the knowledge possessed and transmitted and transformed by the rabbis in their writings, even when certainly textual, it is not presented as being derived from books.50 We do not see rabbis in libraries, sharing volumes with a neighbor, consulting a tome outside of Scripture. And most significant, the tannaim simply do not cite extra-biblical texts.

26"Outside books " (sefarim ha-ḥitsonim51) is a rabbinic category of opprobrium, but outlawed books are never listed or named, a classic rabbinic move of dismissing the external by denying it articulation and specificity (think, in a similar vein, of the stable of vague stock figures that populate the edges of the rabbinic world from the min and epikoros52 to the philosopher and emperor).53 Moreover, when the rabbis understand themselves to be engaged with their neighbors, friend or foe, their encounters are depicted as conversations. A rabbi speaks to a philosopher, emperor, or hegemon, he does not read him (not insignificantly, these encounters are often with a pagan of some sort who has himself been reading Torah!54). In short, wisdom gleaned from any outside source is detextualized before its inclusion.55

27We know precious little about book circulation and storage for the rabbis. That said nothing comparable to what we find in public and private collections in Rome and throughout the East is recorded for rabbinic circles.56 Since the rabbis so rarely talk about books, we are even further removed from the evidence. Ironic then is that the particular rabbinic repackaging of "Torah" into a clearly demarcated book is part of a set of choices in medium, language, and genre by which the rabbis signal their awareness of and draw on, and yet at the same time remove themselves from, the vibrant book economy of the empire; this in a time of rapid expansion and dissemination of writing across the eastern Mediterranean.57

28Paralleling their near contemporaries in a range of genres (Pliny’s [23-79 CE] Natural History, Gaius’s [130–180] commentaries, Origen’s [184–254] Hexapla), the rabbis encounter a complex book landscape in a self-consciously belated posture. By selecting and explicitly acknowledging only some books and effacing others, they function to guide their ideal readers through an always expanding world of texts, manipulating medium as a way to control what is worth knowing, and what worth preserving. Censorship and citation may in this context be seen as strategies for managing this proliferation58 – setting them on a continuum with others such strategies for controlling knowledge such as epitome, paideia, or Alexandrian critical scholarship. As such rabbinic book-phobia participates in more pervasive Roman phenomena.

29We might imagine that rival books are silenced so that rabbinic books may replace them, but neither is this the case, since the rabbis refuse to write as well.59 They commit to oral circulation and officially, at least, avoid inscription.60 Second Temple literatures commonly depict a holy person reading and writing – acts signaling his authority and connection with the divine. Yet writing is erased from the rabbinic performance of authority, restricting the symbolic and iconic potencies of the book to Scripture alone. There is nothing necessary about this strategy. In case the cessation of writing seems a natural outgrowth of canonical closure, remember that Christians similarly define a corpus of sacred texts which they patrol with a similarly ruthless rigor – but they do so using a deluge of written words. The Christian Bible catalyzes an explosion in supporting genres: polemic, sermon, letter, theological tract, martyrology and more. Origen alone writes 1.3 million words, adding a cool 2 percent to all preserved Greek literature to this time. Christians were, "impresarios of the scriptorium and the library" say Grafton and Williams of the scholars of Caesarea, deploying everything in their religio-cultural arsenal to "impose order on highly varied forms of information".61

30Clearly the Church fathers jump into book culture with both feet, and embrace the pagan library in both form and content as a way to variously exploit the classical inheritance. They make it work toward their own ends as they define, protect, and expound their own Scriptural canon.62 Early Christian writing is full of citation of rival texts and the careful delineation of external threats and heresies.63 Christian prolixity and rabbinic silence are two sides of the same coin. Local bibliophilia evident in centers from Rome, Alexandria, and Caesarea are all well within the rabbinic ambit, making only more stark the rabbis’ choices.

The sage (writing)

31Which brings us to the rabbinic sage. Where the book is authoritative, and citation a primary language, reading practices inevitably yield to atomization and repurposing – transforming the authorized exegete into the new master.64 For authors, Roman and Jewish, operating with a powerful inherited set of books, we find privileged books transformed into the raw material of a lexicon, whose syntax and grammar only the most thoroughly versed reader can decipher.65 It follows then that in tannaitic materials, the only recognized, and citable, authority outside of Scripture is the rabbinic sage himself. In early rabbinic literature only two authorities are worthy of direct verbatim citation: Scripture and the rabbi. "Rabbi X said in the name of Rabbi Y…" "The words of Rabbi Z…". By applying a technical (and arguably inimitable) hermeneutic to this circumscribed corpus, the rabbi orders and controls="true" what counts as knowledge. In the absence of other books, the sage-exegete himself bears valuable knowledge and tradition, augments, clarifies, and propounds the textual wisdom of Scripture.

32This regular use of proper names of living men represents yet another discordant element vis-a-vis the antecedent Hebrew literary imagination. Until the rabbis, and with the exception perhaps of Ben Sira66, the scribal interpreter strove to elude detection. The scribe was an anonymous figure often buried behind a pseudonym, if named at all (the Chronicler and Deuteronomist are but two who managed an inherited tradition anonymously).67 Elsewhere the writer writes pseudepigraphically, in the name of an antediluvian wise man or angel, Baruch or Gabriel, the Teacher of Righteousness or Daniel. Rabbinic literature, by contrast, features living men, named sages, developing their personalities, emphasizing their expertise, and placing them in space and relationship.68 This real scholar (even if a character, he is meant to indicate an actual individual69) masters his corpus and teaches us how to read it; as a walking text he is master and curriculum in one.70 The rabbi, like other literary experts, stands on the boundary between the library and the curriculum – an ideal embodiment of the texts he also curates and teaches.71

33The rabbi fits rather easily into recent scholarship exploring the horizon of the Greek and Roman expert and how he positions himself culturally. Combining legal and exegetical virtuosity with proper paideia and moral authority, the tanna is in his way a familiar figure on the intellectual landscape. His "self-assertive" self-representation places him, in the words of König in his study of scientific experts in antiquity, among the social and moral elite, "typical of a set of wider phenomena within ancient scientific and knowledge ordering culture".72 His role as a character in active debate echoes the sympotic colloquies common to Greek and Roman literatures, though shorn of their rhetorical flourishes. This emergent rabbi persona is an expert claiming an expansive authority to broker and embody the divine word – but it is an authority distinctly tempered by the pious discipline of study, a humble relationship to tradition, a sense of the stakes of conflict, the restrictions and conventions of legal discourse, Scripture, and one’s teacher.

34Like the book in Rome, the rabbi and his adages were in broad circulation. The mobile sage – abetted in part by Roman infrastructure73 – and his embodied wisdom moved, and his words were shared among his cohort. It is clear from the literary remains that his teachings reached colleagues and acolytes, were collected, copied into notebooks, memories and curricula, excerpted, combined, and reissued in a range of new contexts. Precise wording of a sage’s teaching was valued and protected. Moreover the deployment of an absent sage’s dictum, even though dramatized orally, treats him like a book – just as a one might say "Homer and Hesiod agree that…" so too the rabbinic dialogue is a citation genre that does not rely on the presence of the person, rather his remembered, excerpted, or repurposed teaching.74 In other words, what seems to be the literary depiction an oral performance of conversation between rabbis ("Rabbi Akiba says…"), when read alongside citation conventions from Greek and Roman citation habits ("Hesiod says…"), may be productively read as its inverse: namely the rabbinic conversation could as well represent the importation of a book based idiom into what is for the rabbis an actual oral practice.

35No matter the authoritativeness or prominence of the depicted sage-teacher-decisor (or the strength of the editorial hand), rabbinic sages are never authors.75 No rabbinic collection is ascribed to an author, speaks in a first person, has an explanatory prologue or even a title. Traditions of patronage, such as Rabbi Judah the Prince and others over Mishnah versions, do not manifest themselves in explicit claims of authorship or even incidental depictions of writing. And so with this ascendant sage we again abruptly turn face (two elements in tension) – for just as the emergent bookolotry of rabbinic Scripture is shadowed by the withering of book culture, so the new authorizing named scholar personality never becomes himself an author.

Conclusion

36In this essay I have been using the word library in a range of ways, to point to the imperial library as a figure of the dynamic reading practices of an imperial elite and to indicate the rabbis’ own relationship to the book. I have played down internal and theological language connected with the rabbis such as Oral Torah, and tried to desacralize the idea of "canon", so as not to occlude points of comparison, parity, and juxtaposition.

37If we think back to Aristeas and Josephus, with whom I opened this essay, it is clear how their political location as subaltern effected their literary decisions, shaping the language they used, their genres – impacting their depiction of the Jews and determining their authorial perspective. Their desire to be included in the imperial library, and read by imperial eyes is obvious. The rabbis, I have argued, are no less affected by the imagined imperial reader and his library. Their language, genres, media, and content are similarly created in relationship to them.76 Key elements distinguishing them from antecedent Jewish literatures are some of the very things that align them with Roman literary culture: (1) a heightened sense of the authority of the book, and a belated and deferential stance to the book-based knowledge manifested in careful attention to wording, precision, and citation as driving mode of creation/composition; (2) a sense of foreboding in the face of the proliferation of knowledge, and an attempt to manage it; and (3) the emergence of the real person from behind the pseudepigraph. The rabbis elevate and authorize the human scholar-expert, who curates and brokers knowledge, and functions in his person as a sort of walking library. In highlighting these parallels I hope to communicate that rabbis are not prima fascia an inversion of the imperial, or a strange outlier or special case. Indeed, they are facing down some of the same challenges that vex Pliny. In this way they are fellow travelers. And yet even as they shape a Roman-inflected literacy, the rabbis curtail imperial logic, and move dramatically to thwart it. The primary gestures of this recusal are three: a contraction of the sacred library and a concomitant censorship of both Jewish and non-Jewish books; a refusal to trace non-Scriptural knowledge to books; and the refusal of authorship – they not only don’t depict themselves writing, but they do not attribute works to individual persons.

38In sum, the rabbis’ unusual focus on the book – the ways they reify the textuality of one small corpus, while stigmatizing the authorizing power of all others – deviates from the book-centered Jewish imagination that preceded them. Looking at the rabbinic project through the lens of the Roman book consumer, for whom the production and circulation of books was central to intellectual life, we might draw together a set of related aversions: in saying no to knowledge derived from books, no to writing, no to authorship, to Greek and Aramaic and most antecedent genres, the early rabbis create a literary cosmos that confronts the Roman reader/librarian with daunting barriers.

39In 1960s America, growing long hair as a man would have been a meaningless, illegible gesture were not close cropped heads symbolic of order, respect, and tradition. The language (inherited from the father) was hair – flowing tresses were the speech act.77 Similarly, abjuring authorship and books in the second century is a political decision, a Jewish speech act built from an imperial lexicon, and a distinct opting out. Rabbinic refusal to acknowledge Rome – to see or be seen by it, read or be read by it, a refusal, in short, to be literate in its shadow – adds up to a rather delicious sort of rebellion. Rabbinic inwardness then is less descriptive of an actual isolation, than is evidence of a prescriptive one. The rabbis as such is a sort of principled scholar-luddite avant la lettre.

40That said, theirs is not an esoteric undertaking – the rabbis have no real interest in hiding their books or their thought, yet nor do they want terribly to be seen.78 Might we gloss this reticence to write as a resistance to being read, collected, copied, quoted, shelved, or epitomized – in short, to becoming a primary source in someone else’s history? It may even indicate a preemptive refusal to be ignored. Books are shorthand for cultures composed of authors, readers, and collectors. The Jewish relationship to their own books and those of others puts us in the way of a certain form of provincial thinking, one contextualizable alongside key Hellenistic Jewish exemplars, and explainable in part as rabbinic responses to Roman questions.